Browse the corpus

Introduction Thyroid hormones have multiple effects on the cardiovascular system [1]. Although many of the characteristic signs and symptoms of hyperthyroidism are cardiovascular, cardiac failure is seen only in up to 6% cases [2] and is more common in older age with underlying ischemic or hypertensive cardiomyopathy [1,2]. There are very few reports in the literature of cardiac failure in juvenile hyperthyroidism in the absence of underlying heart disease [3]. A state of cardiac failure is known to cause changes in the thyroid hormone profile, especially low total tri-iodothyronine (T3) [4,5]. This phenomenon of non-thyroidal illness syndrome (NTIS) may be attributable to various mechanisms including changes in hypothalamic-pituitary axis, altered thyroid hormone binding and altered de-iodinase activity [6]. The same phenomenon has not been well documented in cardiac failure due to hyperthyroidism. Herein we describe an adolescent boy who presented in cardiac failure due to Graves' disease and had a paradoxical euthyroid profile. Case Report A 13 year old boy presented with palpitations of six months duration, fever and hyperdefecation for a month and generalized edema since three days. Fatigue, diaphoresis, tremors, polyphagia and weight loss were present for six months. He was diagnosed to have hyperthyroidism five months before presentation to us. He was started on carbimazole 15 mg daily at the time of diagnosis, which was increased to 45 mg daily one week prior to presentation at our hospital.

s. Fatigue, diaphoresis, tremors, polyphagia and weight loss were present for six months. He was diagnosed to have hyperthyroidism five months before presentation to us. He was started on carbimazole 15 mg daily at the time of diagnosis, which was increased to 45 mg daily one week prior to presentation at our hospital. On examination, the patient was febrile with a heart rate of 130 per minute and a blood pressure of 140/60 mm Hg. There was generalized edema and raised JVP (12 cm of water). He had exophthalamos. The thyroid gland was diffusely enlarged to approximately 60 grams and a bruit was heard over the thyroid. Cardiomegaly was present as well as a grade 3/6 apical ejection systolic murmur. There was mild weakness (grade 4 power) of hips, knees and shoulders, with hyperreflexia. Hepatosplenomegaly was present. Hemoglobin was 96 g/L (normal, 130-160 g/L), total leukocytes 4.1 × 109/L (normal, 4.5-13.5 × 109/L), and platelets 51 × 109/L (normal, 150-400 × 109/L). He had hyponatremia (serum Na 121 mEq/L; normal,135-145 mEq/L) and hypoalbuminemia (serum albumin 25 g/L; normal, 35-55 g/L). Blood and urine cultures, Widal test and smear for malarial parasite were negative. The ECG showed sinus tachycardia, normal QRS voltages and T wave inversion in precordial leads V2-V6. The chest radiograph was normal except for mild cardiomegaly (cardiothoracic ratio 54%). The echocardiogram showed mild pulmonary arterial hypertension, dilated right ventricle and tricuspid regurgitation with normal contractility of both the ventricles. There was no evidence of underlying congenital or acquired heart disease.

t radiograph was normal except for mild cardiomegaly (cardiothoracic ratio 54%). The echocardiogram showed mild pulmonary arterial hypertension, dilated right ventricle and tricuspid regurgitation with normal contractility of both the ventricles. There was no evidence of underlying congenital or acquired heart disease. Thyroid function tests revealed low T3 (0.77 nmol/L; normal, 1.3-2.8 nmol/L), normal total T4 (104.1 nmol/L; normal, 60-160 nmol/L) and free T4 (22.6 pmol/L; normal, 10-25 pmol/L) with a suppressed TSH (<0.15 mU/L; normal, 0.3-5 mIU/L). Thyrotropin receptor antibody titer was 28.5 IU/L by ELISA (normal, <1.5 IU/L). In addition to the supportive care, the patient was started on prednisolone 60 mg/day, propranolol 40 mg/day and carbimazole was continued. After three days of treatment, the signs of heart failure subsided; however, fever and tachycardia were persistent. Repeat T4 and free T4 now rose to hyperthyroid levels (Table 1), with serum albumin of 31 g/L. He was given potassium iodide drops for further symptomatic improvement. During the hospital stay he developed hyperglycemia probably caused by the combined effect of hyperthyroidism and glucocorticoid therapy, requiring insulin for two weeks. Table 1 Serial thyroid functions, clinical features and treatment

In addition to the supportive care, the patient was started on prednisolone 60 mg/day, propranolol 40 mg/day and carbimazole was continued. After three days of treatment, the signs of heart failure subsided; however, fever and tachycardia were persistent. Repeat T4 and free T4 now rose to hyperthyroid levels (Table 1), with serum albumin of 31 g/L. He was given potassium iodide drops for further symptomatic improvement. During the hospital stay he developed hyperglycemia probably caused by the combined effect of hyperthyroidism and glucocorticoid therapy, requiring insulin for two weeks. Table 1 Serial thyroid functions, clinical features and treatment Day T4 (60-160 nmol/L) Free T4 (10-25 pmol/L) T3 (1.3-2.8 nmol/L) TSH (0.3-5 mIU/L) Albumin (35-55 g/L) Clinical event Treatment 1 104.8 22.6 0.7 <0.15 25 CHF* Carbimazole Prednisolone Propranolol 4 205.3 58.0 1.7 31 Resolution of CHF* 8 135.0 49.3 Persistent fever and tachycardia KI† started 12 1.0 Fever resolved Steroid tapering started 4 week 80.8 36 Off KI† and steroids 9 week 35.0 Clinically hypothyroid Carbimazole dose reduced RaIA¶ done 4 month 119.8 <0.15 49 Clinically euthyroid 9 month** <12.9 >60 Radiation induce Hypothyroidism Thyroxine started *CHF: congestive heart failure, † KI: Potassium iodide, ¶ RaIA: Radioactive Iodine Ablation **: The patient was off carbimazole for 2 months at the time of this follow up.

Day T4 (60-160 nmol/L) Free T4 (10-25 pmol/L) T3 (1.3-2.8 nmol/L) TSH (0.3-5 mIU/L) Albumin (35-55 g/L) Clinical event Treatment 1 104.8 22.6 0.7 <0.15 25 CHF* Carbimazole Prednisolone Propranolol 4 205.3 58.0 1.7 31 Resolution of CHF* 8 135.0 49.3 Persistent fever and tachycardia KI† started 12 1.0 Fever resolved Steroid tapering started 4 week 80.8 36 Off KI† and steroids 9 week 35.0 Clinically hypothyroid Carbimazole dose reduced RaIA¶ done 4 month 119.8 <0.15 49 Clinically euthyroid 9 month** <12.9 >60 Radiation induce Hypothyroidism Thyroxine started *CHF: congestive heart failure, † KI: Potassium iodide, ¶ RaIA: Radioactive Iodine Ablation **: The patient was off carbimazole for 2 months at the time of this follow up. After three days of starting potassium iodide, the fever subsided and there was significant improvement in signs of thyrotoxicosis. The steroids and potassium iodide were tapered and omitted sequentially. Thyroid hormone levels gradually normalized after four weeks of treatment. A repeat echocardiogram showed mild mitral and tricuspid regurgitation, normal left ventricular contractility and right ventricular systolic pressure of 33 mm Hg (normal, <30 mm of Hg) suggestive of mild pulmonary hypertension.

ted sequentially. Thyroid hormone levels gradually normalized after four weeks of treatment. A repeat echocardiogram showed mild mitral and tricuspid regurgitation, normal left ventricular contractility and right ventricular systolic pressure of 33 mm Hg (normal, <30 mm of Hg) suggestive of mild pulmonary hypertension. Because of severe presentation and poor access to medical assistance from his native place, he was subjected to radio-iodine ablation after two months of presentation to us, with 10 mCi of radioactive I131. The thyroid scan done at this time revealed diffuse increase in tracer uptake. In subsequent follow up he was diagnosed to have radio-iodine induced hypothyroidism requiring thyroxine replacement (Table 1). Discussion Hyperthyroidism has multiple effects on the cardiovascular system including decreased systemic vascular resistance and increased resting heart rate, left ventricular contractility and blood volume leading to a state of high cardiac output [1]. A small percentage of patients with hyperthyroidism may present with clinical manifestations of cardiac failure [2].

the cardiovascular system including decreased systemic vascular resistance and increased resting heart rate, left ventricular contractility and blood volume leading to a state of high cardiac output [1]. A small percentage of patients with hyperthyroidism may present with clinical manifestations of cardiac failure [2]. Our patient was admitted with signs and symptoms suggestive of hyperthyroidism and cardiac failure. There was no evidence of underlying heart disease on clinical examination and echocardiography. The chest X-ray and ECG findings did not suggest primary pulmonary pathology or viral myocarditis. In a study of 591 adults with hyperthyroidism, heart failure was reported in 5.8% cases [2]. Atrial fibrillation increases the risk of cardiac failure but it can also occur in sinus rhythm [7]. There are very few reports of cardiac failure in juvenile hyperthyroidism. In a retrospective analysis of 21 children with hyperthyroidism, cardiac failure was observed in 2 patients [3]. There is some evidence of presence of thyrotoxic cardiomyopathy in children [8].

failure but it can also occur in sinus rhythm [7]. There are very few reports of cardiac failure in juvenile hyperthyroidism. In a retrospective analysis of 21 children with hyperthyroidism, cardiac failure was observed in 2 patients [3]. There is some evidence of presence of thyrotoxic cardiomyopathy in children [8]. In our patient, in the presence of normal left ventricular contractility, edema, raised JVP and hepatomegaly could be attributed to the pulmonary hypertension and right ventricular failure. In contrast to the fall in systemic vascular resistance, pulmonary arterial hypertension has been reported in as many as 47% of patients with hyperthyroidism [9]. The various underlying mechanisms are related to increased venous return due to increased blood volume, high output or autoimmune pulmonary endothelial injury or increased metabolism of certain vasodilator substances [10]. The increased blood volume (postulated to occur as a result of activation of renin-angiotensin-aldosterone system and increased red cell mass) causes volume overload of the right ventricle, which when unaccompanied by fall in pulmonary vascular resistance, leads to right ventricular failure and systemic venous congestion [10,11].

ed blood volume (postulated to occur as a result of activation of renin-angiotensin-aldosterone system and increased red cell mass) causes volume overload of the right ventricle, which when unaccompanied by fall in pulmonary vascular resistance, leads to right ventricular failure and systemic venous congestion [10,11]. The thyroid hormone profile of the patient revealed low T3 and paradoxically normal total and free T4. The hormones rose to hyperthyroid levels when the cardiac failure subsided. This can be explained by mechanisms similar to those causing thyroid function abnormalities in non-thyroidal illnesses, also known as sick euthyroid syndrome. It is a well known phenomenon in various diseases including cardiac failure [4,5]. The various underlying mechanisms include impaired hypothalamic TRH secretion, decreased binding proteins, decreased de-iodinase activity, presence of inhibitors of thyroid hormone binding, impaired tissue uptake and altered receptor expression [6]. The most common manifestation of this is a low T3. However, low T4 can also be seen in severe or prolonged illness [12]. Another reason for lower than expected T3 and T4 in our case could be the low albumin levels. Per se albumin has little influence on hormone levels, unless associated with changes in thyroglobulin and transthyretin. As all three are synthesized in liver, albumin level may serve as a surrogate marker for thyroglobulin [13]. Free T4 results in non thyroidal illness are variable and are dependent on the assay method used [6]. In our case the free T4 was measured by a solid phase radioimmunoassay and equilibrium dialysis method was not used. Free T3 measurements have been shown to have a modest decline in levels, however reliable estimation of free T3 is difficult [6].

idal illness are variable and are dependent on the assay method used [6]. In our case the free T4 was measured by a solid phase radioimmunoassay and equilibrium dialysis method was not used. Free T3 measurements have been shown to have a modest decline in levels, however reliable estimation of free T3 is difficult [6]. The phenomenon similar to sick euthyroid syndrome in hyperthyroidism has been previously documented in relation to myocardial infarction, diabetic ketoacidosis, pneumonia and fulminant hepatitis [14]. However to the best of our knowledge this is the first report of this phenomenon in cardiac failure due to juvenile hyperthyroidism. In addition to non thyroidal illness, other factors like prior use of carbimazole and concurrent iodine deficiency can influence the results of the thyroid function. Patients with hyperthyroidism on carbimazole therapy tend to normalize T4 ahead of T3, the latter being a better indicator of the clinical status. Hence, some patients may have normal T4 despite being clinically hyperthyroid [15]. However, T3 is usually high in such cases (isolated T3 toxicosis). The presence of low T3 and the rise of T4 and free T4 to hyperthyroid range on the same dose of carbimazole ruled out this possibility in our patient. The finding of normal T4 may also be seen in hyperthyroidism with co-existing iodine deficiency [16]. These patients have higher frequency of isolated T3 toxicosis. Since our patient consumed iodized salt, this mechanism was not invoked to explain the lower than expected T4.

In addition to non thyroidal illness, other factors like prior use of carbimazole and concurrent iodine deficiency can influence the results of the thyroid function. Patients with hyperthyroidism on carbimazole therapy tend to normalize T4 ahead of T3, the latter being a better indicator of the clinical status. Hence, some patients may have normal T4 despite being clinically hyperthyroid [15]. However, T3 is usually high in such cases (isolated T3 toxicosis). The presence of low T3 and the rise of T4 and free T4 to hyperthyroid range on the same dose of carbimazole ruled out this possibility in our patient. The finding of normal T4 may also be seen in hyperthyroidism with co-existing iodine deficiency [16]. These patients have higher frequency of isolated T3 toxicosis. Since our patient consumed iodized salt, this mechanism was not invoked to explain the lower than expected T4. The severity of the case required the use of glucocorticoids and potassium iodide for control of the symptoms. Glucocorticoids reduce de-iodination of T4 to T3. In addition, T4 secretion may be reduced due to direct thyroidal action and reduced thyroid stimulating antibodies [17]. Potassium iodide rapidly reduces the release of T4 and T3 from the thyroid gland and suppresses iodide oxidation and organification. When used for short term, it is useful for rapid symptomatic relief. To avoid stimulation of thyroid hormone synthesis, iodide should be given not earlier than 60 minutes after the start of the anti-thyroid medication [18].

release of T4 and T3 from the thyroid gland and suppresses iodide oxidation and organification. When used for short term, it is useful for rapid symptomatic relief. To avoid stimulation of thyroid hormone synthesis, iodide should be given not earlier than 60 minutes after the start of the anti-thyroid medication [18]. Some other less common features of Graves' disease noted in this case were leukopenia, thrombocytopenia [19], splenomegaly [20] and persistent fever [21]. Key Messages • Congestive cardiac failure occurring in the absence of atrial fibrillation or underlying heart disease is a rare manifestation of hyperthyroidism. • Occurrence of cardiac failure in hyperthyroidism may be associated with a paradoxically euthyroid hormone profile due to the phenomenon of sick euthyroid syndrome. Consent Written informed consent was obtained from the patient for publication of this case report. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Competing interests The authors declare that they have no competing interests. Authors' contributions GJ prepared the manuscript. VB critically reviewed and modified the manuscript. All authors were closely involved in care of the patient and interpretation of the unusual presentation and course. All authors read and approved the final manuscript.

Introduction Graves' disease (GD) is the most common cause of hyperthyroidism in children, adolescents and adults [1-3]. Treatments available for GD include anti-thyroid medications (methimazole or propylthiouracil), surgery and radioactive iodine (RAI) [4,5]. There is ongoing debate worldwide regarding the most suitable therapy for GD in pediatric patients. Although anti-thyroid medications are commonly used as first-line therapy for pediatric GD, long-term remission occurs in only 20% to 30% of pubertal cases and 15% of pre-pubertal cases treated pharmacologically [3,6-8]. Consequently, either surgery or RAI is needed to achieve a long-term cure in most pediatric GD patients. RAI therapy is generally considered to be safe, inexpensive and effective, with relatively few side effects [8-10]. Radioiodine was introduced for the treatment of GD more than 50 years ago [11], and at present is the most commonly used treatment for adult GD in the North America [12]. In 107 young GD patients who had been treated with RAI before age 20 years, no increased risk of adverse events was reported [13]. In some facilities, RAI is becoming the first-line therapy for GD in children and adolescents [14,15].

at present is the most commonly used treatment for adult GD in the North America [12]. In 107 young GD patients who had been treated with RAI before age 20 years, no increased risk of adverse events was reported [13]. In some facilities, RAI is becoming the first-line therapy for GD in children and adolescents [14,15]. The goal of iodine-131 therapy for pediatric GD is to induce hypothyroidism [16,17]. When children are treated with 330 μCi/g of iodine-131, hypothyroidism is achieved in nearly 95% of patients [18]. Higher dose ablative therapy (13.8 to 15.6 mCi) is effective in nearly all children with GD [19]. The use of high dose iodine-131 will destroy most thyroid tissue, thereby decreasing the risk of RAI-induced thyroid tumors, and is thus preferable especially in children [20]. The long-term risks of thyroid cancer appear to be lower when the thyroid gland is largely ablated than when residual thyroid tissue remains [21,22]. Changes in post-RAI thyroid volume have been investigated in adult GD patients [10,23-26], but not in pediatric and/or adolescent patients [3]. The objective of this retrospective study was to investigate changes in post-radioiodine thyroid volume in adolescent GD patients (< 20 years old) and also to examine whether these changes predict post-treatment hypothyroidism.

Introduction The association between vitiligo and autoimmune thyroid disease, especially Hashimoto's thyroiditis, has been characterized in adults [1-3]. Studies indicate an autoimmune etiology for vitiligo, and genes have been identified that cause both vitiligo and autoimmune thyroid disease [1-11]. However, little data are available on the association of vitiligo and thyroid disease in children [1-3]. In addition, reported studies of vitiligo and thyroid disease are based on studies of children who were identified as having vitiligo and subsequently evaluated for the presence of thyroid function abnormalities [1-7]. Graves' disease is an autoimmune condition with an estimated incidence of 1 in 10,000 children [12,13]. Graves' disease most commonly occurs in teenagers, and less than 5% of pediatric patients present at less than 5 years of age [14,15]. Hashimoto's thyroiditis is also an autoimmune disease associated with the presence of anti-thyroid antibodies that can result in either hypothyroidism or a euthyroid state [16,17]. Like Graves' disease, the peak prevalence is in the teenage years [14,15,18,19]. The incidence of antithyroid antibodies in the pediatric population is as high as 1% in some studies [16-20].

se associated with the presence of anti-thyroid antibodies that can result in either hypothyroidism or a euthyroid state [16,17]. Like Graves' disease, the peak prevalence is in the teenage years [14,15,18,19]. The incidence of antithyroid antibodies in the pediatric population is as high as 1% in some studies [16-20]. Both Graves' disease and Hashimoto's thyroiditis are associated with other autoimmune conditions including diabetes mellitus, inflammatory bowel disease, celiac disease, autoimmune hepatitis, adrenal insufficiency, and vitiligo [14,17-20]. To better characterize the association between vitiligo and thyroid disease in pediatric patients, we performed a retrospective analysis of patients diagnosed with thyroid disease. We now report the prevalence and clinical characteristics of children with vitiligo in the setting of Graves' disease and Hashimoto's thyroiditis.

er characterize the association between vitiligo and thyroid disease in pediatric patients, we performed a retrospective analysis of patients diagnosed with thyroid disease. We now report the prevalence and clinical characteristics of children with vitiligo in the setting of Graves' disease and Hashimoto's thyroiditis. Methods A retrospective chart review was performed on 333 children who had been evaluated for autoimmune thyroid disorders at the Yale Pediatric Thyroid Center over the last 5 years. Patient ages ranged from 1 to 21 years. Standard clinical diagnostic criteria were used to establish the diagnosis of Graves' disease or Hashimoto's thyroiditis [21]. As part of routine clinical evaluation, all children cared for at the Yale Pediatric Thyroid Center are evaluated for eye disease, liver disease, thyroid disease, and vitiligo. Vitiligo was diagnosed by standard criteria [4], and confirmed by a dermatologist. In the patients noted to have vitiligo, demographical and clinical features of the thyroid disease were recorded. The age at diagnosis of vitiligo and the timing in relationship to the diagnosis of thyroid disease were recorded. Results Of a total 333 children and adolescents with autoimmune thyroid disease, 9 (2.7%) were noted to have vitiligo (Tables 1 and 2). Eighty-seven had Graves' disease, and 4 of these patients with Graves' disease (4.6%) had vitiligo. Two hundred forty-six children had Hashimoto's thyroiditis, and 5 of these patients (2.0%) had vitiligo. Table 1 Patients with Graves' Disease and Vitiligo

Results Of a total 333 children and adolescents with autoimmune thyroid disease, 9 (2.7%) were noted to have vitiligo (Tables 1 and 2). Eighty-seven had Graves' disease, and 4 of these patients with Graves' disease (4.6%) had vitiligo. Two hundred forty-six children had Hashimoto's thyroiditis, and 5 of these patients (2.0%) had vitiligo. Table 1 Patients with Graves' Disease and Vitiligo Age of onset thyroid disease (yrs) Age of onset vitiligo (yrs) Gender 3.5 Unknown (present at presentation) M 3.5 Unknown (present at presentation) F 4 4.5 F 5 Unknown (present at presentation) M Table 2 Patients with Hashimoto's Thyroiditis and Vitiligo Age of onset thyroid disease (yrs) Gender TSH (uU/ml) 9.75 M 2 12 F 5.14 13 M 4.48 14.25 M 25 17.25 F 2.3 For the children with Graves' disease and vitiligo (Table 1), the average age of onset of thyroid disease was 4 ± 0.7 years (range 3.5 to 5 years). In 3 of the 4 children, vitiligo presented prior to the thyroid disease. One child developed vitiligo 6 months after Graves' disease was diagnosed. Males and females were represented equally among the patients with Graves' disease and vitiligo.

e age of onset of thyroid disease was 4 ± 0.7 years (range 3.5 to 5 years). In 3 of the 4 children, vitiligo presented prior to the thyroid disease. One child developed vitiligo 6 months after Graves' disease was diagnosed. Males and females were represented equally among the patients with Graves' disease and vitiligo. Of the patients with Hashimoto's thyroiditis and vitiligo (Table 2), all were clinically and biochemically euthyroid at the time of evaluation with the exception of one child with an elevated TSH level. Three of the 5 patients (60%) had positive anti-thyroid peroxidase (TPO) antibodies, and 2 (40%) had anti-thyroglobulin (Tg) antibodies. The average age at diagnosis of thyroid disease was 13.25 ± 2.8 years (range 9.75 to 17.25 years). Three of the children were males and 2 were females. Discussion Studying a large cohort of children with Graves' disease and Hashimoto's thyroiditis, we find that 4.6% and 2.0% of children, respectively, have vitiligo. Interestingly, the children with vitiligo and Graves' disease were much younger than those with Hashimoto's disease and vitiligo. Graves' disease is rare in the pediatric population, with a prevalence of about 1 per 10,000 individuals [12,13]. In children, the peak age of Graves' disease is 11-15 years, and it is three to five times more common in females than in males. In young children, no gender differences have been found [14,15].

Discussion Studying a large cohort of children with Graves' disease and Hashimoto's thyroiditis, we find that 4.6% and 2.0% of children, respectively, have vitiligo. Interestingly, the children with vitiligo and Graves' disease were much younger than those with Hashimoto's disease and vitiligo. Graves' disease is rare in the pediatric population, with a prevalence of about 1 per 10,000 individuals [12,13]. In children, the peak age of Graves' disease is 11-15 years, and it is three to five times more common in females than in males. In young children, no gender differences have been found [14,15]. When Hashimoto's thyroiditis occurs in pediatric patients, it is most common in adolescence, and is rare before the age of 3 [19]. The incidence of Hashimoto's in adolescence ranges from 1-2% of individuals [18-20]. However, an NHANES study found that 6.3% of adolescents from 12 to 19 years of age had positive anti-thyroglobulin antibodies and 4.8% had anti-thyroid peroxidase antibodies [16]. Another study of 160 children with antibodies consistent with euthyroid autoimmune thyroiditis had a mean age of 9.1 years [17].

ls [18-20]. However, an NHANES study found that 6.3% of adolescents from 12 to 19 years of age had positive anti-thyroglobulin antibodies and 4.8% had anti-thyroid peroxidase antibodies [16]. Another study of 160 children with antibodies consistent with euthyroid autoimmune thyroiditis had a mean age of 9.1 years [17]. As T-cell-mediated autoimmune diseases, Graves' disease and Hashimoto's thyroiditis have lymphocytic infiltration into thyroid parenchyma [10,14,19]. In Graves' disease, the antibodies bind to thyrotropin receptors, stimulating thyroid hormone production [10,14,19]. In Hashimoto's thyroiditis, lymphocytic infiltration leads to thyroid destruction [10]. Similarly, skin biopsies from vitiligo patients show dermal and epidermal lymphocytic infiltrate, consisting of activated T cells, which are thought to cause melanocyte destruction [8].

ulating thyroid hormone production [10,14,19]. In Hashimoto's thyroiditis, lymphocytic infiltration leads to thyroid destruction [10]. Similarly, skin biopsies from vitiligo patients show dermal and epidermal lymphocytic infiltrate, consisting of activated T cells, which are thought to cause melanocyte destruction [8]. Vitiligo is the most common acquired pigmentary disorder in children and adults, with an incidence of approximately 1% in the general population [1-5]. Vitiligo results from loss of melanocytes, which leads to well-demarcated depigmentation in macules or patches on skin, overlying hair, and/or mucus membranes. Onset of vitiligo prior to age 20 years occurs in approximately 50% of cases and before 10 years in approximately 25%, with a nearly equal gender ratio [1,4-7]. Several etiologies for vitiligo have been proposed, and significant evidence supports an autoimmune pathogenesis, with circulating autoantibodies that target melanocyte antigens and subsequently attack and destroy melanocytes [1,3-5]. This theory is supported by the recent identification of genes linked to vitiligo that are involved in innate immunity [5,6]. Studies have shown an association between vitiligo and thyroid disease with an 8-25% incidence of autoimmune thyroid disease in patients with vitiligo [2,3,5,6]. Hashimoto's thyroiditis is seen in the majority of adult patients with vitiligo and autoimmune thyroid disease [2,3,22]. As in our cohort, vitiligo was diagnosed before thyroid disease in most patients in these studies [1,3].

ase with an 8-25% incidence of autoimmune thyroid disease in patients with vitiligo [2,3,5,6]. Hashimoto's thyroiditis is seen in the majority of adult patients with vitiligo and autoimmune thyroid disease [2,3,22]. As in our cohort, vitiligo was diagnosed before thyroid disease in most patients in these studies [1,3]. The age of onset of vitiligo has been found to be earlier in families with a history of multiple autoimmune diseases [22]. This may be at least in part due to variants of genes including NALP1 that have been associated with susceptibility to vitiligo and autoimmune thyroid disease [11]. This gene regulates the immune system, including activating the inflammatory cascade, and is expressed on T cells and Langerhans' cells [11]. Studies have also identified immunogenetic associations with vitiligo and autoimmune thyroid diseases. Vitiligo is associated with HLA-DR4, and thyroid disease is associated with Class 1 and Class II HLA including HLA-DR [8-10]. Graves' disease has been associated with HLA-DR3 [9,14,20]. In comparison, Hashimoto's thyroiditis has not had a consistent HLA-association, but HLA-DR has been associated [9]. Other studies have shown an increase in CD4+ T-lymphocytes in addition to an elevated CD4+/CD8+ ratio in vitiligo patients [8], a finding that is present in patients with autoimmune thyroid disease [9,10].

comparison, Hashimoto's thyroiditis has not had a consistent HLA-association, but HLA-DR has been associated [9]. Other studies have shown an increase in CD4+ T-lymphocytes in addition to an elevated CD4+/CD8+ ratio in vitiligo patients [8], a finding that is present in patients with autoimmune thyroid disease [9,10]. Similarly, patients with vitiligo and at least one other autoimmune condition, including either Graves' disease or Hashimoto's thyroiditis, have been found to have a cytotoxic T lymphocyte antigen-4 (CTLA-4) polymorphism, which is involved in T cell apoptosis [8-10]. In Graves' disease, a specific CTLA-4 polymorphism has been associated with an early age of onset and severity of presentation [9]. Thus in the future, it will be interesting to characterize the genetic associations of vitiligo and other autoimmune diseases in the pediatric population. Overall, our observations of a significant incidence of vitiligo in children with thyroid disease shows that children with thyroid disease should be screened for vitiligo, especially young children with Graves' disease. Thyroid screening is already recommended annually for patients with vitiligo [1-6]. The occurrence of Graves' disease and vitiligo in young children further supports the notion that the nature of the autoimmune disorder in the young population differs from that seen in older children. Competing interests The authors declare that they have no competing interests.

Overall, our observations of a significant incidence of vitiligo in children with thyroid disease shows that children with thyroid disease should be screened for vitiligo, especially young children with Graves' disease. Thyroid screening is already recommended annually for patients with vitiligo [1-6]. The occurrence of Graves' disease and vitiligo in young children further supports the notion that the nature of the autoimmune disorder in the young population differs from that seen in older children. Competing interests The authors declare that they have no competing interests. Authors' contributions BP carried out the chart review and performed the data analysis. SR conceived of the study, and participated in its design and coordination. All authors read and approved the final manuscript.

The goal of iodine-131 therapy for pediatric GD is to induce hypothyroidism [16,17]. When children are treated with 330 μCi/g of iodine-131, hypothyroidism is achieved in nearly 95% of patients [18]. Higher dose ablative therapy (13.8 to 15.6 mCi) is effective in nearly all children with GD [19]. The use of high dose iodine-131 will destroy most thyroid tissue, thereby decreasing the risk of RAI-induced thyroid tumors, and is thus preferable especially in children [20]. The long-term risks of thyroid cancer appear to be lower when the thyroid gland is largely ablated than when residual thyroid tissue remains [21,22]. Changes in post-RAI thyroid volume have been investigated in adult GD patients [10,23-26], but not in pediatric and/or adolescent patients [3]. The objective of this retrospective study was to investigate changes in post-radioiodine thyroid volume in adolescent GD patients (< 20 years old) and also to examine whether these changes predict post-treatment hypothyroidism. Patients and Methods The medical records of all adolescent patients (< 20 years old) at Tajiri Thyroid Clinic who received a single RAI treatment for GD during the decade from January 2000 to January 2010 were examined retrospectively. The present study was approved by the Institutional Review Board of our clinic.

The objective of this retrospective study was to investigate changes in post-radioiodine thyroid volume in adolescent GD patients (< 20 years old) and also to examine whether these changes predict post-treatment hypothyroidism. Patients and Methods The medical records of all adolescent patients (< 20 years old) at Tajiri Thyroid Clinic who received a single RAI treatment for GD during the decade from January 2000 to January 2010 were examined retrospectively. The present study was approved by the Institutional Review Board of our clinic. GD was diagnosed based on elevated free thyroxine and suppressed thyrotropin concentrations, elevated TSH receptor antibodies (TRAb), and diffuse, elevated uptake of radioiodine or technetium-99 m within the thyroid. Thyrotropin, free thyroxine and TSH receptor antibody were measured by electrochemiluminescence immunoassay (Cobas e601; Roche Diagnostics, Tokyo, Japan).

nd suppressed thyrotropin concentrations, elevated TSH receptor antibodies (TRAb), and diffuse, elevated uptake of radioiodine or technetium-99 m within the thyroid. Thyrotropin, free thyroxine and TSH receptor antibody were measured by electrochemiluminescence immunoassay (Cobas e601; Roche Diagnostics, Tokyo, Japan). The iodine-131-absorbed radiation dose was calculated from RAIU and thyroid weight, using the formula: dose (μCi/g) = oral iodine-131 dose (mCi) × estimated 24 h RAIU (%) × 10/thyroid weight (g). Twenty-four hour RAI uptake was estimated using 4-hour uptake of iodine-123 [27] or 20-minute uptake of technetium-99 m [28]. Thyroid volume was estimated by ultrasound (SSA-350A; Toshiba Inc. Ltd., Tokyo, Japan) as previously reported [29]. Thyroid function (free thyroxine and thyrotropin) and ultrasonographic thyroid volume were determined at 1, 3, 5, 8 and 12 months after RAI therapy. When free thyroxine values dropped below 0.8 ng/dL and/or thyrotropin levels rose above 20 μIU/mL, replacement therapy with levothyroxine was initiated. Statistical analyses were performed using Student's t test and the chi-squared test. Values are shown as means ± standard deviation (SD). A P-value of less than 0.05 was considered to indicate a statistically significant difference.

The iodine-131-absorbed radiation dose was calculated from RAIU and thyroid weight, using the formula: dose (μCi/g) = oral iodine-131 dose (mCi) × estimated 24 h RAIU (%) × 10/thyroid weight (g). Twenty-four hour RAI uptake was estimated using 4-hour uptake of iodine-123 [27] or 20-minute uptake of technetium-99 m [28]. Thyroid volume was estimated by ultrasound (SSA-350A; Toshiba Inc. Ltd., Tokyo, Japan) as previously reported [29]. Thyroid function (free thyroxine and thyrotropin) and ultrasonographic thyroid volume were determined at 1, 3, 5, 8 and 12 months after RAI therapy. When free thyroxine values dropped below 0.8 ng/dL and/or thyrotropin levels rose above 20 μIU/mL, replacement therapy with levothyroxine was initiated. Statistical analyses were performed using Student's t test and the chi-squared test. Values are shown as means ± standard deviation (SD). A P-value of less than 0.05 was considered to indicate a statistically significant difference. Results There were 10 males and 39 females ranging in age from 12 to 19 years (mean ± SD, 16.4 ± 1.8 years old). All 49 patients were initially treated with anti-thyroid medications for 1 to 108 months (80% with methimazole, 20% with propylthiouracil). RAI therapy was performed due to lack of remission after 14 to 108 months (39%) of medical treatment, the development of a presumed toxic reaction to anti-thyroid drugs (44%; rash, arthralgia, hepatitis, neutropenia), or a desire for definitive therapy (17%). Anti-thyroid drugs were discontinued 3 to 5 days before RAI treatment. After administration of iodine-131, patients were treated with anti-thyroid drugs with or without propranolol to control symptoms of hyperthyroidism until hyperthyroxinemia abated.

, hepatitis, neutropenia), or a desire for definitive therapy (17%). Anti-thyroid drugs were discontinued 3 to 5 days before RAI treatment. After administration of iodine-131, patients were treated with anti-thyroid drugs with or without propranolol to control symptoms of hyperthyroidism until hyperthyroxinemia abated. The mean RAI dose for our 49 patients was 184 ± 84 μCi/g (range 44 - 393 μCi/g) and mean thyroid volume decreased significantly from 34.5 ml to 8.3 ml during the one year period after RAI (P < 0.00001). Based on thyroid functions at one year after RAI, patients were divided into two groups: 29 (59.2%) with overt hypothyroidism (hypothyroid patients) requiring levothyroxine replacement therapy (mean time until hypothyroidism: 4 ± 1.5 months, range 1 - 8 months) and 20 (40.8%) without hypothyroidism (non-hypothyroid patients) taking no medication at one year after RAI. The 20 non-hypothyroid patients consisted of 8 euthyroid and 12 hyperthyroid patients. Euthyroid patients included two who experienced transient hypothyroidism at 3 months after I-131 but had recovered without intervention at 1 year and one with subclinical hypothyroidism (TSH 6.53 μIU/ml (normal range: 0.20 - 3.30 μIU/ml)) at 1 year. Hyperthyroidism was subclinical in 8 patients and mild in 4 with serum free T4 levels of 1.93 - 2.03 ng/dl (normal range: 0.90 - 1.80 ng/dl). None of the hyperthyroid patients took anti-thyroid drugs at one year after RAI because all were asymptomatic. There were no statistically significant differences between euthyroid and hyperthyroid patients with respect to pre- and post-treatment thyroid volumes, or in percent volume reductions at 1, 3, 5, 8 and 12 months. The characteristics of adolescent GD patients who received RAI therapy, divided into hypothyroid and non-hypothyroid groups, are summarized in Table 1. There were no significant differences between the two groups in gender (P = 0.20), age (P = 0.18), pre-treatment thyroid volume (P = 0.30) or pre-treatment TRAb values measured by cosmic TRAb coated-tube kit (P = 0.45). The two groups differed only in the RAI dose administered (P = 0.048).

s, are summarized in Table 1. There were no significant differences between the two groups in gender (P = 0.20), age (P = 0.18), pre-treatment thyroid volume (P = 0.30) or pre-treatment TRAb values measured by cosmic TRAb coated-tube kit (P = 0.45). The two groups differed only in the RAI dose administered (P = 0.048). Table 1 Characteristics of Adolescent GD Patients Receiving Iodine-131 Therapy Hypothyroid (n = 29) Non-hypothyroid (n = 20) P-value Gender, female/male 20/9 19/1 NS (0.06) Age (yr) 16.1 ± 1.9 16.8 ± 1.6 NS (0.18) Pre-treatment thyroid volume (ml) 36.3 ± 17.5 31.9 ± 11.6 NS (0.30) Pre-treatment TRAb (%) 60.2 ± 27.8 (n = 22) 53.5 ± 29.1 NS (0.45) I-131 dose (μCi/g) 202.7 ± 88.6 156.3 ± 71.2 0.048 NS: not significant. All values are means ± SD.

-value Gender, female/male 20/9 19/1 NS (0.06) Age (yr) 16.1 ± 1.9 16.8 ± 1.6 NS (0.18) Pre-treatment thyroid volume (ml) 36.3 ± 17.5 31.9 ± 11.6 NS (0.30) Pre-treatment TRAb (%) 60.2 ± 27.8 (n = 22) 53.5 ± 29.1 NS (0.45) I-131 dose (μCi/g) 202.7 ± 88.6 156.3 ± 71.2 0.048 NS: not significant. All values are means ± SD. We compared changes in post-RAI administration thyroid volume during the one year period after treatment between hypothyroid and non-hypothyroid patients. As shown in Figure 1, post-treatment thyroid volume was significantly smaller in hypothyroid than in non-hypothyroid patients, especially after 8 months (8.2 ml vs. 13.1 ml at 8 months; P = 0.003, 6.3 ml vs. 10.9 ml at 12 months; P = 0.0005). The mean percent reduction in thyroid volume was significantly greater in hypothyroid than in non-hypothyroid patients at all measurement time points (P < 0.005) (Figure 2). We examined whether changes in thyroid volume predict post-treatment hypothyroidism at one year. The optimal cut-off point for predicting post-treatment hypothyroidism is considered to be a 50% reduction, as compared to the original volume of the thyroid gland, at 3 months after iodine-131 administration (Table 2). The sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, and negative likelihood ratio were 75.9%, 85.0%, 88.0%, 70.8%, 5.1 and 0.3, respectively.

uction, as compared to the original volume of the thyroid gland, at 3 months after iodine-131 administration (Table 2). The sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, and negative likelihood ratio were 75.9%, 85.0%, 88.0%, 70.8%, 5.1 and 0.3, respectively. Figure 1 Changes in thyroid volume during the one year period after iodine-131 therapy in hypothyroid and non-hypothyroid patients. Mean thyroid volumes in hypothyroid vs. non-hypothyroid patients were 36.3 ml vs. 31.9 ml at 0 months (m), 20.3 ml vs. 23.2 ml at 1 m, 13.6 ml vs. 18.8 ml at 3 m, 10.9 ml vs. 14.8 ml at 5 m, 8.2 ml vs. 13.1 ml at 8 m, and 6.3 ml vs. 10.9 ml at 12 m after iodine-131 therapy. Figure 2 Percent reductions in thyroid volume during the one year period after iodine-131 therapy in hypothyroid and non-hypothyroid patients. Mean percent reductions in thyroid volume in hypothyroid vs. non-hypothyroid patients were 42.6% vs. 24.2% at 1 month (m), 61.2% vs. 37.6% at 3 m, 70.1% vs. 46.5% at 5 m, 76.6% vs. 56.9% at 8 m, and 82.2% vs. 62.4% at 12 m after iodine-131 therapy. Table 2 Sensitivity and Specificity of Percent Reductions in Thyroid Volume at Each Measurement Time Point after Iodine-131 Therapy Month(s) after iodine-131 therapy 1 3 5 8 12 % reduction* Sens Spec Sens Spec Sens Spec Sens Spec Sens Spec 30% 75.9 55.0 40% 55.2 80.0 79.3 60.0 50% 41.4 95.0 75.9 85.0 92.6 53.3 60% 55.2 90.0 73.3 74.1 96.2 57.9 70% 51.9 93.3 73.1 78.9 85.2 60.0 80% 46.2 94.7 74.1 85.0 90% 18.5 100.0 *% reduction: cut-off point for percent reduction in thyroid volume

1 3 5 8 12 % reduction* Sens Spec Sens Spec Sens Spec Sens Spec Sens Spec 30% 75.9 55.0 40% 55.2 80.0 79.3 60.0 50% 41.4 95.0 75.9 85.0 92.6 53.3 60% 55.2 90.0 73.3 74.1 96.2 57.9 70% 51.9 93.3 73.1 78.9 85.2 60.0 80% 46.2 94.7 74.1 85.0 90% 18.5 100.0 *% reduction: cut-off point for percent reduction in thyroid volume Sens: sensitivity for predicting hypothyroidism at one year after iodine-131 therapy using %reduction Spec: specificity for predicting hypothyroidism at one year after iodine-131 therapy using %reduction We also examined the relationship between iodine-131 doses and post-RAI administration thyroid volume at 3 months in all 49 patients. Three iodine-131 doses (mean ± SD) were compared: 120 ± 32 μCi/g (n = 25, range 44 - 171 μCi/g), 200 ± 20 μCi/g (n = 11, range 174 - 224 μCi/g) and 300 ± 62 μCi/g (n = 13, range 225 - 393 μCi/g). When doses of 120 μCi/g, 200 μCi/g and 300 μCi/g were used, post-RAI administration thyroid volumes at 3 months were less than 50% of the pre-treatment thyroid volume in 36% (9/25), 55% (6/11) and 85% (11/13) of patients, respectively. We also found that doses of 120 μCi/g, 200 μCi/g and 300 μCi/g resulted in hypothyroidism at 1 year after RAI in 48% (12/25), 55% (6/11) and 85% (11/13) of the patients, respectively. Discussion This retrospective study showed mean post-treatment thyroid volume to be significantly decreased, from 34.5 ml to 8.3 ml (P < 0.00001), at one year after RAI in 49 adolescent GD patients (age range: 12 - 19 years), as has been demonstrated in adult GD patients (10,23-26).

We also examined the relationship between iodine-131 doses and post-RAI administration thyroid volume at 3 months in all 49 patients. Three iodine-131 doses (mean ± SD) were compared: 120 ± 32 μCi/g (n = 25, range 44 - 171 μCi/g), 200 ± 20 μCi/g (n = 11, range 174 - 224 μCi/g) and 300 ± 62 μCi/g (n = 13, range 225 - 393 μCi/g). When doses of 120 μCi/g, 200 μCi/g and 300 μCi/g were used, post-RAI administration thyroid volumes at 3 months were less than 50% of the pre-treatment thyroid volume in 36% (9/25), 55% (6/11) and 85% (11/13) of patients, respectively. We also found that doses of 120 μCi/g, 200 μCi/g and 300 μCi/g resulted in hypothyroidism at 1 year after RAI in 48% (12/25), 55% (6/11) and 85% (11/13) of the patients, respectively. Discussion This retrospective study showed mean post-treatment thyroid volume to be significantly decreased, from 34.5 ml to 8.3 ml (P < 0.00001), at one year after RAI in 49 adolescent GD patients (age range: 12 - 19 years), as has been demonstrated in adult GD patients (10,23-26). The goal of iodine-131 therapy for pediatric GD is to ablate the thyroid gland, in order to decrease the risk of RAI-induced thyroid tumors [18]. However, changes in post- RAI administration thyroid volume have not been investigated in pediatric and/or adolescent patients [3]. We found thyroid volume at one year after RAI administration for adolescent Graves' hyperthyroidism to be significantly smaller in hypothyroid than in non-hypothyroid patients (mean 6.3 ml vs. 10.9 ml; P < 0.001). As the post- RAI administration thyroid volume is smaller in hypothyroid patients, apparently conferring a lower risk of thyroid neoplasm development, this underscores the need for hypothyroidism to be a goal of therapy when using iodine-131 to treat GD in children [18].

d patients (mean 6.3 ml vs. 10.9 ml; P < 0.001). As the post- RAI administration thyroid volume is smaller in hypothyroid patients, apparently conferring a lower risk of thyroid neoplasm development, this underscores the need for hypothyroidism to be a goal of therapy when using iodine-131 to treat GD in children [18]. A correlation between changes in thyroid volume and thyroid function outcome in adult patients with GD has been described (10,23-25). However, these studies did not investigate the relationship between the degree of thyroid volume reduction and thyroid function outcome. Chiovato et al reported that the degree of thyroid volume reduction after RAI administration was the best predictor of early (within 1 year) thyroid function outcome in adult Graves' hyperthyroidism [26]. In fact, we found that about 90% of our patients had become hypothyroid at one year when thyroid volume was less than 50% of the original volume at 3 months after iodine-131 administration (positive predictive value 88%, sensitivity 75.9%, specificity 85.0%). We also found that 85% of patients treated with a dose of 300 μCi/g (range: 225 - 393 μCi/g) showed remarkable thyroid gland shrinkage (< 50% of the original thyroid gland volume at 3 months) and 85% were hypothyroid at one year. These data indicate that doses of approximately 300 μCi/g are needed to insure ablation of thyroid tissue. Our findings are thus consistent with those reported by Rivkees et al [18].

showed remarkable thyroid gland shrinkage (< 50% of the original thyroid gland volume at 3 months) and 85% were hypothyroid at one year. These data indicate that doses of approximately 300 μCi/g are needed to insure ablation of thyroid tissue. Our findings are thus consistent with those reported by Rivkees et al [18]. High thyroid-stimulating antibody levels before iodine-131 seem to be associated with a relative resistance to therapy (24, 26). On the other hand, TRAb levels did not show any predictive value for iodine-131 therapeutic outcome [25]. In our present study, pre-treatment TRAb values were not correlated with iodine-131 therapeutic outcome. In conclusion, thyroid volume progressively diminished for one year after iodine-131 administration for adolescent GD. Decreases were more significant in hypothyroid than in non-hypothyroid patients. We also demonstrated that approximately 90% of patients became hypothyroid within one year when thyroid volume was less than 50% of the original volume at 3 months after RAI therapy. We believe ultrasonographic thyroid volume measurement at 3 months after iodine-131 administration to be clinically useful for predicting post-treatment hypothyroidism. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors contributed to the development and writing of this manuscript and each has many years of clinical experience in the care of individuals with Graves' disease. All authors read and approved the final manuscript.

Background Children with type 1 diabetes (DM1) require multiple daily injections of insulin to maintain good glycemic control. The Diabetes Control and Complications Trial (DCCT) has shown that intensive insulin treatment using at least three times daily (TID) injections achieves superior blood glucose control with decreased risk of long term complications than conventional insulin treatment using once daily or twice daily (BID) injections [1,2]. However, this study was done when long acting insulin analogues (LAIA) were not available which limited the types of insulin regimens and there are limited randomized controlled trials assessing analogue insulins in children. Multiple daily injection regimens are not consistently superior in children and other factors including patient support and team cohesion play large roles in glycemic control [3]. Many patients find it difficult to adhere to TID injections since it is an invasive and painful therapy, which results in frequent insulin omission. By replacing bedtime intermediate acting insulin with a LAIA at dinner, this would allow children with DM1 to maintain their glycemic control with only a BID insulin injection regimen.

find it difficult to adhere to TID injections since it is an invasive and painful therapy, which results in frequent insulin omission. By replacing bedtime intermediate acting insulin with a LAIA at dinner, this would allow children with DM1 to maintain their glycemic control with only a BID insulin injection regimen. The pharmacokinetic properties of the LAIA, detemir, have some potential benefits in the treatment of children with DM1 [4-8]. The longer duration of action would allow this long acting insulin analogue to be incorporated into a BID insulin regimen that could potentially offer equivalent glycemic control to a TID injection regimen. Intermediate and rapid acting insulin could still be given in the morning to avoid a lunchtime injection, while rapid acting and long acting insulin could be given at dinner to cover for the meal plus the background insulin required until the morning. Therefore, children only need to take insulin twice a day, which may result in greater compliance and improved quality of life. Satisfaction with diabetes treatment may also be improved because of more predictable glycemic control and less frequent adverse events. The risk of hypoglycemia may be decreased with detemir [4], because of the flat and protracted pharmacodynamic profile compared to the peak in insulin activity seen with intermediate acting insulins. Less frequent episodes of hypoglycemia may also result in less weight gain that can be seen in intensive insulin therapy.

he risk of hypoglycemia may be decreased with detemir [4], because of the flat and protracted pharmacodynamic profile compared to the peak in insulin activity seen with intermediate acting insulins. Less frequent episodes of hypoglycemia may also result in less weight gain that can be seen in intensive insulin therapy. The primary objective of this pilot, randomized controlled trial is to compare the glycemic control as measured by HbA1c in children with DM1 treated with a BID regimen of insulin using a LAIA overnight versus a TID insulin injection regimen with intermediate acting insulin. Secondary objectives included assessing the satisfaction with treatment of diabetes in each group using the Diabetes Quality of Life Measure for Youths (DQOLY) and determining the frequency of adverse events (severe hypoglycemia, nocturnal hypoglycemia, mild hypoglycemia, diabetic ketoacidosis, and change in body mass index (BMI)). Methods Study Design The study was an open-labelled, randomized controlled trial design with two groups: control group (TID) and intervention group (BID). The trial was registered at ClinicalTrials.gov Identifier: NCT00522210.

The primary objective of this pilot, randomized controlled trial is to compare the glycemic control as measured by HbA1c in children with DM1 treated with a BID regimen of insulin using a LAIA overnight versus a TID insulin injection regimen with intermediate acting insulin. Secondary objectives included assessing the satisfaction with treatment of diabetes in each group using the Diabetes Quality of Life Measure for Youths (DQOLY) and determining the frequency of adverse events (severe hypoglycemia, nocturnal hypoglycemia, mild hypoglycemia, diabetic ketoacidosis, and change in body mass index (BMI)). Methods Study Design The study was an open-labelled, randomized controlled trial design with two groups: control group (TID) and intervention group (BID). The trial was registered at ClinicalTrials.gov Identifier: NCT00522210. Subjects Subjects were recruited from children with DM1 currently being followed at the Alberta Children's Hospital (Calgary, Alberta, Canada). Inclusion criteria were: children aged 6-17 years old, diagnosed with DM1 for at least 1 year and currently being treated with a TID regimen of insulin with rapid acting insulin and intermediate acting insulin. Exclusion criteria were: HbA1c ≥ 10% at enrolment, chronic underlying medical conditions that could affect glycemic control (examples: uncontrolled hypothyroidism, hyperthyroidism, celiac disease, etc.), current participants of other clinical trials, language or psychosocial barrier preventing the family from completing the study.

iteria were: HbA1c ≥ 10% at enrolment, chronic underlying medical conditions that could affect glycemic control (examples: uncontrolled hypothyroidism, hyperthyroidism, celiac disease, etc.), current participants of other clinical trials, language or psychosocial barrier preventing the family from completing the study. Protocol A sample of size of 65 subjects per group was initially calculated based on an estimated baseline HbA1c of 8.4% (SD 1.3%) in each group, a clinically acceptable difference of 10% (absolute difference of 0.84), a power of 90%, and a dropout rate of 20%. Therefore, if one assumed no drop-outs a sample size of 52 patients per arm would have been sufficient. Subjects were randomized into the control group or intervention group using a computer generated randomization sequence (the sequence was generated in blocks to keep groups as balanced as possible and to help ensure allocation concealment). Subjects were stratified by age groups (6 to 10 years, and greater than 10 years old). Given the nature of the intervention, it was not possible to blind patients, their parents, and caregivers to the treatment allocation.

ated in blocks to keep groups as balanced as possible and to help ensure allocation concealment). Subjects were stratified by age groups (6 to 10 years, and greater than 10 years old). Given the nature of the intervention, it was not possible to blind patients, their parents, and caregivers to the treatment allocation. In the control group (ie. the TID regimen), subjects were asked to continue on their usual insulin regimen (intermediate acting insulin and rapid acting analogue at breakfast, rapid acting analogue at dinner and intermediate acting insulin at bedtime). In the intervention group (ie. the BID regimen), the subjects' usual bedtime dose of intermediate acting insulin was discontinued and replaced with a dose of insulin detemir at dinner time. The intermediate acting insulin used was neutral protamine hagedorn (either Novolin NPH or Humulin N depending on what the patient was currently using). The rapid acting analogues used were lispro insulin (Humalog) or aspart insulin (Novorapid) depending on what the patient was currently using. The detemir was not mixed with the rapid acting insulin and was given as a separate injection. The dose of detemir was approximately 50% of the total daily dose of insulin, with the remaining 50% being comprised of the subject's breakfast dose of intermediate acting insulin and rapid acting analogue and dinner rapid acting analogue. A run-in period of 1 month, with a minimum of a weekly phone contact, was used to facilitate the change in insulin regimen and dose finding for the intervention group and to optimize insulin doses in the control group. No changes were made to the subjects' usual diet and exercise routines.

dinner rapid acting analogue. A run-in period of 1 month, with a minimum of a weekly phone contact, was used to facilitate the change in insulin regimen and dose finding for the intervention group and to optimize insulin doses in the control group. No changes were made to the subjects' usual diet and exercise routines. Throughout the study, monthly phone contact for insulin adjustments was done for both groups. In addition, subjects were assessed in clinic at baseline, 3 months and 6 months (Figure 1). This included height, weight, HbA1c, current insulin doses, episodes of severe hypoglycemia (glucose less than 4 mmol/L associated with a decreased level of consciousness, seizure, or coma), reported nocturnal hypoglycemia, mild hypoglycemia (glucose less than 4 mmol/L where the patient is able to self treat) and diabetic ketoacidosis (hyperglycemia and ketonuria associated with a pH < 7.3 and/or bicarbonate level < 15 mmol/l). Figure 1 Study design and subject follow up.

Throughout the study, monthly phone contact for insulin adjustments was done for both groups. In addition, subjects were assessed in clinic at baseline, 3 months and 6 months (Figure 1). This included height, weight, HbA1c, current insulin doses, episodes of severe hypoglycemia (glucose less than 4 mmol/L associated with a decreased level of consciousness, seizure, or coma), reported nocturnal hypoglycemia, mild hypoglycemia (glucose less than 4 mmol/L where the patient is able to self treat) and diabetic ketoacidosis (hyperglycemia and ketonuria associated with a pH < 7.3 and/or bicarbonate level < 15 mmol/l). Figure 1 Study design and subject follow up. The Diabetes Quality of Life Measure for Youths (DQOLY) was administered at baseline and again at 6 months. This questionnaire was initially used by the DCCT group and was later revised by Ingersoll et al [9]. It has been validated in youths aged 10-21 years. This instrument has three Likert scales including a 17 item diabetes life satisfaction scale (range of scores 17-85), 23 item disease impact scale (range of scores 23-115), and an 11 item disease related worries scale (range of scores 11-55) [9]. In this study, reverse scores were recorded for the impact and worries scales so that a higher score indicated a better quality of life. For the satisfaction scale, a higher score indicates higher satisfaction.

act scale (range of scores 23-115), and an 11 item disease related worries scale (range of scores 11-55) [9]. In this study, reverse scores were recorded for the impact and worries scales so that a higher score indicated a better quality of life. For the satisfaction scale, a higher score indicates higher satisfaction. Data Analysis Baseline demographic and clinical variables are presented as means with standard deviations (SD) for numerical variables and as proportions for categorical variables. The 95% confidence interval for the difference between groups is presented too. At each time point, HbA1c between the two groups was compared using a two sample t-test and confidence interval for the difference is provided to help assess non-inferiority. The true expected difference in HbA1c between the control and treatment group was taken to be zero. Based on previous follow up data from the ACH Diabetes Clinic, the mean HbA1C is estimated at 8.4% with a standard deviation of 1.3 for each group. A 10% relative difference was considered a clinically acceptable difference. Therefore, the non-inferiority margin was set at 0.84. Ethical approval The protocol, including subject information, informed consent, recruitment procedure, interventions and data collection has been approved by the Conjoint Health Research Ethics Board of the Faculty of Medicine, University of Calgary (Calgary, Alberta, Canada) in accordance with the Declaration of Helsinki and Tri-Council Guidelines.

including subject information, informed consent, recruitment procedure, interventions and data collection has been approved by the Conjoint Health Research Ethics Board of the Faculty of Medicine, University of Calgary (Calgary, Alberta, Canada) in accordance with the Declaration of Helsinki and Tri-Council Guidelines. Results Table 1 shows the enrolment characteristics of the subjects. There were no significant differences between the groups at baseline. In total, 18 subjects were enrolled (10 control, 8 intervention). The mean age at diagnosis of DM1 was 6.31 years (SD 2.91) for control and 7.76 years (SD 3.22) for intervention. The mean duration of DM1 was 5.96 years (SD 4.95) for control and 3.76 years (SD 3.37) for intervention. Table 1 Enrolment characteristics of subjects. Characteristic Control (TID) N = 10 Intervention (BID) N = 8 Difference between groups (95% confidence interval for the difference) Gender (Female/Male) 4/6 (40% female) 4/4 (50% female) 10% (-50.02% - 33.36%) Age groups (children < 10 years old/children ≥ 10 years old) 3/7 (30% children < 10 years old) 3/5 (37.5% children < 10 years old) 7.5% (-47.08% - 33.74%) Age at study enrolment (years) 12.26 (3.40) 11.52 (2.08) -0.74 (-3.52 - 2.03) Age at type 1 diabetes diagnosis (years) 6.31 (2.91) 7.76 (3.22) 1.45 (-1.68 - 4.58) Duration of diabetes (years) 5.96 (4.95) 3.76 (3.37) -2.20 (-6.37 - 1.98) Last HbA1c prior to enrolment (%) 8.54 (0.70) 8.70 (0.58) 0.16 (-0.48 - 0.80) Insulin dose (units/kg/day) 1.02 (0.40) 0.94 (0.24) -0.08 (-0.41 - 0.25)

Age at study enrolment (years) 12.26 (3.40) 11.52 (2.08) -0.74 (-3.52 - 2.03) Age at type 1 diabetes diagnosis (years) 6.31 (2.91) 7.76 (3.22) 1.45 (-1.68 - 4.58) Duration of diabetes (years) 5.96 (4.95) 3.76 (3.37) -2.20 (-6.37 - 1.98) Last HbA1c prior to enrolment (%) 8.54 (0.70) 8.70 (0.58) 0.16 (-0.48 - 0.80) Insulin dose (units/kg/day) 1.02 (0.40) 0.94 (0.24) -0.08 (-0.41 - 0.25) Body Mass Index (kg/m2) 20.62 (4.07) 20.99 (3.60) 0.38 (-3.46 - 4.22) No significant differences were found between the groups at baseline. Table shows Number (%) or Mean (Standard deviation). Difference is for intervention minus control. There were no significant differences in the mean HbA1c between control and intervention groups at 0 months [8.48 (SD 0.86) vs 8.57 (SD 1.13)], 3 months [8.47 (SD 0.50) vs 7.99 (SD 0.61)], or 6 months [8.42 (SD 0.63) vs 8.30 (SD 0.76)] (Table 2). Adverse events, such as DKA and reported hypoglycemic episodes, were similar in frequency in the control and intervention groups. There were no significant differences in body mass index or quality of life scales between groups (Table 2). The width of the confidence intervals for each of the outcome measures was large, likely due to the small sample size. However, they do indicate that the intervention was non-inferior when compared to the control group. Table 2 Results Baseline 3 Months 6 Months Control Intervention Difference (95% confidence interval) Control Intervention Difference (95% confidence interval) Control Intervention Difference (95% confidence interval)

There were no significant differences in the mean HbA1c between control and intervention groups at 0 months [8.48 (SD 0.86) vs 8.57 (SD 1.13)], 3 months [8.47 (SD 0.50) vs 7.99 (SD 0.61)], or 6 months [8.42 (SD 0.63) vs 8.30 (SD 0.76)] (Table 2). Adverse events, such as DKA and reported hypoglycemic episodes, were similar in frequency in the control and intervention groups. There were no significant differences in body mass index or quality of life scales between groups (Table 2). The width of the confidence intervals for each of the outcome measures was large, likely due to the small sample size. However, they do indicate that the intervention was non-inferior when compared to the control group. Table 2 Results Baseline 3 Months 6 Months Control Intervention Difference (95% confidence interval) Control Intervention Difference (95% confidence interval) Control Intervention Difference (95% confidence interval) HbA1c % 8.48 (0.86) 8.57 (1.13) 0.095 (-0.95 - 1.14) 8.47 (0.50) 7.99 (0.61) -0.48 (-1.06 - 0.095) 8.42 (0.63) 8.30 (0.76) -0.12 (-0.84 - 0.60) Body Mass Index (kg/m2) 20.62 (4.07) 20.99 (3.60) 0.38 (-3.46 - 4.22) 21.81 (6.40) 22.21 (3.48) 0.39 (-4.67 - 5.46) 21.04 (3.87) 21.75 (3.58) 0.71 (-3.03 - 4.45) DKA (episodes in last 3 months) 0 0 0 0 0 0 0 0 0 Severe Hypoglycemia (episodes in last 3 months) 0.10 (0.32) 0 -0.10 (-0.33 - 0.13) 0.30 (0.67) 0 -0.30 (-0.78 - 0.18) 0.1 (0.32) 0 -0.10 (-0.33 - 0.13) Mild/Moderate Hypoglycemia #/week 2.11 (1.40) 2.16 (1.65) 0.046 (-1.53 - 1.62) 2.17 (1.76) 1.87 (1.34) -0.30 (-1.85 - 1.25) 2.12 (1.31) 2.54 (1.36) 0.41 (-0.94 - 1.77)

DKA (episodes in last 3 months) 0 0 0 0 0 0 0 0 0 Severe Hypoglycemia (episodes in last 3 months) 0.10 (0.32) 0 -0.10 (-0.33 - 0.13) 0.30 (0.67) 0 -0.30 (-0.78 - 0.18) 0.1 (0.32) 0 -0.10 (-0.33 - 0.13) Mild/Moderate Hypoglycemia #/week 2.11 (1.40) 2.16 (1.65) 0.046 (-1.53 - 1.62) 2.17 (1.76) 1.87 (1.34) -0.30 (-1.85 - 1.25) 2.12 (1.31) 2.54 (1.36) 0.41 (-0.94 - 1.77) Nocturnal Hypoglycemia #/week 0.15 (0.34) 0.62 (1.03) 0.47 (-0.39 - 1.34) 0.12 (0.32) 0.72 (0.91) 0.59 (-0.18 - 1.36) 0.26 (0.57) 0.50 (0.72) 0.24 (-0.43 - 0.91) QOL Impact 83.78 (7.38) 90.37 (8.14) 6.60 (-1.50 - 14.70) --- --- --- 83.70 (17.57) 91.00 (8.45) 7.30 (-6.28 - 20.88) QOL Worries 41.62 (6.30) 42.00 (5.76) 0.37 (-6.10 - 6.85) --- --- --- 37.90 (8.55) 39.37 (7.29) 1.47 (-6.45 - 9.40) QOL Satisfaction 67.33 (8.37) 67.00 (10.42) -0.33 (-10.28 - 9.61) --- --- --- 65.00 (12.62) 72.50 (8.37) 7.50 (-3.06 - 18.06) Comparison of glycemic control, adverse events and quality of life scores between the control and intervention groups. No significant differences were detected between the groups at baseline, 3 months or 6 months. Dashed line indicates data was not collected at that time point. Control (TID) N = 10. Intervention (BID) N = 8. Difference is for intervention minus control. Table shows Mean (Standard Deviation).

ol and intervention groups. No significant differences were detected between the groups at baseline, 3 months or 6 months. Dashed line indicates data was not collected at that time point. Control (TID) N = 10. Intervention (BID) N = 8. Difference is for intervention minus control. Table shows Mean (Standard Deviation). Discussion Currently, a standard TID injection regimen with intermediate acting insulin at breakfast and bedtime, and rapid acting insulin at breakfast and dinner is often used in children. Families that opt for more intensive therapy can choose a continuous subcutaneous insulin infusion or a basal bolus regimen. However, these regimens are costly and require a significant amount of skill and effort from the family. In addition, patient compliance with multiple invasive and painful injections can be an issue when using multiple injections. Another challenge with exogenous insulin administration is hypoglycemia. While the results of the DCCT clearly demonstrated the importance of maintaining a near normal glucose level, the major adverse event reported in the intensive insulin treated subset of patients aged 13-17 years was a nearly three fold increase in severe hypoglycemic events [1,2]. The development of new LAIA offers the opportunity to simplify insulin regimens while achieving similar glycemic control.

near normal glucose level, the major adverse event reported in the intensive insulin treated subset of patients aged 13-17 years was a nearly three fold increase in severe hypoglycemic events [1,2]. The development of new LAIA offers the opportunity to simplify insulin regimens while achieving similar glycemic control. Detemir is a LAIA that has prolonged insulin absorption with less intra-patient variability in peak insulin activity as well as very minimal peak activity, in comparison with intermediate acting insulin [10,11]. The pharmacokinetic properties of detemir have been studied in patients with DM1 using a euglycemic glucose clamp technique [5,12]. In comparison to neutral protamine hagedorn (NPH) insulin, detemir resulted in a more stable serum concentration of insulin without the peak seen in NPH [5,12]. In addition, there were less fluctuations in the glucose infusion rates required with detemir in steady state compared to NPH in steady state [5,12]. Detemir has been shown to have a consistent pharmacokinetic profile in children, adolescents, and adults with DM1[7,8].

on of insulin without the peak seen in NPH [5,12]. In addition, there were less fluctuations in the glucose infusion rates required with detemir in steady state compared to NPH in steady state [5,12]. Detemir has been shown to have a consistent pharmacokinetic profile in children, adolescents, and adults with DM1[7,8]. Recently, a retrospective study by Cengiz et al [13] analyzed the same BID regimen (NPH and rapid acting insulin analogue at breakfast with insulin detemir and rapid acting insulin analogue at dinner) in children with new onset DM1 as an option prior to initiation of insulin pump therapy. They found that by 12 months after diagnosis of DM1, 49 of the patients had changed to pump therapy with a median HbA1c of 6.9% while 59 remained on the BID injection regimen with a median HbA1c of 7.2% [13]. The authors concluded that this BID regimen was effective in children with new onset diabetes and had similar glycemic control [13]. The findings of this retrospective study are consistent with our results, as we also did not find a difference in glycemic control for patients on TID versus the BID regimen. In contrast to Cengiz et al's study [13], our randomized control trial was aimed at assessing the effectiveness of a BID regimen in children with DM1 greater than 12 months as an option for families where intensive diabetes therapy was not feasible and improved quality of life could be achieved by simplifying the diabetes regimen.

to Cengiz et al's study [13], our randomized control trial was aimed at assessing the effectiveness of a BID regimen in children with DM1 greater than 12 months as an option for families where intensive diabetes therapy was not feasible and improved quality of life could be achieved by simplifying the diabetes regimen. HbA1c changes when using insulin detemir has been studied in children and adolescents with conflicting results [14-17]. Braun et al [16] reported improved HbA1c and fewer severe hypoglycemic episodes in a chart review study of children who were switched from evening NPH insulin to detemir. Interestingly, in this study a subset of children under 12 years of age were treated with a BID regimen and showed an improvement in HbA1c from 8.3% to 7.6% after changing from evening NPH to detemir [16]. Dundar et al [14] reported an improvement in HbA1c in children who changed from a basal bolus regimen with NPH to either glargine or detemir. This study was limited by the fact that it was retrospective and was small with only 15 patients in the detemir treated group [14]. In a large, prospective, 26 week, randomized study of 347 children, Robertson et al [15] reported no difference in HbA1c in children changed to basal bolus therapy with either NPH or detemir compared to pre-basal bolus therapy. Although HbA1c was not better in the detemir group, the risk of nocturnal hypoglycemia was 26% lower [15]. Kurtoglu et al [17] retrospectively assessed children that were initially on basal bolus regimens with NPH or glargine then switched to detemir. After 12 weeks of using detemir, HbA1c was improved and the frequency of hypoglycemic episodes was decreased [17]. Although we did not see a similar improvement in HbA1c in our patients using detemir, these studies examined detemir in basal bolus regimens rather than the BID regimen used in our study. It is also reassuring that our study did not find a worsening of glycemic control despite simplifying the insulin regimen.

sed [17]. Although we did not see a similar improvement in HbA1c in our patients using detemir, these studies examined detemir in basal bolus regimens rather than the BID regimen used in our study. It is also reassuring that our study did not find a worsening of glycemic control despite simplifying the insulin regimen. Hypoglycemia can be a side effect of intensive insulin therapy. Several studies have demonstrated that detemir is associated with a decreased frequency of hypoglycemia since it does not have a peak activity [4,18]. In a randomized, open, cross-over trial [18], detemir has been compared to NPH insulin in a basal bolus regimen in adults with DM1 and has been found to be as effective as NPH in maintaining glycemic control. Fewer patients experienced hypoglycemia with detemir compared to NPH [18]. Vague et al [4] compared detemir and NPH in 448 adult patients with DM1 in a basal-bolus regimen using twice daily detemir or NPH for basal coverage and rapid acting insulin with each meal. In their detemir group, more predictable glycemic control was seen during night-time plasma glucose monitoring using an intravenous line [4]. A significant reduction in the frequency of hypoglycemia as well as weight gain was also seen in their detemir group [4]. Interestingly, there was no difference in the HbA1c between the two groups after 6 months [4]. In our study, no differences in reported episodes of hypoglycemia were seen between groups.

s line [4]. A significant reduction in the frequency of hypoglycemia as well as weight gain was also seen in their detemir group [4]. Interestingly, there was no difference in the HbA1c between the two groups after 6 months [4]. In our study, no differences in reported episodes of hypoglycemia were seen between groups. Weight gain can be a concern when children are on intensified insulin regimens or have frequent hypoglycemia. Home et al [6] conducted a 16 week, randomized control trial of 408 patients with type 1 diabetes. HbA1c improved by 0.18% in the group using insulin detemir as the basal insulin compared to NPH insulin [6]. In addition, there was a decreased frequency of hypoglycemia and no weight gain in the group using insulin detemir compared to the NPH group which did have some weight gain [6]. Although the insulin regimen used in our study was not a basal bolus one, we did not find any significant changes in BMI between the groups. Diabetic ketoacidosis (DKA) was not seen in either the control or intervention group. This is not necessarily surprising given the short duration of this study and small sample size. Karges et al [19], compared the incidence of DKA in a cohort of 10 682 children and adolescents with DM1 being treated with either NPH insulin or a LAIA. They found that the incidence of DKA was significantly higher in patients using glargine or detemir compared to those using NPH [19]. However, all of the patients studied were on at least three or more insulin injections per day while our patients using detemir were on a BID regimen.

with either NPH insulin or a LAIA. They found that the incidence of DKA was significantly higher in patients using glargine or detemir compared to those using NPH [19]. However, all of the patients studied were on at least three or more insulin injections per day while our patients using detemir were on a BID regimen. Quality of life measures were not significantly different between the two groups using the DQOLY. A limitation of using the DQOLY was that this questionnaire has only been validated in youths aged 10-21 years [9]. Our study included children aged 6 years and older. However, only 1 subject was 7 years old and 5 subjects were 8 to 9 years old at enrolment and the DQOLY was the most practical and accessible measure to use at the time of this study. At the time the study was conducted, it was recommended that families could continue mixing the intermediate insulin and rapid insulin analogue in one syringe in the morning. However, at supper time the LAIA and rapid insulin were to be given in two separate injections. Recently, Nguyen et al [20] published a study of 14 children with type 1 diabetes who underwent continuous glucose monitoring and found that mixing insulin detemir with aspart had equivalent effects on blood glucose when compared with giving them as separate injections. The ability to mix insulin detemir with aspart at supper time would again simplify the regimen for families and potentially have a greater impact on satisfaction and compliance.

d found that mixing insulin detemir with aspart had equivalent effects on blood glucose when compared with giving them as separate injections. The ability to mix insulin detemir with aspart at supper time would again simplify the regimen for families and potentially have a greater impact on satisfaction and compliance. A significant limitation of this study is the small sample size. Interestingly, recruitment was difficult for this study. Once families received a description of the alternate BID insulin regimen compared to the traditional TID regimen, many did not want to risk being randomized to the control group. The theoretical benefits of decreased nocturnal hypoglycemia and twice daily insulin injection times was very attractive to families; particularly those that struggle with compliance. Conclusions The results of this pilot study demonstrate that using a BID insulin regimen incorporating a LAIA allows for maintenance of glycemic control despite a less intensive injection regimen. Ideally, a basal bolus regimen with multiple daily injections or an insulin pump would mimic physiologic insulin secretion most closely, but practically this is often difficult to achieve in young children who are dependent on a responsible adult to be available for injections.

spite a less intensive injection regimen. Ideally, a basal bolus regimen with multiple daily injections or an insulin pump would mimic physiologic insulin secretion most closely, but practically this is often difficult to achieve in young children who are dependent on a responsible adult to be available for injections. Simplifying to BID insulin regimens incorporating LAIA may be possible with no increase in adverse events and comparable HbA1c compared to standard TID regimens used in children, although larger clinical studies would be required to confirm this finding. Although no significant improvements were seen in DQOLY and nocturnal hypoglycemia, it is important that HbA1c remained stable, and suggests that this regimen is a viable option for families. Competing interests The authors declare that they have no competing interests. Authors' contributions JH designed the study; collected, analyzed and interpreted the data; and drafted the manuscript. CH, AN and DP significantly contributed in the study design, data analysis, data interpretation and revising of the manuscript. All authors read and approved the final manuscript. Acknowledgements This study was funded by the Alberta Children's Hospital Foundation. The Alberta Children's Hospital Foundation was not involved in the study design; data collection, analysis or interpretation; writing of the manuscript; or in the decision to publish the manuscript.

Introduction The CHARGE syndrome (MIM 214800) is an autosomal dominant or sporadic disorder of variable multisystemic congenital anomalies that occurs with an incidence of approximately 1 in 10,000 [1,2]. Heterozygous CHD7 (chromodomain helicase DNA-binding protein 7, MIM 608892) mutations have been identified in approximately 60%-70% of patients with clinically diagnosed CHARGE Syndrome and are most commonly due to de novo truncating mutations. Furthermore, CHD7 mutations are reported throughout the entire coding sequence of the gene without an apparent pattern or cluster, and meaningful genotype-phenotype correlations have not been recognized [1-3].

patients with clinically diagnosed CHARGE Syndrome and are most commonly due to de novo truncating mutations. Furthermore, CHD7 mutations are reported throughout the entire coding sequence of the gene without an apparent pattern or cluster, and meaningful genotype-phenotype correlations have not been recognized [1-3]. The first descriptions of this syndrome were provided by Hall and Hitner independently in 1979 [4,5], though it was in 1981 that Pagan and colleagues coined the acronym CHARGE to summarize its dominant features: coloboma, heart defects, atresia choanae, retarded growth and development, genital and/or urinary abnormalities, ear anomalies and/or deafness [6]. It was only in 2004 that the CHD7 gene was established as a genetic etiology for CHARGE syndrome by Vissers et al [7]. A wider spectrum of associated features has since emerged, albeit less consistently, and includes hyposmia or anosmia, cleft lip and palate, hypocalcemia [8,9], and tracheoesophageal fistula [10]. As such, CHARGE Syndrome has several overlapping clinical characteristics with DiGeorge syndrome [11], Kallmann syndrome, and Hypoparathyroidism, Sensorineural Deafness, and Renal Disease (HDR) (Barakat's syndrome) [12,13]. As the CHARGE phenotype continues to expand, particularly into the clinical purview of other conditions, its diagnosis becomes more challenging as well as increasingly inclusive.

eorge syndrome [11], Kallmann syndrome, and Hypoparathyroidism, Sensorineural Deafness, and Renal Disease (HDR) (Barakat's syndrome) [12,13]. As the CHARGE phenotype continues to expand, particularly into the clinical purview of other conditions, its diagnosis becomes more challenging as well as increasingly inclusive. Herein, we report a patient with a unique presentation of CHARGE syndrome, including primary hypoparathyroidism, bilateral multicystic dysplastic kidneys (MCDK), and an atypical limb anomaly; he carried a CHD7 mutation that has not been previously characterized. Our report expands the spectrum of phenotypes associated with CHD7 mutations.

patient with a unique presentation of CHARGE syndrome, including primary hypoparathyroidism, bilateral multicystic dysplastic kidneys (MCDK), and an atypical limb anomaly; he carried a CHD7 mutation that has not been previously characterized. Our report expands the spectrum of phenotypes associated with CHD7 mutations. Clinical Report The proband, an 18 year old African American male, presented to us initially for management of refractory hypocalcemia. He was born weighing 795 grams at 25 weeks gestation to a 31-year old woman with severe hypertension, who died in the postpartum period of a myocardial infarction. In infancy, he underwent cardiac surgery for a ventricular septal defect, and another to correct a right eyelid coloboma. Additionally, he had congenital hypothyroidism, bilateral sensorineural hearing loss, and severe global developmental delay; he began walking during his third year of life, remains unable to independently dress or tie shoe laces, and he has a vocabulary of fewer than 5 words. At a young age, the family had been told that he had Down syndrome, a diagnosis he carried until our meeting. Our initial consultation for hypocalcemia was during a hospitalization for complications of end stage renal disease secondary to MCDK, which was diagnosed in early life based on X-ray computed tomography (CT) showing dysplastic kidneys with multiple cysts. Interestingly, both a full sister and maternal half sister had severe hearing deficits and renal disease (Figure 1A). Unfortunately, neither was available for further characterization.

dary to MCDK, which was diagnosed in early life based on X-ray computed tomography (CT) showing dysplastic kidneys with multiple cysts. Interestingly, both a full sister and maternal half sister had severe hearing deficits and renal disease (Figure 1A). Unfortunately, neither was available for further characterization. Figure 1 A patient with a unique presentation of CHARGE syndrome and a G744S CHD7 mutation. A. Pedigree of a CHARGE patient with a novel CHD7 mutation. circle: female; square: male; arrow: proband; +: wild type allele. B. Photographs of the facies (I), eyes (II), and hands (III) of the CHD7 G744S heterozygous proband. C. Evolutionary conservation of the residue Gly744. ClustalW multiple alignment of partial protein sequence of CHD7 orthologs. The position of residue G744 altered by one heterozygous nucleotide change is marked by arrow and red letters in the corresponding segment of the multiple alignment. The amino acid residues that differ from the sequence of the human CHD7 protein are indicated blue. Gly744 residue is evolutionarily fully conserved in all fifteen available CHD7 orthologs. D. Structural model of the amino acid regions spanning 651-794 of the CHD7 protein obtained by sequence homology to a bacterial flagellar filament. The site of the mutation (indicated in magenta), which is predicted to be detrimental by POLYPHEN, lies on a protein interaction surface as indicated by SSPIDER.

orthologs. D. Structural model of the amino acid regions spanning 651-794 of the CHD7 protein obtained by sequence homology to a bacterial flagellar filament. The site of the mutation (indicated in magenta), which is predicted to be detrimental by POLYPHEN, lies on a protein interaction surface as indicated by SSPIDER. His physical examination revealed an overweight and short young man (weight 65.9kg, height 150.5cm or -3.58 SDS, body mass index 29.3 kg/m2), communicating primarily with hand gestures including pointing. He had brachycephaly with a flat facial profile, short forehead and facial asymmetry, slightly upslanted eyes with mild right ptosis, a scar from eyelid coloboma corrective surgery, and mildly prominent ears with low nasal bridge, upturned nasal tip, and smooth philtrum (Figure 1B.I and 1B.II). He had no cleft lip or palate. The chest was symmetrical. He had scars from cardiac surgery on the left anterior and posterior walls, and on the anterior abdominal wall at the site of a neonatal gastric tube placement. On examination of the extremities, his palmar creases were normal and he had a flexion deformity of the right thumb whereby this digit was fixed in the adducted position (Figure 1B.III). His genital examination revealed Tanner V pubic hair, and bilaterally descended testes consistent in size with early to mid-pubertal range (5 mL right and 8 mL left, using Prader Orchidometer).

ormal and he had a flexion deformity of the right thumb whereby this digit was fixed in the adducted position (Figure 1B.III). His genital examination revealed Tanner V pubic hair, and bilaterally descended testes consistent in size with early to mid-pubertal range (5 mL right and 8 mL left, using Prader Orchidometer). A CT scan of the head showed bilateral basal ganglia calcifications and scattered calcification in the frontal white matter and cerebellum, likely related to chronic hypcalcemia. Olfactory structures could not be evaluated. X-ray of his hands revealed no osseous abnormalities to explain his right thumb deformity. The initial laboratory results included low serum calcium (6.3 mg/dL, normal 8.3-10 mg/dL) normal albumin (4.2 g/dL, normal 3.5-5.8 g/dL), low phosphorus (1.1 mg/dL, normal 2.5-5 mg/dL), normal 25-hydroxy vitamin D (38 ng/mL, normal 20-100 ng/mL), slightly low magnesium (1.6 mg/dL, normal 1.9-2.7 mg/dL) coincident with low parathyroid hormone level (4.9 pg/mL, normal 15-65 pg/mL), indicating primary hypoparathyroidism. His low phosphorus was due to phosphate wasting associated with polyuria of end stage renal disease (ESRD). He had normal complete blood cell counts with no evidence of white cell line depression. Additionally, despite having early to mid-pubertal sized testes, his LH, FSH and Testosterone were not in the hypogonadal range (LH 4.1 mIU/mL, FSH 6.8 mIU/mL, Testosterone 505 ng/dL), which is not consistent with Idiopathic Hypogonadotropic Hypogonadism.

lood cell counts with no evidence of white cell line depression. Additionally, despite having early to mid-pubertal sized testes, his LH, FSH and Testosterone were not in the hypogonadal range (LH 4.1 mIU/mL, FSH 6.8 mIU/mL, Testosterone 505 ng/dL), which is not consistent with Idiopathic Hypogonadotropic Hypogonadism. In view of his dysmorphic features and questionable history of Down Syndrome (MIM 190685), a karyotype was obtained which showed a normal 46, XY configuration in 200 stimulated peripheral lymphocytes. Fluorescent in situ hybridization (FISH) for a chromosome 22q11.2 deletion, associated with DiGeorge Syndrome (MIM 188400), was negative. To exclude a Calcium-Sensing Receptor defect (MIM 601199), sequencing of the CASR gene was conducted of the coding regions and exon/intron boundaries and did not reveal a mutation. In light of the patient's constellation of hypoparathyroidism, renal disease, as well as deafness, we performed microarray and then sequencing analyses for mutations in GATA3 gene associated with the HDR syndrome (MIM 146255) [13], and none were found. Finally, CHD7 mutation analysis was performed on a genomic sample after PCR amplification of exons 2 to 38. The primers for all exons flanked the respective intron-exon junctions and direct sequencing was performed in both forward and backward directions using automated fluorescence di-deoxy sequencing methods. This revealed 6 heterozygous unclassified variants in the CHD7 gene including 2 missense and 4 silent changes (Table 1). Among the missense changes, a G744S (NP_060250) resulting from a c.2230G→A nucleotide change (NM-017780), was found to be conserved in available CHD7 orthologs (Figure 1C) and was not identified in 192 control subjects. The subject's asymptomatic father, the only family member available for genetic analysis, was found not to carry the G744S substitution. An informed consent for all genetic testing and for image publication was obtained from father and from the subject's legal guardian.

) and was not identified in 192 control subjects. The subject's asymptomatic father, the only family member available for genetic analysis, was found not to carry the G744S substitution. An informed consent for all genetic testing and for image publication was obtained from father and from the subject's legal guardian. Table 1 Sequence Analysis Revealed Six Unclassified Variants in the CHD7 Gene Nucleotide Change NM_017780 Amion Acid Change NP_060250 SNP ID* Missense changes c.2230G→A p.G744S - c.6478G→A p.A2160T rs61753399 Silent Changes c.309G→A p.S103S rs41272435 c.657C→T p.G219G - c.2124T→C p.S708S - c.7590A→G p.K2530K rs61742801 *Reported in the Single Nucleotide Polymorphism database dbSNP:http:www.ncbi.nlm.nih.gov//SNP

e NM_017780 Amion Acid Change NP_060250 SNP ID* Missense changes c.2230G→A p.G744S - c.6478G→A p.A2160T rs61753399 Silent Changes c.309G→A p.S103S rs41272435 c.657C→T p.G219G - c.2124T→C p.S708S - c.7590A→G p.K2530K rs61742801 *Reported in the Single Nucleotide Polymorphism database dbSNP:http:www.ncbi.nlm.nih.gov//SNP Structural Modeling of the CHD7 Protein The G744S mutation, which is located near chromodomain 1 in exon 4, was hypothesized to possess functional implications in the pathogenesis of CHARGE Syndrome. We constructed a structural model for the portion of the CHD7 protein containing the mutation based on homology to a bacterial flagellar filament (pdb-code 1UCU) [14] (See Figure 1D and additional file 1 for methods and alignment). Based on this model, SPPIDER [15] identified the site of the mutation as a possible binding site. G744 is located at the protein surface, where mutations are most likely to affect protein function and signaling compared to mutations in the protein interior. In agreement with this observation, the bioinformatics based approach POLYPHEN predicted a possible damaging effect of the mutation with a PISC SCORE: 1.58 (values below 0.5 are considered benign and values above 1.0 possibly/probably damaging) [16]. The PISC score contains a sequence based estimate of the accessibility of the mutation site, which is underestimated in comparison with the predictions of the structural model. Therefore, the sequence based PSIC score likely underestimates the damaging effect of our mutation.

above 1.0 possibly/probably damaging) [16]. The PISC score contains a sequence based estimate of the accessibility of the mutation site, which is underestimated in comparison with the predictions of the structural model. Therefore, the sequence based PSIC score likely underestimates the damaging effect of our mutation. Chd7 Murine Expression Studies Throughout murine development Chd7 is broadly expressed, including organs classically anomalous in CHARGE Syndrome, such as the eyes, heart, and ears. To investigate the distribution of Chd7 transcripts in the developing limbs, we performed in situ hybridization analysis using whole mount preparations. The following primers were used for amplification of the probe spanning nucleotides 8305 to 9140 of the murine Chd7 transcript (NM_001081417): 5′-CAGGTGGCTGGAGGAGAACCC-3′ and 5′-CTTTACAGGGCCCTCCCTCGGCC-3′. Amplicons were ligated to a Topo-TA vector (Invitrogen) and subcloned into pBluescript via EcoRI and XbaI restriction sites. The probe was labeled with Digoxigenin (DIG) for whole-mount in situ hybridization following standard procedures. No specific signals were detected using the respective sense probes. By embryonic day E11.5, the limb buds are clearly divided into proximal and distal elements. By E12, the handplate showed evidence of angular contours at its peripheral margin, corresponding to the location of the future digits. At E12.5-E13 early evidence of digital rays that are separated by the digital interzones were apparent. We found high expression levels of Chd7 throughout these steps of limb development in DIG-labelled whole mount embryos. Pronounced Chd7 expression was noted at the limb bud apical ectodermal ridge (AER) (Figure 2), a thickened layer of ectodermal cells at the distal tip of the developing limb bud, which is a crucial organizing region during limb formation. This Chd7 expression pattern supports a role for this gene in limb development.

. Pronounced Chd7 expression was noted at the limb bud apical ectodermal ridge (AER) (Figure 2), a thickened layer of ectodermal cells at the distal tip of the developing limb bud, which is a crucial organizing region during limb formation. This Chd7 expression pattern supports a role for this gene in limb development. Figure 2 Chd7 expression in the developing mouse. At E11.5 Chd7 is broadly expressed and shows prominent signals with the Chd7 antisense probe in DIG labeled whole mount in-situ hybridization in the region of the distal tip of the limb buds (see also enlargements on the right). T: telencephalic vesicle, E: corneal ectoderm overlying lens vesicle, M: maxillary component of first branchial arch, H: heart, FL: forelimb, HL: hindlimb.

ith the Chd7 antisense probe in DIG labeled whole mount in-situ hybridization in the region of the distal tip of the limb buds (see also enlargements on the right). T: telencephalic vesicle, E: corneal ectoderm overlying lens vesicle, M: maxillary component of first branchial arch, H: heart, FL: forelimb, HL: hindlimb. Discussion CHARGE is a phenotypically heterogeneous autosomal dominant disorder that has been recognized as a non-random cohesive syndrome since the identification of CHD7 mutations as an underlying etiology [1,2,7]. CHD7, at 8q12.1, encodes a protein of the chromodomain family [17,18]. The exact function of CHD7 has not been elucidated, however, in situ hybridization analysis during human development has demonstrated expression of this gene in the central nervous system, semicircular canals and the neural crest of the pharyngeal arches; thereby implicating the embryologic role of CHD7 in the development of the respective organs [19,20]. Furthermore, CHD7 mRNA expression was documented in the hypothalamus, pituitary and olfactory bulb in the rat and also demonstrated in both migratory and post-migratory GnRH neuronal cell lines [21]. Our subject carried a heterozygous G744S, not seen in controls; this missense change has been reported in one previous series and designated a polymorphism, though no functional studies were conducted and controls were not tested [22]. In this instance, the G744S change was seen in a family with three children with clinical CHARGE born to two different women and one man. A larger CHD7 rearrangement was additionally seen in all affected children, and found in mosaic form in the father's sperm cell DNA but not in lymphocytes. The G744S change, however, was found in two of the CHARGE children (the third was not tested) and in the completely unaffected father. Since the father was heterozygous and not mosaic for G744S, the authors considered G744S to be a non-pathogenic variant (Kohlhase J, unpublished data). CHD7 mutations have clinically variable expression, and clear genotype-phenotype correlations are not observed, even among patients with identical CHD7 mutations [1-3]. Asymptomatic carriers are reported as well, particularly in inherited forms [3].

d G744S to be a non-pathogenic variant (Kohlhase J, unpublished data). CHD7 mutations have clinically variable expression, and clear genotype-phenotype correlations are not observed, even among patients with identical CHD7 mutations [1-3]. Asymptomatic carriers are reported as well, particularly in inherited forms [3]. We can speculate that the disparate putative effect of this mutation is subject to yet undefined secondary genetic, epigenetic or environmental influences; this has been demonstrated in other genetic disease models, such as idiopathic hypogonadotropic hypogonadism (IHH) and Kallmann syndrome [23-25]. Our bioinformatics based structural analysis, protein alignment, and DNA sequencing in normal controls, provide supportive evidence that the G744S is a true deleterious mutation involving a highly conserved amino acid, likely disrupting a crucial protein interaction site.

pogonadism (IHH) and Kallmann syndrome [23-25]. Our bioinformatics based structural analysis, protein alignment, and DNA sequencing in normal controls, provide supportive evidence that the G744S is a true deleterious mutation involving a highly conserved amino acid, likely disrupting a crucial protein interaction site. In this age of sophisticated and readily available genetic analysis, the diagnosis of CHARGE appears to be transforming from the rigid fulfillment of conglomerates of specific major and minor criteria, towards an approach more inclusive of patients with atypical or attenuated phenotypes, as was demonstrated by Kim et al [26]. CHARGE and Kallmann syndromes (KS, KAL5, MIM 612370), though distinct developmental disorders, were noted to share features of impaired olfaction and hypogonadism; thus CHD7 was hypothesized to be involved in the pathogenesis of KS even in the absence of the CHARGE phenotype. Among 197 patients, they identified seven heterozygous mutations (two splice and five missense, absent in ≥ 180 controls) in three sporadic KS and four sporadic normosmic IHH patients. Thus, sporadic CHD7 mutations occurred in 6% of IHH/KS patients studied, allowing them to conclude that IHH/KS can represent a milder allelic variant of CHARGE syndrome. Furthermore, Jongmans et al also identified de novo CHD7 mutations in 3 of 56 mixed KS and nIHH subjects. Interestingly, in retrospect, their IHH patients with CHD7 mutations had some CHARGE features, including colobomas, deafness, ear anomalies, cleft palate and short stature; however, not to the degree that would fulfill traditional criteria [27]. Supported by this rational, we can concluded that our subject, with an atypical eyelid coloboma, hearing loss, severe developmental delay, ventricular septal defect, short stature, and abnormal facies, in addition to other more recently described features, such as a limb anomaly, primary hypoparathyroidism and interrupted pubertal development, may be included in this designation.

with an atypical eyelid coloboma, hearing loss, severe developmental delay, ventricular septal defect, short stature, and abnormal facies, in addition to other more recently described features, such as a limb anomaly, primary hypoparathyroidism and interrupted pubertal development, may be included in this designation. CHARGE syndrome was initially not considered to involve the limb. Several subsequent reports have shown associated limb anomalies and one series reported at least one limb anomaly in over one third of 172 CHARGE patients; furthermore, there did not appear to be a common limb anomaly in their cohort and minor abnormalities were included [28]. Prasad et al in 1997 were among the first to report severe limb abnormalities, including camptodactyly, tibial hemimelia and severe club-foot, in a patient with clinical CHARGE Syndrome [29]. It is notable that "club-foot", or congenital talipes equinovarus, is typically not due to osseous malformation. In 2007, Van de Laar et al reported 3 patients with heterozygous CHD7 truncating mutations in distinct exons, who displayed several limb malformations, including tibial aplasia, monodactyly and bifid femora [30]. Other limb defects, including triphalangeal thumb, polydactyly of the foot, ectrodactyly, and radial aplasia, have been reported as well [1,19,29,31]. Sanlaville et al in 2006, studied expression of CHD7 during human embryonic development detecting a weak signal in limb bud mesenchym at C14 [20]. Our patient, added to those previously described, highlighted by our demonstration of strong Chd7 expression in murine limb buds, further supports the suggestion that limb abnormalities should be a more recognized feature within the phenotypic spectrum of CHARGE syndrome. Moreover, homology modeling revealed a similarity to a bacterial flagellae protein [14], which in turn has high homology with the human TCN gene, encoding the human cytoskeletal protein titin; in concert with actin, titin plays a dominant role in human cytoskeletal development. Interestingly, direct comparison of titin and 1UCU sequences revealed a 19 amino acid gap of highly charge amino acids (spanning aa 741-759, EDPGVQKRRSSRQVKRKRY), which are most likely involved in functional differences between the two proteins and hence confer particular vulnerability to functional changes upon mutation.

erestingly, direct comparison of titin and 1UCU sequences revealed a 19 amino acid gap of highly charge amino acids (spanning aa 741-759, EDPGVQKRRSSRQVKRKRY), which are most likely involved in functional differences between the two proteins and hence confer particular vulnerability to functional changes upon mutation. In CHARGE Syndrome, anomalies of the urinary tract are reported in 10-40%, and include neurogenic bladder, duplex kidneys, renal ectopia or agenesis, horseshoe kidneys, and ureteral anomalies. [1,32,33]. To our knowledge, our patient is the first CHARGE patient to be reported with the severe phenotype of MCDK, requiring renal replacement therapy with dialysis followed by transplant. Cystic renal dysplasia is an anomaly of differentiation of the fetal kidney, whereby the kidney contains primitive ducts and non-renal tissues such as cartilage, fat, hematopoietic tissue, and often cysts. The most severe form of cystic renal dysplasia is MCDK, and most cases are unilateral. In subjects with CHARGE, only few reports of simple renal cysts exist [1,2,30]. Interestingly, one CHARGE subject with a unilateral right-sided dysplastic kidney also had significant limb anomalies, including right tibia aplasia, left tibia hypoplasia, and bilateral club feet [30]. The molecular events involved in multicystic dysplastic kidneys, in general, remain to be elucidated, though studies have suggested involvement of WNT-1 [34], FGFR3 [35], and PAX2 [36]. The PAX2 gene is associated with the renal coloboma syndrome (MIM 120330), a syndrome characterized by renal hypoplasia and insufficiency, vesicoureteric reflux, and optic disc coloboma. Interestingly, multicystic dysplastic kidney has been reported in one family with Renal Coloboma Syndrome [36]. In a study of the distribution pattern of the PAX2 gene in human embryos, Tellier et al demonstrated that PAX2 gene expression occurs in the primordia affected with CHARGE syndrome. Therefore, PAX2 was further analyzed in 34 patients fulfilling the diagnostic criteria of the CHARGE syndrome, though no deletions or nucleotide variations of the coding sequence were detected, suggesting that mutations of the PAX2 gene was not a cause of the CHARGE [37].

in the primordia affected with CHARGE syndrome. Therefore, PAX2 was further analyzed in 34 patients fulfilling the diagnostic criteria of the CHARGE syndrome, though no deletions or nucleotide variations of the coding sequence were detected, suggesting that mutations of the PAX2 gene was not a cause of the CHARGE [37]. Considering the embryonic expression of PAX2 reported, and the common clinical features of Renal Coloboma Syndrome with CHARGE, one can hypothesize that CHD7 may have a role in regulating PAX2 gene and therefore this overlapping pathway might be explored in CHARGE etiology, and perhaps contributes to the variable expression observed.

embryonic expression of PAX2 reported, and the common clinical features of Renal Coloboma Syndrome with CHARGE, one can hypothesize that CHD7 may have a role in regulating PAX2 gene and therefore this overlapping pathway might be explored in CHARGE etiology, and perhaps contributes to the variable expression observed. The significant clinical overlap and inherently variable features of CHARGE and DiGeorge Syndromes can make differentiating these initial diagnoses particularly challenging. Hypocalcemia has been reported in CHARGE, though hypoparathyroidism, specifically, has been implicated in only few cases [38,39]. A study comparing 25 CHARGE subjects with CHD7 mutations to a large cohort of subjects with 22q11.2 deletion syndrome, noted that features found more commonly in CHARGE syndrome included coloboma, choanal atresia, facial nerve palsy, tracheoesophageal fistula, and genital hypoplasia in boys. Interestingly, a high incidence of marked hypocalcemia was observed in their CHARGE study group (72%), and a pronounced spectrum of cell-mediated immunodeficiency ranging from lymphopenia (60%) to severe combined immunodeficiency (8%), was seen as well. Defects in humoral immunity were documented in 4 CHARGE patients and included severe hypogammaglobulinemia with decreased T-cell numbers, transient hypogammaglobulinemia during infancy, and immunoglobulin A deficiency [40]. An accurate distinction between these two entities can, therefore, be challenging but will influence genetic counseling; CHD7 mutations more typically occur sporadically, whereas 22q11.2 deletions are familial in 10% of cases [2,41].

transient hypogammaglobulinemia during infancy, and immunoglobulin A deficiency [40]. An accurate distinction between these two entities can, therefore, be challenging but will influence genetic counseling; CHD7 mutations more typically occur sporadically, whereas 22q11.2 deletions are familial in 10% of cases [2,41]. Conclusion In summary, we report an 18 year old male with CHARGE syndrome and a unique phenotype, including primary hypoparathyroidism, bilateral MCDK, a limb anomaly, disrupted testicular growth, and an atypical eyelid coloboma, who harbored a heterozygous G744S CHD7 mutation. Our case emphasizes that CHARGE features are perhaps even more heterogeneous than previously described and should include limb anomalies more universally. Additionally, the stringent fulfillment of the conventional CHARGE criteria should not strictly guide genetic analysis. Furthermore, our report highlights that the clinical overlap of CHARGE with DiGeorge, HDR, and Kallmann Syndromes can pose a diagnostic challenge to the clinician, but the correct designation can have a critical impact on treatment, anticipatory guidance, and genetic counseling. Competing interests The authors declare that they have no competing interests.

Conclusion In summary, we report an 18 year old male with CHARGE syndrome and a unique phenotype, including primary hypoparathyroidism, bilateral MCDK, a limb anomaly, disrupted testicular growth, and an atypical eyelid coloboma, who harbored a heterozygous G744S CHD7 mutation. Our case emphasizes that CHARGE features are perhaps even more heterogeneous than previously described and should include limb anomalies more universally. Additionally, the stringent fulfillment of the conventional CHARGE criteria should not strictly guide genetic analysis. Furthermore, our report highlights that the clinical overlap of CHARGE with DiGeorge, HDR, and Kallmann Syndromes can pose a diagnostic challenge to the clinician, but the correct designation can have a critical impact on treatment, anticipatory guidance, and genetic counseling. Competing interests The authors declare that they have no competing interests. Authors' contributions SJ lead and participated in the phenotyping and genotyping of our proband and in the characterization of the pedigree members; he also contributed to writing the manuscript. HK and LL lead the characterization of the CHD7 G744S mutation. HK also contributed to writing this manuscript. IM and WW performed the structural modeling of the CHD7 Protein. FL and ST guided the phenotyping and genotyping of the proband. IK conducted the Chd7 murine expression studies. JS and MS conducted phenotyping of the proband's renal pathology. EJD conceived of this study, oversaw the phenotyping and genotyping of the proband and his family, and was the supervising writer of this manuscript. All authors read and approved the final manuscript.

nd. IK conducted the Chd7 murine expression studies. JS and MS conducted phenotyping of the proband's renal pathology. EJD conceived of this study, oversaw the phenotyping and genotyping of the proband and his family, and was the supervising writer of this manuscript. All authors read and approved the final manuscript. Supplementary Material Additional file 1 Additional file 1includes further elaboration of chd7 protein modeling and alignement methods, as well as one figure of the protein sequence alignement. Click here for file Acknowledgements We would like to acknowledge support of the German Academic Exchange Service (DAAD) and the Baden-Wurttemberg Stiftung GmbH, for the bioinformatic analysis.

Introduction First described by Prader, Labhart and Willi in 1956 [1], this syndrome represents the most common genetic cause of obesity with an estimated incidence of 1:15,000 to 1:25,000 live births [2,3]. Reported prevalence rates vary among countries but both sexes appear to be equally affected. Prader-Willi syndrome (PWS) is the first human syndrome identified with genomic imprinting [4]. The original descriptions of this syndrome included short stature, hypotonia, hypogonadism and mental retardation [1]. As infants grow to age 2-4 years, failure to thrive related, at least in part, to poor muscle tone and poor suck are replaced by increased appetite and food intake resulting in obesity and its comorbidities. Early diagnosis and intervention to prevent obesity and the associated complications are critical. Genetic testing and genetic counseling Candidate genes for Prader-Willi syndrome have been located on the long arm of chromosome 15q11-q13. These genes are physiologically imprinted and silenced on the maternally inherited chromosome. PWS arises when the paternally derived genes are missing, defective or silenced. The frequencies of each are shown in Table 1. Table 1 Frequency of genetic subtypes associated with PWS Subtype Frequency Paternal deletion of chromosome 15q11-q13 (type I or II) 75% Maternal uniparental disomy (UPD) 24% Imprinting center defects (ID) 1%

Genetic testing and genetic counseling Candidate genes for Prader-Willi syndrome have been located on the long arm of chromosome 15q11-q13. These genes are physiologically imprinted and silenced on the maternally inherited chromosome. PWS arises when the paternally derived genes are missing, defective or silenced. The frequencies of each are shown in Table 1. Table 1 Frequency of genetic subtypes associated with PWS Subtype Frequency Paternal deletion of chromosome 15q11-q13 (type I or II) 75% Maternal uniparental disomy (UPD) 24% Imprinting center defects (ID) 1% Translocation < 1% High resolution chromosomal analysis (HRCA) is done along with the fluorescence in situ hybridization (FISH) to detect deletions and translocation of chromosome 15 [5]. Deletion has been divided in type I (TI) and II (TII) according to the size. Studies indicate that individuals with the TI (~500 kb larger than TII) generally have more behavioral and psychological problems than individuals with the TII and UPD [6]. Negative FISH or karyotype analysis does not exclude the diagnosis and thus if done first should be followed by DNA methylation analysis. DNA methylation analysis is the only technique which can both confirm and reject the diagnosis of PWS, and therefore should typically be the investigation of choice. This is most commonly done using DNA methylation-specific techniques at the SNURF-SNRPN locus [7,8]. If DNA methylation analysis shows only a maternal pattern, then PWS is confirmed. Further methods may then be performed to determine the genetic subtype and allow appropriate genetic counseling. DNA methylation analysis has a sensitivity exceeding 99%; however, it does not differentiate between deletion, UPD and imprinting defect. In order to distinguish a maternal UPD from an imprinting defect, further DNA polymorphism analysis should be performed on the proband and parents [9,10].

opriate genetic counseling. DNA methylation analysis has a sensitivity exceeding 99%; however, it does not differentiate between deletion, UPD and imprinting defect. In order to distinguish a maternal UPD from an imprinting defect, further DNA polymorphism analysis should be performed on the proband and parents [9,10]. Most cases of Prader-Willi syndrome occur sporadically. The overall recurrence risk is dependent on the type of molecular defect. In families where the proband has either maternal disomy or deletion, the recurrence risk is small (less than 1%). Patients with an imprinting defect warrant further investigation in a specialized laboratory to determine whether an imprinting center deletion is present. Those families with a child with an imprinting center deletion have a recurrence risk of up to 50% if the father of the child is a carrier for the imprinting center deletion [11]. When a deletion is the result of a translocation or structural rearrangement involving chromosome 15, then the recurrence risk can be high. The actual risk in individual families depends upon the rearrangement which they carry. Overall, the risk of recurrence in the case of chromosomal translocations has been estimated up to 15%. In the future the methylation-specific multiplex ligation PCR amplification may be more widely used because it has the advantage of combining dosing and DNA methylation analysis in one assay, thus distinguishing different subtypes [12].

Most cases of Prader-Willi syndrome occur sporadically. The overall recurrence risk is dependent on the type of molecular defect. In families where the proband has either maternal disomy or deletion, the recurrence risk is small (less than 1%). Patients with an imprinting defect warrant further investigation in a specialized laboratory to determine whether an imprinting center deletion is present. Those families with a child with an imprinting center deletion have a recurrence risk of up to 50% if the father of the child is a carrier for the imprinting center deletion [11]. When a deletion is the result of a translocation or structural rearrangement involving chromosome 15, then the recurrence risk can be high. The actual risk in individual families depends upon the rearrangement which they carry. Overall, the risk of recurrence in the case of chromosomal translocations has been estimated up to 15%. In the future the methylation-specific multiplex ligation PCR amplification may be more widely used because it has the advantage of combining dosing and DNA methylation analysis in one assay, thus distinguishing different subtypes [12]. Clinical Presentation Infants with Prader-Willi syndrome present with neonatal hypotonia, hypoplasia of the clitoris/labia minora in girls and small penis and undescended testis in boys. Their hypotonia is associated with poor suck and feeding, often resulting in failure to thrive. Mothers may report decreased fetal activity and infants are often found in the breech position at the time of delivery. Clinical features include increased neonatal head:chest circumference ratio, narrow bifrontal diameter, dolichocephaly, almond shaped eyes, downturned angles of the mouth with abundant and thick saliva, small hands and feet with straight borders of the ulnar side of the hands and inner side of the legs. The presence of some of these features associated with neonatal hypotonia should alert physicians for early diagnosis of PWS during infancy. These features may become more prominent by age 2-3 years (Figures 1 and 2). Excessive eating and obsession with food generally begins in the preschool age group and will lead to morbid obesity if not controlled.

tures associated with neonatal hypotonia should alert physicians for early diagnosis of PWS during infancy. These features may become more prominent by age 2-3 years (Figures 1 and 2). Excessive eating and obsession with food generally begins in the preschool age group and will lead to morbid obesity if not controlled. Figure 1 Typical Facial Features of Child with Prader-Willi syndrome (Photograph with Permission). Figure 2 Typical Findings of Hands and Legs in Individuals with Prader-Willi syndrome. As these individuals age, manifestations, such as obesity, short stature, hypogonadism, skin picking, learning disabilities, behavioral and psychiatric problems become more evident. Consensus criteria for the clinical diagnosis of PWS were first established in 1993 by Holm et al [13]. These criteria were used until the introduction of the highly sensitive genetic testing, described above. Currently, these criteria are used as a screening tool for determining the need for further PWS specific genetic testing. In many infants poor cry and unexplained hypotonia may be the only clear clinical manifestations and indication for genetic testing. As many as 16.7% of patients diagnosed with molecular testing do not meet these clinical diagnostic criteria, therefore a revised clinical criteria to help identify the appropriate patients for DNA testing was proposed in 2001 [14] and modified in 2008 [15]. See Table 2 for composite, including additional features suggested by authors. Table 2 Indications for DNA testing

As many as 16.7% of patients diagnosed with molecular testing do not meet these clinical diagnostic criteria, therefore a revised clinical criteria to help identify the appropriate patients for DNA testing was proposed in 2001 [14] and modified in 2008 [15]. See Table 2 for composite, including additional features suggested by authors. Table 2 Indications for DNA testing Age at assessment Features sufficient to prompt DNA testing Birth to 2 yr Hypotonia with poor suck 2-6 yr Hypotonia with a history of poor suck Global developmental delay Short stature and/or growth failure associated with accelerated weight gain 6-12 yr Hypotonia with a history of poor suck (hypotonia often persists) Global developmental delay Excessive eating (hyperphagia, obsession with food) with central obesity if uncontrolled Short stature and/or decreased growth velocity* 13 yr through adulthood Cognitive impairment, usually mild mental retardation Excessive eating (hyperphagia, obsession with food) with central obesity if uncontrolled Short stature and/or decreased growth velocity* Hypothalamic hypogonadism and/or typical behavior problems (including temper tantrums and obsessive-compulsive features) *Features added by the authors

usually mild mental retardation Excessive eating (hyperphagia, obsession with food) with central obesity if uncontrolled Short stature and/or decreased growth velocity* Hypothalamic hypogonadism and/or typical behavior problems (including temper tantrums and obsessive-compulsive features) *Features added by the authors Endocrine Issues Hypogonadism Hypothalamic and pituitary dysfunction is most commonly manifested as hypogonadism, obesity and short stature. Hypogonadism with genital hypoplasia (cryptorchidism, scrotal or clitoral hypoplasia) can be identified in the newborn period. Cryptorchidism is present up to 86% of boys from birth [16,17] Undescended testes should be treated within the first year of life. There is evidence that early damage to the germ cells that produce sperm begins at this age. Scrotal hypoplasia and small penis however can make orchiopexy and circumcision difficult in infants with PWS. Repeat surgical interventions are frequently required, especially in those infants with underdeveloped scrotal sacs.

There is evidence that early damage to the germ cells that produce sperm begins at this age. Scrotal hypoplasia and small penis however can make orchiopexy and circumcision difficult in infants with PWS. Repeat surgical interventions are frequently required, especially in those infants with underdeveloped scrotal sacs. The most effective treatment for undescended testes is surgery. The Committee on Genetics, American Academy of Pediatrics however, recommends a therapeutic trial of human chorionic gonadotropin (hCG) for treatment of undescended testes before surgery, because avoidance of general anesthesia is desirable for infants with low muscle tone and potential for underlying respiratory compromise [18]. The precise mechanism of action in regards to testicular descent is unknown but benefits of a course of hCG may include increased scrotal size and partial normalization of phallus length, thereby improving surgical outcomes for undescended testes and facilitating later standing micturition. Premature adrenarche (PA) is the precocious appearance of pubic and/or axillary hair and less commonly an apocrine odor, comedones, and acne without other signs of puberty or virilization. PA is usually seen before age 8 and 9 years, in girls and boys respectively. PA has been reported in 57% of children receiving GH therapy [19] but in general, pubertal development in PWS is characterized by normal adrenarche, pubertal arrest, and hypogonadism due to variable combinations of a unique primary gonadal defect and hypothalamic dysfunction [20,21].

rs, in girls and boys respectively. PA has been reported in 57% of children receiving GH therapy [19] but in general, pubertal development in PWS is characterized by normal adrenarche, pubertal arrest, and hypogonadism due to variable combinations of a unique primary gonadal defect and hypothalamic dysfunction [20,21]. At some stage almost all subjects will require sex hormone replacement therapy. Mental retardation should not be a contraindication to allow normal pubertal development or preclude sex hormone replacement at any age in those affected individuals. Regardless of body weight, patients with PWS have increased body fat content. Those individuals with low body weight or significant low sex hormone levels during adolescence and adulthood should be considered for sex hormone treatment. There is no consensus as to the most appropriate regimen for sex hormone replacement therapy in PWS. Intramuscular testosterone is given every 3 -4 weeks. Testosterone gel preparations can be useful in selected cases, although precautions must be taken to avoid cross-contamination. Whatever preparation is preferred, the initial dose should be one third to one half of the normally recommended androgen dose to prevent the aggressive behavior occasionally seen in some individuals. In females with PWS, the use of gonadal hormone replacement should be considered if there is amenorrhea/oligomenorrhea or decreased bone mineral density (BMD) in the presence of reduced estradiol levels. Hypogonadism is a common but not necessarily universal finding in adults with PWS [16,17]. Sexual counseling and contraceptive treatment should be used as appropriate, especially in the presence of complete sexual maturation, including regular menses. There are a few case reports of pregnancy in females with PWS [22,23]. Their cognitive dysfunction, social and emotional immaturity and the risk of Angelman syndrome in offspring of PWS deletion mothers prompt us to advise against pregnancy. At present there are no reports of paternity in PWS. Estrogen and androgen status should be monitored yearly during adolescence and adulthood and BMD assessed as indicated by dual-energy x-ray photon absorptiometry.

he risk of Angelman syndrome in offspring of PWS deletion mothers prompt us to advise against pregnancy. At present there are no reports of paternity in PWS. Estrogen and androgen status should be monitored yearly during adolescence and adulthood and BMD assessed as indicated by dual-energy x-ray photon absorptiometry. Adrenal insufficiency When the pituitary begins to fail, there is generally a specific sequential failure of pituitary hormones, starting with growth hormone (GH), continuing through luteinizing (LH) and follicle stimulating hormone (FSH) deficiency, and culminating in the loss of thyrotropin stimulating hormone (TSH) and adrenocorticotropic hormone (ACTH). Generally ACTH is the last to be affected. Hypothalamic dysfunction is characteristic of individuals with PWS, therefore the clinical manifestations of pituitary hormone deficiency are expected. Short stature and hypogonadism, as a result of GH and gonadotropin (LH and FSH) deficiencies are seen in most individuals with this genetic syndrome. Under normal conditions the secretion of cortisol, the main adrenal glucocorticoid in humans is under the dominant control of pituitary ACTH. Clinical manifestations of adrenal insufficiency, however are uncommon in individuals with PWS.

(LH and FSH) deficiencies are seen in most individuals with this genetic syndrome. Under normal conditions the secretion of cortisol, the main adrenal glucocorticoid in humans is under the dominant control of pituitary ACTH. Clinical manifestations of adrenal insufficiency, however are uncommon in individuals with PWS. The circadian peak serum cortisol usually occurs around 0800 hours. An extremely low basal serum cortisol at this time, below 100 nmol/liter (3.62 mcg/dl), may be assumed to demonstrate true cortisol deficiency [24]. However, levels at other times have little diagnostic utility. For this reason, various dynamic tests including insulin tolerance test (ITT), ACTH stimulation and metyrapone test have been devised to assess whether the patient can provide a stress-induced rise in cortisol similar to a normal person. The insulin tolerance test evaluates the integrity of the entire hypothalamic-pituitary axis (HPA) by inducing symptomatic and biochemical hypoglycemia, with cortisol then measured over 120 minutes. Peak cortisol values greater than 550 nmol/liter (19.75 mcg/dl) is considered a normal response. Under supervision by a nurse or physician, the ITT (0.15 U/kg administered intravenously (IV) is relatively safe [24]. The GH reserve can also be estimated with ITT.

poglycemia, with cortisol then measured over 120 minutes. Peak cortisol values greater than 550 nmol/liter (19.75 mcg/dl) is considered a normal response. Under supervision by a nurse or physician, the ITT (0.15 U/kg administered intravenously (IV) is relatively safe [24]. The GH reserve can also be estimated with ITT. Due to the relative inconvenience of the ITT, suggestions have been made to use a simpler and less invasive surrogate. The most widely performed is the short ACTH (Cortrosyn ™) stimulation test, where ACTH 0.25 mg is injected IV or IM and serum cortisol is measure at 0, 30 and 60 minutes. A peak cortisol is defined as normal if it is greater than 550 nmol/liter (19.75 mcg/dl) at any of these time points.

er and less invasive surrogate. The most widely performed is the short ACTH (Cortrosyn ™) stimulation test, where ACTH 0.25 mg is injected IV or IM and serum cortisol is measure at 0, 30 and 60 minutes. A peak cortisol is defined as normal if it is greater than 550 nmol/liter (19.75 mcg/dl) at any of these time points. In 2008 a study using the overnight metyrapone test reported a 60% prevalence of central adrenal insufficiency (CAI) in children with PWS [25]. Based on the high prevalence of CAI, the authors suggested treatment with hydrocortisone during acute illness in patients with PWS unless CAI has recently been ruled out with a metyrapone test. ITT, as the gold standard dynamic test suggests that metyrapone test with an ACTH cut off of 33 pmol/l yields a high false positive rate. In our experience in the PWS center at Winthrop University Hospital, New York, we have not found any abnormal response to ITT or low cortisol levels during surgical stress of different natures. Three recent studies using a more sensitive stimulation and spontaneous acute stress in larger numbers of patients did not find high prevalence of central adrenal insufficiency in Prader-Willi syndrome [26-28]. Thus rather than common, CAI seems to be a rare event in children and adults with PWS, however, they should be evaluated and treated accordingly.

nsitive stimulation and spontaneous acute stress in larger numbers of patients did not find high prevalence of central adrenal insufficiency in Prader-Willi syndrome [26-28]. Thus rather than common, CAI seems to be a rare event in children and adults with PWS, however, they should be evaluated and treated accordingly. Growth & Growth hormone deficiency Length in newborns with PWS is normal but there is significant decrease in growth velocity after age 2-3 years with final adult height ~ 2 standard deviations (SD) below the mean for the normal population [29-31]. Only a small percentage of children with PWS are GH sufficient, thus provocative testing is not required in the face of reduced growth velocity. Multiple studies have documented the benefits of GH therapy in individuals with PWS including, but not limited to, improvements in lean body mass, decreased body fat, increased bone mineral density, and normalization of adult height [19,32-37]. The benefits of starting GH treatment as early as age 2 years are well established, but there is increasing evidence of additional benefit to starting therapy between ages 6-12 months, particularly in terms of motor development, muscle, head circumference, and possibly cognition [35-38].

ion of adult height [19,32-37]. The benefits of starting GH treatment as early as age 2 years are well established, but there is increasing evidence of additional benefit to starting therapy between ages 6-12 months, particularly in terms of motor development, muscle, head circumference, and possibly cognition [35-38]. It should be stressed that GH therapy should be used in conjunction with appropriate nutritional intake and physical activity. GH treatment should not be viewed as a substitute for diet and exercise. Treatment should commence using standard dose guidelines (0.18-0.3 mg/kg/week), given as a daily subcutaneous injection with careful monitoring of clinical status at regular intervals. Careful history and assessment of nutritional status, scoliosis, respiratory and sleep abnormalities should be evaluated prior to and during GH therapy. Recent studies indicate that adults with Prader-Willi syndrome may also benefit from GH replacement therapy, with improvements in body composition, bone mineral density, exercise capacity, quality of life and well-being [39-45]. Treatment doses are typically started at 0.2 mg/day and increased by 0.2 mg increments as necessary to maintain IGF-1 levels within the normal range for age and gender. At the present time, documentation of GH deficiency by provocative testing is required for adults with PWS to receive insurance authorization for GH treatment in the United States. These patients should be monitored with IGF-1, glucose, insulin, lipid profile, BMD and cardiac evaluation during GH treatment [46].

ender. At the present time, documentation of GH deficiency by provocative testing is required for adults with PWS to receive insurance authorization for GH treatment in the United States. These patients should be monitored with IGF-1, glucose, insulin, lipid profile, BMD and cardiac evaluation during GH treatment [46]. Central hypothyroidism as a result of hypothalamic dysfunction can also be seen in individuals with PWS. Periodic monitoring of thyroid function, fasting plasma glucose and insulin level is strongly recommended regardless of growth hormone therapy. Neurocognition and Behavior Decreased intellectual functioning was among the four original defining characteristics of PWS [1]. Subsequent studies document a typical neurobehavioral profile that includes altered intellectual functioning and centrally driven maladaptive behaviors, including the hallmark hyperphagia that exists in the context of a more extensive food related behavioral constellation, an age related emotional and behavioral profile, altered sensory processing, social deficits and for many a predictable psychiatric profile [47-51].

ioning and centrally driven maladaptive behaviors, including the hallmark hyperphagia that exists in the context of a more extensive food related behavioral constellation, an age related emotional and behavioral profile, altered sensory processing, social deficits and for many a predictable psychiatric profile [47-51]. Intellectual Functioning Following the original description, early studies of intellectual development documented a wide range of intellectual abilities, although most affected individuals tested in the borderline to mildly slow IQ ranges. As more sensitive genetic testing has become available, the population of individuals with PWS has become more clearly defined. Table 3 highlights studies of individuals with PWS who had genetic confirmation of their diagnosis, and who received age appropriate and properly administered cognitive testing, supplemented with measures of adaptive functioning. Table 3 Intelligence Quotient (IQ) Intelligence Quotient Degree of Mental Retardation (%) Investigator Year Number enrolled Mean Age* Normal -Borderline Mild Moderate Severe Einfelda [47] 1999 46 17.7 21.6 64.9 13.5 0 Gross-Tsurb [51] 2001 18 14.3 73 27 0 0 Deschee-Maeker [52] 2002 55 14.1 25.4 27.3 40 7.3 Whittington [53] 2004 55 21.0 31 41.8 27.2 0 Copet [54] 2010 85 24.2 7 54 39 0 Roof [55] 2000 47 23.2 24 38 30 8 a. Only 1/2 of subjects genetically confirmed, most IQ's from records b. Did not give a measure of adaptive functioning *Age in Years

Intelligence Quotient Degree of Mental Retardation (%) Investigator Year Number enrolled Mean Age* Normal -Borderline Mild Moderate Severe Einfelda [47] 1999 46 17.7 21.6 64.9 13.5 0 Gross-Tsurb [51] 2001 18 14.3 73 27 0 0 Deschee-Maeker [52] 2002 55 14.1 25.4 27.3 40 7.3 Whittington [53] 2004 55 21.0 31 41.8 27.2 0 Copet [54] 2010 85 24.2 7 54 39 0 Roof [55] 2000 47 23.2 24 38 30 8 a. Only 1/2 of subjects genetically confirmed, most IQ's from records b. Did not give a measure of adaptive functioning *Age in Years The Israeli data are notable for the number of individuals testing in a normal range, and represents a distribution of IQ scores that is quite different from the remaining four studies. The reasons for this are unclear. Setting aside the Israeli data and averaging across the remaining studies, all with approximately the same number of participants, Full Scale IQ ranges are as follows: ≥ 70 in 21%; mild cognitive impairment in 47%; moderate cognitive impairment in 32% and severe to profound cognitive impairment in 2%. An earlier report by Curf and Fryns [56] reported a greater proportion of subjects both in the > 70 and in the mildly impaired range, however their population included many subjects for whom no genetic testing was available and thus may have included individuals who did not have PWS.

rofound cognitive impairment in 2%. An earlier report by Curf and Fryns [56] reported a greater proportion of subjects both in the > 70 and in the mildly impaired range, however their population included many subjects for whom no genetic testing was available and thus may have included individuals who did not have PWS. Separate from the overall range of functioning among an affected population is the question of subtype differences in intellectual functioning. Such differences may be relevant in understanding the role of various genes in the overall clinical features and phenotype of this disorder. While most studies have not found significant subtype differences in overall IQ scores, at least 2 studies have reported a greater number of UPD subjects with normal IQ scores when compared to those with deletion [57-59]. Indeed Torrado et al reported that 61.5% of those with UPD had a Full scale IQ > 70, while only 10.5% of the subjects with deletion scored in that range. However, the mean age of Torrado's subject population was 4.09 years (range 12 days-17 years), so that the significance and overall stability of the obtained IQ scores is open to question. Statistically significant subtype differences have been reported for overall Verbal vs. Performance IQ scores with at least 2 studies reporting that those with UPD have higher verbal IQ scores and those with a deletion subtype have higher performance IQ scores [49,54], although more recently Copet et al [54] found that only the greater performance IQ of the deletion group vs a disomy group was statistically significant. Keep in mind that even when subtype scores are statistically significant, in no case have those differences ever reached the level of 1 SD for the test in question. Thus, whether these statistical differences are reflected as clinically relevant functional differences between subtypes is a question that must be raised.

Keep in mind that even when subtype scores are statistically significant, in no case have those differences ever reached the level of 1 SD for the test in question. Thus, whether these statistical differences are reflected as clinically relevant functional differences between subtypes is a question that must be raised. In addition to mild cognitive deficits which are seen in most individuals with PWS, the overall cognitive profile at all ages includes cognitive rigidity, attentional deficits, problems with short term memory, auditory processing, sequential processing, arithmetic and social cognition. Relative strengths include long term memory, visual spatial performance, simultaneous processing, unusual abilities with jigsaw puzzles, particularly in the deletion subtype and for some reading decoding (devoid of comprehension).

term memory, auditory processing, sequential processing, arithmetic and social cognition. Relative strengths include long term memory, visual spatial performance, simultaneous processing, unusual abilities with jigsaw puzzles, particularly in the deletion subtype and for some reading decoding (devoid of comprehension). Neuro-behavioral Profile While there are a number of clinical descriptions of a typical behavior profile among those with PWS, from the earliest efforts behavioral studies have primarily focused on describing and quantifying the development of problem behaviors and psychiatric difficulties. Despite calls to include investigations of strength and adaptive behaviors [60], these remain a rare study focus. Moreover many studies include such a wide age range, often including infants through late adulthood measured at a single point in time. Parceling out developmental aspects of the behavior profile requires a critical combination of clinical and empirical evidence. Nonetheless studies across time, taken together, yield a general behavior picture that is remarkably consistent across affected individuals, despite variation in severity and intensity across individuals and within the same individual across time. Foremost among these behaviors is the hyperphagia and associated food related behavior constellation. In addition, most clinical and empirical studies document the commonality of hoarding; cognitive rigidity along with the need for sameness, temper outbursts and emotional lability, repetitive and perseverative behaviors and skin-picking.

e behaviors is the hyperphagia and associated food related behavior constellation. In addition, most clinical and empirical studies document the commonality of hoarding; cognitive rigidity along with the need for sameness, temper outbursts and emotional lability, repetitive and perseverative behaviors and skin-picking. Hyperphagia remains the cardinal defining feature of PWS. Nonetheless, the hyperphagia is only one aspect of a larger food-related behavior constellation that included preoccupations surrounding food; food seeking/foraging; sneaking, hiding and hoarding food; eating unusual food-related items (sticks of butter, used cooking grease, decaying food, garbage), food flavored items, such as shampoos and for many, manipulative and sometimes illegal behaviors designed to acquire food. While hyperphagia is found in other genetic syndromes (e.g. WAGR syndrome, Bardet-Biedel syndrome), the development of the hyperphagia and eating patterns associated with PWS, distinguish the hyperphagia associated with PWS from other disorders. Primary among these is the relatively late age of emergence, the rapid escalation and intensification of the hyperphagia following several years of poor to relatively normal eating, often accompanied early on by failure to thrive. Additional distinguishing characteristics include the duration of eating, amount of food eaten and a delayed to absent deceleration of eating, leading to gorging when both physiologic satiation and volume induced discomfort should preclude additional intake.

rmal eating, often accompanied early on by failure to thrive. Additional distinguishing characteristics include the duration of eating, amount of food eaten and a delayed to absent deceleration of eating, leading to gorging when both physiologic satiation and volume induced discomfort should preclude additional intake. In the daily run of life, this is reflected as constant talking about food and unrelenting requests and demands of parents and other caregivers for food, that when denied often precipitate a tantrum. This frequently happens at the grocery or while shopping in other stores that may also have food or candy aisles and at restaurants. The denial-related tantrums can be of such a nature that parents give in as a method of avoiding the behavior, thus creating a pattern that escalates in severity and intensity over time. In addition affected individuals display a constant preoccupation with food leading to extraordinary vigilance for detecting food anywhere in the environment often resulting in stealing other's lunches at school or work, food from teacher's desks or caregiver's purses, stealing food at home or in shopping areas, begging others for food, foraging in garbage cans, entering another's home in search of food and manipulating others to obtain food. It is a rare parent who has not received a call from the school or vocational site indicating that the affected individual has been obtaining extra food by convincing caregivers that parents are ill or haven't had time to feed them, often for an extended period of time.

h of food and manipulating others to obtain food. It is a rare parent who has not received a call from the school or vocational site indicating that the affected individual has been obtaining extra food by convincing caregivers that parents are ill or haven't had time to feed them, often for an extended period of time. The etiology of the hyperphagia remains elusive. Long attributed to a hypothalamically mediated failure of satiety control [61,62], current studies suggest a far more complex etiology than previously hypothesized, including, for many a theoretical reorientation that views the hyperphagia as reflecting a starvation syndrome rather than an obesity syndrome. From this vantage point, the obesity associated with PWS is seen as resulting from a physiologic signaling defect indicating that the body is in a constant state of starvation similar to that of malnourished infants, thus leading to the constant drive to obtain food.

starvation syndrome rather than an obesity syndrome. From this vantage point, the obesity associated with PWS is seen as resulting from a physiologic signaling defect indicating that the body is in a constant state of starvation similar to that of malnourished infants, thus leading to the constant drive to obtain food. To date there is no effective pharmacologic intervention. Management is environmental and behavioral, requiring restricted access to food in all environments, locks on cabinets and refrigerators, constant supervision, as well as measures to prevent obesity which include calorie restrictive diets, consistently scheduled meals and snacks and regularly scheduled physical activity. While simple in concept, the number of environments encountered in any given day, along with the cooperation needed from the individuals in those environments presents challenges that may be insurmountable for some families. Accounts from both parents and individuals with PWS support that strict limit setting with regard to foraging and food access is associated with reduced anxiety and a sense of safety [63].

the cooperation needed from the individuals in those environments presents challenges that may be insurmountable for some families. Accounts from both parents and individuals with PWS support that strict limit setting with regard to foraging and food access is associated with reduced anxiety and a sense of safety [63]. Behavioral Disturbances Separate from the food related behavioral issues, multiple studies document that affected individuals are more prone to behavioral disturbances including hoarding; inflexibility of thinking and behavior; repetitive and perseverative behaviors; the need for sameness; tantrums and emotional lability; and skin picking [62]. Furthermore, the overall rate, severity and chronicity of these disturbances are frequently more intense than those associated with comparable genetic disorders or cognitive impairments or other obese groups [43,64,65]. Like the hyperphagia, the behavioral patterns appear to evolve over time with predictable epochs. Most authors agree that, on the whole, infants and young toddlers with PWS are affectionate, placid and generally cheerful, largely compliant and usually cooperative. However as the hyperphagia emerges, a separate and distinctly negative behavioral shift is also observed including an emergence and escalation of both food and non food related tantrums, a shorter tolerance for frustration combined with an overreaction to frustration; repetitive and ritualistic behavior as well as becoming " stuck " or perseverating on issues both in thought and speech; and other behavior problems including increasing oppositional tendencies, a lessened ability to " go with the flow " along with a drive for sameness and " increasing stubbornness and rigidity ". Comparison studies indicate that typically developing children and other children with mental retardation also exhibit the emergence of such behaviors, but they occur only transiently, that is, the problem appears and then subsides. However the emergence of such behaviors in those with PWS not only is persistent, but appears to escalate with age, increasing in severity and intensity, independent of intellectual, language or motor abilities [66,67].

of such behaviors, but they occur only transiently, that is, the problem appears and then subsides. However the emergence of such behaviors in those with PWS not only is persistent, but appears to escalate with age, increasing in severity and intensity, independent of intellectual, language or motor abilities [66,67]. Chronic behavior disturbances, including emotional lability accompanied by unbridled displays of temper; repetitive, ritualistic and compulsive like behaviors and hoarding become particularly prevalent in adolescence and persist well into adulthood, distinguishing these individuals from both younger children and older individuals with PWS, as well as from typical adolescents. In addition there has recently been an increasing recognition of accompanying social cognition and social interaction deficits among affected individuals, including an inability to read facial expressions of emotion and difficulty interpreting visually presented social information, such as those inherent in any social interaction [68,69]. Indeed several authors sum up the behavioral profile of those with PWS as " egocentric " and who argue, lie, manipulate and confabulate to change rules, obtain their wishes or justify behavior. Their social judgement is poor, even considering their intellectual ability; and interpretations of visually presented social information is on a level with children who have pervasive developmental disorder [62]. Although some behavioral modulation is often seen in later ages, nonetheless problematic behaviors still exceed those seen in other comparison groups.

ring their intellectual ability; and interpretations of visually presented social information is on a level with children who have pervasive developmental disorder [62]. Although some behavioral modulation is often seen in later ages, nonetheless problematic behaviors still exceed those seen in other comparison groups. The expression of this behavioral phenotype does appear to depend, at least to some degree, on genetic subtype, with hoarding and overt behavioral expressions of frustration, anger and aggression more common among those with a deletion, as is a greater likelihood of modulation in middle adulthood. Internalizing and autistic spectrum behaviors are more common among those with UPD, and appear to be unremitting with little age related modulation [70,71].

behavioral expressions of frustration, anger and aggression more common among those with a deletion, as is a greater likelihood of modulation in middle adulthood. Internalizing and autistic spectrum behaviors are more common among those with UPD, and appear to be unremitting with little age related modulation [70,71]. Sensory Issues in the form of an altered sensitivity to pain, failure to exhibit fevers when expected and high rates of skin picking and gouging other body areas are extremely problematic among this group of individuals. While little research has been done around the issues of pain and lack of fever when expected, nonetheless blunted pain sensitivity and lack of appropriate fever response and the inherent dangers these present are clinically well documented. In this same spectrum, skin picking and other similar self injurious behavior occurs with increased prevalence in PWS when compared to a general intellectually impaired population [67,72]. When looking specifically at a population of those with PWS, skin picking is ubiquitous and when quantified, is as prevalent and problematic and in some studies even more so than hyperphagia [47]. It is the source of significant behavior and medical concerns and management challenges. Management is directed towards minimizing both the occurrence and impact of the behavior. To this end, a recent survey of 67 affected children and adolescents documented skin picking in 96% of respondents, which were directly associated with measures of anxiety, inattention, oppositional behaviors, function and quality of life [73]. Thus separate from medical management, behavior management must be focused on decreasing anxiety and boredom while eliminating opportunities for picking.

d skin picking in 96% of respondents, which were directly associated with measures of anxiety, inattention, oppositional behaviors, function and quality of life [73]. Thus separate from medical management, behavior management must be focused on decreasing anxiety and boredom while eliminating opportunities for picking. A number of case series across time have alluded to a small subset of individuals for whom seizures were problematic; however, it was generally thought that these represented incidental findings rather than a risk associated with PWS. A recent report by Fan et al [74] documented seizure activity in 10 of 56 subjects between the ages of 1-37 years, with suspicion in yet another 6 subjects. Among the ten subjects with documented seizure, one youngster's seizure disorder was attributed to sequelae of a grade II intraventricular hemorrhage associated with an early pre-term birth. Among the other nine cases, eight occurred in those with a deletion subtype and the other in a subject whose etiology was a presumed imprinting center defect; none were found among those with a disomy. After reviewing prior studies in which seizures were reported, the authors conclude that the overall prevalence of seizures in PWS is 16 -17%. Further they suggest that among those with a deletion, the risk for seizure in a PWS population is three to four fold times greater than that expected in a general pediatric population.

viewing prior studies in which seizures were reported, the authors conclude that the overall prevalence of seizures in PWS is 16 -17%. Further they suggest that among those with a deletion, the risk for seizure in a PWS population is three to four fold times greater than that expected in a general pediatric population. Psychiatric Illness For many, this wide ranging problematic, behavioral profile can become sufficiently impairing that hospitalization is needed, while for others it evolves into frank psychiatric difficulties. In fact, Cassidy found behavioral concerns to be the most frequent cause of hospitalization [75]. By late adolescence 15-17% will evidence a diagnosable mood disorder [76]. This appears to be especially true for those with UPD. Separate from a categorical psychiatric diagnosis, studies consistently document that the level of behavioral and thought psychopathology, such as delusions, paranoid ideation, common in adolescents and adults with PWS exceeds that of others with an intellectual disability of other origins or of a typical population [65,67], and is the primary source of residential and vocational failure and family stress among affected adolescents and adults. While pharmacologic intervention can be helpful and in the case of psychosis is mandatory, environmental restructuring and positive behavior support programs are even more critical for facilitating recovery and preventing further difficulties.

vocational failure and family stress among affected adolescents and adults. While pharmacologic intervention can be helpful and in the case of psychosis is mandatory, environmental restructuring and positive behavior support programs are even more critical for facilitating recovery and preventing further difficulties. The proliferation of less invasive and more available brain imaging techniques during the past decade offers the possibility of new insights into the central origin of the behavioral picture associated with PWS. Mantoulan [77] compared MRI and PET scans in PWS and non-PWS individuals. MRI images did not show evidence of anatomic abnormalities. However the PET scans showed hypoperfused brain regions, particularly in the anterior cingulum and superior temporal regions. The authors went on to correlate regional cerebral blood flow (rCBF) in the hypoperfused regions with results from the Child Behavior Check list (CBCL) and identified significant correlations, which suggested that the functional consequences of these perfusion abnormalities in specific brain regions might help to explain the social and behavioral issues observed in PWS. Similarly, a number of studies looking at brain processing of food related concerns have yielded mixed findings [78-80]. Functional findings must be considered tentative as the technology is sufficiently challenging that few affected individuals can tolerate the technology nor cooperate with the necessary tasks. Nonetheless, as the technology evolves, the possibility for future studies holds great promise.

yielded mixed findings [78-80]. Functional findings must be considered tentative as the technology is sufficiently challenging that few affected individuals can tolerate the technology nor cooperate with the necessary tasks. Nonetheless, as the technology evolves, the possibility for future studies holds great promise. Sleep Disturbance General Sleep Disturbances Sleep disturbance is frequent in all patients with Prader-Willi syndrome independent of age and weight. PWS patients with normal weight have been shown to have multiple sleep disturbances including daytime sleepiness, disrupted sleep organization, prolonged nocturnal sleep and sleep disordered breathing (SDB). In infants, SDB consists primarily of central apneas and absent, reduced or delayed ventilatory responses and arousal to hypoxia and hypercapnia [81-84]. Adult individuals with PWS-related morbid obesity may have the preceding sleep disturbances as well as an obesity-hypoventilation syndrome and obstructive sleep apnea. Clinical Features Abnormal sleep-wake organization, daytime sleepiness and sleep disordered breathing are the most common sleep related complaints. Irregular REM cycle and sleep disordered breathing appear as early as during infancy [85-87]. Sleep wake organization Early surveys of sleep in PWS reported long nocturnal sleep (> 8 hours) as a common finding [88]. Early morning awakenings and sleep fragmentation have also been reported [89].

Clinical Features Abnormal sleep-wake organization, daytime sleepiness and sleep disordered breathing are the most common sleep related complaints. Irregular REM cycle and sleep disordered breathing appear as early as during infancy [85-87]. Sleep wake organization Early surveys of sleep in PWS reported long nocturnal sleep (> 8 hours) as a common finding [88]. Early morning awakenings and sleep fragmentation have also been reported [89]. The most consistent finding found in polysomnographic studies has been altered rhythm of REM sleep. Studies have shown a tendency towards shorter REM latency, increased number of REM periods and shorter intervals between REM cycles. Total percentage of REM sleep appears to be normal [90-92]. REM sleep alterations appear to be unrelated to the patient genotype [93]. Excessive daytime sleepiness Excessive daytime sleepiness (EDS) is an almost universal characteristic of individuals with PWS [88,89,94]. Clarke et al [88] reported EDS in more than 90% of their surveyed patients. Those patients who reported EDS were more likely to exhibit temper tantrums during the day.

The most consistent finding found in polysomnographic studies has been altered rhythm of REM sleep. Studies have shown a tendency towards shorter REM latency, increased number of REM periods and shorter intervals between REM cycles. Total percentage of REM sleep appears to be normal [90-92]. REM sleep alterations appear to be unrelated to the patient genotype [93]. Excessive daytime sleepiness Excessive daytime sleepiness (EDS) is an almost universal characteristic of individuals with PWS [88,89,94]. Clarke et al [88] reported EDS in more than 90% of their surveyed patients. Those patients who reported EDS were more likely to exhibit temper tantrums during the day. Early studies, using daytime polysomnographic recordings, confirmed the presence of pathologic sleepiness in > 95% of the patients studied [95]. Later studies employed the multiple sleep latency test (MSLT). This test, also called a " nap test " is used to measure the time elapsed to sleep onset. It consists of 4 or 5 nap opportunities during the day. The MSLT is the gold standard to quantify sleepiness and diagnose disorders of excessive sleepiness. Studies with MSLT in individuals with PWS have shown abnormally short sleep latencies and frequent sleep onset REM periods (SOREMPs) [90,92,96]. Daytime sleepiness, as measured by MSLT, appears to be independent of the degree of sleep related breathing disorders [90,92,96-98], additionally suggesting that daytime sleepiness reflects a central, possibly hypothalamic hypoarousal.

lly short sleep latencies and frequent sleep onset REM periods (SOREMPs) [90,92,96]. Daytime sleepiness, as measured by MSLT, appears to be independent of the degree of sleep related breathing disorders [90,92,96-98], additionally suggesting that daytime sleepiness reflects a central, possibly hypothalamic hypoarousal. EDS and the atypical REM sleep findings bear resemblance to features of narcolepsy. Indeed, preliminary evidence in a small number of patients showed that hypocretin deficiency, a characteristic finding in narcolepsy, was also found in individuals with PWS who were severely sleepy [99]. However, in a postmortem study, there was no significant difference in the number of hypothalamic hypocretin containing neurons between patients with PWS and age matched controls [100]. Several preliminary studies have suggested a link between EDS and disruptive behavior in PWS. Hertz et al [101] reported a significant correlation between daytime sleepiness and disruptive behavior as measured by care taker's ranking. Similarly, Richdale [89] reported increased behavioral disturbance in children and adolescents with PWS who also reported EDS. Finally, in Clarke et al's [88] survey, adult patients who reported EDS were more likely to exhibit temper tantrums during the day. In contrast, Maas et al [94], reported no significant correlation between sleep disturbance and behavioral disturbance in a group of adults with PWS.

Several preliminary studies have suggested a link between EDS and disruptive behavior in PWS. Hertz et al [101] reported a significant correlation between daytime sleepiness and disruptive behavior as measured by care taker's ranking. Similarly, Richdale [89] reported increased behavioral disturbance in children and adolescents with PWS who also reported EDS. Finally, in Clarke et al's [88] survey, adult patients who reported EDS were more likely to exhibit temper tantrums during the day. In contrast, Maas et al [94], reported no significant correlation between sleep disturbance and behavioral disturbance in a group of adults with PWS. Sleep Disordered Breathing (SDB) Infants with Prader-Willi syndrome, as young as 4 months old, already demonstrate evidence of sleep disordered breathing. The most frequent type of SDB in infants with PWS are central apneas and periodic breathing [85,86]. Hypotonia and central control abnormalities likely play an important role. As obesity develops, around age 2 years, sleep apnea of the obstructive type becomes more common. Obstructive sleep apnea in both children and adults is directly associated with the degree of obesity and is inversely associated with age [97]. Oxygen desaturation is commonly seen even when the Apnea-hypopnea index (AHI) is only mildly elevated. The degree of sleep related oxygen desaturation may be severe, especially during REM sleep related hypotonia. Its severity is significantly increased with greater body mass index (91).

ersely associated with age [97]. Oxygen desaturation is commonly seen even when the Apnea-hypopnea index (AHI) is only mildly elevated. The degree of sleep related oxygen desaturation may be severe, especially during REM sleep related hypotonia. Its severity is significantly increased with greater body mass index (91). Management of sleep disorders A sleep evaluation of all patients with Prader-Willi syndrome should be routinely considered because of the high prevalence of sleep disturbances. Patients who are habitual snorers and/or sleepy during the day may require a polysomnogram to rule out sleep disordered breathing. In recent years, as more patients are treated with growth hormone (GH), there has been a growing concern over the potentially adverse effects on sleep related breathing. GH may exacerbate OSA in PWS, especially in the presence of other respiratory complications [87]. Review of the death records from the French database of patients with PWS showed an association with respiratory tract infections in both GH and non GH treated patients, highlighting the need for added vigilance during these periods. In patients who were receiving GH treatment the concern for an adverse outcome of SDB and respiratory tract infection is particularly salient during the first nine months of treatment, more so among males [102]. Therefore, an overnight sleep study is recommended before GH therapy is instituted to rule out sleep disordered breathing.

o were receiving GH treatment the concern for an adverse outcome of SDB and respiratory tract infection is particularly salient during the first nine months of treatment, more so among males [102]. Therefore, an overnight sleep study is recommended before GH therapy is instituted to rule out sleep disordered breathing. The gold standard for the treatment of sleep apnea in adults is Continuous Positive Airways Pressure (CPAP) or BiPAP. In children adenoidectomy, tonsillectomy or adenotonsillectomy is often first line of treatment. Supplemental oxygen therapy may be added in the presence of obesity hypoventilation syndrome. The management of other sleep disturbances may include implementation of adequate sleep hygiene, sleep wake schedule regulation and even circadian rhythm modification. In patients who present with excessive daytime sleepiness, a Multiple Sleep Latency Test (MSLT), is also indicated. Once diagnosed, daytime sleepiness can be managed pharmacologically or with behavioral intervention. The pharmacological management of daytime sleepiness has been controversial because of the potential side effects of stimulant medication. Additional research is needed to assess the effects of stimulant medication on daytime alertness, disruptive behavior and the general well being of the sleepy patient with PWS. Behavioral management of EDS focuses on improving nighttime sleep and scheduling daytime naps when needed.

ide effects of stimulant medication. Additional research is needed to assess the effects of stimulant medication on daytime alertness, disruptive behavior and the general well being of the sleepy patient with PWS. Behavioral management of EDS focuses on improving nighttime sleep and scheduling daytime naps when needed. Gastrointestinal Issues Abnormal surges in ghrelin may precede the characteristic hyperphagia seen in PWS. Whether this causes or is the result of the lack of satiety in PWS is not clear. Left unchecked, lack of appetite control can lead to morbid obesity. Low calorie and well balanced diets with rigorous supervision and restriction of food access combined with regularly scheduled meals and activities are recommended [15]. Reduced energy requirements have been reported for children with PWS as compared to healthy, age matched controls [103-105]. Those initiating growth hormone replacement therapy may require increased caloric load during the initial muscle building phase, but once lean mass accretion has stabilized, a reduced caloric limit may again be needed.

ents have been reported for children with PWS as compared to healthy, age matched controls [103-105]. Those initiating growth hormone replacement therapy may require increased caloric load during the initial muscle building phase, but once lean mass accretion has stabilized, a reduced caloric limit may again be needed. Poor oromotor control, muscle hypotonia and voracious eating with a limited time for mastication of food may lead to choking episodes. Choking accounts for approximately 8% of all PWS deaths. Binge eating has been seen in both obese and lean individuals with PWS. Acute gastric distention with necrosis and death has been reported with and without binging behavior. While acute gastric distention is frequently accompanied by vomiting in the general population, individuals with PWS have a decreased ability to vomit and may be missed due to lack of this important finding. These issues may be further complicated by their increased tolerance to pain which may be in part responsible for delays in seeking medical attention related to these episodes. Musculoskeletal Issues The prevalence of scoliosis in PWS is high (30% before 10 yr of age; 80% after age 10 years) [106-109]. Many patients shows progression of scoliosis with age irrespective of the use of GH and therefore scoliosis should no longer be considered a contraindication for GH treatment in children with PWS.

Musculoskeletal Issues The prevalence of scoliosis in PWS is high (30% before 10 yr of age; 80% after age 10 years) [106-109]. Many patients shows progression of scoliosis with age irrespective of the use of GH and therefore scoliosis should no longer be considered a contraindication for GH treatment in children with PWS. Most published reports of scoliosis in children with PWS have been retrospective. Recent evaluation of concerns regarding worsening of scoliosis in patients currently receiving GH have not been substantiated [110,111]. Prospective studies however are warranted. Studies by Shim et al [111] showed a high prevalence of spinal deformity, limb malalignment and foot abnormalities. This group found correlations between various musculoskeletal abnormalities, independent of obesity, but noted that obesity may conceal some of these abnormalities, especially in the early stage. At this time annual musculoskeletal evaluations are recommended for scoliosis, hip dysplasia, foot abnormalities and lower limb malalignments [112]. Slipped femoral capital epiphysis is seen with increased frequency in otherwise healthy, obese children. This has not been reported with increased frequency in children with PWS.

Most published reports of scoliosis in children with PWS have been retrospective. Recent evaluation of concerns regarding worsening of scoliosis in patients currently receiving GH have not been substantiated [110,111]. Prospective studies however are warranted. Studies by Shim et al [111] showed a high prevalence of spinal deformity, limb malalignment and foot abnormalities. This group found correlations between various musculoskeletal abnormalities, independent of obesity, but noted that obesity may conceal some of these abnormalities, especially in the early stage. At this time annual musculoskeletal evaluations are recommended for scoliosis, hip dysplasia, foot abnormalities and lower limb malalignments [112]. Slipped femoral capital epiphysis is seen with increased frequency in otherwise healthy, obese children. This has not been reported with increased frequency in children with PWS. Early work [113] compared gait strategies in patients with PWS with those of both obese and non obese healthy patients. Adults with PWS in their study were found to walk slower, with shorter stride length, lower cadence and longer stance phases compared to non PWS controls. Range of motion at the level of the knee and ankle and plantar-flexor activity were significantly reduced. Spatio-temporal gait parameters in adults with PWS were further evaluated. Using 3 D gait analysis in an attempt to develop rehabilitation therapies, Cimolin et al [113] found that participating adults with PWS showed cautious abnormal gait strategies characterized by longer stance duration, reduced anterior step length and lower velocity of progression. Hip flexion with a forward pelvic tilt was present throughout the gait cycle. Investigators felt that this reflected an attempt to achieve balance and stability in the face of excessive body weight.

gait strategies characterized by longer stance duration, reduced anterior step length and lower velocity of progression. Hip flexion with a forward pelvic tilt was present throughout the gait cycle. Investigators felt that this reflected an attempt to achieve balance and stability in the face of excessive body weight. Prognosis While there is no cure for Prader-Willi syndrome, major strides to improve quality of life have been made since the introduction of more sensitive genetic testing modalities which has allowed early diagnosis and intervention. The early use of GH has improved final adult height, body composition and muscle strength. Obesity and the consequences of obesity continue to be major risk factors for mortality in persons with PWS, even after correction for the effect associated with intellectual disability [114]. Consent Written informed consent was obtained from the parent/guardian of the patient for publication of the accompanying image. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors contributed to the development and writing of this manuscript and each has many years of clinical experience in the care of individuals with Prader-Willi syndrome. All authors read and approved the final manuscript.

Dear Editor, we read with interest the article by Bener and Hoffmann on the incidence of nutritional rickets in a sun rich country like Qatar [1], where decreased vitamin D was a major risk factor. Hypovitaminosis D is highly prevalent in children throughout the world [2,3] but it is still not clear what is the best practice in pediatric primary care settings. Michael Holick, a recognized expert on the topic, has stated that "there is no need to measure everybody's blood 25-hydroxyvitamin D" [25(OH)D] and that only patients with particular diseases should be screened for vitamin D insufficiency/deficiency [4]. Although the literature has shown that patients with deficiency are much less frequent than those with insufficiency, it is also remarkable that vitamin D deficiency is often subclinical and depending on local situations; for example it may be associated with overweight [5] or underweight [1]. To our knowledge there are only a few studies on children living in northeastern Italy [6-8]: they have been conducted retrospectively [6] or examining patients afferent to a Pediatric Department [7] or asthmatic [8]. On this basis an analysis of vitamin D status was prospectively conducted in children cared by a "family pediatrician" in a rural area near Padua (45° N latitude). In 65 patients the vitamin D test was included in exams ordered for different reasons (suspected anemia, fatigue, poor growth, etc.) between November 2010 and June 2011. Results were retrieved from 58 children (age range 1.1-15.3 years, median age 6.75 years). Serum 25(OH)D was dosed by chemiluminescence; the laboratory normal range was 75-250 nmol/l (30-99 ng/ml); insufficiency was defined as 25-74 nmol/l (10-29 ng/ml), deficiency as < 25 nmol/l (< 10 ng/ml).

ember 2010 and June 2011. Results were retrieved from 58 children (age range 1.1-15.3 years, median age 6.75 years). Serum 25(OH)D was dosed by chemiluminescence; the laboratory normal range was 75-250 nmol/l (30-99 ng/ml); insufficiency was defined as 25-74 nmol/l (10-29 ng/ml), deficiency as < 25 nmol/l (< 10 ng/ml). Most of the children (77%) had low serum 25(OH)D levels: 38 of them (66% of all patients) had an insufficiency and 7 (12%) had a deficiency. Moreover, 29 children (50%) had 25(OH)D < 50 nmol/l (< 20 ng/ml) that is the cut-off recently suggested to diagnose vitamin D deficiency [5,9]. None among the 9 young teens (11-15 years) had a normal value of 25(OH)D and 6 of the 7 children with 25(OH)D < 25 nmol/l (< 10 ng/ml) were between ages 2 and 5 years; this deficiency was asymptomatic in 5/6 cases. Moreover, our children with 25(OH)D ≥ 75 nmol/l (≥ 30 ng/ml) and those with deficiency didn't differ for exposure to sunlight, food consumption, gender, ethnicity or BMI. Vitamin D status was also irrespective of other blood test results (see Table 1). Table 1 Vitamin D status of cases based on laboratory reference ranges (plain text columns) and on recent literature [5,9] (bold columns) n Sufficiency Insufficiency Deficiency Insufficiency Deficiency 75-250 nmol/l 25-74 nmol/l < 25 nmol/l 50-74 nmol/l < 50 nmol/l 30-99 ng/ml 10-29 ng/ml < 10 ng/ml 20-29 ng/ml < 20 ng/ml 1-5 years 27 8 (30%) 13 6 7 12 6-10 years 22 5 (23%) 17 0 5 12 11-15 years 9 0 8 1 4 5 Total 58 13 (23%) 38 7 16 29 Gender (M/F) 40/18 9/4 31/14 Overweight 9 3 6 Underweight 7 0 7 Immigrant 11 3 8

n Sufficiency Insufficiency Deficiency Insufficiency Deficiency 75-250 nmol/l 25-74 nmol/l < 25 nmol/l 50-74 nmol/l < 50 nmol/l 30-99 ng/ml 10-29 ng/ml < 10 ng/ml 20-29 ng/ml < 20 ng/ml 1-5 years 27 8 (30%) 13 6 7 12 6-10 years 22 5 (23%) 17 0 5 12 11-15 years 9 0 8 1 4 5 Total 58 13 (23%) 38 7 16 29 Gender (M/F) 40/18 9/4 31/14 Overweight 9 3 6 Underweight 7 0 7 Immigrant 11 3 8 Normal exams 23 2 21 Holick has suggested that "it would be much more cost-effective to implement a vitamin D supplementation program for all children and adults" [4] but the question now is how much vitamin D should be given. If 400 IU cholecalciferol per day may be sufficient in the first year of life [10], much more is needed in older children, assuming that most of them have less or much less than the minimum desirable [11]. Moreover, recommended doses of 600 IU per day [12] probably offer no advantage to children with 25(OH)D < 25 nmol/l (< 10 ng/ml) [13]. Although authoritative guidelines state that routine vitamin D testing is not warranted in the average risk population, the Holick's D-lemma [4] is far from being resolved. Abbreviations 25(OH)D: 25-hydroxyvitamin D. Competing interests The authors declare that they have no competing interests. Authors' contributions MS performed the statistical analysis and interpretation of data and drafted the manuscript. TD participated in the design of the study, in interpretation of data and final approval of the manuscript. BC participated in the collection of data, in the statistical analysis and in the drafting of manuscript. All authors read and approved the final manuscript.

Introduction Klinefelter syndrome (KS, also known as 47,XXY) and 47,XYY syndrome are the two most common sex chromosome aneuploidies in humans with prevalence of approximately 1 in 600–1000 males [1-4]. Individuals with KS are usually tall adolescents and adults who have hypergonadotrophic hypogonadism with small testicles. However, the KS phenotype is highly variable and individuals may not show these physical features to a degree that distinguishes them from the general male population. Males with 47,XYY are also taller than average, but in contrast to KS, they usually do not have phenotypic characteristics to differentiate them from 46,XY males. Compared to 46,XY males, individuals with KS or 47,XYY syndrome exhibit a greater incidence of behavioral problems, psychiatric disorders and neuropsychological characteristics including developmental delays and difficulties in cognitive, verbal and social skills [5]. Yet individuals with both syndromes often fail to be ascertained. Newborn screening studies estimate that only 25% of all individuals with KS, and 10% of all individuals with 47,XYY are diagnosed during their lifetime [6,7]. More complex male sex chromosome aneuploidies, such as 48,XXYY and 48,XXXY, are less common than KS with prevalences ranging from 1 in 18,000 to 1 in 100,000 or greater [1]. While some phenotypic characteristics of 48,XXYY and 48,XXXY syndromes overlap with KS, the unique and significant differences in physical appearance, cognitive function, social and adaptive skills observed in affected individuals differentiate these aneuploidies from KS [8,9].

in 18,000 to 1 in 100,000 or greater [1]. While some phenotypic characteristics of 48,XXYY and 48,XXXY syndromes overlap with KS, the unique and significant differences in physical appearance, cognitive function, social and adaptive skills observed in affected individuals differentiate these aneuploidies from KS [8,9]. The gold standard for detection of chromosome aneuploidies is karyotype analysis, an invasive, time-consuming and labor-intensive process. Yet despite the widespread availability of karyotyping, most males with sex chromosome aneuploidy are never diagnosed during their lifetime [6,7]. Thus, the development of more convenient methods for detection of sex chromosome aneuploidies should facilitate identification of these individuals, allowing them to receive early evaluation and therapeutic intervention as indicated. To address this need, we developed a two-stage Pyrosequencing based assay which measures Y:X and X:autosome chromosome ratios. Using this approach, we demonstrated 100% sensitivity in the identification of males with sex chromosome aneuploidy.

them to receive early evaluation and therapeutic intervention as indicated. To address this need, we developed a two-stage Pyrosequencing based assay which measures Y:X and X:autosome chromosome ratios. Using this approach, we demonstrated 100% sensitivity in the identification of males with sex chromosome aneuploidy. Materials & methods DNA samples De-identified karyotype-confirmed DNA samples from individuals with male sex chromosome aneuploidies (n = 117) were obtained from Children's Hospital Colorado and the UC Davis MIND Institute (Dr. Flora Tassone). Additional de-identified karyotype-confirmed DNA samples from subjects with 45,X (n = 1), 46,XX (n = 4), 47,XXX (n = 11), and 46,XY (n = 206) were provided by the above sources, plus the Yale Cytogenetics Lab (Dr. Peining Li) and the Genetics Diagnostic Lab of Children’s Hospital Boston (Dr. Bai-Lin Wu). Before use, the DNA samples were diluted 20-fold with nuclease-free water. The concentration of diluted DNA was determined by real time PCR using the human specific probe WIAF699 as described [10]. Only samples with a concentration of diluted DNA ≥1 ng/μl were used as templates for PCR of the XYM and XA markers (see below). DNA isolation from buccal swabs Buccal swabs were obtained from patients of Children’s Hospital Colorado after informed consent. DNA was isolated from buccal cells using the Qiagen EZ1 robot and EZ1 DNA tissue kit according to the manufacturer’s protocol. Extracted DNA was quantified by real time PCR as above.

Materials & methods DNA samples De-identified karyotype-confirmed DNA samples from individuals with male sex chromosome aneuploidies (n = 117) were obtained from Children's Hospital Colorado and the UC Davis MIND Institute (Dr. Flora Tassone). Additional de-identified karyotype-confirmed DNA samples from subjects with 45,X (n = 1), 46,XX (n = 4), 47,XXX (n = 11), and 46,XY (n = 206) were provided by the above sources, plus the Yale Cytogenetics Lab (Dr. Peining Li) and the Genetics Diagnostic Lab of Children’s Hospital Boston (Dr. Bai-Lin Wu). Before use, the DNA samples were diluted 20-fold with nuclease-free water. The concentration of diluted DNA was determined by real time PCR using the human specific probe WIAF699 as described [10]. Only samples with a concentration of diluted DNA ≥1 ng/μl were used as templates for PCR of the XYM and XA markers (see below). DNA isolation from buccal swabs Buccal swabs were obtained from patients of Children’s Hospital Colorado after informed consent. DNA was isolated from buccal cells using the Qiagen EZ1 robot and EZ1 DNA tissue kit according to the manufacturer’s protocol. Extracted DNA was quantified by real time PCR as above. Assay design Pyrosequencing (PSQ) assays were designed to interrogate two types of markers. The first class of markers, designated XYM, consisted of regions of the X and Y chromosomes with nearly identical sequence that differ by a chromosome-specific biallelic single nucleotide polymorphism, such that one allele is present only on the X chromosome and the other allele is located on the Y chromosome. Candidate sequences were identified by examining closely related genes present on both X and Y chromosomes outside of the terminal pseudoautosomal regions (see Table 1 of reference [11] and Table 2 of reference [12]). BLAST [13] searches of the human reference genome sequence were used to confirm a single match to the X and Y chromosomes for all sequences entered into the PSQ assay design software (version 1.0.6).

romosomes outside of the terminal pseudoautosomal regions (see Table 1 of reference [11] and Table 2 of reference [12]). BLAST [13] searches of the human reference genome sequence were used to confirm a single match to the X and Y chromosomes for all sequences entered into the PSQ assay design software (version 1.0.6). Table 1 Statistical values calculated from the percent Y allele signals of the three XYM markers for samples grouped together by karyotype Karyotype 47, XXY (n = 42) 47, XYY (n = 26) 48, XXXY (n = 4) 48, XXYY (n = 45) 46, XY (n = 206) Marker XYM1 XYM2 XYM3 XYM1 XYM2 XYM3 XYM1 XYM2 XYM3 XYM1 XYM2 XYM3 XYM1 XYM2 XYM3 Average 32.3 33.3 33.7 63.7 62.5 64.6 24.4 26.7 26.0 47.4 48.1 49.3 48.2 49.7 49.3 Median 32.3 33.4 33.7 63.6 62.2 64.7 24.1 26.7 25.7 47.6 48.0 49.4 48.2 49.8 49.2 Std Dev 1.53 2.56 1.39 1.55 1.53 0.83 1.54 0.57 0.90 1.17 2.70 1.15 1.46 1.61 1.55 Maximum 38.9 40.6 39.0 67.8 65.9 66.8 26.4 27.3 27.3 50.6 58.3 51.8 51.9 53.2 53.3 Minimum 30.1 27.9 31.0 61.5 57.8 62.3 23.0 25.9 25.3 45.4 44.2 46.1 43.1 41.1 42.8 Table 2 Statistical values calculated from the percent X allele signals of the four XA markers for samples grouped together by karyotype

15 1.46 1.61 1.55 Maximum 38.9 40.6 39.0 67.8 65.9 66.8 26.4 27.3 27.3 50.6 58.3 51.8 51.9 53.2 53.3 Minimum 30.1 27.9 31.0 61.5 57.8 62.3 23.0 25.9 25.3 45.4 44.2 46.1 43.1 41.1 42.8 Table 2 Statistical values calculated from the percent X allele signals of the four XA markers for samples grouped together by karyotype Karyotype 48, XXYY (n = 45) 46, XY (n = 206) Marker XA1 XA2 XA3 XA4 XA1 XA2 XA3 XA4 Average 56.8 54.1 48.5 50.4 42.3 35.3 33.0 32.0 Median 56.4 54.0 48.9 50.8 42.1 35.1 33.3 31.7 Std Dev 1.74 2.57 2.43 3.47 2.51 2.03 2.99 3.56 Maximum 59.7 59.1 51.1 57.8 60.2 42.6 37.2 51.3 Minimum 52.2 49.5 35.3 43.8 35.5 27.4 0.0 20.8 The second class of markers, designated XA, represent regions of the X chromosome that are nearly identical with an autosome except for a single chromosome-specific base. Candidate sequences for assay design were identified by BLAST searching the human genome reference sequence with the set of all X-chromosome transcripts obtained from the ENSEMBL database [14]. High scoring matches where at least 200 bases were >95% identical with one locus on the X-chromosome and one on an autosome were used to generate consensus sequences for assay design by the PSQ software.

uman genome reference sequence with the set of all X-chromosome transcripts obtained from the ENSEMBL database [14]. High scoring matches where at least 200 bases were >95% identical with one locus on the X-chromosome and one on an autosome were used to generate consensus sequences for assay design by the PSQ software. PCR and extension primers for high scoring assays were synthesized using standard methods by the W. M. Keck Foundation Biotechnology Resource Lab of Yale University. One of the PCR primers for each assay was labeled at the 5´ end with biotin. For all assays, the extension primer had the same orientation as the forward PCR primer and is complementary to the biotinylated template strand generated with the reverse PCR primer.

k Foundation Biotechnology Resource Lab of Yale University. One of the PCR primers for each assay was labeled at the 5´ end with biotin. For all assays, the extension primer had the same orientation as the forward PCR primer and is complementary to the biotinylated template strand generated with the reverse PCR primer. PCR and pyrosequencing A minimum of 2.5 ng genomic DNA was used as template in a 25 μl PCR reaction. Each PCR reaction contained 1 X Hotstar buffer (Qiagen), 2.5 mM MgCl2, 200 μM each dNTP, 1 μM each PCR primer, 0.5 U Hotstar Taq Plus (Qiagen). Reactions were performed as follows: initial incubation of 5 min at 95°C; followed by 45 cycles of 30 sec at 95°C, 45 sec at 56°C, and 60 sec at 72 °C; then 5 min at 72 °C and a final hold at 4°C. Upon completion of PCR, the biotinylated template strand was purified using Streptavidin-Sepharose and the Filter Prep tool (Qiagen) according to the manufacturer’s instructions. The resulting single stranded template was annealed to the appropriate extension primer and Pyrosequencing was performed using Pyromark Q96 reagents and PSQ96MA instrument (Qiagen). The allele percentage was calculated by the PSQ software (version 2.1) and exported for analysis by Microsoft Excel.

turer’s instructions. The resulting single stranded template was annealed to the appropriate extension primer and Pyrosequencing was performed using Pyromark Q96 reagents and PSQ96MA instrument (Qiagen). The allele percentage was calculated by the PSQ software (version 2.1) and exported for analysis by Microsoft Excel. Results Principle of assay The assay measures both Y:X and X:A (X:Autosome) chromosome ratios by Pyrosequencing, a quantitative short-read DNA sequencing technology [15]. For the first step, three XYM loci are PCR amplified with specific primer pairs and the resulting amplicons subjected to Pyrosequencing, yielding the percent of Y-chromosome-specific allele signal for each locus and hence, the ratio of the Y and X chromosomes. Figure 1 shows Pyrosequencing data using the XYM3 marker with DNA from a 46,XX female (A), 46,XY male (B), a 47,XXY individual (C), and a 47,XYY subject (D). The percent Y allele signal has close agreement with the value predicted for each karyotype. The locations of the three XYM markers on the X and Y chromosomes are shown in Figure 2.

sequencing data using the XYM3 marker with DNA from a 46,XX female (A), 46,XY male (B), a 47,XXY individual (C), and a 47,XYY subject (D). The percent Y allele signal has close agreement with the value predicted for each karyotype. The locations of the three XYM markers on the X and Y chromosomes are shown in Figure 2. Figure 1 Pyrograms for DNA from a 46,XX female (A), a 46,XY male (B), a 47,XXY KS male (C), and a 47,XYY male (D) using the XYM3 assay. The box encloses the two nucleotide dispensations which define the C/T SNP; the C-allele is derived from the X-chromosome and the T-allele is from the Y-chromosome. The percent of each allele is shown in the shaded box above each pyrogram. The y-axis depicts the intensity of light signal in arbitrary units and the x-axis shows the time of addition of Pyrosequnecing enzymes (E), substrates (S), and each individual nucleotide dispensation (A, C, G, or T). Figure 2 Locations of the XYM markers on the X and Y chromosomes. X and Y chromosome ideograms are drawn to scale using data from the UCSC genome browser. X chromosome is 154.9 Mb and Y chromosome is 57.8 Mb. In the second step of the assay, Pyrosequencing of the four XA markers determines the X:autosome (X:A) ratio, which permits differentiation between individuals with 46,XY and 48,XXYY karyotypes.

Figure 2 Locations of the XYM markers on the X and Y chromosomes. X and Y chromosome ideograms are drawn to scale using data from the UCSC genome browser. X chromosome is 154.9 Mb and Y chromosome is 57.8 Mb. In the second step of the assay, Pyrosequencing of the four XA markers determines the X:autosome (X:A) ratio, which permits differentiation between individuals with 46,XY and 48,XXYY karyotypes. Overall performance of XYM markers To evaluate the ability of the three XYM markers to detect sex chromosome aneuploidies, PCR and Pyrosequencing were performed on 339 DNA samples from females (n = 16) and males (n = 323), with the technician blinded to each individual’s karyotype. Following completion of Pyrosequencing, the karyotype of each sample was matched to the data for all three markers. All samples from phenotypic females, including individuals with a 46,XX karyotype (n = 4) and those with sex chromosome aneuploidies (45,X, n = 1 and 47,XXX, n = 11), did not display detectable Y allele for any of the three XYM markers (maximum Y allele signal of 1.9%).

le was matched to the data for all three markers. All samples from phenotypic females, including individuals with a 46,XX karyotype (n = 4) and those with sex chromosome aneuploidies (45,X, n = 1 and 47,XXX, n = 11), did not display detectable Y allele for any of the three XYM markers (maximum Y allele signal of 1.9%). Data generated with the three XYM markers were analyzed for the 323 DNA samples from males with karyotype-confirmed sex chromosome aneuploidies (n = 117) or 46,XY (n = 206). Scatter plots of the percent Y allele signal for both the XYM2 and XYM3 markers versus the percent Y allele signal of the XYM1 marker are shown in Figure 3. For all three XYM markers, the data for individuals with the 48,XXXY, 47,XXY, and 47,XYY karyotypes were tightly clustered and distinguishable from males with a 46,XY karyotype. As expected from the known Y:X chromosome ratio, the percent Y allele signals for the subjects with 48,XXYY karyotype overlapped with those for the 46,XY males. The percent Y allele signal for both XYM2 and XYM3 was strongly correlated with XYM1, with correlation coefficients (r2) of 0.90 and 0.94, respectively. Table 1 summarizes the statistical data calculated for the measured percent Y allele signal of each marker for samples grouped according to karyotype; the average percent Y allele signal for each group was close to the predicted value based on the known karyotype.

ation coefficients (r2) of 0.90 and 0.94, respectively. Table 1 summarizes the statistical data calculated for the measured percent Y allele signal of each marker for samples grouped according to karyotype; the average percent Y allele signal for each group was close to the predicted value based on the known karyotype. Figure 3 Percent Y allele signal of XYM2 (A) and XYM3 (B) plotted against the percent Y allele signal of XYM1 for 323 DNA samples from male subjects grouped by karyotype. For each marker, the percent Y allele signal is the ratio of signal from the Y-chromosome specific allele divided by the total signal from both the Y- and X-chromosome specific alleles, expressed as a percentage.

ed against the percent Y allele signal of XYM1 for 323 DNA samples from male subjects grouped by karyotype. For each marker, the percent Y allele signal is the ratio of signal from the Y-chromosome specific allele divided by the total signal from both the Y- and X-chromosome specific alleles, expressed as a percentage. Detection of 47,XXY (KS), 48,XXXY and 47,XYY syndromes To identify 47,XXY, 48,XXXY, or 47,XYY karyotypes, receiver operator characteristic (ROC) curves were constructed by varying the percent Y allele signal lower and upper thresholds for samples scored as normal 46,XY karyotype (Figure 4). For this analysis, the average percent Y allele signal for the three XYM markers was calculated; the lower threshold was increased from 34% to 48% by 1% increments while sensitivity and false positive rates for detecting 47,XXY (KS) and 48,XXXY karyotypes were calculated. All samples with either 47,XXY or 48,XXXY karyotypes and average percent Y allele signals less than the lower threshold value were classified as true positives. False negative samples had the same karyotypes and the average percent Y allele signal above the threshold. True negatives were defined as a 46,XY subject with the average percent Y allele signals greater than the lower limit, while a false positive was defined as having the average percent Y allele signal less than the lower limit.

samples had the same karyotypes and the average percent Y allele signal above the threshold. True negatives were defined as a 46,XY subject with the average percent Y allele signals greater than the lower limit, while a false positive was defined as having the average percent Y allele signal less than the lower limit. Figure 4 Receiver operator characteristic curves for detection of 47,XXY (KS) and 48,XXXY (A) and 47,XYY syndromes (B). The average of the three XYM marker data points for each sample was calculated and compared to the detection thresholds. The threshold for detecting XXY and XXXY syndromes was varied by 1% increments from 34-48% Y allele signal; that for detecting XYY syndrome was incremented by 1% from 49-63%. The sensitivity (TP/(TP + FN)) and false positive rate (FP/(FP + TN)) were calculated for each value of the appropriate threshold. For XXY and XXXY, true positives (TP) have either a 47,XXY or 48,XXXY karyotype and average Y allele signal less than the threshold; false negatives (FN) have the same karyotypes and average Y allele signal greater than or equal to the threshold. True negatives (TN) for XXY and XXXY are samples with a 46,XY karyotype and average Y allele signal greater than or equal to the threshold; false positives (FP) have a 46,XY karyotype and average Y allele signal below the threshold. For XYY syndrome, true positives have 47,XYY karyotype and average Y allele signal greater than the threshold; false negatives have the same karyotype and average Y allele signal less than or equal to the threshold. True negatives for XYY are samples with a 46,XY karyotype and average Y allele signal less than or equal to the threshold; false positives have a 46,XY karyotype and average Y allele signal above the threshold.

old; false negatives have the same karyotype and average Y allele signal less than or equal to the threshold. True negatives for XYY are samples with a 46,XY karyotype and average Y allele signal less than or equal to the threshold; false positives have a 46,XY karyotype and average Y allele signal above the threshold. Conversely for 47,XYY syndrome, the upper threshold was examined from 49% to 63% by 1% increments, and sensitivity and false positive rates were calculated. True positives had 47,XYY karyotype and an average percent Y allele signal above the upper threshold; any 47,XYY sample with the average percent Y allele signal below the upper limit was scored as false negative. True negatives (46,XY) had the average percent Y allele signal less than or equal to the upper limit; false positives had the average percent Y allele signal above the upper limit. Examination of the ROC curves indicated that a percent Y allele signal scoring threshold of 43% for 47,XXY and 48,XXXY syndromes and 57% for 47,XYY syndrome yielded 100% sensitivity with a 0% false positive rate. Combining the two thresholds gave a percent Y allele signal range of 43-57% for 46,XY samples. Separate analysis of the individual XYM marker data and the median of the three XYM values generated ROC curves which overlapped the curve generated with the average XYM values; thresholds of 43% and 57% allow the detection of KS and 47,XYY syndrome with 100% sensitivity and specificity using the data from either each individual marker or the median value.

M marker data and the median of the three XYM values generated ROC curves which overlapped the curve generated with the average XYM values; thresholds of 43% and 57% allow the detection of KS and 47,XYY syndrome with 100% sensitivity and specificity using the data from either each individual marker or the median value. Detection of 48,XXYY syndrome As noted above, the percent Y allele signals from 48,XXYY individuals (n = 45) and 46,XY males (n = 206) showed nearly complete overlap (Figure 3 and Table 1). To distinguish these karyotypes, four XA markers were amplified by PCR and the resulting products examined by Pyrosequencing. Figure 5 is a plot of the average percent X allele for the four XA markers graphed versus the sample ID. The average percent X allele signal differed for the samples with 48,XXYY and 46,XY karyotypes and was close to the expected values of 50% and 33.3%, respectively, based on the known chromosome ratios (Table 2). ROC analysis (Figure 6) performed by comparing the average percent X allele for the four XA markers with a threshold varying from 33-52% in 1% increments indicated that a threshold of 43% gave 100% detection of 48,XXYY karyotype with a 0% false positive rate. Figure 5 Average percent X allele signal of four XA markers versus sample ID for male samples with 46,XY and 48,XXYY karyotypes. The average percent X allele signal is the average for all four markers of the ratio of signal from the X-chromosome specific allele divided by the total signal from both the X-chromosome and autosome specific alleles, expressed as a percentage.

rkers versus sample ID for male samples with 46,XY and 48,XXYY karyotypes. The average percent X allele signal is the average for all four markers of the ratio of signal from the X-chromosome specific allele divided by the total signal from both the X-chromosome and autosome specific alleles, expressed as a percentage. Figure 6 Receiver operator characteristic curves for detection of 48,XXYY syndrome. The average and median of the four XA marker data points for each sample was calculated: either the average, median or individual marker data was compared to the threshold for detecting 48,XXYY syndrome. The threshold was varied by 1% increments from 33-52% X allele signal. The sensitivity (TP/(TP + FN)) and false positive rate (FP/(FP + TN)) were calculated for each value of the appropriate threshold. True positives (TP) have a 48,XXYY karyotype and X allele signal greater than the threshold; false negatives (FN) have the same karyotype and X allele signal less than or equal to the threshold. True negatives (TN) have a 46,XY karyotype and X allele signal less than or equal to the threshold; false positives (FP) have a 46,XY karyotype and X allele signal above the threshold.

ignal greater than the threshold; false negatives (FN) have the same karyotype and X allele signal less than or equal to the threshold. True negatives (TN) have a 46,XY karyotype and X allele signal less than or equal to the threshold; false positives (FP) have a 46,XY karyotype and X allele signal above the threshold. As a confirmatory approach to distinguish males with 48,XXYY and 46,XY karyotypes, 18 X-linked markers were PCR amplified and the amplicons subjected to Pyrosequencing using DNA from all 48,XXYY males. The relative allele strength for each of the X-chromosome specific markers was scored as homozygous, heterozygous, or out-of-range as described for a Turner Syndrome assay [16]. Of the 45 individuals with the 48,XXYY karyotype, 43 demonstrated definitive evidence for the presence of two distinct X-chromosomes, with the number of heterozygous markers ranging from 4 to 12 out of 18 total. Thus, most 48,XXYY individuals either inherit one X-chromosome from each parent or inherit two distinct X chromosomes as a result of nondisjunction in maternal meiosis I. The remaining two individuals were homozygous for all 18 markers, and therefore, appear to have inherited both X chromosomes from their mother due to nondisjuction in maternal meiosis II or as a result of nondisjunction of a 46,XY embryo in early mitotic cell divisions. Thus, for the latter group, only signal from the XA markers distinguished them from 46,XY individuals.

18 markers, and therefore, appear to have inherited both X chromosomes from their mother due to nondisjuction in maternal meiosis II or as a result of nondisjunction of a 46,XY embryo in early mitotic cell divisions. Thus, for the latter group, only signal from the XA markers distinguished them from 46,XY individuals. Analysis of Coriell samples with a 47,XXY karyotype In an additional test of the sex chromosome aneuploidy assay, three XYM markers were measured on DNA obtained from immortalized lymphocyte cultures of 16 individuals with a 47,XXY karyotype (Coriell Institute of Medical Research). For 15 of 16 individuals, the percent Y allele signal for all three markers clustered tightly around the expected 33.3%. One individual had percent Y allele signals between 53.6 and 59% and is known to have a complex mosaic karyotype: 47,XXY [17].ish X (DXZ1x2).ish Y (SRYx1)/47,XYY [28].ish X (DXZ1x1).ish Y (SRYx2)/46,XY [5].ish X (DXZ1x1).ish Y (SRYx1), with prominent contributions from both 47,XXY and 47,XYY cell populations. Based on the mosaic estimate, this subject is expected to have percent Y allele signals near 50%.

complex mosaic karyotype: 47,XXY [17].ish X (DXZ1x2).ish Y (SRYx1)/47,XYY [28].ish X (DXZ1x1).ish Y (SRYx2)/46,XY [5].ish X (DXZ1x1).ish Y (SRYx1), with prominent contributions from both 47,XXY and 47,XYY cell populations. Based on the mosaic estimate, this subject is expected to have percent Y allele signals near 50%. Analysis of buccal swab samples As a final test for assay performance, buccal swabs were obtained from 29 males with known karyotypes. Buccal cell DNA was extracted and all seven markers amplified by PCR. Following Pyrosequencing, the data from all 29 samples were classified using the previously described threshold values for the XYM and XA markers. Table 3 presents the aggregate statistical data for the average allele signal of the buccal swab samples grouped by karyotype. Buccal swab samples demonstrated 100% sensitivity and specificity for detection of male sex chromosome aneuploidies. Table 3 Statistical values calculated from the average percent allele signals of the XYM and XA marker sets for buccal swab samples grouped together by karyotype

Analysis of buccal swab samples As a final test for assay performance, buccal swabs were obtained from 29 males with known karyotypes. Buccal cell DNA was extracted and all seven markers amplified by PCR. Following Pyrosequencing, the data from all 29 samples were classified using the previously described threshold values for the XYM and XA markers. Table 3 presents the aggregate statistical data for the average allele signal of the buccal swab samples grouped by karyotype. Buccal swab samples demonstrated 100% sensitivity and specificity for detection of male sex chromosome aneuploidies. Table 3 Statistical values calculated from the average percent allele signals of the XYM and XA marker sets for buccal swab samples grouped together by karyotype Karytoype 47, XXY (n = 8) 47, XYY (n = 3) 46, XY (n = 5) 48, XXYY (n = 13) Marker Set Average Y Allele XYM Average X Allele XA Average Y Allele XYM Average X Allele XA Average Y Allele XYM Average X Allele XA Average Y Allele XYM Average X Allele XA Average 35.1 52.1 65.2 36.3 50.6 38.1 50.1 52.3 Median 35.0 51.5 65.1 36.1 50.7 37.7 50.3 52.1 Std Dev 0.9 1.7 0.2 0.9 0.8 1.3 1.0 1.9 Maximum 36.8 54.3 65.4 37.3 51.6 39.5 52.1 56.3 Minimum 34.0 50.2 65.0 35.6 49.6 36.6 48.0 49.7 Discussion We report the development of a rapid, high-throughput Pyrosequencing assay for detecting sex chromosome aneuploidies in males. The assay initially interrogates three XYM markers, yielding the percent Y allele signal which is directly related to the Y:X chromosome ratio. Next, to distinguish 46,XY and 48,XXYY karyotypes, the assay utilizes four XA markers to determine the X:A ratio. Using this approach, our assay identifies males with 47,XXY, 47,XYY, 48,XXXY and 48,XXYY karyotypes at 100% sensitivity and specificity (Table 4).

gnal which is directly related to the Y:X chromosome ratio. Next, to distinguish 46,XY and 48,XXYY karyotypes, the assay utilizes four XA markers to determine the X:A ratio. Using this approach, our assay identifies males with 47,XXY, 47,XYY, 48,XXXY and 48,XXYY karyotypes at 100% sensitivity and specificity (Table 4). Table 4 Sensitivity and false positive rate for detection of sex chromosome aneuploidies in males Karyotype Total Samples Sensitivity False Positive Rate 47,XXY (KS) 65 100% 0% 47,XYY 29 100% 0% 48,XXXY 4 100% 0% 48,XXYY 58 100% 0% Combination of results for DNAs isolated from buccal swabs (n = 29) or obtained from Colorado Children’s Hospital, UC Davis MIND Institute, Children’s Hospital Boston, Yale University and Coriell Institute (n = 338). One Coriell sample was omitted from this table due to its complicated mosaic karyotype: 47,XXY [17].ish X (DXZ1x2).ish Y (SRYx1)/47,XYY [28].ish X (DXZ1x1).ish Y (SRYx2)/46,XY [5].ish X (DXZ1x1).ish Y (SRYx1).

ospital, UC Davis MIND Institute, Children’s Hospital Boston, Yale University and Coriell Institute (n = 338). One Coriell sample was omitted from this table due to its complicated mosaic karyotype: 47,XXY [17].ish X (DXZ1x2).ish Y (SRYx1)/47,XYY [28].ish X (DXZ1x1).ish Y (SRYx2)/46,XY [5].ish X (DXZ1x1).ish Y (SRYx1). Previous studies of the parental origin of the sex chromosomes in males with 48,XXYY are limited to a total of eight individuals [17-21]. These studies concluded that the extra sex chromosomes are paternally derived, resulting from two sequential nondisjunction events in meiosis I and II of spermatogenesis. Our data for 96% of subjects with 48,XXYY (n = 43) are consistent with this mechanism since the DNA samples demonstrated heterozygosity for between 4 and 12 of a total 18 X-linked biallelic SNP markers. However, the data cannot rule out an alternative mechanism whereby the additional X chromosome is maternally derived from nondisjunction in meiosis I of oogenesis and the supernumerary Y is due to nondisjunction in meiosis II of spermatogenesis. Only detailed molecular genetic analysis of the parents of 48,XXYY males can ascertain the relative contribution of these two mechanisms; still, the statistical likelihood of an X aneuploid oocyte being fertilized by a Y aneuploid sperm is quite low. The remaining 4% of subjects with 48,XXYY (n = 2) were shown to have identical X chromosomes due to complete homozygosity of the 18 X-linked markers and thus likely result from nondisjunction during early mitotic divisions of a 46,XY embryo. Alternatively, but less likely, this subset of individuals may result from nondisjunction in meiosis II of both maternal and paternal gametes. The current study has a large enough population (n = 45) to detect this novel mechanism for the chromosomal origin of the supernumerary sex chromosomes in 48,XXYY males.

embryo. Alternatively, but less likely, this subset of individuals may result from nondisjunction in meiosis II of both maternal and paternal gametes. The current study has a large enough population (n = 45) to detect this novel mechanism for the chromosomal origin of the supernumerary sex chromosomes in 48,XXYY males. Our Pyrosequencing based assay is robust and readily interpretable allowing the reliable detection of male sex chromosome aneuploidies with 100% sensitivity and specificity. This particular methodology serves as a rapid, high-throughput screen, and the accuracy of detection for KS and other sex chromosome aneuploidies translates to an extremely low likelihood of discrepant karyotypic analysis if utilized for confirmation. The assay may be completed in 8–10 hrs. which is considerably faster than the time required for either fluorescent in situ hybridization (FISH) or karyotype analysis. Individuals with KS may present clinically at many points during their lifetime [1], and yet, because of variable and often subtle phenotype, they are not recognized and in most cases, never diagnosed [22]. The current assay provides a non-invasive molecular test applicable for rapid diagnosis, thus allowing for earlier assessments and interventions in all facets of therapy for KS, 47,XYY, 48,XXYY and 48,XXXY, including androgen replacement and cognitive and behavioral treatments [23].

in most cases, never diagnosed [22]. The current assay provides a non-invasive molecular test applicable for rapid diagnosis, thus allowing for earlier assessments and interventions in all facets of therapy for KS, 47,XYY, 48,XXYY and 48,XXXY, including androgen replacement and cognitive and behavioral treatments [23]. For male children suspected of KS or another sex chromosome aneuploidy, the ability to make the diagnosis using DNA isolated from buccal swabs is an advantage over invasive, often traumatic, blood testing. Diagnosis during infancy/childhood, especially for KS, allows for early interventional speech/language therapy and educational planning, as well as promotion of physical activity to inhibit the development of dyspraxia. Endocrine monitoring and early management can be instituted to eventually support puberty, preserve fertility, and determine the timing of androgen replacement [23]. With respect to 47,XYY syndrome, rapid and efficient detection similarly permits initiation of appropriate cognitive and behavioral therapy. Current trials of pharmaceuticals for general developmental disorders in male children that overlap with KS and other supernumerary X chromosome syndromes may also benefit from diagnostic specificity for these relevant aneuploidies.

detection similarly permits initiation of appropriate cognitive and behavioral therapy. Current trials of pharmaceuticals for general developmental disorders in male children that overlap with KS and other supernumerary X chromosome syndromes may also benefit from diagnostic specificity for these relevant aneuploidies. Even making the delayed diagnosis of KS or other sex chromosome aneuploidies in adulthood offers specific treatment goals for their related complications. KS is one of the most frequent causes of male infertility [24]. For adult males being evaluated for failure to conceive, making this diagnosis earlier offers specific and better options to preserve fertility [25], and this rapid methodology may decrease the stress and anxiety associated with waiting for the diagnosis of KS (or other sex chromosome aneuploidy) by karyotyping. As with male children, making the diagnosis of KS in adulthood is also important for instituting specific endocrine therapies to prevent gynecomastia and osteopenia [23]. The clinical application of this sensitive and specific Pyrosequencing based assay will enable the rapid, efficient and high-throughput detection of sex chromosome aneuploidies in males and allow for early, appropriate assessment and therapeutic interventions for individuals with these common but under-diagnosed genetic conditions.

application of this sensitive and specific Pyrosequencing based assay will enable the rapid, efficient and high-throughput detection of sex chromosome aneuploidies in males and allow for early, appropriate assessment and therapeutic interventions for individuals with these common but under-diagnosed genetic conditions. Competing interests Karl Hager, Kori Jennings & Seiyu Hosono are employees of JS Genetics, Inc.. Karl Hager, Seiyu Hosono, Jeffrey R. Gruen, Scott A. Rivkees, and Henry M. Rinder hold equity interest in JS Genetics, Inc.. Susan Howell and Nicole R. Tartaglia have no competing interests. Authors’ contributions KH contributed to study design, developed the hypothesis, analyzed data, and wrote the manuscript; KJ analyzed data and contributed to writing of the manuscript; SH contributed to study design and writing of the manuscript; SH contributed to study design, writing of the manuscript, and collection of patient samples; JG contributed to study design and writing of the manuscript; SR contributed to study design and writing of the manuscript; NT contributed to study design, writing of the manuscript, and collection of patient samples; HR developed the hypothesis, analyzed data, and contributed to study design and writing the manuscript. All authors read and approved the final manuscript.

Background This paper is a commentary from the Pediatric Endocrine Society Drugs and Therapeutics Committee. Its goal is provide background information and guidance to both pediatric endocrinologists, and also to primary care practictioners on the current use of HbA1c for the diagnosis of Type 2 diabetes in children. In addition, it has the goal of pointing out some of the flaws in current methodologies being used to address this question. Diagnostic criteria for diabetes were initially was based on variation from normal [1]. As information about the degrees of glycemia that lead to diabetic complications became available, criteria were revised. In 1997, the Expert Committee on Diagnosis and Classification of Diabetes Mellitus examined population data for retinopathy, and noted that for Fasting Plama Glucose (FPG), 2-hour postload glucose during an oral glucose tolerance test (OGTT), and hemoglobin A1c (HbA1c), the diabetes-related complication of retinopathy increased linearly above a certain cutpoints; for FPG and OGTT, those cutpoints became the basis for the diagnosis of diabetes [2].

ted that for Fasting Plama Glucose (FPG), 2-hour postload glucose during an oral glucose tolerance test (OGTT), and hemoglobin A1c (HbA1c), the diabetes-related complication of retinopathy increased linearly above a certain cutpoints; for FPG and OGTT, those cutpoints became the basis for the diagnosis of diabetes [2]. In 2010, it was felt that with the increasing adherence to the National Glycohemoglobin Standardization Program (NGSP), laboratory-based HbA1c is measured in a standardized fashion in the majority of labs in the U.S. Furthermore, the American Diabetes Association (ADA) noted that review of epidemiologic data supported a relationship between HbA1c and the risk of retinopathy similar to what had been shown for FPG and OGTT. Thus, in 2010, the ADA added A1c of 6.5 percent or greater as a diagnostic criterion [3]. The ADA required that a laboratory-based HbA1c assay method certified by the NGSP be used. This ensures that the assay used is standardized or traceable to the Diabetes Control And Complications Trial [4,5]. Obtaining screening laboratories for diabetes in the at-risk population has long been established practice in adults. However, in the pediatric population, data have thus far largely been extrapolated from adult studies, and screening practices vary. The issue has become more pertinent with the rise of pediatric obesity. While FPG and OGTT thresholds, as extrapolated from adult populations, have largely been accepted by the community of pediatric practictioners, the more recent recommendation of use of HbA1c has met resistance.

tudies, and screening practices vary. The issue has become more pertinent with the rise of pediatric obesity. While FPG and OGTT thresholds, as extrapolated from adult populations, have largely been accepted by the community of pediatric practictioners, the more recent recommendation of use of HbA1c has met resistance. Main text and discussion An important benefit of use of HbA1c is that patient does not need to be fasting, and testing does not require a return visit [6]. HbA1c has less variability and is more reproducible than FPG and OGTT [7]. However, HbA1c also has several potential disadvantages. Hb A1c may miss cases of Type 1 diabetes in which hyperglycemia develops over a short period of time. Furthermore, HbA1c is not a perfect estimation of mean blood glucose, and varies by ethnicity [8,9]. In addition, diseases such as iron deficiency anemia, sickle-cell disease, thalassemia, and other hemoglobinopathies, can alter HbA1c [10]. Recent studies When FPG was used to diagnose diabetes, HbA1c of 6.5% had sensitivity of 75.0% and specificity of 99.9% [11]. The authors examined the diagnosis of pre-diabetes noted a low sensitivity but a good specificity (98.3%) for a HbA1c of 5.7%, and also for HbA1c of 6.0% (specificity 99.4%).

However, HbA1c also has several potential disadvantages. Hb A1c may miss cases of Type 1 diabetes in which hyperglycemia develops over a short period of time. Furthermore, HbA1c is not a perfect estimation of mean blood glucose, and varies by ethnicity [8,9]. In addition, diseases such as iron deficiency anemia, sickle-cell disease, thalassemia, and other hemoglobinopathies, can alter HbA1c [10]. Recent studies When FPG was used to diagnose diabetes, HbA1c of 6.5% had sensitivity of 75.0% and specificity of 99.9% [11]. The authors examined the diagnosis of pre-diabetes noted a low sensitivity but a good specificity (98.3%) for a HbA1c of 5.7%, and also for HbA1c of 6.0% (specificity 99.4%). Another study [12] compared HbA1c to OGTT in over 1000 obese patients and concluded that A1c has low sensitivity and specificity for diabetes when diabetes is defined by OGTT results; 9 of 893 patients with an HbA1c less than 5.7% were determined to have diabetes using OGTT criteria. In addition, a larger number of cases of prediabetes defined by OGTT were not identified using an HbA1c cut-off of 5.7%; 240 of the 347 (69%) cases of prediabetes in this high-risk population had HbA1c <5.7%. In a smaller study, HBA1c cut-off of 6.5% had a sensitivity of 40% and a specificity of 96% in accurately diagnosing patients with type 2 diabetes, when using OGTT as a gold standard [13].

Another study [12] compared HbA1c to OGTT in over 1000 obese patients and concluded that A1c has low sensitivity and specificity for diabetes when diabetes is defined by OGTT results; 9 of 893 patients with an HbA1c less than 5.7% were determined to have diabetes using OGTT criteria. In addition, a larger number of cases of prediabetes defined by OGTT were not identified using an HbA1c cut-off of 5.7%; 240 of the 347 (69%) cases of prediabetes in this high-risk population had HbA1c <5.7%. In a smaller study, HBA1c cut-off of 6.5% had a sensitivity of 40% and a specificity of 96% in accurately diagnosing patients with type 2 diabetes, when using OGTT as a gold standard [13]. A study of 254 overweight or obese adolescents also raised questions about HbA1c use for diagnosis [14]. In this study, there were 99 (39%) cases of prediabetes and 3 (1.2%) cases of diabetes using FPG and OGTT as gold standards. Test performance was assessed using receiver operating characteristic (ROC) curves and calculations of area under the ROC curve (AUC). HbA1c (AUC 0.54 [95% CI 0.47-0.61]) displayed poor discrimination for identifying children with dysglycemia that had been identified on OGTT. In fact, in this study, random glucose (AUC 0.66 [0.60-0.73]) had better correlation with OGTT.

ng characteristic (ROC) curves and calculations of area under the ROC curve (AUC). HbA1c (AUC 0.54 [95% CI 0.47-0.61]) displayed poor discrimination for identifying children with dysglycemia that had been identified on OGTT. In fact, in this study, random glucose (AUC 0.66 [0.60-0.73]) had better correlation with OGTT. The potential advantages of use of HbA1c, however, were born out by a study before and after a change in recommendation to allow use of HbA1c in screening of adolescents [15]. Rates of screening for diabetes increased from 39 to 47% as a result, and this led to twice as many incident T2DM diagnoses during a similar time period. HbA1c threshold of 6% indicated progression to diabetes in 18% of patients, while the 5.7% threshold resulted in only 1.3% progression to diabetes, over about 3 years . In a separate study with 468 subjects, HbA1c cut-off of 6% had greater correlation with OGTT results than A1c threshold of 5.7% [16]. This study showed sensitivity and specificity of 86% and 85%, respectively, for HbA1c threshold 5.7%, but 99% and 96%, respectively, for HbaA1c threshold of 6%. Comparing unvalidated methodologies: a pitfall of many studies Studies that compare HbA1c to other methods of diagnosing diabetes in pediatric populations are handicapped by the fact that the other methods – FPG and OGTT – are themselves not validated in the pediatric population. A truly validated definition of diabetes in pediatric populations requires insight into the relationship of the proposed definitions to relevant aspects of medium and long-term health [17].

ions are handicapped by the fact that the other methods – FPG and OGTT – are themselves not validated in the pediatric population. A truly validated definition of diabetes in pediatric populations requires insight into the relationship of the proposed definitions to relevant aspects of medium and long-term health [17]. Assuming that OGTT or FPG are better than HbA1c for determining risk of complications is unfounded; in adults, even though HbA1c, FPG, and OGTT are often discrepant in individuals, all 3 markers are shown to correlate very well with risk of complications [2,3]. Conclusions The diagnostic thresholds of glycemia in the adult population were formulated because these are theh thresholds at which retinopathy increases. Such thresholds have never been defined in the pediatric population. In this sense, for diagnosis of diabetes in asymptomatic or minimally symptomatic children, there are no validated methodologies. OGTT, which is being used as a ‘gold standard’ in many studies, in addition to lacking validation as noted above, also suffers from having low reproducibility [18]. Therefore, dismissal of HbA1c 6.5% or greater for diagnosis of diabetes at this time, because of lack of correlation with OGTT, is a flawed approach.

gies. OGTT, which is being used as a ‘gold standard’ in many studies, in addition to lacking validation as noted above, also suffers from having low reproducibility [18]. Therefore, dismissal of HbA1c 6.5% or greater for diagnosis of diabetes at this time, because of lack of correlation with OGTT, is a flawed approach. The lack of correlation between studies has already been acknowledged and discussed in adult populations, including in the ADA’s clinical practice guidelines [3,4]. While the correlation appears to be lower in pediatrics, it is not yet clear which of the studies, if any, are ‘faulty,’ and obtaining such data would require comprehensive, multi-center, long-term studies on incidence of retinopathy; such studies appear unlikely to occur at this time. Therefore, conclusions that dismiss HbA1c use for the diagnosis of diabetes in children are based on incomplete data. Considering that the demographics of Type 2 diabetes skew towards disadvantated populations, we should not dismiss a valuable, flexible tool that, put into widespread use, may in fact increase, not decrease, early detection of this disease [15]. Recommendations 1. ADA criteria for the diagnosis of diabetes, though formulated from data in adults, are useful for screening the asymptomatic, or minimally symptomatic, at-risk pediatric population at this time, and this includes the ADA’s more recent recommendations regarding HbA1c.

Therefore, conclusions that dismiss HbA1c use for the diagnosis of diabetes in children are based on incomplete data. Considering that the demographics of Type 2 diabetes skew towards disadvantated populations, we should not dismiss a valuable, flexible tool that, put into widespread use, may in fact increase, not decrease, early detection of this disease [15]. Recommendations 1. ADA criteria for the diagnosis of diabetes, though formulated from data in adults, are useful for screening the asymptomatic, or minimally symptomatic, at-risk pediatric population at this time, and this includes the ADA’s more recent recommendations regarding HbA1c. 2. When test results do not correlate with each other, or appear to give information conflicting to the clinical situation, practictioners must use clinical judgement. Amongst FPG, OGTT, or HbA1c, one test is not validated to a greater extent than another in the pediatric population. 3. There are some questions about HbA1c use in those with cystic fibrosis [19], though some data shows it may be useful [20]. 4. HbA1c is likely reliable in those with sickle-cell carrier status, as long as an assay without interference from abnormal hemoglobins is used [5,21]. HbA1c is not reliable in those known to have a hemoglobinopathy or any other disorder resulting in significant increases in red blood cell turnover, or in pregnancy [22]. Assay interference information is available on the NGSP website [5].

long as an assay without interference from abnormal hemoglobins is used [5,21]. HbA1c is not reliable in those known to have a hemoglobinopathy or any other disorder resulting in significant increases in red blood cell turnover, or in pregnancy [22]. Assay interference information is available on the NGSP website [5]. 5. Laboratory-based hemoglobin A1c using a methodology and assay certified by the National Glycohemoglobin Standardization Program is the preferred method for diagnostic purposes [3-5]. Abbreviations ADA: American Diabetes Association; FPG: Fasting Plasma Glucose; OGTT: Oral Glucose Tolerance Test; HbA1c: Hemoglobin A1c; NGSP: National Glycohemoglobin Standardization Program. Competing interests There are no competing interest for any participating authors or collaborators. Authors’ contributions CK and PZ are the primary authors and editors of this commentary. The remaining collaborators from the PES Drugs and Therapeutics Committee helped shape the article through discussion at meetings and conference calls, and provided editing of the manuscript. All authors read and approved the final manuscript. Authors’ information Pediatric Endocrine Society Drugs and Therapeutics Committee for 2011–2012: Kapadia Ca, Zeitler Pb, Divall, Sc, Grimberg, Ad, Gitelman SEe, Quintos JBf, Draznin Mg, Nebesio Th, Myers SEi, Potter Aj, Vogiatzi Mk, Raman Sl, Waguespack SGm, Eugster Eh, Boney CMf, Rosenthal SMe, Silverstein Jn a. Phoenix Children’s Hospital, Phoenix, AZ b. Department of Pediatrics, University of Colorado, Denver, CO c. Johns Hopkins University, Baltimore, MD

Kapadia Ca, Zeitler Pb, Divall, Sc, Grimberg, Ad, Gitelman SEe, Quintos JBf, Draznin Mg, Nebesio Th, Myers SEi, Potter Aj, Vogiatzi Mk, Raman Sl, Waguespack SGm, Eugster Eh, Boney CMf, Rosenthal SMe, Silverstein Jn a. Phoenix Children’s Hospital, Phoenix, AZ b. Department of Pediatrics, University of Colorado, Denver, CO c. Johns Hopkins University, Baltimore, MD d. Children’s Hospital of Philadelphia, Philadelphia, PA e. University of California-San Francisco, San Francisco, CA f. Rhode Island Hospital, The Warren Alpert Medical School of Brown University, Providence, RI g. Michigan State University Kalamazoo Center for Medical Studies, Kalamazoo, MI h. Riley Children’s Hospital, Indianapolis, IN i. Cardinal Glennon Children’s Hospital, St Louis, MO j. Vanderbilt University Medical Center, Nashville, TN k. Weill Medical Center of Cornell University, NY, NY l. Children’s Mercy Hospital, Kansas City, KS m. University of Texas M.D. Anderson Cancer Center, Houston, TX n. University of Florida College of Medicine, Gainesville, FL

Background The stimulant medications dexamphetamine and methylphenidate are used for the treatment of children with attention deficit hyperactivity disorder (ADHD) because they reduce the level of hyperactivity and enhance the attention span. Weight loss and reduced height velocity are well-recognized collateral effects [1-3], particularly in the first 6 months of treatment. In a previous study we demonstrated that the growth rates for height and weight progressively normalized over 2–3 years on treatment, but the period of slower growth was associated with long-term reductions in the age and sex corrected Z-scores for height and weight [4]. These observations raise the following important questions: what is the nature of the tissue being lost and what is the effect on bone development? The aims of this prospective study were to monitor growth in children with ADHD starting treatment with stimulant medication and to look for changes in body composition and markers of bone metabolism. The hypotheses informing the study design were (1) that acute changes in hormones and bone biomarkers during weight loss would be observed in the first 3 months of treatment; (2) that changes in body composition would be detectable after 6 months and (3) a new steady state of normal growth on stimulant medication would be reached after 3 years of continuous treatment.

that acute changes in hormones and bone biomarkers during weight loss would be observed in the first 3 months of treatment; (2) that changes in body composition would be detectable after 6 months and (3) a new steady state of normal growth on stimulant medication would be reached after 3 years of continuous treatment. Methods Design and setting This was a clinic based, prospective, cohort study with up to three years follow up. Recruitment was mainly from a single pediatric private practice in western Sydney; two other practices each contributed one patient. Recruitment began in July 2003 and continued until April 2008. Ethical approval was granted by the Human Research Ethics Committee of Wentworth Area Health Service in western Sydney (02/013). Study participants We invited parents of consecutive children aged <9 years who were newly diagnosed with ADHD and had a clinical indication for starting treatment with stimulant medication to enroll their children in the study. The diagnosis of ADHD was based on clinical interview with additional information on functioning at school or preschool. All children met the diagnostic criteria of the American Psychiatric Association (DSM-IV) [5]. Children with a history of previous treatment with psychotropic medication or with medical conditions likely to impact on growth were excluded. Informed consent was given by the parents of all participants; all participating children assented to the study.

criteria of the American Psychiatric Association (DSM-IV) [5]. Children with a history of previous treatment with psychotropic medication or with medical conditions likely to impact on growth were excluded. Informed consent was given by the parents of all participants; all participating children assented to the study. Treatment protocol All children were initially trialed on immediate release methylphenidate or dexamphetamine; some children were subsequently prescribed extended release formulations. Treatment allocation was based on cost, with dexamphetamine being the preferred medication until methylphenidate also became subsidised. The dose was titrated to give the maximum therapeutic benefit at the lowest possible dose of medication, in line with recommended practice parameters [6]. Children who were stable and functioning well on stimulant medication were reviewed every 6 months and the dose adjusted when clinically indicated. This decision was based on parental reports of the child’s functioning, the child’s opinion, school reports and other documentation from the school, including the IOWA Conner’s rating scale [7]. If at any stage there was no clear advantage to stimulant medication and the child was functioning adequately, medication was ceased and the child’s functioning reassessed off medication. No further data were collected on children who had ceased medication.

from the school, including the IOWA Conner’s rating scale [7]. If at any stage there was no clear advantage to stimulant medication and the child was functioning adequately, medication was ceased and the child’s functioning reassessed off medication. No further data were collected on children who had ceased medication. Data collection We measured height to the nearest 1mm using a wall mounted stadiometer and weight to the nearest 0.1kg using electronic scales. All measurements were made without shoes or outdoor clothing. Measurements were taken at every clinic visit and without prior reference to any previous measurements. Body mass index (BMI) was calculated as the weight in kg divided by the square of the height in meters and the height, weight and BMI were corrected for age and sex by conversion to Z-scores based on Centers for Disease Control and Prevention (CDC) reference data [8].

Data collection We measured height to the nearest 1mm using a wall mounted stadiometer and weight to the nearest 0.1kg using electronic scales. All measurements were made without shoes or outdoor clothing. Measurements were taken at every clinic visit and without prior reference to any previous measurements. Body mass index (BMI) was calculated as the weight in kg divided by the square of the height in meters and the height, weight and BMI were corrected for age and sex by conversion to Z-scores based on Centers for Disease Control and Prevention (CDC) reference data [8]. Dual-energy X-ray absoptiometry (DXA) scans of the Total Body and lumbar spine were performed using a GE-Lunar Prodigy (GE Lunar Corp, Madison, WI). Daily quality assurance and quality control (spine phantom in water bath) were performed.We measured total body (including head) and body subregions (arms, legs, trunk) lean tissue and fat masses, bone mineral content (BMC), bone mineral density (BMD) and fat distribution (ratio of central to total fat; fat C/T). For estimation of the central (truncal) fat, the trunk region (which includes the pelvis) was delineated by a single assessor using standard manufacturer analysis guidelines. The assessor (JB) has shown good inter-rater reliability with a colleague [9]. Lumbar spine (L2-L4), BMD and BMC were also determined. Raw DXA values were converted into Z-scores using DXA control data.

gion (which includes the pelvis) was delineated by a single assessor using standard manufacturer analysis guidelines. The assessor (JB) has shown good inter-rater reliability with a colleague [9]. Lumbar spine (L2-L4), BMD and BMC were also determined. Raw DXA values were converted into Z-scores using DXA control data. DXA control data measured using a Lunar DPX (GE Lunar Corp, Madison, WI) were available from 241 healthy children (105 boys), aged 4.2-12.0 years (mean 8.54±1.93 years). Parents of these children had consented for them to have a single DXA to provide reference data for research purposes. Although there was no direct comparison of the Lunar DPX and the Lunar Prodigy used for the subjects, our data from 97 children comparing another Lunar Prodigy with the Lunar DPX showed no significant differences in BMC or BMD. There were small differences in lean tissue mass (mean 28.6 vs 29 kg), and fat (mean 11.4 vs 12.3 kg) but the results were all highly correlated (all r2 > 0.98).

r Prodigy used for the subjects, our data from 97 children comparing another Lunar Prodigy with the Lunar DPX showed no significant differences in BMC or BMD. There were small differences in lean tissue mass (mean 28.6 vs 29 kg), and fat (mean 11.4 vs 12.3 kg) but the results were all highly correlated (all r2 > 0.98). Blood samples were collected unmedicated after an 8 hour overnight fast. The following bone turnover markers and calciotropic hormones were measured: osteocalcin, 25-hydroxy vitamin D, parathyroid hormone (PTH), amino-terminal propeptide of type I collagen (P1NP) and C-telopeptides (CTX). In addition, appetite regulating hormones (leptin and ghrelin) and the growth hormone dependent growth factors insulin-like growth factor 1 (IGF-1) and insulin-like growth factor binding protein 3 (IGFBP-3), together with fasting insulin, glucose albumin, pre-albumin, ferritin and transferrin were measured to look for changes in growth and appetite regulation and for biochemical evidence of undernutrition. Vitamin D and osteocalcin were analysed by chemiluminescence immunoassays on a Diasorin Liaisonanalyzer (Diasorin, Italy). CTX and P1NP were measured with commercial electrochemiluminescence immunoassays (Modular E170, Roche Diagnostics, Australia). Albumin was analysed using an automated colorimetric assay using a Vitros Fusion analyser (Ortho Clinical Diagnostics, Australia). Pre-albumin used an immunoturbid metric assay on an Integra (Roche Diagnostics, Australia). Ferritin was assayed using an immunometric assay on a Vitros ECI (Ortho Clinical Diagnostics, Australia). Iron deficiency was defined as ferritin <20 ug/L. Transferrin used the TRNF method, which is a quantitative turbid metric assay (Dimension Xpand, Siemens Medical Solutions Diagnostics Pty Ltd, Sydney, Australia). Glucose was analysed using routine laboratory procedures. Leptin and total ghrelin levels were analysed using radioimmunoassay (Millipore, MA) and insulin by chemiluminescence (IMMULITE 2000®, Siemens Medical Solutions Diagnostics Pty Ltd, Sydney, Australia). PTH, IGF-I and IGFBP-3 were measured by ELISA using the IMMULITE 1000® analyser (Siemens Medical Solutions Diagnostics Pty Ltd, Sydney, Australia).

ls were analysed using radioimmunoassay (Millipore, MA) and insulin by chemiluminescence (IMMULITE 2000®, Siemens Medical Solutions Diagnostics Pty Ltd, Sydney, Australia). PTH, IGF-I and IGFBP-3 were measured by ELISA using the IMMULITE 1000® analyser (Siemens Medical Solutions Diagnostics Pty Ltd, Sydney, Australia). Sample size calculations These were based on the growth data from our previous study [4]. The estimated sample sizes required to detect a change in height Z-score with 90% power at 5% two sided significance were 16 and 11 after 6 months and 3 years respectively; for weight Z-score these were 8 and 13 respectively. Therefore we aimed to recruit 30 children, to allow for 50% attrition at 3 years. Outcome measures Growth Growth was analyzed as changes in height, weight and BMI Z-scores. Body composition Body composition was analysed as follows: 1. Longitudinal changes in lean tissue, fat, BMC, BMD and fat distribution (fat C/T). 2. Cross sectional comparison of lean tissue, fat, BMC, BMD and fat C/T at each timepoint with the DXA control data, controlling for age, sex and height. 3. Longitudinal changes in lean tissue, fat, BMC, BMD and fat C/T at 6 months and 3 years in relation to growth in height using sex and height corrected Z-scores based on the DXA control data. Height was used instead of age as we considered this to be the more clinically relevant measure if height was not changing proportionally with age due to stimulant associated changes in the height velocity.

nd 3 years in relation to growth in height using sex and height corrected Z-scores based on the DXA control data. Height was used instead of age as we considered this to be the more clinically relevant measure if height was not changing proportionally with age due to stimulant associated changes in the height velocity. Statistical analysis We used paired t-tests for longitudinal data analysis and independent sample t-tests for cross sectional comparison. The DXA data from the subjects were compared to the DXA control data using general linear modeling with age, sex and height as independent variables. Correlations between variables used the Pearson correlation. All analyses were 2-tailed and statistical significance was taken as p<0.05.

for cross sectional comparison. The DXA data from the subjects were compared to the DXA control data using general linear modeling with age, sex and height as independent variables. Correlations between variables used the Pearson correlation. All analyses were 2-tailed and statistical significance was taken as p<0.05. Results Participant characteristics Thirty-four children aged 4.7 to 9.1 years (mean 7.3±1.3 years, 29 boys) were recruited. Of these, 29 had baseline DXA scans and 31 had baseline blood tests (Figure 1). All met a DSM-IV diagnosis of ADHD and the majority (78%) had hyperactivity-impulsivity as well as inattention (combined type). The majority of children were treated with methylphenidate: at 6 months there were 23 on methylphenidate (mean 24.3±6.2 mg/day; 0.91±0.19 mg/kg/day) and 7 on dexamphetamine (mean 11.1±2.8 mg/day; 0.42±0.08 mg/kg/day). The attrition rate for investigations was 53% (18/34) at 3 years. Eight of these children were still being treated with stimulant medication and had growth data available; the remaining 10 (29% of the cohort) had either ceased medication (n=3), had moved away (n=1) or were otherwise lost to follow up (n=6). There were no significant differences in age or growth parameters between those who did and those who did not complete their investigations, therefore the growth data of all the children who remained on treatment have been included in the tables and figures. Figure 1 Number of subjects and attrition in each part of the study.

Results Participant characteristics Thirty-four children aged 4.7 to 9.1 years (mean 7.3±1.3 years, 29 boys) were recruited. Of these, 29 had baseline DXA scans and 31 had baseline blood tests (Figure 1). All met a DSM-IV diagnosis of ADHD and the majority (78%) had hyperactivity-impulsivity as well as inattention (combined type). The majority of children were treated with methylphenidate: at 6 months there were 23 on methylphenidate (mean 24.3±6.2 mg/day; 0.91±0.19 mg/kg/day) and 7 on dexamphetamine (mean 11.1±2.8 mg/day; 0.42±0.08 mg/kg/day). The attrition rate for investigations was 53% (18/34) at 3 years. Eight of these children were still being treated with stimulant medication and had growth data available; the remaining 10 (29% of the cohort) had either ceased medication (n=3), had moved away (n=1) or were otherwise lost to follow up (n=6). There were no significant differences in age or growth parameters between those who did and those who did not complete their investigations, therefore the growth data of all the children who remained on treatment have been included in the tables and figures. Figure 1 Number of subjects and attrition in each part of the study. Growth rates At baseline the mean Z-scores for height, weight and BMI were 0.49 ± 0.99, 0.62 ± 0.97 and 0.52 ± 1.02 respectively (Table 1). Over 3 years there were significant reductions from baseline in the Z-scores for height, weight and BMI (all p<0.001). This was predominantly due to slower growth in the first 6 months (Table 2); the growth rate showed normalization with time (p<0.01 and p<0.001 for the difference in the rate of change from 0–6 months and from 6 months to 3 years for height and weight Z-scores respectively). The growth pattern is illustrated in Figure 2.

is was predominantly due to slower growth in the first 6 months (Table 2); the growth rate showed normalization with time (p<0.01 and p<0.001 for the difference in the rate of change from 0–6 months and from 6 months to 3 years for height and weight Z-scores respectively). The growth pattern is illustrated in Figure 2. Table 1 Growth data of the subjects compared to the controls All subjects All subjects All subjects DXA Controls Baseline n=34 6 months n=30 36 months n=24 n=241 Mean ± sd Range Mean ± sd Range Mean ± sd Range Mean ± sd Range Age (years) 7.27 ± 1.30*** 4.71-9.12 7.99 ± 1.21* 5.33-9.64 10.49 ± 1.22*** 7.87-12.21 8.54 ± 1.93 4.02-11.99 Height (cm) 125.5 ± 9.1* 110.3-142.2 128.6 ± 9.2 112.6-145.6 140.6 ± 9.0*** 125.5-156.0 130.5 ± 12.7 94.5-161.6 Weight (kg) 27.0 ± 6.1 19.1-42.1 26.8 ± 5.9 19.3-42.2 34.4 ± 8.4** 25.0-56.0 29.0 ± 8.2 13.1-60.4 BMI (kg/m2) 17.0 ± 2.3 13.7-23.7 16.1 ± 2.3 12.9-23.0 17.3 ± 3.1 13.9-26.5 16.7 ± 2.2 13.3-25.6 Height Z-score 0.49 ± 0.99** −1.08-3.02 0.22 ± 0.95 −1.19-2.76 −0.07 ± 0.81 −1.18-1.67 −0.06 ± 0.97 −2.24-2.07 Weight Z-score 0.62 ± 0.97*** −1.60-2.95 0.04 ± 1.04 −1.86-2.63 −0.12 ± 0.87 −1.48-1.58 −0.05 ± 0.92 −2.25-2.23 BMI Z-score 0.52 ± 1.02** −1.88-2.31 −0.17 ± 1.21 −2.92-2.04 −0.16 ± 1.11 −2.45-1.97 0.03 ± 0.88 −2.11-2.09 *p<0.05; **p<0.01; ***p<0.001 compared to DXA controls (independent samples t-test). DXA: Dual-energy X-ray absorptiometry. Table 2 Growth rates of the subjects over different time periods on medication

All subjects All subjects All subjects DXA Controls Baseline n=34 6 months n=30 36 months n=24 n=241 Mean ± sd Range Mean ± sd Range Mean ± sd Range Mean ± sd Range Age (years) 7.27 ± 1.30*** 4.71-9.12 7.99 ± 1.21* 5.33-9.64 10.49 ± 1.22*** 7.87-12.21 8.54 ± 1.93 4.02-11.99 Height (cm) 125.5 ± 9.1* 110.3-142.2 128.6 ± 9.2 112.6-145.6 140.6 ± 9.0*** 125.5-156.0 130.5 ± 12.7 94.5-161.6 Weight (kg) 27.0 ± 6.1 19.1-42.1 26.8 ± 5.9 19.3-42.2 34.4 ± 8.4** 25.0-56.0 29.0 ± 8.2 13.1-60.4 BMI (kg/m2) 17.0 ± 2.3 13.7-23.7 16.1 ± 2.3 12.9-23.0 17.3 ± 3.1 13.9-26.5 16.7 ± 2.2 13.3-25.6 Height Z-score 0.49 ± 0.99** −1.08-3.02 0.22 ± 0.95 −1.19-2.76 −0.07 ± 0.81 −1.18-1.67 −0.06 ± 0.97 −2.24-2.07 Weight Z-score 0.62 ± 0.97*** −1.60-2.95 0.04 ± 1.04 −1.86-2.63 −0.12 ± 0.87 −1.48-1.58 −0.05 ± 0.92 −2.25-2.23 BMI Z-score 0.52 ± 1.02** −1.88-2.31 −0.17 ± 1.21 −2.92-2.04 −0.16 ± 1.11 −2.45-1.97 0.03 ± 0.88 −2.11-2.09 *p<0.05; **p<0.01; ***p<0.001 compared to DXA controls (independent samples t-test). DXA: Dual-energy X-ray absorptiometry. Table 2 Growth rates of the subjects over different time periods on medication Time period (months: mean ± SD) Pre-treatment 0-6 months 6-12 months 12-24 months 24-36 months (16.6 ± 9.4) (6.3 ± 1.5) (7.3 ± 2.0) (11.4 ± 2.1) (10.9 ± 2.5) n=11 n=30 n=30 n=28 n=24 Height velocity (cm/year) 6.5 ± 1.1 4.3 ± 1.9 4.5 ± 1.3 4.7 ± 1.2 5.3 ± 1.2 Weight velocity (kg/year) 2.7 ± 1.8 −1.5 ± 2.8 3.2 ± 3.1 2.9 ± 2.3 4.1 ± 2.3 Change in BMI/year 0.08 ± 0.86 −2.09 ± 1.89 0.81 ± 1.74 0.45 ± 1.05 0.83 ± 0.95 Δ Height Z-score/year −0.04 ± 0.17 −0.32 ± 0.38*** −0.22 ± 0.21*** −0.11 ± 0.19** −0.02 ± 0.17 Δ Weight Z-score/year −0.09 ± 0.36 −1.04 ± 0.74*** 0.03 ± 0.63 −0.13 ± 0.27* 0.06 ± 0.29 Δ BMI Z-score/year −0.08 ± 0.44 −1.29 ± 0.94*** 0.26 ± 0.96 −0.10 ± 0.39 0.12 ± 0.43 *p<0.05; **p<0.01; ***p<0.001 for the change in Z-score over the time period using single sample t-tests against a test value of zero.

0.02 ± 0.17 Δ Weight Z-score/year −0.09 ± 0.36 −1.04 ± 0.74*** 0.03 ± 0.63 −0.13 ± 0.27* 0.06 ± 0.29 Δ BMI Z-score/year −0.08 ± 0.44 −1.29 ± 0.94*** 0.26 ± 0.96 −0.10 ± 0.39 0.12 ± 0.43 *p<0.05; **p<0.01; ***p<0.001 for the change in Z-score over the time period using single sample t-tests against a test value of zero. Statistical analysis was applied to the changes in Z-scores only and used single sample t-tests with a test value of zero. Figure 2 Growth chart showing the average height and weight calculated from the changes in Z-scores. The error bars denote the standard deviation of the change in Z-scores going forwards and backwards from the baseline data, standardised for the average baseline age. The data show initial weight loss with simultaneous slowing of the height velocity on starting stimulant medication. Reference data: Centres for Disease Control and Prevention (CDC) 2000; mean ± 2SD. Body composition Baseline DXA data were available for 29 children, repeated at 6 months (median 6.8 months, mean 7.3±1.8 months) in 24 (19 boys) and at 3 years (median 2.9 years, mean 2.7±0.6 years) in 14 (12 boys).

Figure 2 Growth chart showing the average height and weight calculated from the changes in Z-scores. The error bars denote the standard deviation of the change in Z-scores going forwards and backwards from the baseline data, standardised for the average baseline age. The data show initial weight loss with simultaneous slowing of the height velocity on starting stimulant medication. Reference data: Centres for Disease Control and Prevention (CDC) 2000; mean ± 2SD. Body composition Baseline DXA data were available for 29 children, repeated at 6 months (median 6.8 months, mean 7.3±1.8 months) in 24 (19 boys) and at 3 years (median 2.9 years, mean 2.7±0.6 years) in 14 (12 boys). Longitudinal changes in tissue mass In the first 6 months the children lost 2.2±3.6% of their total tissue mass (p=0.007), which equated to an average loss of 0.65 ± 1.07 kg of tissue (p=0.009). This was associated with a significant reduction in their fat mass (5.7±3.6 to 4.3±3.1kg, change −1.40 ± 0.96 kg, p<0.001) (Figure 3). The average fat loss was greater than the weight loss because of a significant rise in lean tissue (20.4±3.0 to 21.2±3.1kg, change 0.76±0.64 kg, p<0.001). Total BMC increased significantly (1.02±0.20 to 1.06±0.20kg, change 0.04±0.03 kg, p<0.001). BMD increased significantly in the arms (0.61±0.04 to 0.62±0.04 kg/m2, change 0.006±0.012 kg/m2, p=0.023) but there was no significant change in the total body, trunk, legs or lumbar spine. Over the period from baseline to 3 years there were significant increases in all tissues.

6±0.20kg, change 0.04±0.03 kg, p<0.001). BMD increased significantly in the arms (0.61±0.04 to 0.62±0.04 kg/m2, change 0.006±0.012 kg/m2, p=0.023) but there was no significant change in the total body, trunk, legs or lumbar spine. Over the period from baseline to 3 years there were significant increases in all tissues. Figure 3 Percentage change in components of body composition. BMC: bone mineral content; BMD: bone mineral density 6 months n=23; 3 years n=14 *** p<0.001; ** p<0.01; * p<0.05 from baseline, paired t-test. There was a significant reduction in fat mass and significant increases in lean tissue and BMC in the first 6 months. Over 3 years there were significant increases in lean tissue, BMC, BMD and fat.

tent; BMD: bone mineral density 6 months n=23; 3 years n=14 *** p<0.001; ** p<0.01; * p<0.05 from baseline, paired t-test. There was a significant reduction in fat mass and significant increases in lean tissue and BMC in the first 6 months. Over 3 years there were significant increases in lean tissue, BMC, BMD and fat. Cross sectional comparison with DXA control data controlling for age, sex and height using general linear modeling After controlling for age, sex and height the subjects’ BMD and fat C/T were significantly greater than the DXA controls’ at all time points (Table 3). The subjects’ BMC was significantly greater at baseline but there was no significant difference at 6 months or 3 years. The subjects’ lean tissue mass was significantly lower than the DXA controls at 3 years. There was no significant difference at any stage between subjects and DXA controls in their total fat corrected for age, sex and height. Table 3, as well as showing the comparison of the body composition and bone density of the subjects and controls, also indicates the direction of change with time. For example lean tissue increased with age in the controls but decreased in the subjects and after 3 years of treatment the subjects had significantly lower sex, height and age corrected lean tissue than the controls. Table 3 Body composition of the subjects at different times compared to the controls

Cross sectional comparison with DXA control data controlling for age, sex and height using general linear modeling After controlling for age, sex and height the subjects’ BMD and fat C/T were significantly greater than the DXA controls’ at all time points (Table 3). The subjects’ BMC was significantly greater at baseline but there was no significant difference at 6 months or 3 years. The subjects’ lean tissue mass was significantly lower than the DXA controls at 3 years. There was no significant difference at any stage between subjects and DXA controls in their total fat corrected for age, sex and height. Table 3, as well as showing the comparison of the body composition and bone density of the subjects and controls, also indicates the direction of change with time. For example lean tissue increased with age in the controls but decreased in the subjects and after 3 years of treatment the subjects had significantly lower sex, height and age corrected lean tissue than the controls. Table 3 Body composition of the subjects at different times compared to the controls Subjects DXA Controls Baseline: n=29 (24 boys) n=241 (105 boys) 6 months: n=23 (19 boys) 36 months n=14 (12 boys) LS mean Compared to DXA controls LS mean Comparator Lean tissue (kg) Baseline 20.82 p=0.22 21.36 6 months 20.79 p=0.14 21.51 36 months 20.00 p=0.007 21.72 BMC (kg) Baseline 1.053 p=0.04 1.001 6 months 1.052 p=0.13 1.011 36 months 1.050 p=0.46 1.024 BMD (kg/m2) Baseline 0.891 p<0.0001 0.849 6 months 0.889 p=0.0002 0.851 36 months 0.889 p=0.006 0.854 Total fat (kg) Baseline 6.28 p=0.24 5.55 6 months 4.84 p=0.25 5.62 36 months 5.35 p=0.69 5.71 Central/total fat Baseline 0.421 p<0.0001 0.349 6 months 0.390 p=0.0009 0.350 36 months 0.383 p=0.047 0.352 LS Mean: Least squares mean and p values compared to DXA controls using general linear model controlling for effects of age, sex and height.

e 6.28 p=0.24 5.55 6 months 4.84 p=0.25 5.62 36 months 5.35 p=0.69 5.71 Central/total fat Baseline 0.421 p<0.0001 0.349 6 months 0.390 p=0.0009 0.350 36 months 0.383 p=0.047 0.352 LS Mean: Least squares mean and p values compared to DXA controls using general linear model controlling for effects of age, sex and height. DXA: Dual-energy X-ray absorptiometry. Statistically significant p-values are given in bold type. Longitudinal changes in sex and height corrected Z-scores After 6 months the sex and height corrected Z-scores for BMC, BMD, fat and fat C/T all showed significant reductions (0.52±0.80 to 0.34±0.79, p<0.001; 0.54±0.79 to 0.39±0.85, p=0.010; 0.76±1.13 to −0.23±0.97, p<0.000001 and 1.10±0.80 to 0.65±0.88, p=0.003 respectively) but there was no significant change in the lean tissue Z-score for height and sex (−0.05±0.96 to 0.19±1.02, p=0.11). After 3 years the sex and height corrected Z-scores for lean tissue, BMC, BMD and fat C/T all showed significant reductions from their baseline values (0.09±0.89 to −0.74±1.10, p=0.003; 0.57±0.60 to 0.02±0.66, p<0.0001; 0.62±0.67 to 0.21±0.70, p<0.0001; and 1.01±0.80 to 0.46±1.06, p=0.006 respectively). By contrast at 3 years the Z-score for the fat mass for height was not significantly different from baseline (0.41±1.06 to 0.26±1.29, p=0.62).

m their baseline values (0.09±0.89 to −0.74±1.10, p=0.003; 0.57±0.60 to 0.02±0.66, p<0.0001; 0.62±0.67 to 0.21±0.70, p<0.0001; and 1.01±0.80 to 0.46±1.06, p=0.006 respectively). By contrast at 3 years the Z-score for the fat mass for height was not significantly different from baseline (0.41±1.06 to 0.26±1.29, p=0.62). Blood tests Of 31 children who provided baseline blood samples, 25 (81%) provided repeat samples after 3 months and 14 (48%) after 3 years (Table 4). At baseline 10 children (32%) had iron deficiency; these children showed a significant rise in ferritin over 3 years (p=0.04). In the first 3 months there were significant falls in concentrations of leptin and P1NP (4.45±4.03 ng/ml to 2.73±2.03 ng/ml, p<0.010 and 554±176 ng/ml to 458±93ng/ml, p=0.019 respectively) and a significant rise in albumin (45.1 ± 3.7 g/L to 46.5±3.2 g/L, p<0.027). Concentrations of osteocalcin, CTX, IGF-1 and IGFBP-3 increased over 3 years (50.9±21.9 ng/mL to 80.9±22.2 ng/mL, p=0.003; 0.545±0.205 pg/mL to 0.886±0.518 pg/mL, p=0.038; 19.7±7.3 nmol/L to 31.1±18.0 nmol/L, p=0.041; and 4.07±1.08 ug/mL to 5.82±2.75 ug/mL, p=0.034 respectively). There were no significant changes in fasting levels of vitamin D, PTH, prealbumin, ferritin, transferrin, insulin, glucose or ghrelin. Table 4 Biochemistry results at baseline and after 3 months and 3 years

Blood tests Of 31 children who provided baseline blood samples, 25 (81%) provided repeat samples after 3 months and 14 (48%) after 3 years (Table 4). At baseline 10 children (32%) had iron deficiency; these children showed a significant rise in ferritin over 3 years (p=0.04). In the first 3 months there were significant falls in concentrations of leptin and P1NP (4.45±4.03 ng/ml to 2.73±2.03 ng/ml, p<0.010 and 554±176 ng/ml to 458±93ng/ml, p=0.019 respectively) and a significant rise in albumin (45.1 ± 3.7 g/L to 46.5±3.2 g/L, p<0.027). Concentrations of osteocalcin, CTX, IGF-1 and IGFBP-3 increased over 3 years (50.9±21.9 ng/mL to 80.9±22.2 ng/mL, p=0.003; 0.545±0.205 pg/mL to 0.886±0.518 pg/mL, p=0.038; 19.7±7.3 nmol/L to 31.1±18.0 nmol/L, p=0.041; and 4.07±1.08 ug/mL to 5.82±2.75 ug/mL, p=0.034 respectively). There were no significant changes in fasting levels of vitamin D, PTH, prealbumin, ferritin, transferrin, insulin, glucose or ghrelin. Table 4 Biochemistry results at baseline and after 3 months and 3 years Baseline 3 months 3 years (n=25-31) (n=17-25) (n=9-15) Albumin (g/L) 45.1 ± 3.7 46.5 ± 3.2* 46.1 ± 4.1 Prealbumin (g/L) 0.19 ± 0.04 0.19 ± 0.04 0.20 ± 0.04 Ferritin (ug/L) 28.0 ± 12.6 33.0 ± 13.5 31.4 ± 9.4 Transferrin (g/L) 2.61± 0.30 2.59 ± 0.39 2.63 ± 0.37 Glucose (mmol/L) 4.55 ± 0.35 4.52 ± 0.39 4.57 ± 0.63 Insulin (uU/mL) 4.23 ± 2.92 3.72 ± 2.37 3.07 ± 3.22 Leptin (ng/mL) 4.45 ± 4.03 2.73 ± 2.03* 3.13 ± 1.77 Ghrelin (pmol/L) 385 ± 143 461 ± 135 378 ± 204 IGF-1 (nmol/L) 19.7 ± 7.3 18.0 ± 7.3 31.1 ± 18.0* IGFBP-3 (ug/mL) 4.07 ± 1.08 4.02 ± 0.87 5.82 ± 2.75* P1NP (ng/mL) 554 ± 176 458 ± 93* 620 ± 264 C telopeptides (pg/mL) 0.545 ± 0.205 0.504 ± 0.169 0.886 ± 0.518* Osteocalcin (ng/mL) 50.9 ± 21.9 54.4 ± 13.7 80.9 ± 22.2** Parathyroid hormone (pmol/L) 1.40 ± 0.96 1.22 ± 0.62 1.99 ± 1.22 Vitamin D (nmol/L) 67.6 ± 23.0 71.6 ± 21.4 69.1 ± 15.8 Mean ± standard deviation shown.

P1NP (ng/mL) 554 ± 176 458 ± 93* 620 ± 264 C telopeptides (pg/mL) 0.545 ± 0.205 0.504 ± 0.169 0.886 ± 0.518* Osteocalcin (ng/mL) 50.9 ± 21.9 54.4 ± 13.7 80.9 ± 22.2** Parathyroid hormone (pmol/L) 1.40 ± 0.96 1.22 ± 0.62 1.99 ± 1.22 Vitamin D (nmol/L) 67.6 ± 23.0 71.6 ± 21.4 69.1 ± 15.8 Mean ± standard deviation shown. * p<0.05 **p<0.01 compared to baseline, paired t-test. P1NP: propeptide of type I collagen; IGF-1: insulin-like growth factor 1; IGFBP-3: insulin-like growth factor binding protein 3; Vitamin D: 25-hydroxy vitamin D; There were significant reductions in P1NP, leptin and albumin in the first 3 months. Over 3 years there were significant increases in IGF-1, IGFBP-3, and C telopeptides and osteocalcin. We found no significant differences between the effects of dexamphetamine and methylphenidate on any of the parameters measured. 0–3 month changes predicting subsequent changes in body composition variables Changes in the concentrations of IGF-1 and IGFBP-3 in the first 3 months were significant predictors of the changes in lean tissue from 6 months to 3 years (r=0.72, p=0.028 and r=0.74, p=0.022 respectively); in addition the changes in IGFBP-3 also correlated inversely with the subsequent changes in fat mass (r=−0.95, p<0.001). The changes in osteocalcin in the first 3 months correlated with the changes in BMD from 6 months to 3 years (r=0.77, p=0.026).

6 months to 3 years (r=0.72, p=0.028 and r=0.74, p=0.022 respectively); in addition the changes in IGFBP-3 also correlated inversely with the subsequent changes in fat mass (r=−0.95, p<0.001). The changes in osteocalcin in the first 3 months correlated with the changes in BMD from 6 months to 3 years (r=0.77, p=0.026). Discussion To our knowledge this is the first study to utilize DXA for determining the actual and relative losses of tissue when children start treatment with stimulant medication and to relate these to biochemical changes. The serum concentrations of leptin were decreased after starting treatment, correlating with the significant loss of fat mass. There was an early relative decrease in bone mass associated with a reduction in bone formation, with a significant recovery after three years of treatment associated with a recovery in cell coupling and bone turnover. In contrast, serum concentrations of vitamin D and PTH remained stable. At baseline the ADHD subjects had greater central adiposity than the DXA controls. The proportion of central fat declined on treatment, in parallel with progressive reductions in weight and BMI Z-scores. Over 3 years the lean tissue including bone increased more slowly than would be expected for growth in height.

able. At baseline the ADHD subjects had greater central adiposity than the DXA controls. The proportion of central fat declined on treatment, in parallel with progressive reductions in weight and BMI Z-scores. Over 3 years the lean tissue including bone increased more slowly than would be expected for growth in height. The main strength of this study is in the combination of body composition analysis with biochemical parameters and regular, careful monitoring of height and weight. This methodology allowed us to relate the changes in growth parameters and body composition with our characterization of changes in serum metabolic markers, calciotropic hormones and bone turnover markers. However, the study was limited by the small number of timepoints for evaluating longitudinal changes. Another limitation of the study is the relatively small size of the cohort, which may have been the reason that the reductions in IGF-1, IGFBP-3 and CTX at 3 months did not reach statistical significance. This study is also limited by the lack of longitudinal control data. Although there were no untreated children with ADHD for longitudinal comparison, we were able to show significant reductions in the rates of increase in height and weight on starting treatment. This is consistent with stimulant associated reductions in the growth rates [4]. Another limitation is that the subjects and controls were not analysed on the same DXA machine. However, we principally used the controls’ data as a reference for longitudinal analysis of changes within subjects. For this, minor differences between machines would be far less important than they would be for the direct comparisons of the subjects with the controls. Using linear modeling controlling for age meant that all the controls could be compared with the subjects at every time point, maximising statistical power. It was not possible to evaluate the changes in relation to physical maturation as only one girl and one boy had entered puberty at the 3 year review. When we included stage of puberty in the analysis there was no significant effect, which we attribute to insufficient numbers.

y time point, maximising statistical power. It was not possible to evaluate the changes in relation to physical maturation as only one girl and one boy had entered puberty at the 3 year review. When we included stage of puberty in the analysis there was no significant effect, which we attribute to insufficient numbers. There has been one previous cross sectional study investigating BMD and bone turnover in 10 boys who had been taking methylphenidate for 1–2 years compared to 10 healthy boys matched for age, height and weight [10]. The boys had had no changes in growth percentiles on treatment. This study reported no significant differences in BMD, serum bone-specific alkaline phosphatase or in urinary deoxypyridinoline in the study group as compared to the controls. In contrast, using more specific methods to quantify bone turnover, we have identified a significant decrease in bone formation (P1NP) during the early phase of the treatment and a “catch up” phenomenon at year 3 for both bone formation and bone resorption.

xypyridinoline in the study group as compared to the controls. In contrast, using more specific methods to quantify bone turnover, we have identified a significant decrease in bone formation (P1NP) during the early phase of the treatment and a “catch up” phenomenon at year 3 for both bone formation and bone resorption. Our finding of greater weight and BMI Z-scores in the subjects than that of the DXA controls is consistent with the findings of Holtkamp et al. who found a higher rate of obesity in untreated children with ADHD than community controls [11]. We were also able to identify a higher proportion of central fat and a higher baseline BMD. As children approach puberty their total fat increases, together with their proportion of central fat [12]. However, in the subjects the proportion of central fat decreased with time. Central body fat distribution in children and adolescents is associated with increased cardiovascular and metabolic risk factors [13]. The relative improvement in fat distribution experienced by the subjects in this study contrasts with the pattern of increasing abdominal fat in relation to age among adolescents within their local community [12].

n in children and adolescents is associated with increased cardiovascular and metabolic risk factors [13]. The relative improvement in fat distribution experienced by the subjects in this study contrasts with the pattern of increasing abdominal fat in relation to age among adolescents within their local community [12]. Our findings of initial preferential loss of central fat are consistent with findings from a systematic review of weight loss in adults that included 61 studies [14]. More weight loss over longer periods showed losses of central and peripheral fat in comparable proportions. The changes in body composition were not specific to the method employed for weight reduction; methods included exercise, diet, medication and surgery. Our study of weight loss on stimulant medication may therefore supply detail about the changes in body composition that might be anticipated if growing children lose weight. Conclusions The effects of treatment in childhood on health in adult life are important. The present study shows that even relatively minor reductions in weight on stimulant medication can be associated not only with acute effects on fat and bone turnover but also with long-term changes in body composition. The improvement we observed in the proportion of central fat for height might reduce the long term cardiovascular and metabolic risk. However, any benefit would need to be evaluated in relation to the slower rate of increase in lean tissue including bone. Further study is required to determine the effects of these changes on adult health.

bserved in the proportion of central fat for height might reduce the long term cardiovascular and metabolic risk. However, any benefit would need to be evaluated in relation to the slower rate of increase in lean tissue including bone. Further study is required to determine the effects of these changes on adult health. Competing interests The authors declare that they have no competing interests. Authors’ contributions AP conceived the study, recruited and enrolled the participants, was involved in data collection and study coordination, statistical analysis and interpretation and writing the paper. JB contributed to data collection and statistical analysis. TM and MH carried out biochemical analysis including the bone biomarkers. EM contributed to the data collection and study coordination. LAB contributed to the study design and data interpretation. GD contributed to the study design and data interpretation. All authors were involved in revision of the manuscript and approved the final version. Acknowledgements The authors would like to thank the children and their families, the Departments of Nuclear Medicine at Nepean Hospital and The Children’s Hospital at Westmead, and Dr Jason Hort for referring patients to the study. Funding This research was supported by the Australian Women & Children’s Research Foundation (OZWAC) and by the Nepean Medical Research Foundation. The funding bodies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Introduction Normal fluid balance requires an intact thirst mechanism and normal free water excretion by the kidneys, mediated by arginine vasopressin (AVP), also known as antidiuretic hormone (ADH). AVP exerts its antidiuretic action via the V2 vasopressin receptor (V2R) in the basolateral membrane of renal collecting duct cells. Binding and activation of the V2R, a G-protein coupled receptor (GPCR), increases intracellular cAMP and mediates shuttling of the water channel aquaporin-2 (AQP2) from cytosolic storage vesicles to the apical membrane of collecting duct cells, resulting in increased water permeability and antidiuresis [1]. A fraction of this AQP2 is excreted in the urine and has been found to be a useful marker of V2R activity [2]. Water loading in a normal individual suppresses plasma AVP levels and attenuates antidiuresis as a result of decreased AQP2 shuttling to the apical membrane of collecting duct cells. Consequently, less AQP2 is shed into the urine [3,4].

excreted in the urine and has been found to be a useful marker of V2R activity [2]. Water loading in a normal individual suppresses plasma AVP levels and attenuates antidiuresis as a result of decreased AQP2 shuttling to the apical membrane of collecting duct cells. Consequently, less AQP2 is shed into the urine [3,4]. We have previously described a novel syndrome of impaired water excretion mediated by gain-of-function mutations in the X-linked gene for V2R in two unrelated male infants who presented with irritability or seizures and hyponatremia [5]. Their clinical and laboratory findings were consistent with the syndrome of inappropriate secretion of antidiuretic hormone (SIADH) yet their AVP levels were undetectable. We have termed this condition "nephrogenic syndrome of inappropriate antidiuresis" (NSIAD). Each patient carries a missense mutation in codon 137 of AVPR2, which results in a change from arginine to cysteine (R137C) in one patient and to leucine (R137L) in the other [5]. Codon 137 is part of a highly conserved DRY/H motif located at the junction of the third transmembrane domain and the second intracellular loop of class 1 GPCRs. This motif is critical for G-protein coupling [6] and the two mutations each resulted in a constitutively active V2R [5].

to leucine (R137L) in the other [5]. Codon 137 is part of a highly conserved DRY/H motif located at the junction of the third transmembrane domain and the second intracellular loop of class 1 GPCRs. This motif is critical for G-protein coupling [6] and the two mutations each resulted in a constitutively active V2R [5]. Since our initial description of NSIAD, many reports from around the world, including individual and family studies, have characterized the clinical course of this syndrome [7-12]. All of these patients have the R137C V2R mutation. Here we report the clinical course of the only known patient with NSIAD caused by the more potent R137L V2R mutation [5]. This patient demonstrated the ability to maintain eunatremia during ad lib dietary intake. However, urine AQP2 levels were elevated and did not suppress normally during a standard water loading test, consistent with a gain-of-function mutation of V2R.

patient with NSIAD caused by the more potent R137L V2R mutation [5]. This patient demonstrated the ability to maintain eunatremia during ad lib dietary intake. However, urine AQP2 levels were elevated and did not suppress normally during a standard water loading test, consistent with a gain-of-function mutation of V2R. Patient and methods Patient CS first presented at 2.5 months of age with seizures and hyponatremia (118-120 mmol/L). His laboratory studies suggested SIADH; however, he had undetectable AVP levels on repeated occasions. DNA sequencing revealed a missense mutation in his AVPR2 gene, resulting in a R137L gain-of-function mutation of the V2R [5]. Fluid restriction and oral urea resulted in eunatremia and normal growth and development. The patient was lost to follow-up between age 20-32 months, during which time urea was given only intermittently and ultimately discontinued by the family at age 30 months. Despite inconsistent or complete lack of intake of urea, periodic measurements of serum sodium concentration by his pediatrician were always within the normal range. He was admitted to our institution at age 36 months to assess his water metabolism and to determine need for ongoing urea treatment.

ily at age 30 months. Despite inconsistent or complete lack of intake of urea, periodic measurements of serum sodium concentration by his pediatrician were always within the normal range. He was admitted to our institution at age 36 months to assess his water metabolism and to determine need for ongoing urea treatment. Water loading test A water loading test [4,13,14] was performed in the Pediatric Clinical Research Center at our institution. A peripherally-inserted central catheter and a Foley catheter were placed 24 hours prior to the test. The patient was monitored without fluid restriction for 24 hr. Baseline urine and blood were collected at 60 and 30 minutes prior to administration of water [20 mL/kg (360 mL) over 30 minutes via nasogastric tube]. Urine and blood samples were collected every 30 and 60 minutes, respectively, for 4 hours for measurement of sodium and osmolality. In addition, serum AVP (Quest diagnostics) [5] and urine AQP2, creatinine, and urine volume were measured. After the test, the patient was observed on ad libitum dietary intake. Intake and output were monitored for an additional 72 hours. Of note, the same reference laboratory for serum AVP measurement was used throughout this child's life.

(Quest diagnostics) [5] and urine AQP2, creatinine, and urine volume were measured. After the test, the patient was observed on ad libitum dietary intake. Intake and output were monitored for an additional 72 hours. Of note, the same reference laboratory for serum AVP measurement was used throughout this child's life. Preparation of urine samples for urinary AQP2 analysis Urine samples were centrifuged at 4,000 g for 5 min to remove any debris. The supernatant was diluted 1:1 with 2 × sodium dodecyl sulfate (SDS) sample buffer containing 2-mercaptoethanol. The samples were heated for 15 min at 65°C and stored at -20°C until analyzed. Just before loading for electrophoresis, samples were warmed to 37°C. Development of AQP2 antibody As described previously [15], a rabbit polyclonal antibody against AQP2 was prepared by Gene-med Biotechnologies Inc., (San Francisco, CA, USA) using a synthetic peptide (VELHSPQSLPRGSKA) from the COOH terminus of AQP2 [16]. The peptide was conjugated to keyhole limpit hemocyanin (KHL) by a cystein sulfhydryl linkage. Test bleedings were screened by ELISA. Final titers were reported to be > 100,000. In western blot analysis, the AQP2 antibody showed no reactivity to BSA.

thetic peptide (VELHSPQSLPRGSKA) from the COOH terminus of AQP2 [16]. The peptide was conjugated to keyhole limpit hemocyanin (KHL) by a cystein sulfhydryl linkage. Test bleedings were screened by ELISA. Final titers were reported to be > 100,000. In western blot analysis, the AQP2 antibody showed no reactivity to BSA. Urine AQP2 western blot AQP2 peptide was conjugated to BSA using the Imject Immunogen EDC conjugation kit (Pierce, Rockford, IL, USA). The AQP2 peptide conjugated BSA (AQP2-BSA) at varying concentrations and 10 μl of the patient's sample at each time point were resolved on a 12% SDS-polyacrylamide gel and transferred to a polyvinylidine difluoride membrane. The membrane was incubated with the rabbit anti-human AQP2 antibody at 4°C overnight, washed, and incubated with anti-rabbit IgG horseradish peroxidase antibody (Amersham, Piscataway, NJ, USA) [15]. Immunoreactive bands were visualized by enhanced chemiluminescence (NEN Life Science). The bands of the AQP2-BSA standards and the nonglycosylated form of AQP2 on the film were scanned and analyzed using NIH image software.

bated with anti-rabbit IgG horseradish peroxidase antibody (Amersham, Piscataway, NJ, USA) [15]. Immunoreactive bands were visualized by enhanced chemiluminescence (NEN Life Science). The bands of the AQP2-BSA standards and the nonglycosylated form of AQP2 on the film were scanned and analyzed using NIH image software. Calculation of urinary AQP2 excretion After densitometry measurements, a standard curve was constructed of known amounts of AQP2-BSA versus densitometry measurements, and the densitometry measurements of the urine samples were converted to a numerical value calculated from the curve (Prism, Graph Pad, San Diego, CA, USA). Numerical values were reported as pmol/mg creatinine. In initial studies, concentrations of AQP2 peptide conjugated BSA included 10, 20, 30, and 50 ng [15]. However, with these concentrations of AQP2-BSA, the patient's samples for urine AQP2 were too high to accurately measure. Repeat immunoblots were performed with AQP2-BSA at 50, 100, 150, and 200 ng as described above. Results Pre-water loading test As initially reported at the time of NSIAD diagnosis, serum AVP was undetectable with concurrent hyponatremia, serum hypoosmolality, and inappropriately concentrated urine (Table 1). Following initiation of water restriction and urea treatment [17], serum sodium and osmolality normalized, associated with measurable serum AVP of 5.5 pg/mL (Table 1). Despite ad libitum fluid intake and subsequent discontinuation of urea, these values remained normal (Table 1).

ropriately concentrated urine (Table 1). Following initiation of water restriction and urea treatment [17], serum sodium and osmolality normalized, associated with measurable serum AVP of 5.5 pg/mL (Table 1). Despite ad libitum fluid intake and subsequent discontinuation of urea, these values remained normal (Table 1). Table 1 Serum sodium (Na) concentration, serum osmolality (Osm), serum AVP, and urine osmolality from initial evaluation at 2.5 months to 3 years of age Age Events Serum Na (134-143 mmol/L) Serum Osm (285-293 mOsm/kg) Serum AVP (1.0-13.3 pg/mL) Urine Osm (300-900 mOsm/kg) 2.5 m Initial presentation 118 247 < 1 390 5.5 m After 1 month of water restriction and urea 142 295 5.5 n/a* 3 yr Off urea for > 6 months; ad lib intake 139 295 3.1 628 *n/a = not available

Table 1 Serum sodium (Na) concentration, serum osmolality (Osm), serum AVP, and urine osmolality from initial evaluation at 2.5 months to 3 years of age Age Events Serum Na (134-143 mmol/L) Serum Osm (285-293 mOsm/kg) Serum AVP (1.0-13.3 pg/mL) Urine Osm (300-900 mOsm/kg) 2.5 m Initial presentation 118 247 < 1 390 5.5 m After 1 month of water restriction and urea 142 295 5.5 n/a* 3 yr Off urea for > 6 months; ad lib intake 139 295 3.1 628 *n/a = not available Water loading test Prior to the water loading test, the patient's serum sodium concentration was 135 mmol/L, and serum and urine osmolality were 283 mmol/L and 857 mmol/kg, respectively. An hour after the 20 mL/kg water load, serum sodium was 128 mmol/L, and serum and urine osmolality were 272 mmol/L and 709 mmol/kg, respectively (Table 2). Over the next three hours, his serum sodium concentration remained stable and was 127 mmol/L at the conclusion of the test, at which time his serum osmolality had decreased to 267 mmol/L. Urine osmolality was inappropriately increased to 884 mmol/kg, exceeding the baseline value. Of note, serum AVP was 2.6 pg/mL at baseline and decreased to 1.1 pg/mL by one hour after water loading. Urine output during the 4-hour test was 140 mL, which was only 39% of the water load (normal ≥ 80-90% of the water load). Table 2 Serum sodium, serum osmolality, and urine osmolality during standard water loading test (20 mL/kg) at 3 years of age Water Loading Test Time (min) Serum Na (134-143 mmol/L) Serum Osm (285-294 mOsm/kg) Urine Osm (300-900 mOsm/kg)

Water loading test Prior to the water loading test, the patient's serum sodium concentration was 135 mmol/L, and serum and urine osmolality were 283 mmol/L and 857 mmol/kg, respectively. An hour after the 20 mL/kg water load, serum sodium was 128 mmol/L, and serum and urine osmolality were 272 mmol/L and 709 mmol/kg, respectively (Table 2). Over the next three hours, his serum sodium concentration remained stable and was 127 mmol/L at the conclusion of the test, at which time his serum osmolality had decreased to 267 mmol/L. Urine osmolality was inappropriately increased to 884 mmol/kg, exceeding the baseline value. Of note, serum AVP was 2.6 pg/mL at baseline and decreased to 1.1 pg/mL by one hour after water loading. Urine output during the 4-hour test was 140 mL, which was only 39% of the water load (normal ≥ 80-90% of the water load). Table 2 Serum sodium, serum osmolality, and urine osmolality during standard water loading test (20 mL/kg) at 3 years of age Water Loading Test Time (min) Serum Na (134-143 mmol/L) Serum Osm (285-294 mOsm/kg) Urine Osm (300-900 mOsm/kg) -60 135 283 857 +60 128 272 709 +120 128 266 753 +180 128 268 842 +240 127 267 884 Intake and output during ad lib fluid intake During the subsequent 3 days following the water loading test, the patient had free access to liquid based on his thirst. His daily intake was between 50-55 mL/kg/24 h (normal fluid intake for weight is 75-80 mL/kg/24 h) [18] whereas his 24 hour urine output remained at 30-35 mL/kg/24 hr. His serum sodium concentration was maintained between 131 and 139 mmol/L during this time.

t, the patient had free access to liquid based on his thirst. His daily intake was between 50-55 mL/kg/24 h (normal fluid intake for weight is 75-80 mL/kg/24 h) [18] whereas his 24 hour urine output remained at 30-35 mL/kg/24 hr. His serum sodium concentration was maintained between 131 and 139 mmol/L during this time. Urine aquaporin 2 Quantitative immunoblot assay for urine AQP2 showed an abnormally elevated baseline of 420 pmol/mg Cr, which did not suppress normally for at least 3 hr after water loading (87-98% of baseline, where normal is 5-17% of baseline) [15] (Figures 1, 2). The AQP2-BSA standard curve for densitometry measurement is shown in Figure 3[5]. Figure 1 Western immunoblot showing urine AQP2 excretion from 60 minutes prior to 240 minutes following oral water load. AQP2-BSA standards included 50, 100, 150, and 200 ng. Figure 2 Quantitative assessment of nonglycosylated AQP2 excretion in pmol/mg creatinine from 60 minutes prior to 240 minutes following oral water load. Sample values for the patient were corrected for urinary creatinine. Figure 3 Standard curve for urine AQP2-BSA measurement. The y-axis represents densitometry measurement and the x-axis represents the corresponding nanogram value for AQP2-BSA.

Figure 2 Quantitative assessment of nonglycosylated AQP2 excretion in pmol/mg creatinine from 60 minutes prior to 240 minutes following oral water load. Sample values for the patient were corrected for urinary creatinine. Figure 3 Standard curve for urine AQP2-BSA measurement. The y-axis represents densitometry measurement and the x-axis represents the corresponding nanogram value for AQP2-BSA. Discussion This is the first demonstration in a patient with NSIAD caused by the R137L V2R mutation of urinary AQP2 excretion which was markedly elevated at baseline and which did not suppress normally in a standard water loading test. Our patient's baseline urine AQP2 excretion exceeded that seen in normal adults following 12 hours of water deprivation (7-24 pmol/mg creatinine; personal communication, F. Umenishi), a condition known to increase urine AQP2 excretion [3]. These findings of increased urinary AQP2 in our patient are consistent with a gain of function mutation in V2R.

P2 excretion exceeded that seen in normal adults following 12 hours of water deprivation (7-24 pmol/mg creatinine; personal communication, F. Umenishi), a condition known to increase urine AQP2 excretion [3]. These findings of increased urinary AQP2 in our patient are consistent with a gain of function mutation in V2R. Persistent urinary excretion of AQP2 during water loading has been reported in one patient with NSIAD [12] caused by the less potent R137C mutation [5]. In addition, altered urinary excretion of AQP2 has been described in several human diseases of pathological AVP secretion. A decrease or increase in urinary AQP2 levels was shown to correspond with diminished or exaggerated levels of AVP, as seen in central diabetes insipidus or SIADH, respectively [3,4]. In normal individuals, water loading reduces antidiuresis and, as expected, urine AQP2 levels [3,15]. In studies of water loading in normal adults, urine AQP2 excretion at 2 hours after an oral water load of 20 mL/kg was reduced to 10% of baseline [15]. In the absence of appropriate diuresis during water loading, NSIAD patients would be predicted to have persistently elevated urine AQP2 levels. This is demonstrated in the R137C V2R mutation patient who received 50% of a standard water load [12] and confirmed by the results of our patient, who received the standard water load, in which urine AQP2 levels did not suppress normally, falling to only 87-98% of baseline for up to 3.5 hours after water loading.

vels. This is demonstrated in the R137C V2R mutation patient who received 50% of a standard water load [12] and confirmed by the results of our patient, who received the standard water load, in which urine AQP2 levels did not suppress normally, falling to only 87-98% of baseline for up to 3.5 hours after water loading. The patient's serum AVP levels merit discussion. AVP levels were undetectable (< 1 pg/ml) on initial presentation with hyponatremia during infancy, leading us to the identification of an activating V2R mutation [5]. Following normalization of the patient's hydration status with fluid restriction and oral urea supplementation, AVP levels were in the normal range (5.5 pg/ml). AVP levels were again in the normal range (3.1 pg/ml) 6 months following self-discontinuation of urea, associated with eunatremia and hypodipsia. Furthermore, during the water loading challenge, serum AVP levels decreased from a baseline of 2.6 pg/mL to 1.1 pg/mL, the approximate assay limit of detectability, by one hour, when the patient had developed hyponatremia and serum hypoosmolality. Water loading in a patient with R137C mutation also resulted in low but detectable serum AVP level [10]. Thus, despite the presence of a constitutively active V2R, these results indicate appropriate regulation of AVP secretion in NSAID. Possible explanations for the low but detectable AVP levels following water loading in NSIAD include contamination with platelet-bound AVP [19] and/or a slight vasovagal stimulus that may have exerted a non-osmotic effect that resulted in an incomplete suppression of AVP secretion.

regulation of AVP secretion in NSAID. Possible explanations for the low but detectable AVP levels following water loading in NSIAD include contamination with platelet-bound AVP [19] and/or a slight vasovagal stimulus that may have exerted a non-osmotic effect that resulted in an incomplete suppression of AVP secretion. It is noteworthy that this patient demonstrated an intact thirst mechanism and relative hypodipsia, allowing him to maintain serum sodium in the 131-139 mmol/L range under ad lib conditions. His relative basal eunatremia is likely a consequence of a diet that is no longer exclusively liquid and is in marked contrast to the severe hyponatremia observed at initial presentation, when the infant was exclusively formula-fed. From a dietary standpoint, our results concur with a previous report in which a child with NSIAD, diagnosed retrospectively, remained eunatremic after transition to solid food and discontinuation of sodium supplementation [7,8,11].

onatremia observed at initial presentation, when the infant was exclusively formula-fed. From a dietary standpoint, our results concur with a previous report in which a child with NSIAD, diagnosed retrospectively, remained eunatremic after transition to solid food and discontinuation of sodium supplementation [7,8,11]. An individual with NSIAD may escape detection during infancy if the activating V2R mutation is mild and/or if overhydration sufficient to induce hyponatremia does not occur. A report by Decaux et al. of a large pedigree of NSIAD patients caused by the R137C mutation in V2R highlights the marked variability in clinical presentation of this disorder [9]. One affected hemizygous male was apparently asymptomatic throughout life, and was discovered only after being administered a water loading test. Similarly, two other hemizygous adult males were discovered only when their 8 week old nephew was diagnosed with the disease and genetic testing was performed on the family [8]. One of the adult males was completely asymptomatic throughout his life and the other apparently has had generalized tonic-clonic seizures early in life of unknown etiology. In addition, several heterozygous females demonstrated inappropriate antidiuresis during water loading [8,10,12]. Thus, NSIAD may be more prevalent in the general population than would have otherwise been predicted based on the relatively small number of patients diagnosed in infancy.

early in life of unknown etiology. In addition, several heterozygous females demonstrated inappropriate antidiuresis during water loading [8,10,12]. Thus, NSIAD may be more prevalent in the general population than would have otherwise been predicted based on the relatively small number of patients diagnosed in infancy. In conclusion, this is the first report that urinary AQP2 levels in a patient with NSIAD caused by the R137L V2R mutation are elevated at baseline and do not suppress appropriately in response to water loading. This study demonstrates that urinary AQP2 can be a useful marker of V2R activity in NSIAD. Thus, increased urinary AQP2 excretion, in the setting of euvolemic hyponatremia and low or undetectable levels of serum AVP, suggests the possibility of NSIAD. This diagnosis should be confirmed by sequencing of AVPR2. Consent Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Competing interests The authors declare that they have no competing interests. Authors' contributions CCC and SMR designed the study, performed the analysis, and drafted the manuscript. MAC performed the western blot analysis. SAR and SEG participated in the design of the study. All authors read and approved the final manuscript. Acknowledgements This paper is supported by a grant (K08 DK076721-01) to Dr. Cheung, a training grant (T32-DK007161) to Dr. Ranadive, and a grant (RR024131-01) to the Pediatric Clinical Research Center.

Dear Editor, A false statement has been published in your journal in an article by P. A. Lee and C. P. Houk. (2010) Article ID 563640. "The Role of Support Groups, Advocacy Groups, and Other Interested Parties in Improving the Care of Patients with Congenital Adrenal Hyperplasia: Pleas and Warnings [1]." These authors say "Confrontational tactics used by "advocacy" groups have included pressuring the medical community to adopt narrow mandates, such as a moratorium on all reproductive system surgery (six), that cannot apply to the broad range of situations encountered in practice or the use of accusations regarding therapy received by patients in the past." In support of their argument they cite our paper (M. Diamond & H. K. Sigmundson, Management of intersexuality. Guidelines for dealing with persons with ambiguous genitalia. Archives of Pediatrics & Adolescent Medicine. 1997;151(10):1046-1050 [2]. We have been wrongly quoted and cited. Here is exactly what we say:

These authors say "Confrontational tactics used by "advocacy" groups have included pressuring the medical community to adopt narrow mandates, such as a moratorium on all reproductive system surgery (six), that cannot apply to the broad range of situations encountered in practice or the use of accusations regarding therapy received by patients in the past." In support of their argument they cite our paper (M. Diamond & H. K. Sigmundson, Management of intersexuality. Guidelines for dealing with persons with ambiguous genitalia. Archives of Pediatrics & Adolescent Medicine. 1997;151(10):1046-1050 [2]. We have been wrongly quoted and cited. Here is exactly what we say: "Perform no major surgery for cosmetic reasons alone; only for conditions related to physical/medical health. This will entail a great deal of explanation needed for the parents who will want their children to "look normal." Explain to them that appearances during childhood, while not typical of other children, may be of less importance than functionality and post pubertal erotic sensitivity of the genitalia. Surgery can potentially impair sexual/erotic function. Therefore such surgery, which includes all clitoral surgery and any sex reassignment, should typically wait until puberty or after when the patient is able to give truly informed consent." [emphasis added]. Our emphasis was on cosmetic surgery and since our statement no evidence has shown any surgery had been of benefit. Many unneeded surgeries have been devastating.

"Perform no major surgery for cosmetic reasons alone; only for conditions related to physical/medical health. This will entail a great deal of explanation needed for the parents who will want their children to "look normal." Explain to them that appearances during childhood, while not typical of other children, may be of less importance than functionality and post pubertal erotic sensitivity of the genitalia. Surgery can potentially impair sexual/erotic function. Therefore such surgery, which includes all clitoral surgery and any sex reassignment, should typically wait until puberty or after when the patient is able to give truly informed consent." [emphasis added]. Our emphasis was on cosmetic surgery and since our statement no evidence has shown any surgery had been of benefit. Many unneeded surgeries have been devastating. Significantly damaging were those surgeries imposed on males that were sex reassigned for different reasons. Often these were instances where the penis was considered too small for appropriate male status; e.g., the John/Joan case [3,4] but there now have been many instances recorded where males were reassigned and raised as girls and then rebelled to live as males; particularly for cases of cloacal exstrophy [5,6]. Cases of micropenis are also currently recognized as not needing surgery or sex reassignment [7,8].

tatus; e.g., the John/Joan case [3,4] but there now have been many instances recorded where males were reassigned and raised as girls and then rebelled to live as males; particularly for cases of cloacal exstrophy [5,6]. Cases of micropenis are also currently recognized as not needing surgery or sex reassignment [7,8]. Creighton (2004), for example, has found "in girls with ambiguous genitalia, vaginoplasty is commonly performed during the first year of life . . . although the child is unlikely to be sexually active until after puberty. There is no good evidence it is justified. [9]" And Alizai et al. (1999) reported, "The outcome of clitoral surgery was unsatisfactory (clitoral atrophy or prominent glans) in [girls] whose genitoplasty had been performed by 3 different specialist pediatric urologists. Additional vaginal surgery was necessary for normal comfortable intercourse in [other] patients. Fibrosis and scarring were most evident in those who had undergone aggressive attempts at vaginal reconstruction in infancy [10]." Minto and others [11,12] echo similar expressions against early genital surgery. Schober has stated "A reliable, successful genitoplasty procedure that can be performed early in childhood for either feminization or masculinization has not yet been developed [13]." Legal and ethical reasons against such cosmetic surgery on infants have been presented [14] and argued against [15].

Background Growth hormone (GH) treatment is approved for treatment of short stature in a number of childhood diagnoses, such as isolated growth hormone deficiency (IGHD) and multiple pituitary hormone deficiency (MPHD). There are other childhood indications, which are not associated with a deficiency of endogenous growth hormone that can be improved by GH treatment, for instance, born small for gestational age (SGA) and idiopathic short stature (ISS) [1,2]. Growth hormone treatment has been shown to increase final adult height in each of these patient populations [3-12]. Moreover, in patients with IGHD, total gain in height SDS was reported to correlate significantly with pre-pubertal gain in height SDS, with younger age at treatment start being a significant predictor of greater treatment response [13,14]. Few studies have directly compared the growth response to GH treatment according to diagnosis and pubertal stage. Some studies suggest that pre-pubertal children with ISS experience less gain in height with GH treatment than short children with growth hormone deficiency [15,16]. The growth response has, however, not been compared in large, observational real-life studies across indications and pubertal stage.

s and pubertal stage. Some studies suggest that pre-pubertal children with ISS experience less gain in height with GH treatment than short children with growth hormone deficiency [15,16]. The growth response has, however, not been compared in large, observational real-life studies across indications and pubertal stage. Treatment outcomes that can be achieved in daily clinical practice may differ widely from the more controlled setting of randomized trials; therefore, post-marketing surveillance studies are required to generate data from large cohorts about the efficacy (and/or safety) of an intervention in the real-world clinical setting, which may then be used to inform changes in indications. While there are obvious advantages to the data provided by observational studies, there are also a number of drawbacks that arise from the necessity of having a heterogeneous, all-comer study population; these include, but are not limited to, non-standardization of laboratory tests, treatment dose and number of injections, missing parameters (e.g. mid-parental height, target height, birth weight), and variation in local practices when studies are multinational.

ity of having a heterogeneous, all-comer study population; these include, but are not limited to, non-standardization of laboratory tests, treatment dose and number of injections, missing parameters (e.g. mid-parental height, target height, birth weight), and variation in local practices when studies are multinational. The aim of this analysis of data, collected from two large ongoing observational outcome studies, was to evaluate growth and insulin-like growth factor-I (IGF-I) response data for children of short stature with IGHD, MPHD, SGA, or ISS following two years of treatment with the recombinant GH product Norditropin® (Novo Nordisk A/S, Bagsværd, Denmark). A secondary focus was to assess the impact of GH therapy in pre-pubertal children compared with the total patient population within each of the indications investigated.

ture with IGHD, MPHD, SGA, or ISS following two years of treatment with the recombinant GH product Norditropin® (Novo Nordisk A/S, Bagsværd, Denmark). A secondary focus was to assess the impact of GH therapy in pre-pubertal children compared with the total patient population within each of the indications investigated. Methods Data for this analysis were obtained from the NordiNet® International Outcome Study (IOS; NCT00960128) launched in 2006 and ongoing in 19 countries (Czech Republic, Denmark, Finland, France, Germany, Hungary, Ireland, Israel, Italy, Lithuania, Montenegro, Netherlands, Norway, Russia, Serbia, Slovenia, Sweden, Switzerland, UK), and from the US observational study NovoNet®/American Norditropin® Studies: Web-enabled Research (ANSWER) Program® (NCT01009905), which began in 2002 and is also ongoing. The IOS and the ANSWER Program® use a similar electronic platform, NordiNet®/NovoNet®, to collect and manage data about the effectiveness and safety of Norditropin® in normal clinical practice. Physicians enter data on patient history, physical examinations, and treatment regimens using the web-based NordiNet®/NovoNet® tool. The use of Norditropin® for patients included in these large observational studies is at the discretion of the participating physicians as part of their routine clinical practice. Both observational studies are operated in accordance with the Declaration of Helsinki and with the approval of local institutional review boards.

e of Norditropin® for patients included in these large observational studies is at the discretion of the participating physicians as part of their routine clinical practice. Both observational studies are operated in accordance with the Declaration of Helsinki and with the approval of local institutional review boards. Data are anonymized in the two observational studies. For all patients, physicians measure a number of variables at the initial visit according to their standard medical practice, including baseline height, weight, bone age, Tanner stage or testicular volume, maximum stimulated serum GH concentration, and serum IGF-I concentrations. At follow-up visits, data gathered include GH dose and injection frequency, height, weight, Tanner stage or testicular volume, and IGF-I levels, among other variables.

cluding baseline height, weight, bone age, Tanner stage or testicular volume, maximum stimulated serum GH concentration, and serum IGF-I concentrations. At follow-up visits, data gathered include GH dose and injection frequency, height, weight, Tanner stage or testicular volume, and IGF-I levels, among other variables. For this analysis, only patients aged <18 years with data collected at baseline AND at both one and two year follow-up visits (±3 months) were included. Data were divided by indication based on investigators’ clinical diagnosis (IGHD, MPHD, SGA, and ISS), gender, and pubertal status. Patients with IGHD were used as the reference group because this population represents the optimal indication for replacement GH treatment, where it is unlikely that other hormone deficiencies act as confounders for outcomes; in addition, there exists a large quantity of efficacy and safety data for GH treatment in IGHD. For this study, the pre-pubertal population group was defined using descriptions of clinical puberty symptoms: girls had Tanner stage 1 breast development, while boys had a testicular volume <4 mL. Children were identified as pre-pubertal when they remained at this stage of development for the entire two-year follow-up period. To be identified as pre-pubertal when information on pubertal status was lacking, girls had to be no more than 6 years of age, while boys had to be no more than 7 years old at the start of the observational period; this ensured that patients with missing information were unlikely to start puberty during the observational period, because mean age −2 SD at start of puberty in the observed patient population was ~8 years for girls and ~9 years for boys.

oys had to be no more than 7 years old at the start of the observational period; this ensured that patients with missing information were unlikely to start puberty during the observational period, because mean age −2 SD at start of puberty in the observed patient population was ~8 years for girls and ~9 years for boys. For all patients included in this analysis, height SDS at baseline and change in height SDS at two years were calculated. For patients in the NovoNet®/ANSWER Program® study, height SDS was determined according to standard formulas provided by the Center for Disease Control and Prevention [17], and for patients in the NordiNet® IOS study by the corresponding country references. Bone age was determined manually or with the automatic software application BoneXpert®, which is included in the NordiNet®, but not in the NovoNet® system. When laboratory IGF-I measurements were available, IGF-I SDS values were calculated according to established age- and sex-specific reference values and models [18] without adjustments for local differences in laboratory procedures or possible variations in IGF-I levels.

in the NordiNet®, but not in the NovoNet® system. When laboratory IGF-I measurements were available, IGF-I SDS values were calculated according to established age- and sex-specific reference values and models [18] without adjustments for local differences in laboratory procedures or possible variations in IGF-I levels. Statistical analysis Descriptive statistics and a simple analysis of variance (ANOVA) model were used to analyze combined patient data from the NordiNet® IOS and the NovoNet®/ANSWER Program®. Variables examined included change in height SDS and change in IGF-I SDS. Analyses were made both on the total population and on the sub-population of pre-pubertal patients within each indication. No adjustment for multiplicity of testing was done. Differences were regarded as statistically significant if the p-value was <0.05. All analyses were performed with SAS software, version 9.1 (SAS Institute Inc., Cary, NC, USA).

total population and on the sub-population of pre-pubertal patients within each indication. No adjustment for multiplicity of testing was done. Differences were regarded as statistically significant if the p-value was <0.05. All analyses were performed with SAS software, version 9.1 (SAS Institute Inc., Cary, NC, USA). Results Baseline characteristics NordiNet® IOS and NovoNet®/ANSWER Program® have yielded data on >11,000 and >11,500 Norditropin®-treated patients, respectively, most of whom were children with a variety of diagnoses, treated for short stature. For this analysis, 4,582 pediatric patients (i.e. the total population in this analysis) with the following indications and with two years of follow-up data were identified in the databases: IGHD, n = 3,298; SGA, n = 678; ISS, n = 334; and MPHD, n = 272. Mean age at treatment start was 10.2, 7.9, 10.9, and 7.9 years for IGHD, SGA, ISS, and MPHD, respectively (Table 1). Children born SGA had the lowest baseline height SDS of −3.1. Mean bone age was delayed relative to chronological age across all indications, ranging from −1.6 to −1.8 years. With the exception of children born SGA who had a mean IGF-I SDS of −0.7, mean values for IGF-I SDS were below −1.5 SDS for all indications at baseline. Mean GH doses by indication ranged from 0.036 (MPHD) to 0.049 (ISS) mg/kg/day (Table 1). Table 1 Baseline characteristics and mean GH dose during two-year treatment period for total and pre-pubertal patient population by indication

Results Baseline characteristics NordiNet® IOS and NovoNet®/ANSWER Program® have yielded data on >11,000 and >11,500 Norditropin®-treated patients, respectively, most of whom were children with a variety of diagnoses, treated for short stature. For this analysis, 4,582 pediatric patients (i.e. the total population in this analysis) with the following indications and with two years of follow-up data were identified in the databases: IGHD, n = 3,298; SGA, n = 678; ISS, n = 334; and MPHD, n = 272. Mean age at treatment start was 10.2, 7.9, 10.9, and 7.9 years for IGHD, SGA, ISS, and MPHD, respectively (Table 1). Children born SGA had the lowest baseline height SDS of −3.1. Mean bone age was delayed relative to chronological age across all indications, ranging from −1.6 to −1.8 years. With the exception of children born SGA who had a mean IGF-I SDS of −0.7, mean values for IGF-I SDS were below −1.5 SDS for all indications at baseline. Mean GH doses by indication ranged from 0.036 (MPHD) to 0.049 (ISS) mg/kg/day (Table 1). Table 1 Baseline characteristics and mean GH dose during two-year treatment period for total and pre-pubertal patient population by indication Children with IGHD Children born SGA Children with ISS Children with MPHD Total Pre-pubertal Total Pre-pubertal Total Pre-pubertal Total Pre-pubertal Total N (pre-pubertal % of total indication) 3,298 1,120 (34%) 678 434 (64%) 334 76 (23%) 272 165 (61%) Male gender, N (%) 2,444 (74%) 778 (70%) 396 (58%) 274 (63%) 240 (72%) 59 (78%) 171 (63%) 106 (64%) Mean chronological age ± SD (years) 10.2 ± 3.6 6.5 ± 2.8 7.9 ± 3.2 6.1 ± 2.1 10.9 ± 2.9 7.1 ± 2.3 7.9 ± 4.9 4.8 ± 3.5 Mean height SDS ± SD −2.3 ± 1.0 −2.7 ± 1.0 −3.1 ± 0.9 −3.3 ± 0.9 −2.3 ± 0.8 −2.6 ± 0.9 −2.0 ± 1.5 −2.3 ± 1.6 Mean bone age delay from chronological age ± SD (years) [N] −1.8 ± 1.4 [2,419] −1.9 ± 1.3 [703] −1.6 ± 1.4 [430] −1.6 ± 1.2 [266] −1.7 ± 1.5 [294] −1.9 ± 1.4 [65] −1.7 ± 1.8 [137] −1.3 ± 1.7 [63] Mean IGF-I SDS ± SD [N] −1.8 ± 1.7 [2,784] −1.4 ± 1.4 [890] −0.7 ± 1.6 [434] −0.5 ± 1.4 [276] −1.6 ± 1.7 [308] −1.2 ± 1.4 [66] −1.9 ± 2.2 [222] −1.4 ± 1.9 [133] Mean GH dose during two-year treatment period ± SD (mg/kg/day) [N] 0.042 ± 0.013 [3,272] 0.038 ± 0.011 [1,110] 0.042 ± 0.013 [678] 0.041 ± 0.013 [434] 0.049 ± 0.012 [334] 0.046 ± 0.011 [76] 0.036 ± 0.013 [268] 0.035 ± 0.011 [163] The percentage of pre-pubertal children in each indication group varied: ISS, 23%; IGHD, 34%; MPHD, 61%; and SGA, 64%. In comparison with the total population, the mean height SDS was lower for pre-pubertal children, although mean IGF-I SDS was higher in this group.

± 0.011 [76] 0.036 ± 0.013 [268] 0.035 ± 0.011 [163] The percentage of pre-pubertal children in each indication group varied: ISS, 23%; IGHD, 34%; MPHD, 61%; and SGA, 64%. In comparison with the total population, the mean height SDS was lower for pre-pubertal children, although mean IGF-I SDS was higher in this group. Change in height In the total patient population, the mean change in height SDS at one year for all indications was +0.57 SDS or higher, except among children with ISS (+0.49 SDS; Figure 1). At one year, children with MPHD and children born SGA, responded with height SDS of +0.67 and +0.64, respectively; these height gains were +0.10 and +0.07 SDS greater than in children with IGHD (p = 0.001 for both). In contrast, height gain at one year with GH treatment in ISS was −0.08 SDS lower than that in IGHD (p = 0.005). Figure 1 Mean height SDS at baseline, one year and two years and change in height SDS at one year and two years by indication. At two years of GH treatment, mean change in height from baseline was +0.99 vs. +0.97 SDS for patients with MPHD and IGHD, respectively, but was +1.03 SDS for patients born SGA (p = 0.047 vs. IGHD) and +0.84 for patients with ISS (p < 0.001 vs. IGHD). Normal height for age and gender (within ±2 SDS) was reached in 78% (IGHD), 45% (SGA), 76% (ISS), and 79% (MPHD) of children. Across all indications, 73% of patients reached normal height after two years of treatment.

3 SDS for patients born SGA (p = 0.047 vs. IGHD) and +0.84 for patients with ISS (p < 0.001 vs. IGHD). Normal height for age and gender (within ±2 SDS) was reached in 78% (IGHD), 45% (SGA), 76% (ISS), and 79% (MPHD) of children. Across all indications, 73% of patients reached normal height after two years of treatment. In pre-pubertal children, the gains in height SDS at one year, and two years, were higher than in the total population (Figure 1). Among pre-pubertal children at one year, gain in height SDS did not differ significantly between IGHD and any other indication. At two years, the gain in height was above +1 SDS for all indications and comparable, except for in ISS compared with in IGHD (+1.04 vs. +1.24 SDS, p = 0.03) (Figure 1). Normal height was reached in 73% (IGHD), 46% (SGA), 72% (ISS), and 75% (MPHD) of children at two years in the pre-pubertal population. Change in IGF-I SDS The mean change in IGF-I SDS for the total population was greater than +2 SDS after one and two years of GH treatment in all indications, except in children born SGA at one year (Figure 2). Mean values stayed within the reference range (±2 SDS) throughout the study. At two years, mean IGF-I SDS was 0.80 (IGHD), 1.15 (SGA), 1.06 (ISS), and 0.58 (MPHD). At one and two years of treatment, children born SGA had a significantly lower IGF-I increase than children with IGHD (+1.80 vs. +2.36 SDS, and +2.00 vs. +2.57 SDS, respectively, p < 0.001). Figure 2 Mean IGF-I SDS at baseline, one year and two years by indication.

Change in IGF-I SDS The mean change in IGF-I SDS for the total population was greater than +2 SDS after one and two years of GH treatment in all indications, except in children born SGA at one year (Figure 2). Mean values stayed within the reference range (±2 SDS) throughout the study. At two years, mean IGF-I SDS was 0.80 (IGHD), 1.15 (SGA), 1.06 (ISS), and 0.58 (MPHD). At one and two years of treatment, children born SGA had a significantly lower IGF-I increase than children with IGHD (+1.80 vs. +2.36 SDS, and +2.00 vs. +2.57 SDS, respectively, p < 0.001). Figure 2 Mean IGF-I SDS at baseline, one year and two years by indication. In the pre-pubertal population, the change in IGF-I only exceeded +2 SDS for ISS at 1 year (Figure 2). Between indications, no significant differences were observed in change in IGF-I SDS after either one or two years of GH treatment. Mean values for IGF-I SDS in pre-pubertal children were within the reference range for the two-year period, as in the total patient populations (Figure 2). At two years, mean IGF-I SDS was 0.45 (IGHD), 1.13 (SGA), 0.61 (ISS), and 0.65 (MPHD).

bserved in change in IGF-I SDS after either one or two years of GH treatment. Mean values for IGF-I SDS in pre-pubertal children were within the reference range for the two-year period, as in the total patient populations (Figure 2). At two years, mean IGF-I SDS was 0.45 (IGHD), 1.13 (SGA), 0.61 (ISS), and 0.65 (MPHD). Discussion This analysis of international data from two large ongoing observational studies (NordiNet® IOS and NovoNet®/ANSWER Program®) is, to our knowledge, the first to compare growth and IGF-I response rates between children with IGHD and children with MPHD, SGA, and ISS treated with GH. In the total patient population, the two-year change in height SDS was approximately +1 SDS for each indication investigated. While the positive treatment responses for children with MPHD and children born SGA exceeded the change in height for children with IGHD at one year, only children born SGA had a higher response compared with IGHD after two years; these results are similar to those from other studies comparing outcomes with GH treatment in these indications [5,13,19]. It should be noted that although height change was expressed in SDS, which is generally considered more robust across gender and age than using change in cm, change in height SDS is not completely age-independent, because less variation is observed in height SD in younger vs. older ages [20]. This may, to a small degree, contribute to our finding that children with MPHD and those born SGA who were considerably younger at GH treatment start (mean 7.9 years) had a better growth response compared with IGHD (mean age of 10.2 years at treatment start). However, a positive effect of early age for growth hormone response is suggested as reported by others [5,13,19,21-25]. Among pre-pubertal children, where the differences in age were less, children with MPHD and children born SGA showed no significant difference in height gain compared with IGHD.

s at treatment start). However, a positive effect of early age for growth hormone response is suggested as reported by others [5,13,19,21-25]. Among pre-pubertal children, where the differences in age were less, children with MPHD and children born SGA showed no significant difference in height gain compared with IGHD. After two years of GH treatment, more than 75% of children with IGHD, MPHD, and ISS reached their normal height range (above −2 SDS). In contrast, although children born SGA experienced the highest growth response, only 45% reached normal height because their baseline height was considerably lower than any other group. The severe short stature of children born SGA observed in this study can be partially explained by the labeling in Europe, where patients must be below −2.5 SDS before treatment is initiated, and in France, where medical reimbursement is possible only if children born SGA are less than or equal to −3 SDS at treatment start. Pre-pubertal children born SGA had the smallest increase in IGF-I SDS of all pre-pubertal indications, but had the highest baseline levels. A positive correlation has been proposed between increasing IGF-I levels and height increase in pre-pubertal children [26]. The results of our study appear to lend some support to this hypothesis, because pre-pubertal children in all other indications than ISS had comparable increases in IGF-SDS and height SDS.

levels. A positive correlation has been proposed between increasing IGF-I levels and height increase in pre-pubertal children [26]. The results of our study appear to lend some support to this hypothesis, because pre-pubertal children in all other indications than ISS had comparable increases in IGF-SDS and height SDS. The smallest two-year height SDS increase occurred in children with ISS (+0.84 SDS), which was significantly lower compared with IGHD. Several factors could have influenced the lower height increase seen in children with ISS: (a) age, the children with ISS in this study had a higher mean age at the start of GH treatment compared with the other indications investigated. Other studies have shown higher age at treatment initiation to be negatively associated with growth in children receiving GH, including ISS [16,21]; (b) pubertal status, the ISS group had the lowest percentage of pre-pubertal children (23%) (see later for discussion); (c) the variable etiology of the disease, because the diagnosis of ISS is based upon short stature due to a variety of unknown causes [27]. In our analysis, among the four indications, children with ISS had the highest mean age in the pre-pubertal subgroup (7.1 years at baseline) and a male gender dominance (78%), with bone age less or equally delayed in ISS compared with the three other indications studied (Table 1), which could suggest an underlying disorder of constitutional growth delay; d) lastly, in the total population, in spite of a higher GH dose given in ISS, the two-year IGF-I change was comparable between ISS and IGHD, while height gain was less in ISS. At baseline, IGF-I deficiency was less in ISS than in IGHD. These facts suggest some degree of GH or IGF-I insensitivity in ISS compared with IGHD, possibly influenced by differences in underlying disease nature, age, gender, and/or pubertal stage. Other studies support a degree of insensitivity to IGF-I in children with ISS [15,28], with the wide range of growth responses to GH in this patient population being consistent with the broad spectrum of genetic and molecular defects that result in IGF-I insensitivity [27].

e nature, age, gender, and/or pubertal stage. Other studies support a degree of insensitivity to IGF-I in children with ISS [15,28], with the wide range of growth responses to GH in this patient population being consistent with the broad spectrum of genetic and molecular defects that result in IGF-I insensitivity [27]. Few other studies have compared the growth and/or IGF-I response to GH treatment between GHD and ISS. In a two-year, open-label trial, 63 children with GHD and 102 children with ISS were randomized to receive GH therapy based on an IGF-I target of 0 SDS, +2 SDS, or a dosing corresponding to the patient’s weight [15]. Children with GHD grew more than those with ISS in both IGF-targeted dosage groups despite having similar IGF-I levels. In the +2 SDS target group, the mean (±SD) change in height SDS for children with GHD was 2.04 (±0.17) compared with 1.33 (±0.09) for children with ISS. In the 0 SDS target group, the change in heights SDS results were 1.41 (±0.13) and 0.84 (±0.07), respectively. The results of a smaller clinical study comparing GHD with ISS were similar [16]. In spite of the real-life, heterogeneous nature of the observational studies reported here, our finding was consistent with a reduced growth response in ISS. However, all indications in this analysis surpassed the one-year response threshold for GH efficacy, which is generally considered to be approximately +0.25–0.5 SDS for change in height [2,22,29], thus demonstrating the ability of GH to stimulate linear growth regardless of indication.

a reduced growth response in ISS. However, all indications in this analysis surpassed the one-year response threshold for GH efficacy, which is generally considered to be approximately +0.25–0.5 SDS for change in height [2,22,29], thus demonstrating the ability of GH to stimulate linear growth regardless of indication. For children with a variety of short stature indications, the height SDS at the onset of puberty correlates strongly with final adult height [9,10,13,21,23,24,27,30]. Correlations with adult height were not examined in this analysis, but pubertal status was found to have a marked influence on growth by both one and two years of treatment. In pre-pubertal children of all indication groups the change in height SDS was greater than in the total population, consistent with the preliminary two-year findings from the NovoNet®/ANSWER Program® [19], whose entirely US patient population overlapped with US patients included in this study when two-year growth data were available. Further long-term follow-up with adult height data is needed to be able to describe fully the benefits of starting GH treatment early, especially in real-life situations, although long-term safety data on modern growth hormone therapy are generally reassuring [31]. To provide high-quality long term data, it is important to monitor the outcome of the populations in large observational studies, such as NordiNet® IOS and NovoNet®/ANSWER Program®, to assess the effectiveness and safety of GH therapy in children, especially in non-GHD indications.

herapy are generally reassuring [31]. To provide high-quality long term data, it is important to monitor the outcome of the populations in large observational studies, such as NordiNet® IOS and NovoNet®/ANSWER Program®, to assess the effectiveness and safety of GH therapy in children, especially in non-GHD indications. Examining data from clinical practice contained in large observational studies like NordiNet® IOS and NovoNet®/ANSWER Program® can provide insights into optimal treatments in actual clinical practice, such as the desirability of starting growth hormone therapy before puberty observed in this study. It must, however, be considered that the differences in patient populations, diagnostic and treatment practices between countries represented in this large combined database may influence the outcome of any analysis. In particular, the results for IGF-I shown here should be treated with caution due to the lack of information on assay and control over local laboratory measurements and possible differences in normal IGF-I levels within the international patient cohort. Furthermore, although only a small variation was observed in mean GH dose in the total population across indications and also between the total population and pre-pubertal groups, it should be noted that dose recommendations for these indications vary from country to country and, indeed, even within-country variation in prescribing habits is likely. Lastly, selection bias from the entire cohort of children receiving GH treatment for any of the four indications studied could not be excluded; for example, drop outs could not be systematically analyzed.

ndications vary from country to country and, indeed, even within-country variation in prescribing habits is likely. Lastly, selection bias from the entire cohort of children receiving GH treatment for any of the four indications studied could not be excluded; for example, drop outs could not be systematically analyzed. Conclusion After two years of GH treatment, short children born SGA showed a greater height response than children with IGHD and MPHD (who experienced comparable growth responses), while children with ISS had a slightly lower response, possibly owing to confounders and/or differences in disease nature. Despite showing the greatest SDS height response, a lower number of SGA children reached a normal height range (above −2 SDS for mean) at the end of the two-year period due to their low baseline height. More than 75% of children with IGHD, ISS, and MPHD achieved a normal height range after two years of treatment. Beginning treatment at least two years before the onset of puberty was associated with an improved height gain, which suggests that GH treatment should start well in advance of puberty to optimize height growth outcomes. Competing interest Viatcheslav Rakov is a former employee of Novo Nordisk Health Care AG. Birgitte Tønnes Pedersen is an employee of Novo Nordisk A/S. Peter A. Lee has participated in data collection for patient registries, with research support from Novo Nordisk, Pfizer, Ipsen, and Eli Lilly, and has served as a consultant for Novo Nordisk and Ipsen.

Competing interest Viatcheslav Rakov is a former employee of Novo Nordisk Health Care AG. Birgitte Tønnes Pedersen is an employee of Novo Nordisk A/S. Peter A. Lee has participated in data collection for patient registries, with research support from Novo Nordisk, Pfizer, Ipsen, and Eli Lilly, and has served as a consultant for Novo Nordisk and Ipsen. Judith Ross has participated in data collection for patient registries, with research support from Novo Nordisk, Pfizer, and Eli Lilly, and has served as a consultant for Novo Nordisk, Eli Lilly, and Abbot. Peter A. Lee, Lars Sävendahl, Isabelle Oliver, Oliver Blankenstein, Judith Ross, Marta Snajderova and Henrik Thybo Christesen are members of the NordiNet® International Outcome Study International Study Committee. Maithé Tauber is a Nordinet® International Outcome Study investigator. Investigators received financial compensation from Novo Nordisk for time spent entering data on the electronic study forms.

Peter A. Lee, Lars Sävendahl, Isabelle Oliver, Oliver Blankenstein, Judith Ross, Marta Snajderova and Henrik Thybo Christesen are members of the NordiNet® International Outcome Study International Study Committee. Maithé Tauber is a Nordinet® International Outcome Study investigator. Investigators received financial compensation from Novo Nordisk for time spent entering data on the electronic study forms. Authors’ contribution PAL-Involved in assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing and revising the manuscript after review of previous related publications, LS- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, IO- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, MT- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, OB- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, JR- Involved in assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, MS- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript, VR-Directed the project and oversaw the data analysis and intrepretation, BTP-Analyses of the data, including assessment of adequacy and statistics. HTC- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript. All authors read and approved the final manuscript.

he data analysis and intrepretation, BTP-Analyses of the data, including assessment of adequacy and statistics. HTC- Involved in collecting data and assessment of database, deciding what factors would be analyzed, evaluating the assessment, and writing the manuscript. All authors read and approved the final manuscript. Funding source This study was funded by Novo Nordisk Health Care AG. Acknowledgements The authors wrote this paper on behalf of the NordiNet® IOS International Study Committee. The authors are grateful to Peter Budka and Michael Maddalena (Watermeadow Medical, Witney, UK) for writing assistance, supported by Novo Nordisk Health Care AG. The authors are grateful to all physicians and study sites who participated in the study.

Background Central precocious puberty is the early onset of pubertal development as a result of gonadotropin release by the pituitary gland. Precocious puberty in a child can be associated with adverse consequences including compromised final adult height and psychosocial problems. Establishing the diagnosis of central precocious puberty requires documenting pubertal physical findings and measuring luteinizing hormone (LH) concentration, which is the key biochemical assessment of pubertal status. Gonadotropin-releasing hormone (GnRH)-stimulated plasma LH concentrations have been the mainstay for establishing the diagnosis of precocious puberty, but it is no longer available in the United States. GnRH analogue (leuprolide acetate) administered subcutaneously is a suitable substitute for GnRH in the diagnosis of central precocious puberty [1-5]. Ibanez et al. reported that a peak serum LH response >8 IU/L occurred in patients with progressive puberty and in patients with Tanner stage II puberty 3 hours post leuprolide acetate challenge [3]. The LH concentrations declined progressively from 3 to 6 hours post-stimulation. In patients with non-progressive puberty and in pre-pubertal controls, the LH peak occurred between 3 and 6 hours after injection [3].

sive puberty and in patients with Tanner stage II puberty 3 hours post leuprolide acetate challenge [3]. The LH concentrations declined progressively from 3 to 6 hours post-stimulation. In patients with non-progressive puberty and in pre-pubertal controls, the LH peak occurred between 3 and 6 hours after injection [3]. Rosenfield et al. measured gonadotropin levels at 0, 2, 4, 8, 16, and 24 hours post leuprolide acetate injection in a dose–response study comparing acute hormonal responses of the GnRHa leuprolide acetate to GnRH in 15 women and 15 men [6]. They reported peak plasma LH concentration at 1 and 4 hours in men and women, respectively. Other investigators using alternative sampling times demonstrated peak sample at 1 hour following leuprolide administration [2]. A previous study by Houk et al. demonstrated that a single LH measurement obtained 30 minutes post GnRHa stimulation provided adequate information to ascertain pubertal status in girls. However, they examined GnRHa stimulation testing with single 30-minute post-stimulus gonadotropin measurements, and none of the patients in their study had slowly progressive puberty [7]. Thus, despite the evidence of the efficacy of subcutaneous leuprolide acetate in the diagnosis of central precocious puberty, the optimal sampling times for LH post leuprolide challenge has not yet been determined based on the published results. The objective of our study is to re-evaluate optimal sampling time for LH post leuprolide acetate challenge in a cohort of patients presenting with signs of early pubertal development.

ocious puberty, the optimal sampling times for LH post leuprolide challenge has not yet been determined based on the published results. The objective of our study is to re-evaluate optimal sampling time for LH post leuprolide acetate challenge in a cohort of patients presenting with signs of early pubertal development. Materials and Methods We conducted a retrospective analysis of the results of leuprolide acetate stimulation tests in children referred for possible precocious puberty to Texas Children’s Hospital from January 2003-December 2006. During this time, we utilized a multiple sampling protocol as described below. Children of both genders and all ethnicities (aged 1–9 yrs) were identified who had undergone this multi-sample leuprolide acetate (20 mcg/kg SQ) stimulation tests for the suspected diagnosis of central precocious puberty [2]. Serum luteinizing hormone (LH) and follicle stimulating hormone (FSH) concentrations were measured at 0, 1, 3 and 6 h post-injection. Serum estradiol and testosterone concentrations were measured at 0 and 6 h. Of the 155 subjects identified, one subject did not have a blood sample taken at 6 hours; the data from this subject was included since the analysis of the overall data including or excluding these data did not affect the results or conclusions and the primary comparisons were between 1 and 3-hour samples.

t 0 and 6 h. Of the 155 subjects identified, one subject did not have a blood sample taken at 6 hours; the data from this subject was included since the analysis of the overall data including or excluding these data did not affect the results or conclusions and the primary comparisons were between 1 and 3-hour samples. The diagnosis of central precocious puberty was established based on the clinical history of onset of pubertal changes (girls <8y, boys <9y), physical examination suggesting puberty based on Tanner-stage breast and pubic hair development in girls, testicular volume in boys, growth velocity over at least 6 months and bone age. Based on these criteria and a follow up of at least 6 months with the exception of a few cases in which thelarche resolved after 4 months, the subjects were divided into three groups: A. Non-progressive puberty with thelarche, B. Non-progressive puberty with adrenarche, and C. Central precocious puberty.

The diagnosis of central precocious puberty was established based on the clinical history of onset of pubertal changes (girls <8y, boys <9y), physical examination suggesting puberty based on Tanner-stage breast and pubic hair development in girls, testicular volume in boys, growth velocity over at least 6 months and bone age. Based on these criteria and a follow up of at least 6 months with the exception of a few cases in which thelarche resolved after 4 months, the subjects were divided into three groups: A. Non-progressive puberty with thelarche, B. Non-progressive puberty with adrenarche, and C. Central precocious puberty. Gonadotropin and sex steroid assays The serum LH and FSH concentrations were analyzed at our clinical laboratory using the ADVIA Centaur immunoanalyzer (a two-site sandwich immunoassay) and direct chemiluminometric technique (ICMA, third-generation assay). The sensitivity of the FSH and LH assays were 0.3 mIU/mL and 0.07 mIU/mL, respectively. The published per cent coefficient of variation for replicate analysis were <4% for both assays in the 0.3-200 mIU/ml for FSH and 0.07-200 mIU/mL for LH, and the precision accuracy of assay was validated according to CAP laboratory accreditation standards. Serum estradiol and testosterone were measured by LCMS at Esoterix, Calabasas Laboratory.

fficient of variation for replicate analysis were <4% for both assays in the 0.3-200 mIU/ml for FSH and 0.07-200 mIU/mL for LH, and the precision accuracy of assay was validated according to CAP laboratory accreditation standards. Serum estradiol and testosterone were measured by LCMS at Esoterix, Calabasas Laboratory. Statistical analysis All data are provided as mean ± SEM. An LH value < 0.1 mIU/mL was assumed to be 0.1mIU/ml for purposes of calculations. The generalized estimating equations method for the binomial distribution and logit link function (SPSS 18.0) was used to estimate and compare the percent with LH >5 at each time point while accounting for repeated measures collected longitudinally on each subject. Time was treated as fixed, AR1 was assumed for the correlation structure, and Fisher’s LSD was used in the pairwise comparison of time points 1 vs 3, 1 vs 6, and 3 vs 6. Correlations of basal and GnRHa-stimulated peak serum LH values were analyzed by Spearman’s rank correlation.

ollected longitudinally on each subject. Time was treated as fixed, AR1 was assumed for the correlation structure, and Fisher’s LSD was used in the pairwise comparison of time points 1 vs 3, 1 vs 6, and 3 vs 6. Correlations of basal and GnRHa-stimulated peak serum LH values were analyzed by Spearman’s rank correlation. Subjects Of the 155 subjects identified with premature sexual development who had undergone a leuprolide stimulation test, 48 were excluded. Thirty eight (38) had inadequate follow-up, 3 were diagnosed with organic disorders of hypothalamic-pituitary axis and 7 had peripheral puberty e.g. congenital adrenal hyperplasia, testotoxicosis and McCune Albright syndrome. Thus the total number of subjects included for analyses were 107. There were 21 girls in Group A with non-progressive puberty with thelarche. Group B, non-progressive puberty with adrenarche, had 15 subjects of which 12 were girls and 3 were boys. Finally, Group C with central progressive puberty included 71 children (58 girls & 13 boys) (Table 1). Table 1 Patient Characteristics Premature Thelarche Premature Adrenarche Central Precocious Puberty Patients# 21 (girls) 15 (12 girls, 3 boys) 71 (58 girls, 13 boys) Chronological Age 5.04 ± 0.46 ** 6.0 ± 0.55 * 7.78 ± 0.18 Bone Age 6.73 ± 0.57 ** 8.64 ± 0.57 10.46 ± 0.25 Values represent “Mean ± SEM”. *P <0.05 (compared to puberty group), **P < 0.005 (compared to puberty group).

Premature Thelarche Premature Adrenarche Central Precocious Puberty Patients# 21 (girls) 15 (12 girls, 3 boys) 71 (58 girls, 13 boys) Chronological Age 5.04 ± 0.46 ** 6.0 ± 0.55 * 7.78 ± 0.18 Bone Age 6.73 ± 0.57 ** 8.64 ± 0.57 10.46 ± 0.25 Values represent “Mean ± SEM”. *P <0.05 (compared to puberty group), **P < 0.005 (compared to puberty group). Baseline hormone concentrations Group A (Premature thelarche) Basal serum LH concentrations were 0.1 ± 0.0 mIU/ml. None had a basal LH concentration > 0.1 mIU/mL. Mean basal serum FSH was 2.18 ± 0.3 mIU/ml. All 21 girls had a basal LH/FSH ratio < 1. The basal serum concentrations of estradiol were 0.37 ± 0.15 ng/dL (pre-pubertal <1.5 ng/dL). Serum testosterone was not measured in this group. Group B (Premature adrenarche) Basal serum LH concentrations were 0.1 ± 0.0 mIU/ml whereas basal FSH concentrations were 1.48 ± 0.31 mIU/ml (Figure 1). All 15 subjects with premature adrenarche had a basal LH/FSH ratio < 1. The basal estradiol concentration in the girls (n = 12) was 0.51 ± 0.17 ng/dL. Although the mean basal estradiol concentration was slightly higher than the patients with premature thelarche, it was statistically insignificant (P = 0.38). Additionally, their estradiol levels were in the pre-pubertal range and clinical follow up confirmed the diagnosis of premature adrenarche. Mean testosterone concentrations in the boys (n = 3) was 2.05 ± 0.34 ng/dL (pre-pubertal <10 ng/dL) (Figure 2).

premature thelarche, it was statistically insignificant (P = 0.38). Additionally, their estradiol levels were in the pre-pubertal range and clinical follow up confirmed the diagnosis of premature adrenarche. Mean testosterone concentrations in the boys (n = 3) was 2.05 ± 0.34 ng/dL (pre-pubertal <10 ng/dL) (Figure 2). Figure 1 Baseline and peak stimulated serum LH (upper panel) and FSH (lower panel) concentrations (mean ± SEM) at 1, 3, and 6 hours after leuprolide injection. LH concentrations in central precocious puberty were significantly higher than in either the children with premature thelarche or premature adrenarche (P< 0.005). Figure 2 Baseline and peak stimulated serum Estradiol (upper panel) and Testosterone (lower panel) concentrations (mean ± SEM) at baseline and 6 hours after leuprolide injection. Estradiol and Testosterone concentrations in central precocious puberty were significantly higher than in either the children with premature thelarche or premature adrenarche (P< 0.005).

stradiol (upper panel) and Testosterone (lower panel) concentrations (mean ± SEM) at baseline and 6 hours after leuprolide injection. Estradiol and Testosterone concentrations in central precocious puberty were significantly higher than in either the children with premature thelarche or premature adrenarche (P< 0.005). Group C (true central precocious puberty) The basal LH and FSH concentrations were 1.96 ± 0.26 and 3.68 ± 0.31 mIU/mL, respectively (Figure 1). Both were higher than in either the children with premature thelarche or premature adrenarche (p <0.005). Of the 58 female subjects with true puberty, fifteen (26%) had a basal LH value < 0.1 mIU/mL. All 13 of male subjects had a basal LH value > 0.1 mIU/mL (ranged from 0.5-5.3 mIU/mL). Basal estradiol concentrations of the 58 female subjects were 2.59 ± 0.46 ng/dl, and basal testosterone concentrations of the 13 male subjects were 187.69 ± 41.84 ng/dl (Figure 2). Leuprolide - Stimulated Hormonal Concentrations Group A (premature thelarche) All 21 girls with a clinical diagnosis of premature thelarche had a peak stimulated LH concentration < 5 mIU/mL (Figure 1). Of these, 3 had a peak LH concentration at 1 h (14%), 18 at 3 h (86%), 0 at 6 h (Figure 3). Two girls had values at 1 and 3 h which were identical. Plasma FSH concentrations increased and peaked at 3 hr and decreased slightly by 6 h. All of these girls had stimulated LH/FSH ratios of < 1 at 1, 3 and 6 hours. Mean stimulated estradiol concentrations of these 21 girls at 6 hours was 2.00 ± 0.5 ng/dL (Figure 2).

Two girls had values at 1 and 3 h which were identical. Plasma FSH concentrations increased and peaked at 3 hr and decreased slightly by 6 h. All of these girls had stimulated LH/FSH ratios of < 1 at 1, 3 and 6 hours. Mean stimulated estradiol concentrations of these 21 girls at 6 hours was 2.00 ± 0.5 ng/dL (Figure 2). Figure 3 Distribution of number of patients with LH concentration peaked at 1, 3 and 6 hours after leuprolide injection.

Two girls had values at 1 and 3 h which were identical. Plasma FSH concentrations increased and peaked at 3 hr and decreased slightly by 6 h. All of these girls had stimulated LH/FSH ratios of < 1 at 1, 3 and 6 hours. Mean stimulated estradiol concentrations of these 21 girls at 6 hours was 2.00 ± 0.5 ng/dL (Figure 2). Figure 3 Distribution of number of patients with LH concentration peaked at 1, 3 and 6 hours after leuprolide injection. Group B (premature adrenarche) Of the 15 children with premature adrenarche (Figure 1) only 1 girl had a stimulated LH concentration > 5 mIU/mL; the child had a baseline LH = 0.1 mIU/mL and peak stimulated plasma LH concentrations of 5.3 and 5.2 mIU/mL at 3 and 6 h, respectively. Clinically, her diagnosis remained premature adrenarche after 23 months follow up. Among these 15 children, two had a peak stimulated LH at 1 h (13%), 13 at 3 h (87%), none at 6 hr (Figure 3). One child had values of LH that were equal at 1 and 3 h. All had a stimulated LH/FSH ratio < 1 at 1, 3 and 6 hours. It was surprisingly observed that children with premature adrenarche had an FSH-predominant response in our cohort. Their plasma FSH concentrations increased following leuprolide injection and peaked at 3 h but the peak concentrations were significantly less than those of the girls with premature thelarche. We noted that two of the patients in this cohort who were significantly younger (15 months and 3 years) than most patients had a significant FSH–predominant response, but our clinical observation confirmed the diagnosis of premature adrenarche. The mean stimulated serum estradiol of the 12 girls was 1.46 ± 0.42 ng/dl at 6 h and the mean stimulated serum testosterone of the 3 boys was 6.6 ± 1.23 ng/dL at 6 hours (Figure 2).

an most patients had a significant FSH–predominant response, but our clinical observation confirmed the diagnosis of premature adrenarche. The mean stimulated serum estradiol of the 12 girls was 1.46 ± 0.42 ng/dl at 6 h and the mean stimulated serum testosterone of the 3 boys was 6.6 ± 1.23 ng/dL at 6 hours (Figure 2). Group C (central precocious puberty) Group C Overall Results Of the 71 children with central precocious puberty (Figure 1) 15 subjects had a maximum plasma concentrations of LH at 1 h (21%), 38 at 3 h (54%), 18 at 6 h (25%) and 2 subjects had a maximum concentration at both 3 & 6 hours (Figure 3). The plasma LH concentrations were higher than in either the children with premature thelarche or premature adrenarche (p < 0.005). The mean stimulated FSH concentrations of girls with premature thelarche (Group A) were higher than children with central precocious puberty or premature adrenarche at both 1 and 3 h, but not at 6 h. Maximal FSH responses were detected 3 hours post stimulation in children with premature adrenarche and thelarche but at 6 h in children with central precocious puberty (Figure 1). All 13 boys with central precocious puberty had a stimulated LH/FSH ratio >1 at 1 & 3 hours. In contrast, only 32 (70%) girls with central precocious puberty, and a stimulated LH > 5 mIU/mL, had a stimulated LH/FSH ratio > 1 at 1 and 3 hours.

The mean stimulated FSH concentrations of girls with premature thelarche (Group A) were higher than children with central precocious puberty or premature adrenarche at both 1 and 3 h, but not at 6 h. Maximal FSH responses were detected 3 hours post stimulation in children with premature adrenarche and thelarche but at 6 h in children with central precocious puberty (Figure 1). All 13 boys with central precocious puberty had a stimulated LH/FSH ratio >1 at 1 & 3 hours. In contrast, only 32 (70%) girls with central precocious puberty, and a stimulated LH > 5 mIU/mL, had a stimulated LH/FSH ratio > 1 at 1 and 3 hours. Mean Stimulated testosterone concentrations of 13 male subjects with central precocious puberty (302 ± 61.3) were higher than the boys with premature adrenarche (6.6 ± 1.2 ng/dl) at 6 h, (P < 0.005). The mean stimulated plasma estradiol concentrations of the 58 girls with central precocious puberty (6.68 ± 0.64 ng/dL) were higher than those of the girls in Groups A and B at 6 h (p < 0.005) (Figure 2). Concordant and Discordant Stimulated LH with Clinical Precocious Puberty Pubertal subjects were divided in two groups based on their responses to leuprolide challenge: a. With a concordant response for the leuprolide challenge (LH > 5 mIU/mL)

Mean Stimulated testosterone concentrations of 13 male subjects with central precocious puberty (302 ± 61.3) were higher than the boys with premature adrenarche (6.6 ± 1.2 ng/dl) at 6 h, (P < 0.005). The mean stimulated plasma estradiol concentrations of the 58 girls with central precocious puberty (6.68 ± 0.64 ng/dL) were higher than those of the girls in Groups A and B at 6 h (p < 0.005) (Figure 2). Concordant and Discordant Stimulated LH with Clinical Precocious Puberty Pubertal subjects were divided in two groups based on their responses to leuprolide challenge: a. With a concordant response for the leuprolide challenge (LH > 5 mIU/mL) Out of 71 children with clinical evidence of progressive puberty, 59 subjects (83%) had a pubertal response to leuprolide challenge (13 boys and 46 girls). Among these 59 subjects, 52 (88%) had a peak LH > 5 mIU/mL at 1 h (95% CI: 80-96%), but all 59 children (100%) had a peak LH >5 mIU/mL at 3 h (95% CI: 94-100%), P = 0.005 (Figure 3). All 13 boys had clinical evidence of true puberty, and a pubertal response (LH > 5mIU/mL) at 1, 3 and 6 h. Out of these 13 boys, 11 had a maximum concentration of LH at 3 h. Eleven (11) boys (85%) and 32 girls (70%) were treated with GnRHa. b. With a discordant response for the leuprolide challenge (LH < 5 mIU/mL)

Out of 71 children with clinical evidence of progressive puberty, 59 subjects (83%) had a pubertal response to leuprolide challenge (13 boys and 46 girls). Among these 59 subjects, 52 (88%) had a peak LH > 5 mIU/mL at 1 h (95% CI: 80-96%), but all 59 children (100%) had a peak LH >5 mIU/mL at 3 h (95% CI: 94-100%), P = 0.005 (Figure 3). All 13 boys had clinical evidence of true puberty, and a pubertal response (LH > 5mIU/mL) at 1, 3 and 6 h. Out of these 13 boys, 11 had a maximum concentration of LH at 3 h. Eleven (11) boys (85%) and 32 girls (70%) were treated with GnRHa. b. With a discordant response for the leuprolide challenge (LH < 5 mIU/mL) Among 71 children with clinical evidence of true puberty, 12 girls (17%) had a pre-pubertal response (LH < 5 mIU/mL) to Leuprolide challenge despite pubertal progression. Among these 12 subjects, the peak stimulated LH concentrations ranged from 0.9 to 4.6 mIU/mL regardless of the sampling time but had a predominant FSH response, such that we know that the leuprolide was in fact administered. Their stimulated estradiol concentrations at 6 h were not different from groups A and B (data not shown).

e 12 subjects, the peak stimulated LH concentrations ranged from 0.9 to 4.6 mIU/mL regardless of the sampling time but had a predominant FSH response, such that we know that the leuprolide was in fact administered. Their stimulated estradiol concentrations at 6 h were not different from groups A and B (data not shown). Diagnostic values of the measured hormones in the evaluation of central precocious puberty Although not the primary purpose of this study, we evaluated and compared the relative diagnostic values of different parameters including basal LH, estradiol and testosterone, stimulated LH, LH/FSH ratio in the prediction of pubertal status. Our data demonstrates that in boys basal LH (>0.1 mIU/mL), testosterone concentrations (≥10 ng/dL), basal and stimulated LH/FSH ratios (at 1 and 3 h) have excellent sensitivity and specificity all to be 100% (Table 2). However, in girls basal LH > 0.1 mIU/ml, basal and stimulated LH/FSH ratios and basal estradiol (≥1.5 ng/dL) have low sensitivity though excellent specificity (Table 2). When serial potential predictors were combined, the sensitivities were reduced even though specificities were improved (Table 2). Compared to stimulated LH concentration at 1 h, the LH concentration > 5mIU/ml at 3 h had better sensitivity (83% vs 73%) without compromising specificity (97% vs 100%). This cut off also has optimal sensitivity (83%) and specificity (97%) when compared to a lower cut off of 3mIU/ml or a higher cut off of 7mIU/ml (Table 2).

red to stimulated LH concentration at 1 h, the LH concentration > 5mIU/ml at 3 h had better sensitivity (83% vs 73%) without compromising specificity (97% vs 100%). This cut off also has optimal sensitivity (83%) and specificity (97%) when compared to a lower cut off of 3mIU/ml or a higher cut off of 7mIU/ml (Table 2). Table 2 Diagnostic values of basal LH, Estradiol, Testosterone, stimulated LH and the ratio of LH/FSH for Puberty Predictors of puberty Sensitivity Specificity PPV NPV Basal LH > 0.1 mIU/ml (M) 100% 100% 100% 100% Basal LH > 0.1 mIU/ml (F) 67% 100% 100% 63% Basal Estradiol ≥ 1.5 ng/dL 50% 94% 94% 52% Basal Testosterone ≥ 10 ng/dL 100% 100% 100% 100% LH at 1 h ≥ 5 mIU/ml 73% 100% 100% 80% LH at 3 h ≥ 5 mIU/ml 83% 97% 98% 74% LH at 3 h ≥ 3 mIU/ml 92% 75% 88% 82% LH at 3 h ≥ 7 mIU/ml 80% 100% 100% 71% Basal LH/FSH >1 (M) 100% 100% 100% 100% Basal LH/FSH >1 (F) 10% 100% 100% 39% LH/FSH at 1 h >1(M) 100% 100% 100% 100% LH/FSH at 1 h >1(F) 50% 100% 100% 53% LH/FSH at 3 h >1(M) 100% 100% 100% 100% LH/FSH at 3 h >1(F) 45% 100% 100% 51% LH at 1 h ≥ 5 mIU/ml and basal LH/FSH >1 61% 100% 100% 56% LH at 3 h ≥ 5 mIU/ml and basal LH/FSH >1 16% 100% 100% 38% LH at 3 h ≥ 5 mIU/ml and LH/FSH at 1 h >1 59% 100% 100% 56% M – Male, F- Female; PPV – Positive predictive value; NPV – Negative predictive value.

0% 100% 100% 100% LH/FSH at 3 h >1(F) 45% 100% 100% 51% LH at 1 h ≥ 5 mIU/ml and basal LH/FSH >1 61% 100% 100% 56% LH at 3 h ≥ 5 mIU/ml and basal LH/FSH >1 16% 100% 100% 38% LH at 3 h ≥ 5 mIU/ml and LH/FSH at 1 h >1 59% 100% 100% 56% M – Male, F- Female; PPV – Positive predictive value; NPV – Negative predictive value. Discussion In the early phase of central sexual precocious puberty, laboratory confirmation is important to provide an accurate diagnosis and appropriate therapy. When random plasma LH concentrations are low in the presence of physical findings suggestive of precocious puberty, GnRHa stimulation testing is recommended to determine activation of hypothalamic-pituitary-gonadal axis. Despite wide utilization of the GnRHa stimulation test, the timing of blood sampling remains controversial if a single sample protocol is used. In our present study, the peak LH response occurred 3 hours post leuprolide stimulation test in those with true central precocious puberty. However, only 59 of 71 of the children with true central precocious puberty had an LH concentration > 5 mIU/mL at 3 hours. When compared to the 1 h value, the 3 h value was higher (p < 0.005) and had better sensitivity in diagnosing central precocious puberty. We recommend that a 3 h sample should be considered for those cases in which clinical presentation and base line laboratory values are not conclusive, despite the practical difficulties posed by a prolonged test protocol, particularly for those families traveling at a distance.

in diagnosing central precocious puberty. We recommend that a 3 h sample should be considered for those cases in which clinical presentation and base line laboratory values are not conclusive, despite the practical difficulties posed by a prolonged test protocol, particularly for those families traveling at a distance. Girls with central precocious puberty in the early phase of activation of the hypothalamic-pituitary-gonadal axis are capable of clinically relevant estradiol production, which may occur in the face of low LH secretion and low LH/FSH ratios [2]. This observation is puzzling and one speculation is that endocrine or paracrine factors other than LH and FSH may play an important role in amplifying the effects of gonadotropins on ovarian E2 secretion in the early phase of sexual precocity [2]. Among our subjects with clinical evidence of precocious puberty, 12 girls with Tanner stage II-III breast development had a pre-pubertal response to leuprolide challenge (LH < 5 mIU/mL), a predominant FSH response and therefore a low LH/FSH ratio. Their laboratory findings were indistinguishable from those of subjects with proven premature thelarche and adrenarche. Interestingly, 10 out of these 12 subjects had both breast and pubic hair development at initial presentation, and only 2 subjects presented with just thelarche. Only one of these 12 girls with discordant response required GnRHa treatment. She was 6.5 y of age at presentation with Tanner III breast and pubic hair, a bone age of 10 y and a normal brain MRI. Her baseline LH was 0.1 mIU/mL, peak stimulated LH 3.9 mIU/mL at 3 h and estradiol concentration of 5.2 ng/dl at 6 h despite continued pubertal development. Therefore, clinical judgment and follow up continues to be of great importance in the evaluation of precocious puberty.

a bone age of 10 y and a normal brain MRI. Her baseline LH was 0.1 mIU/mL, peak stimulated LH 3.9 mIU/mL at 3 h and estradiol concentration of 5.2 ng/dl at 6 h despite continued pubertal development. Therefore, clinical judgment and follow up continues to be of great importance in the evaluation of precocious puberty. In our study, the basal plasma LH concentration differentiated the pubertal and pre-pubertal boys, without overlap, and is entirely consistent with the findings of Resende et al. in normal male subjects [8]. However, this was not true for the girls. Twenty six percent of the girls ultimately diagnosed with central precocious puberty (Tanner breast stage II-III at presentation) and pubertal responses to leuprolide had basal serum LH concentrations in the pre-pubertal range (LH <0.1 mIU/mL). Basal LH concentrations in excess of 0.1 mIU/mL were strongly correlated with a pubertal stimulated LH concentrations (>5 mIU/mL at 3 h) in pubertal subjects (r = 0.842, P, 0.0001) (data not shown). This finding is in agreement with others [9]. Thus, a random LH concentration measured by third-generation assays such as immunochemiluminometric assay is a useful tool in screening for central precocious puberty. However, our experience suggests that a GnRHa stimulation test should be considered when a basal serum LH is inconclusive or does not fit with the clinical presentation. This conclusion is further strengthened in that none of our children with either premature thelarche or premature adrenarche had a random serum LH > 0.1 mIU/mL.

ur experience suggests that a GnRHa stimulation test should be considered when a basal serum LH is inconclusive or does not fit with the clinical presentation. This conclusion is further strengthened in that none of our children with either premature thelarche or premature adrenarche had a random serum LH > 0.1 mIU/mL. Although mean spontaneous serum FSH concentrations were greater in children with central precocious puberty (p < 0.005) and provided fair sensitivity and specificity, subjects in groups A and B had predominant FSH responses to leuprolide challenge and mean stimulated serum FSH concentrations in girls with premature thelarche were higher than pubertal children at both 1 and 3 h (P < 0.005). These observations further strengthen the findings of others [9] that the stimulated FSH is of limited utility in partitioning the children with central precocious puberty from those without central stimulation. We also demonstrate that both basal LH/FSH > 1 and the stimulated LH/FSH ratio >1 at 1 and 3 hours are excellent predictors in diagnosing central precocious puberty in boys. This was not true for girls however (Table 2).

n partitioning the children with central precocious puberty from those without central stimulation. We also demonstrate that both basal LH/FSH > 1 and the stimulated LH/FSH ratio >1 at 1 and 3 hours are excellent predictors in diagnosing central precocious puberty in boys. This was not true for girls however (Table 2). Conclusion This is the largest group of children reported who have undergone a 6 h leuprolide acetate stimulation test for the evaluation of central precocious puberty. We conclude that in our study a single basal LH measurement using third-generation assays was adequate to diagnose central precocious puberty in boys. In addition, basal LH is adequate to diagnose central precocious puberty in most but not all girls, indicating the need for GnRHa test when a basal LH is inconclusive. Basal testosterone (measured by LCMS) in conjunction with clinical correlation is diagnostic of central precocious puberty in boys. In contrast, a basal estradiol (measured by LCMS) is helpful in most girls, but not all. However, when a GnRHa stimulation test is undertaken, our data demonstrates that a single sample at 3 h is superior in sensitivity and specificity to that of the 1 h sampling time in diagnosing central precocious puberty in girls, and provides the optimal sample to ascertain a diagnosis of central precocious puberty. Obviously, clinical judgment and follow up continues to be essential in that a quarter of our girls with central precocious puberty had discordant clinical findings with those of either the basal or stimulated LH values.

y in girls, and provides the optimal sample to ascertain a diagnosis of central precocious puberty. Obviously, clinical judgment and follow up continues to be essential in that a quarter of our girls with central precocious puberty had discordant clinical findings with those of either the basal or stimulated LH values. Abbreviations GnRH, Gonadotropin-releasing hormone; LH, Luteinizing hormone; FSH, Follicle stimulating hormone; GnRHa, Gonadotropin-releasing hormone analogue; ICMA, Immunochemiluminometric assay; SQ, Subcutaneous. Acknowledgments We would like to thank E. O’Brian Smith, PhD for his statistical help and Arman Sadeghpour PhD, Andrea Balazs, MD, and Rachel Edelen, MD for their support in the development of this protocol. Financial disclosure The authors have no financial relationship relevant to disclose regarding this article.

Adolescents and young adults with type 1 diabetes (T1D) are at high risk for developing early cardiovascular disease [1]. Current recommendations to consider pharmacologic treatment of elevated low-density lipoprotein cholesterol (LDL-c) are based on limited evidence and extrapolation from data in middle age and older adults. We proposed a trial of lipid-lowering medications (simvastatin, a statin, compared to Vytorin, a combination of simvastatin and ezetimibe, a medication that blocks cholesterol absorption) in our patients ages 12–21 years with LDL ≥ 130 mg/dl, consistent with current American Diabetes Association (ADA) guidelines. In this study, we hypothesized that simvastatin and Vytorin would be safe in adolescents with T1D and that in a two-arm design, Vytorin would lower LDL-c more than monotherapy with simvastatin at 6 months compared to baseline. In this report, we describe observations from a trial of lipid-lowering medications in T1D, age 12–21 years.

we hypothesized that simvastatin and Vytorin would be safe in adolescents with T1D and that in a two-arm design, Vytorin would lower LDL-c more than monotherapy with simvastatin at 6 months compared to baseline. In this report, we describe observations from a trial of lipid-lowering medications in T1D, age 12–21 years. In 2004 we ascertained that the Barbara Davis Center for Childhood Diabetes (BDC) followed 1,528 patients with T1D age 12–21 years. The SEARCH for Diabetes in Youth study reported LDL-c >130 mg/dl in 15% of T1D subjects [2]. Based on these data, we estimated that 229 T1D patients age 12–21 years seen at the BDC would have LDL-c >130 mg/dl. A randomized, double-blind, placebo-controlled study of lipid-lowering medications required 82 enrolled subjects which assumed a 36% participation rate. Inclusion criteria were age 12–21 years with T1D diagnosed by positive islet autoantibodies or provider diagnosis of T1D, and LDL-c > 130 mg/dl. Patients with familial hypercholesterolemia, triglycerides > 400 mg/dl, T1D of less than three-months duration, HbA1c > 9.5%, abnormal thyroid function, abnormal creatine kinase values, abnormal liver function tests (ALT/AST), pregnancy or potential to become pregnant during the study, and patients on oral contraceptives were excluded.

ilial hypercholesterolemia, triglycerides > 400 mg/dl, T1D of less than three-months duration, HbA1c > 9.5%, abnormal thyroid function, abnormal creatine kinase values, abnormal liver function tests (ALT/AST), pregnancy or potential to become pregnant during the study, and patients on oral contraceptives were excluded. Findings We identified 105 potential subjects with clinically measured LDL-c >130 mg/dl or non-HDL-c >160 mg/dl. Of these, 42 patients proved to be ineligible (A1c > 9.5%, LDL-c < 130 mg/dl on a repeat test), 26 declined invitation to a screening visit, 16 expressed one time interest but were not able to be scheduled, and 3 were interested but were unable to improve their glycemic control to be eligible for the study. Therefore, 18 agreed to be in the study and 17 subjects attended a study screening visit of which 9 were enrolled in the study (15.8 ± 2.8 years, 67% male, A1c = 8.3 ± 1.1%, TC = 224 ± 42, HDL-c = 51 ± 11, TG = 112 ± 66, LDL-c = 151 ± 29 mg/dl, BMI = 25.4 ± 5.0 kg/m2) (Figure 1). Reasons for poor recruitment included elevated A1c (>9.5%), improved LDL-c from clinic lipid panels to screening visit (n = 42, 40%), lack of interest in taking a lipid-lowering medication and/or long distance to travel to study site (n = 21, 20%). A positive result of our study was improved compliance with lipid screening in our clinic population (a 55% reduction, from 416 to 188) in subjects eligible for, but lacking screening lipids over 2.5 years. Figure 1 Study Recruitment and Participation.

Findings We identified 105 potential subjects with clinically measured LDL-c >130 mg/dl or non-HDL-c >160 mg/dl. Of these, 42 patients proved to be ineligible (A1c > 9.5%, LDL-c < 130 mg/dl on a repeat test), 26 declined invitation to a screening visit, 16 expressed one time interest but were not able to be scheduled, and 3 were interested but were unable to improve their glycemic control to be eligible for the study. Therefore, 18 agreed to be in the study and 17 subjects attended a study screening visit of which 9 were enrolled in the study (15.8 ± 2.8 years, 67% male, A1c = 8.3 ± 1.1%, TC = 224 ± 42, HDL-c = 51 ± 11, TG = 112 ± 66, LDL-c = 151 ± 29 mg/dl, BMI = 25.4 ± 5.0 kg/m2) (Figure 1). Reasons for poor recruitment included elevated A1c (>9.5%), improved LDL-c from clinic lipid panels to screening visit (n = 42, 40%), lack of interest in taking a lipid-lowering medication and/or long distance to travel to study site (n = 21, 20%). A positive result of our study was improved compliance with lipid screening in our clinic population (a 55% reduction, from 416 to 188) in subjects eligible for, but lacking screening lipids over 2.5 years. Figure 1 Study Recruitment and Participation. Despite current recommendations for intensified glycemic control, diet, and exercise as initial therapies for dyslipidemia, some adolescents with T1D will not reach the LDL-c goal of <100 mg/dl. For these patients, data are needed on the safety, efficacy, and ultimately the risk/benefit of dyslipidemia medications initiated in adolescence. We are aware of only one published short-term clinical trial of dyslipidemia medication in adolescents with T1D [3], although a Juvenile Diabetes Research Foundation sponsored trial is on-going in the United Kingdom [4]. Increased awareness of lipid health and treatment options in this patient population are needed as some patients and families required months of deliberation before deciding to have the patient take a dyslipidemia medication in addition to intensification of glycemic control, diet, and exercise. Based on our experience, future studies will require multiple centers to enroll a sufficient number of participants for adequate data to direct treatment guidelines for pharmacologic treatment of dyslipidemia in adolescents with T1D.

mia medication in addition to intensification of glycemic control, diet, and exercise. Based on our experience, future studies will require multiple centers to enroll a sufficient number of participants for adequate data to direct treatment guidelines for pharmacologic treatment of dyslipidemia in adolescents with T1D. Authors’ contributions Author Contributions: FB, RPW, DMM researched data. FB wrote manuscript. FB, RPW, DMM, MR, SE reviewed/edited manuscript. FB, RPW, DMM, MR, SE contributed to discussion, reviewed/edited manuscript. Acknowledgments Support for this investigator-initiated study was provided by Merck/Schering-Plough Pharmaceuticals. NCT #00477204. The study was performed at the Barbara Davis Center for Childhood Diabetes in Aurora, CO and the Children’s Clinical Translational Research Center at the University of Colorado Denver supported by the NIH M01 RR000051.

Background It is well known that the prevalence of obesity in children has reached epidemic proportions: during the past decade, the prevalence of children with a Body Mass Index (BMI) > 95th percentile has tripled in all pediatric age-range groups [1]. Pediatric obesity is associated with significant medical complications during childhood [2], and it is a significant risk factor for morbidity and mortality in adulthood [2,3]. Most of the comorbidities of childhood obesity share insulin resistance as a common underlying mechanism. Such comorbidities (dyslipidemia, nonalcoholic fatty liver disease, type 2 diabetes mellitus (DM), hypertension, obstructive sleep apnea, polycystic ovary syndrome) tend to cluster in what is known as the metabolic syndrome [1,4]. Pediatric metabolic syndrome is a predictor of the metabolic syndrome and type 2 DM in adulthood [5]; in addition, obesity-associated atherosclerosis begins in childhood [6] and its rate of progression is greatly increased by lipid abnormalities. As a result, detecting and correcting obesity and its associated metabolic abnormalities in childhood may help prevent cardiovascular morbidity and mortality in adulthood.

5]; in addition, obesity-associated atherosclerosis begins in childhood [6] and its rate of progression is greatly increased by lipid abnormalities. As a result, detecting and correcting obesity and its associated metabolic abnormalities in childhood may help prevent cardiovascular morbidity and mortality in adulthood. A number of studies have demonstrated the positive effects of intensive weight-loss programs in children [7-11]. Yet, intensive programs are based on frequent interactions between children, their families, and a multi-disciplinary team of providers; thus, they are necessarily expensive and short-term. In addition, most of the beneficial effects of an intensive short-term intervention often do not persist once the program is completed [12-14]. Since little is known of the long-term impact of a non-intensive, conventional weight management program in obese children [15,16], we evaluated the effects of our Weight Management Program over a 4-year period at the Section of Endocrinology and Diabetes at St. Christopher’s Hospital for Children in a subset of obese patients who maintained ongoing periodic follow-up visits. The goals of our study were: 1) to analyze the changes of BMI z-score, glycemic measures and lipid profiles at the end of the 4-year follow-up period, and 2) to correlate these changes with the frequency of the follow-up visits.

or Children in a subset of obese patients who maintained ongoing periodic follow-up visits. The goals of our study were: 1) to analyze the changes of BMI z-score, glycemic measures and lipid profiles at the end of the 4-year follow-up period, and 2) to correlate these changes with the frequency of the follow-up visits. Methods We conducted a retrospective analysis of the medical records of obese (BMI > 95th percentile for age and sex) children and adolescents evaluated between 2001 and 2008 at the Weight Management Center in the Section of Endocrinology and Diabetes at St. Christopher’s Hospital for Children. All children were identified by tracking patients for whom the ICD-9 Code 783.1 was used (“abnormal weight gain”). Most of these patients were referred to us by their pediatricians for an evaluation of overweight/obesity and/or high serum insulin/abnormal lipid levels. Inclusion criteria were the following: [1] children with a BMI > 95th percentile, [2] 1–18 years of age [3] ≥ 2 years of follow-up, and [4] fasting lipid panel, glucose and insulin obtained at the beginning and at the end of the follow-up period. Exclusion criteria included: 1) diagnosis of DM (type 1 or type 2) and 2) use of medications known to affect insulin sensitivity or glucose/lipid metabolism (metformin, insulin, growth hormone). The Institutional Regulatory Board of Drexel University College of Medicine approved the retrospective analysis of the medical records.

Inclusion criteria were the following: [1] children with a BMI > 95th percentile, [2] 1–18 years of age [3] ≥ 2 years of follow-up, and [4] fasting lipid panel, glucose and insulin obtained at the beginning and at the end of the follow-up period. Exclusion criteria included: 1) diagnosis of DM (type 1 or type 2) and 2) use of medications known to affect insulin sensitivity or glucose/lipid metabolism (metformin, insulin, growth hormone). The Institutional Regulatory Board of Drexel University College of Medicine approved the retrospective analysis of the medical records. At the initial visit, a physician screened the obese subjects for metabolic comorbidities, while a registered dietitian assessed their dietary habits and amount of physical activity. Generalized handouts geared towards healthy eating were provided to each subject; the dietitian reviewed each topic of the handout with the subjects and their parent/guardian. The topics included: making healthy food choices (increasing whole grains, lean meats and lower fat foods into the diet), eating three balanced meals per day using the plate method, portion control (measuring portions and learning about food labels), eliminating beverages containing more than 5 calories except for low fat milk, and making sensible snack choices. Thirty minutes of daily physical activity was recommended to all subjects with a focus on an activity that the child would enjoy.

method, portion control (measuring portions and learning about food labels), eliminating beverages containing more than 5 calories except for low fat milk, and making sensible snack choices. Thirty minutes of daily physical activity was recommended to all subjects with a focus on an activity that the child would enjoy. At the subsequent visits (scheduled every 2 to 4 months), obese children and adolescents met with a pediatric nurse practitioner and the dietitian for approximately 20 minutes with each of the two providers; in order to measure implementation of the previous recommendations, during each of the follow-up visits, the child’s dietary intake and physical activities were reassessed by patient history and diet and exercise recall. Patients who reported adhering to the dietary or physical activity changes were praised and encouraged to continue. If there was poor adherence to previous recommendations, or if there was need for further improvement, efforts were made by the dietitian and nurse practioner to detect barriers to change and to help determine alternative ways of engaging patients to adhere.

cal activity changes were praised and encouraged to continue. If there was poor adherence to previous recommendations, or if there was need for further improvement, efforts were made by the dietitian and nurse practioner to detect barriers to change and to help determine alternative ways of engaging patients to adhere. At each visit, body weight was measured with a balance scale, and height was measured with a wall-mounted stadiometer by a trained medical assistant. BMI was calculated as weight in kilograms divided by the height in meters squared, and expressed as a z-score by using the Centers for Disease Control and Prevention 2000 program [17]. Body weight, height and BMI were compared to the measurements/calculations of the previous visit by the nurse practitioner and dietitian in order to monitor each patient’s weight loss/gain and change in the BMI as standard of care, and as an objective way to measure the likelihood that the patient was adhering to the dietary and physical activity-related recommendations.

urements/calculations of the previous visit by the nurse practitioner and dietitian in order to monitor each patient’s weight loss/gain and change in the BMI as standard of care, and as an objective way to measure the likelihood that the patient was adhering to the dietary and physical activity-related recommendations. A fasting lipid panel (total cholesterol, low-density lipoprotein [LDL] cholesterol, high-density lipoprotein [HDL] cholesterol, and triglyceride [TG] levels) was requested yearly. A 2-hour Oral Glucose Tolerance test (OGTT) was recommended to all children: impaired glucose tolerance (IGT) and DM were defined as a 2-hour glucose level of 140–199 mg/dL and ≥ 200 mg/dL, respectively [18]. Insulin resistance was estimated by using the Homeostasis Model Assessment of Insulin Resistance (HOMA-IR), calculated as fasting plasma glucose (mg/dL) x fasting insulin (μU/mL) ÷ 405 [19]. Dyslipidemia was defined as high TG (≥ 90th percentile for age and sex) and/or low HDL-cholesterol (≤ 10th percentile for age and sex) concentrations [4]. Data were analyzed with SPSS software version 17.0 for Windows (SPSS, Chicago, IL). All data were expressed as the mean plus or minus SD or range. A p-value < 0.05 was considered to be statistically significant. Differences in the mean values between groups were evaluated using a Student’s t-test or analysis of variance.

Data were analyzed with SPSS software version 17.0 for Windows (SPSS, Chicago, IL). All data were expressed as the mean plus or minus SD or range. A p-value < 0.05 was considered to be statistically significant. Differences in the mean values between groups were evaluated using a Student’s t-test or analysis of variance. Results A total of 61 children and adolescents met our inclusion criteria. 39 children were females and 24 were prepubertal; the mean age was 11.1 ± 2.6 years (mean ± SD). With respect to their ethnicity, 25 were African American, 26 were Hispanic, 7 were Caucasian, 2 Asian and 1 identified himself as other. The duration of the follow-up period was 47.3 ± 11.1 months while the number of outpatient visits per year (OV/yr) was 2.9 ± 0.9. At the end of the follow-up, children exhibited a significant decrease in their BMI z-score (p = 0.03) and LDL-cholesterol (p = 0.022) (Table 1). In the subset of children in whom OGTT was performed both at the beginning and at the end of the follow-up period (n = 42), there was a significant decrease in both the 2-hour glucose (p = 0.004) and peak insulin (p = 0.043) (Table 1). Table 1 Anthropometric and metabolic characteristics of the population sample

At the end of the follow-up, children exhibited a significant decrease in their BMI z-score (p = 0.03) and LDL-cholesterol (p = 0.022) (Table 1). In the subset of children in whom OGTT was performed both at the beginning and at the end of the follow-up period (n = 42), there was a significant decrease in both the 2-hour glucose (p = 0.004) and peak insulin (p = 0.043) (Table 1). Table 1 Anthropometric and metabolic characteristics of the population sample Initial Visit Last Visit p-value BMI z-score 2.49 ± 0.4 2.33 ± 0.4 0.03 Fasting glucose (mmol/L)a 4.8 ± 0.5 4.6 ± 0.5 0.08 Fasting insulin (pmol/L)b 145.1 ± 90.3 135.4 ± 100.7 0.56 HOMA-IR 4.5 ± 3 4.1 ± 2.9 0.38 HDL-cholesterol (mmol/L)c 1.2 ± 0.3 1.2 ± 0.3 0.61 LDL-cholesterol (mmol/L)c 2.9 ± 0.9 2.5 ± 0.6 0.022 Triglycerides (mmol/L)d 2.8 ± 1.8 2.3 ± 0.9 0.057 2-hour glucose (mmol/L)a 2.9 ± 0.6 2.5 ± 0.7 0.004 Peak insulin (pmol/L)b 1445.9 ± 869.5 1070.9 ± 748.7 0.043 Results are expressed as mean ± SD. p-values < 0.05 are in bold. aTo convert to mg/dL divide by 0.0555. bTo convert to μIU/mL divide by 6.945. cTo convert to mg/dL divide by 0.0259. dTo convert to mg/dL divide by 0.0113.

Initial Visit Last Visit p-value BMI z-score 2.49 ± 0.4 2.33 ± 0.4 0.03 Fasting glucose (mmol/L)a 4.8 ± 0.5 4.6 ± 0.5 0.08 Fasting insulin (pmol/L)b 145.1 ± 90.3 135.4 ± 100.7 0.56 HOMA-IR 4.5 ± 3 4.1 ± 2.9 0.38 HDL-cholesterol (mmol/L)c 1.2 ± 0.3 1.2 ± 0.3 0.61 LDL-cholesterol (mmol/L)c 2.9 ± 0.9 2.5 ± 0.6 0.022 Triglycerides (mmol/L)d 2.8 ± 1.8 2.3 ± 0.9 0.057 2-hour glucose (mmol/L)a 2.9 ± 0.6 2.5 ± 0.7 0.004 Peak insulin (pmol/L)b 1445.9 ± 869.5 1070.9 ± 748.7 0.043 Results are expressed as mean ± SD. p-values < 0.05 are in bold. aTo convert to mg/dL divide by 0.0555. bTo convert to μIU/mL divide by 6.945. cTo convert to mg/dL divide by 0.0259. dTo convert to mg/dL divide by 0.0113. All children with an initially normal OGTT (n = 37) maintained a normal OGTT by the end of the follow-up period, with the exception of 1 child who developed IGT (2-hr glucose, 141 mg/dL); his HOMA-IR increased and BMI z-score decreased. 5 children were found with IGT on their initial OGTT; their BMI z-score and HOMA-IR were similar to those of children with normal OGTT. In 4 children with IGT, the OGTT normalized by the end of the follow-up, while the 5th child developed DM. Of the 4 children with normalized OGTT, 1 child experienced increased BMI z-score and HOMA-IR, 1 child decreased BMI z-score and increased HOMA-IR, and 2 children increased BMI z-score and decreased HOMA-IR. The child who became diabetic had decreased BMI z-score and HOMA-IR by the end of the follow-up period.

oped DM. Of the 4 children with normalized OGTT, 1 child experienced increased BMI z-score and HOMA-IR, 1 child decreased BMI z-score and increased HOMA-IR, and 2 children increased BMI z-score and decreased HOMA-IR. The child who became diabetic had decreased BMI z-score and HOMA-IR by the end of the follow-up period. When children were grouped according to changes in BMI z-score [increased vs. same/decreased BMI z-score (0.2 ± 0.15 vs. -0.32 ± 0.3, p < 0.001)], those with the same or decreased BMI z-score exhibited decreased fasting insulin (136.7 ± 98.6 vs. 153.5 ± 91 pmol/L, last vs. initial visit, p < 0.001) and LDL-cholesterol (2.4 ± 0.5 vs. 2.7 ± 0.9 mmol/L, last vs. initial visit, p = 0.02). Among the 60 subjects for whom a fasting lipid panel was obtained at the initial visit, 25 children had dyslipidemia: 5 with low HDL and high TG, 11 with isolated low HDL and 9 with isolated high TG; when compared to those with normal lipid levels, children with abnormal levels had similar BMI z-scores (Table 2). At the end of the follow-up period, lipid levels normalized in 15 children and remained abnormal in 10 children: the 15 children with normalized lipid levels exhibited a significantly decreased BMI z-score (Table 3), unlike those with persistently abnormal lipid profile. Table 2 Comparison of BMI z-score and metabolic parameters according to the presence of dyslipidemia at the initial visit

Among the 60 subjects for whom a fasting lipid panel was obtained at the initial visit, 25 children had dyslipidemia: 5 with low HDL and high TG, 11 with isolated low HDL and 9 with isolated high TG; when compared to those with normal lipid levels, children with abnormal levels had similar BMI z-scores (Table 2). At the end of the follow-up period, lipid levels normalized in 15 children and remained abnormal in 10 children: the 15 children with normalized lipid levels exhibited a significantly decreased BMI z-score (Table 3), unlike those with persistently abnormal lipid profile. Table 2 Comparison of BMI z-score and metabolic parameters according to the presence of dyslipidemia at the initial visit Dyslipidemia (n=25) Normal Lipids (n=35) p-value BMI z-score 2.42 ± 0.4 2.51 ± 0.4 0.42 HDL-cholesterol (mmol/L)a 1.0 ± 0.2 1.3 ± 0.2 < 0.001 Triglycerides (mmol/L)b 1.6 ± 1.0 0.9 ± 0.3 < 0.001 HOMA-IR 4 ± 1.8 4.6 ± 3.1 0.45 Results are expressed as mean ± SD. p-values < 0.05 are in bold. aTo convert to mg/dL divide by 0.0259. bTo convert to mg/dL divide by 0.0113. Table 3 Changes of BMI z-score and metabolic parameters at the last visit according to changes of lipid levels

Dyslipidemia (n=25) Normal Lipids (n=35) p-value BMI z-score 2.42 ± 0.4 2.51 ± 0.4 0.42 HDL-cholesterol (mmol/L)a 1.0 ± 0.2 1.3 ± 0.2 < 0.001 Triglycerides (mmol/L)b 1.6 ± 1.0 0.9 ± 0.3 < 0.001 HOMA-IR 4 ± 1.8 4.6 ± 3.1 0.45 Results are expressed as mean ± SD. p-values < 0.05 are in bold. aTo convert to mg/dL divide by 0.0259. bTo convert to mg/dL divide by 0.0113. Table 3 Changes of BMI z-score and metabolic parameters at the last visit according to changes of lipid levels Abnormal HDL & TG ↓ Abnormal HDL & TG (n=10) Abnormal HDL & TG ↓ Normal HDL & TG (n=15) Normal HDL & TG ↓ Abnormal HDL & TG (n=5) Normal HDL & TG ↓ Normal HDL & TG (n=30) Initial Visit Last Visit Initial Visit Last Visit Initial Visit Last Visit Initial Visit Last Visit BMI z-score 2.36 ± 0.4 2.32 ± 0.5 2.49 ± 0.4 2.19 ± 0.3* 2.47 ± 0.5 2.47 ± 0.5 2.52 ± 0.4 2.37 ± 0.4 HDL-cholesterol (mmol/L)a 0.9 ± 0.1 0.9 ± 0.1 1.1 ± 0.3 1.1 ± 0.2 1.2 ± 0.1 1 ± 0.1* 1.3 ± 0.2 1.3 ± 0.3 Triglycerides(mmol/L)b 1.9 ± 1.4 1.3 ± 32 1.6 ± 0.7 1 ± 0.3* 1 ± 0.3 1.3 ± 0.4 0.9 ± 0.3 0.9 ± 0.3 HOMA-IR 4.2 ± 1.9 5.3 ± 3.3 4.7 ± 3.4 3.7 ± 2 4 ± 2.4 3.3 ± 2.4 4.7 ± 3.2 3.9 ± 3.2 Results are expressed as mean ± SD. *p-value < 0.05, last vs. initial visit. aTo convert to mg/dL divide by 0.0259. bTo convert to mg/dL divide by 0.0113.

Abnormal HDL & TG ↓ Abnormal HDL & TG (n=10) Abnormal HDL & TG ↓ Normal HDL & TG (n=15) Normal HDL & TG ↓ Abnormal HDL & TG (n=5) Normal HDL & TG ↓ Normal HDL & TG (n=30) Initial Visit Last Visit Initial Visit Last Visit Initial Visit Last Visit Initial Visit Last Visit BMI z-score 2.36 ± 0.4 2.32 ± 0.5 2.49 ± 0.4 2.19 ± 0.3* 2.47 ± 0.5 2.47 ± 0.5 2.52 ± 0.4 2.37 ± 0.4 HDL-cholesterol (mmol/L)a 0.9 ± 0.1 0.9 ± 0.1 1.1 ± 0.3 1.1 ± 0.2 1.2 ± 0.1 1 ± 0.1* 1.3 ± 0.2 1.3 ± 0.3 Triglycerides(mmol/L)b 1.9 ± 1.4 1.3 ± 32 1.6 ± 0.7 1 ± 0.3* 1 ± 0.3 1.3 ± 0.4 0.9 ± 0.3 0.9 ± 0.3 HOMA-IR 4.2 ± 1.9 5.3 ± 3.3 4.7 ± 3.4 3.7 ± 2 4 ± 2.4 3.3 ± 2.4 4.7 ± 3.2 3.9 ± 3.2 Results are expressed as mean ± SD. *p-value < 0.05, last vs. initial visit. aTo convert to mg/dL divide by 0.0259. bTo convert to mg/dL divide by 0.0113. 5 children with normal HDL and TG at baseline developed dyslipidemia at the end of the follow-up period: 1 child developed both elevated TG and low HDL, 3 children had low HDL and 1 developed high TG. These 5 children did not experience any significant change of the mean BMI-z-score or HOMA-IR at the end of the follow-up period (Table 3).

al HDL and TG at baseline developed dyslipidemia at the end of the follow-up period: 1 child developed both elevated TG and low HDL, 3 children had low HDL and 1 developed high TG. These 5 children did not experience any significant change of the mean BMI-z-score or HOMA-IR at the end of the follow-up period (Table 3). When children were divided in two groups according to their mean number of OV per year (≤ 2 vs. >2; 1.53 ± 0.5 vs. 3.19 ± 0.7, p < 0.001), there were no differences at the initial visit for any of the metabolic variables analyzed or for BMI z-score (Table 4). At the end of the follow-up, children with > 2 OV/yr exhibited a significant decrease in their LDL-cholesterol (Table 4) and BMI z-score (Table 4), while children with ≤ 2 OV/yr did not. In the group of children who underwent OGTT, those with > 2 OV/yr experienced a significant decrease of both the 2-hour glucose (n = 33), (Table 4, p = 0.0001) and peak insulin (n = 32), (Table 4, p = 0.0036), while children with ≤ 2 OV/yr did not (n = 8 and 6, respectively). Table 4 Changes of BMI z-score and metabolic parameters according to the frequency of clinic visits

When children were divided in two groups according to their mean number of OV per year (≤ 2 vs. >2; 1.53 ± 0.5 vs. 3.19 ± 0.7, p < 0.001), there were no differences at the initial visit for any of the metabolic variables analyzed or for BMI z-score (Table 4). At the end of the follow-up, children with > 2 OV/yr exhibited a significant decrease in their LDL-cholesterol (Table 4) and BMI z-score (Table 4), while children with ≤ 2 OV/yr did not. In the group of children who underwent OGTT, those with > 2 OV/yr experienced a significant decrease of both the 2-hour glucose (n = 33), (Table 4, p = 0.0001) and peak insulin (n = 32), (Table 4, p = 0.0036), while children with ≤ 2 OV/yr did not (n = 8 and 6, respectively). Table 4 Changes of BMI z-score and metabolic parameters according to the frequency of clinic visits ≤ 2 Outpatient visits (n=13) p-value >2 Outpatient visits (n=48) p-value Initial Visit Last Visit   Initial Visit Last Visit Age (years) 10.9 ± 1.64     11.2 ± 2.9   0.71 Sex (male/female) 5/8     17/31   0.84 BMI z-score 2.44 ± 0.4 2.37 ± 0.4 0.66 2.5 ± 0.4 2.32 ± 0.4 0.03 Fasting glucose (mmol/L)a 4.8 ± 0.32 4.5 ± 0.4 0.09 4.8 ± 0.6 4.7 ± 0.5 0.21 Fasting insulin (pmol/L)b 151.4 ± 59.0 150.01 ± 95.8 0.97 143.8 ± 97.2 131.3 ± 102.1 0.54 HOMA-IR 4.7 ± 1.8 4.4 ± 2.9 0.8 4.5 ± 3.2 4 ± 2.9 0.4 HDL-cholesterol (mmol/L)c 1.2 ± 0.2 1.1 ± 0.3 0.25 1.2 ± 0.3 1.2 ± 0.3 0.98 LDL-cholesterol (mmol/L)c 106.2 ± 84.4 97.9 ± 12.3 0.16 111.3 ± 36.7 96.4 ± 23.5 0.022 Triglycerides (mmol/L)d 1.6 ± 1 1.2 ± 0.4 0.20 1.1 ± 0.7 1 ± 0.4 0.13 2-hour glucose (mmol/L)a (n=8) 5.8 ± 1.2 5.3 ± 1.3 0.44 (n=34) 6.3 ± 1.2 5.2 ± 1.0 0.0001 Peak insulin (pmol/L)a (n=6) 1486.9 ± 1039.0 1474.4 ± 1206.3 0.98 (n=32) 1472.3 ± 817.4 952.9 ± 563.2 0.0036 Results are expressed as mean ± SD. p-values < 0.05 are in bold.

(mmol/L)d 1.6 ± 1 1.2 ± 0.4 0.20 1.1 ± 0.7 1 ± 0.4 0.13 2-hour glucose (mmol/L)a (n=8) 5.8 ± 1.2 5.3 ± 1.3 0.44 (n=34) 6.3 ± 1.2 5.2 ± 1.0 0.0001 Peak insulin (pmol/L)a (n=6) 1486.9 ± 1039.0 1474.4 ± 1206.3 0.98 (n=32) 1472.3 ± 817.4 952.9 ± 563.2 0.0036 Results are expressed as mean ± SD. p-values < 0.05 are in bold. aTo convert to mg/dL divide by 0.0555. bTo convert to μIU/mL divide by 6.945. cTo convert to mg/dL divide by 0.0259. dTo convert to mg/dL divide by 0.0113. Discussion In our multi-ethnic population sample, a non-intensive (~ 3 visits per year) weight management program that reinforced healthy dietary modifications and regular daily activity over a 4-year period resulted in a statistically significant reduction of BMI z-score and LDL-cholesterol, and improvement of glucose tolerance.

In our multi-ethnic population sample, a non-intensive (~ 3 visits per year) weight management program that reinforced healthy dietary modifications and regular daily activity over a 4-year period resulted in a statistically significant reduction of BMI z-score and LDL-cholesterol, and improvement of glucose tolerance. Extensive evidence previously published supports the effectiveness of intensive weight-loss programs in children. In a study conducted by Wilfley et al., 204 overweight children were enrolled to determine the short-term and long-term efficacy of weight-loss and weight maintenance programs [7]. After 5 months of intensive weekly meetings focused on weight-loss treatment with a multi-disciplinary team, almost 90% of children exhibited a decreased BMI z-score. At the end of the weight-loss intervention, the 2 active maintenance groups experienced a mean change in BMI z-score of – 0.22 from baseline to 2-year follow-up versus the control group. Such BMI z-score reduction is similar to the one shown in our study by the end of the 4-year follow-up (−0.16); however, Wilfley et al. did not evaluate the impact of weight loss on metabolic parameters. Savoye et al. studied a population of 209 obese children to evaluate the effects of a 12-month weight management program on adiposity and metabolic parameters [8]. The program included exercise, nutrition, and behavior modification: intervention occurred biweekly the first 6 months and bimonthly thereafter. At the end of study, the weight-management group experienced a significant decrease of BMI and HOMA-IR compared to the control group; conversely, no difference was found relative to changes in fasting glucose, HDL-cholesterol, LDL-cholesterol, or blood pressure. Reinehr et al. studied changes in weight status and cardiovascular disease (CVD) risk factors in 203 obese children who attended a 1-year outpatient intervention program; enrolled subjects were then evaluated 1 year after the end of the intervention [20]. The program included weekly meetings with an exercise physiologist as well as once to twice monthly with a dietitian and a psychologist. Children who experienced a reduction of BMI SDS (72% of the group) at the end of the 12-month intervention maintained this reduction 1 year later. In addition, children with reduced BMI SDS (but not those without) showed improved HDL-cholesterol, LDL-cholesterol, blood pressure, and HOMA-IR.

etitian and a psychologist. Children who experienced a reduction of BMI SDS (72% of the group) at the end of the 12-month intervention maintained this reduction 1 year later. In addition, children with reduced BMI SDS (but not those without) showed improved HDL-cholesterol, LDL-cholesterol, blood pressure, and HOMA-IR. Although the positive effects of all these studies were sustained for a relatively long period of time, the high costs associated with the frequent utilization of a team of dietitians, social workers, and exercise physiologists render this format not widely applicable. In contrast, our findings suggest that weight management programs based on less frequent encounters with a smaller team (pediatric nurse practitioner and a registered dietitian) may result in a similarly effective and lasting reduction of obesity and obesity-associated metabolic abnormalities.

at not widely applicable. In contrast, our findings suggest that weight management programs based on less frequent encounters with a smaller team (pediatric nurse practitioner and a registered dietitian) may result in a similarly effective and lasting reduction of obesity and obesity-associated metabolic abnormalities. The importance of preventing or reducing the severity of overweight in childhood is supported by a number of studies demonstrating the link between pediatric obesity and morbidity and mortality in adulthood. Three previous studies have identified an association between overweight in children and adolescents with increased rates of death due to coronary heart disease [21,22] and with death from all causes [22,23]. In a large cohort of American-Indian subjects followed since childhood [3], the rate of premature death (before 55 years of age) from endogenous causes among children in the highest quartile of BMI was more than double than that in children in the lowest quartile. Of note, the association between BMI and premature death was attenuated but remained significant after adjustment for glucose level, cholesterol level, and blood pressure: thus, some of the effects of overweight on the risk of premature death may not depend on abnormal glucose and lipid metabolism, or on hypertension.

Of note, the association between BMI and premature death was attenuated but remained significant after adjustment for glucose level, cholesterol level, and blood pressure: thus, some of the effects of overweight on the risk of premature death may not depend on abnormal glucose and lipid metabolism, or on hypertension. In our cohort of 61 children, 5 were found with IGT at baseline; in 4 of these children, the OGTT normalized by the end of the 4-year follow-up period, while one child became diabetic. In a similar study, Weiss et al. identified 33 children with IGT in a population sample of 117 obese children [24]. By the end of a 2-year period, 15 of those with IGT reverted to normal glucose tolerance while 8 developed Type 2 DM. When compared to those who reverted to normal glucose tolerance, subjects who developed DM were significantly more obese at baseline and increased their BMI during the follow-up period; in our study, the relationship between OGTT results and initial BMI z-score and/or change in BMI z-score overtime is less clear. While the association between IGT and risk to develop DM has not been well defined in children, it has been clearly demonstrated in adults [25,26]; in addition, IGT in adults appears to be linked to an increased risk for cardiovascular disease and mortality [27,28].

and/or change in BMI z-score overtime is less clear. While the association between IGT and risk to develop DM has not been well defined in children, it has been clearly demonstrated in adults [25,26]; in addition, IGT in adults appears to be linked to an increased risk for cardiovascular disease and mortality [27,28]. In the present study, 25 of the 60 subjects had dyslipidemia at the initial visit: by the end of the 4-year follow-up period, in 15 of these 25 subjects the abnormal lipid levels normalized: unlike those with persistently abnormal lipid panel, the 15 children with normalized lipid levels exhibited a significantly decreased BMI z-score during the 4-year follow-up period. Previous cross-sectional studies have shown a high prevalence of low HDL-cholesterol and elevated TG in obese children [29,30]. A longitudinal study conducted in the United Kingdom in more than 5,000 children showed that 1 SD greater BMI at age 9–12 years was associated with high TG and low HDL-cholesterol at age 15–16 years [31]; in the same study, changing from overweight/obese at age 9–12 to normal weight at age 15–16 was associated with better cardiovascular risk profiles than remaining overweight/obese from childhood through adolescence. Data from 4 prospective cohorts have demonstrated that cardiovascular risk factors in childhood (including high TG) significantly predict subclinical atherosclerosis as early as 9 years of age [32], thus justifying sustained efforts to correct obesity and lipid abnormalities in children.

childhood through adolescence. Data from 4 prospective cohorts have demonstrated that cardiovascular risk factors in childhood (including high TG) significantly predict subclinical atherosclerosis as early as 9 years of age [32], thus justifying sustained efforts to correct obesity and lipid abnormalities in children. There are some limitations of our study, such as the lack of a control group and the relatively small sample size. However, the fact that our results are consistent with those of studies including a larger number of subjects and a control group supports the validity of our findings. In addition, there may have been a selection bias regarding the subjects included in the retrospective analysis, since only a small number of children initially evaluated at our Weight Management Center were eventually followed for 2 or more years. We can speculate that the effects of the program were less significant for those subjects followed for less than 2 years. Those subjects having a longer duration of follow-up may have experienced weight loss early in the program, greater adherence to the lifestyle changes, or more family involvement. Our results demonstrate that a non-intensive weight management program offers potential medical benefits to children and adolescents who are sufficiently motivated to continue their follow-up visits.

ve experienced weight loss early in the program, greater adherence to the lifestyle changes, or more family involvement. Our results demonstrate that a non-intensive weight management program offers potential medical benefits to children and adolescents who are sufficiently motivated to continue their follow-up visits. The long duration of our retrospective study, and the non-intensive approach of our intervention, has rendered unfeasible the concomitant evaluation of a control, completely untreated, group of obese children. To circumvent such limitation, we have used two historical control groups followed longitudinally by Reinehr et al. [33] and by D’Hondt et al. [34]. In the former study, 100 overweight children [BMI-SDS 1.92 (1.27-2.75)] with a mean age of 9 years (6–15 years) were periodically evaluated during a 2-year period without any intervention. This control group did not experience any significant change in their BMI-SDS. In the latter study by D’Hondt et al., at baseline 50 overweight children (8 of which were obese) had a mean age of 11.6 ± 0.8 years and a baseline BMI z-score range of 1.55 ± 0.39 (1.00; 2.64). 2 years later, even these children’s BMI z-scores remained unchanged. These finding suggests that the significant reduction of BMI-SDS observed in our study likely depends on the lifestyle modifications reinforced by our team, rather than simply reflecting a physiological change in adiposity.

1.55 ± 0.39 (1.00; 2.64). 2 years later, even these children’s BMI z-scores remained unchanged. These finding suggests that the significant reduction of BMI-SDS observed in our study likely depends on the lifestyle modifications reinforced by our team, rather than simply reflecting a physiological change in adiposity. In conclusion, our study suggests that a non-intensive, long-term weight management program may significantly improve the degree of obesity and some cardiovascular risk factors in childhood. In addition, this non-intensive treatment (a small team approach) is more likely to be reimbursed by 3rd party payors making it more financially sustainable. Prospective studies with a larger population sample and comparison to a control group are warranted to confirm these findings. Abbreviations BMI: Body Mass Index; OV/yr: Outpatient visits per year; DM: Diabetes Mellitus; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; TG: Triglycerides; OGTT: Oral Glucose Tolerance Test; IGT: Impaired glucose tolerance; HOMA-IR: Homeostasis Model Assessment of Insulin Resistance; CVD: Cardiovascular disease. Competing interests The authors disclose no potential, perceived, or real conflict of interest. Authors’ contributions DN and CD initiated data collection; CD continued data collection and started literature search. RAK continued data collection and completed literature search with FDL. CD, RAK and FDL analyzed data. RAK and FDL as well as CD wrote the paper. All authors had final approval of the submitted version.

Definition and epidemiology of small for gestational age (SGA) Despite past inconsistencies in defining small for gestational age (SGA) (as reviewed by Saenger et al [1]) the International Societies of Pediatric Endocrinology and the Growth Hormone Research Society, as well as the International Small for Gestational Age Advisory Board, recently recommended that the term refer to infants whose birth weights and/or lengths are at least two standard deviation (SD) units less than the mean for gestational age [2,3]. According to this definition, approximately 3%–10% of newborns are considered SGA at birth, although it should be noted that new intrauterine growth curves created with a more contemporary, larger, and more racially diverse population suggest that many SGA patients are often misclassified as appropriate for gestational age (AGA) [4]. While most of these infants undergo catch-up growth, 10%–15% remain small for their age at the age of 2 years [5-8]. In 2001, human growth hormone (hGH) therapy using dose regimens up to 48 mcg/kg/week [3] was approved by the United States (US) Food and Drug Administration (FDA) to treat SGA patients greater than 2 years old. However, because the causes of SGA are diverse, hGH treatment outcomes vary among patients. Thus, identifying the underlying mechanisms for SGA births may help predict patient response to hGH therapy. Causes for SGA births, which are summarized in Table 1[3,4], involve environmental factors, placental factors such as abnormal uteroplacental blood flow, or inherited genetic mutations [3,4]. Over the last two decades, significant research related to genetic mutations that influence SGA has been conducted, and this article reviews the results of these studies and summarizes the success of hGH therapy in treating this condition. It should be mentioned at the beginning of this review, however, that the number of genetic variations for any particular gene that has been associated with SGA birth does not necessarily correlate with the number of patients who have this defect.

es and summarizes the success of hGH therapy in treating this condition. It should be mentioned at the beginning of this review, however, that the number of genetic variations for any particular gene that has been associated with SGA birth does not necessarily correlate with the number of patients who have this defect. For instance, four different genetic mutations in the distal region of the terminal long arm of chromosome 15 linked with SGA birth size will be described, while only two mutations are illustrated for patients born SGA with Silver-Russell syndrome (SRS). However, this does not mean that SGA patients are two times more likely to have a mutation in the distal region of chromosome 15 than a mutation associated with SRS. Of note, a website from the Growth Genetics Consortium, an international collaboration, gathers all current information about genetic syndromes disrupting the growth hormone and insulin-like growth factor (IGF) axis [9]. Cases reported involve the following genes: GHR, GHRHR, STAT5B, IGF1, IGF2, IGFALS, and IGF1R. Forty-eight cases have been approved for inclusion into the database so far. This paper aims to illustrate the variety of genetic mutations that are associated with SGA births while concurrently describing how other phenotype characteristics of the patient, such as motor or mental development, can vary depending on which mutation was inherited. Table 1 Factors associated with increased incidence of SGA birth

For instance, four different genetic mutations in the distal region of the terminal long arm of chromosome 15 linked with SGA birth size will be described, while only two mutations are illustrated for patients born SGA with Silver-Russell syndrome (SRS). However, this does not mean that SGA patients are two times more likely to have a mutation in the distal region of chromosome 15 than a mutation associated with SRS. Of note, a website from the Growth Genetics Consortium, an international collaboration, gathers all current information about genetic syndromes disrupting the growth hormone and insulin-like growth factor (IGF) axis [9]. Cases reported involve the following genes: GHR, GHRHR, STAT5B, IGF1, IGF2, IGFALS, and IGF1R. Forty-eight cases have been approved for inclusion into the database so far. This paper aims to illustrate the variety of genetic mutations that are associated with SGA births while concurrently describing how other phenotype characteristics of the patient, such as motor or mental development, can vary depending on which mutation was inherited. Table 1 Factors associated with increased incidence of SGA birth Fetal Maternal Uterine/Placental Demographic Karyotypic Medical conditions Gross structural placental factors Maternal age -Trisomy 21 -Hypertension -Single umbilical artery -Very young age -Trisomy 18 -Renal disease -Placental hemangiomas -Older age -Monosomy X -Diabetes mellitus -Infarcts, focal lesions Maternal height -Trisomy 13 -Collagen vascular diseases Insufficient uteroplacental perfusion Maternal weight Chromosomal abnormalities -Maternal hypoxemia -Suboptimal implantation site Maternal and paternal race -Autosomal deletion Infection Placenta previa History of SGA -Ring chromosomes -Toxoplasmosis Low-lying placenta Genetic diseases -Rubella Placental abruption -Achondroplasia -Cytomegalovirus -Bloom syndrome -Herpesvirus Congenital anomalies -Malaria -Potter syndrome -Trypanosomiasis -Cardiac abnormalities -HIV Nutritional status -Low prepregnancy weight -Low pregnancy weight Substance use/abuse -Cigarette smoking -Alcohol -Illicit drugs -Therapeutic drugs HIV, human immunodeficiency virus; SGA, small for gestational age.

-Herpesvirus Congenital anomalies -Malaria -Potter syndrome -Trypanosomiasis -Cardiac abnormalities -HIV Nutritional status -Low prepregnancy weight -Low pregnancy weight Substance use/abuse -Cigarette smoking -Alcohol -Illicit drugs -Therapeutic drugs HIV, human immunodeficiency virus; SGA, small for gestational age. Reprinted with permission from [3]. Genetic factors influencing SGA GH/IGF-1 axis IGF-1 gene Transcription of the gene for IGF-1 is mediated by the binding of pituitary GH to specific GH receptors on hepatocytes. The secretion of IGF-1 from the liver then stimulates cell growth (particularly bone) and inhibits secretion of GH from the pituitary [10]. Consequently, mutations of the IGF-1 gene affect growth and GH secretion and have been correlated with SGA births. A homozygous partial deletion of exons 4 and 5 of IGF-1 was observed for one patient born SGA. The mutation truncated the IGF-1 peptide sequence from 70 to 25 amino acids and was followed by an out-of-frame nonsense sequence and stop codon. In addition to growth defects, the patient suffered from bilateral sensorineural deafness and mental retardation, a feature indicating the importance of IGF-1 in central nervous system development [11]. When treated with hGH at a 0.1 U/kg dose for 4 days, no detectable IGF-1 level could be observed in the patient. However, when treated with recombinant human IGF-1 (rhIGF-1) therapy for one year (three months at 40 mcg/kg/day, nine months at 80 mcg/kg/day), insulin sensitivity, bone mineral density, and line growth of this patient were improved [12].

/kg dose for 4 days, no detectable IGF-1 level could be observed in the patient. However, when treated with recombinant human IGF-1 (rhIGF-1) therapy for one year (three months at 40 mcg/kg/day, nine months at 80 mcg/kg/day), insulin sensitivity, bone mineral density, and line growth of this patient were improved [12]. A second patient born SGA with sensorineural deafness and mental retardation was evaluated for IGF-1 defects. Investigators observed a T→A transversion in the 3′-untranslated region of exon 6 that caused the expression of a truncated version of exon 6 and an altered E domain of the IGF-1 prohormone. hGH therapy for this patient (200 mcg hGH/day intramuscular doses for seven days) afforded no improvement in IGF-1 levels [13]. It should be noted that a second research group later sequenced IGF-1 (exons 1–6) in 53 children born SGA and determined that none of the mutations in the coding region of IGF-1 correlate with SGA stature [14].

is patient (200 mcg hGH/day intramuscular doses for seven days) afforded no improvement in IGF-1 levels [13]. It should be noted that a second research group later sequenced IGF-1 (exons 1–6) in 53 children born SGA and determined that none of the mutations in the coding region of IGF-1 correlate with SGA stature [14]. Similarly, an IGF-1 defect was observed for a third patient who had initially been evaluated at the age of 21 years for SGA birth size [15]. In addition to SGA size, the patient initially presented with bilateral hearing loss, microcephaly, and severe mental retardation. When investigators re-evaluated the patient at 55 years of age, unusually high serum levels of IGF-1 were noted, which varied from the patient phenotypes described by Woods and Bonapace [11-13]. Furthermore, both insulin-like growth factor binding protein-3 (IGFBP-3) and insulin-like growth factor-1 receptor (IGF-1R) levels were normal. However, by sequencing IGF-1, investigators detected a nucleotide substitution at position 274 (G→A) in the sequence, which caused an amino acid substitution at position 44 of the IGF-1 protein (V44M). The modified IGF-1 protein displayed a 90-fold-lower binding affinity for IGF-1R than the wild-type derivative, although the mutated protein had normal binding capacity for IGFBPs. This reduced affinity of IGF-1 for IGF-1R resulted in diminished phosphorylation of IGF-1R and downstream-acting signaling proteins, particularly Akt/PKB [16,17].

ein displayed a 90-fold-lower binding affinity for IGF-1R than the wild-type derivative, although the mutated protein had normal binding capacity for IGFBPs. This reduced affinity of IGF-1 for IGF-1R resulted in diminished phosphorylation of IGF-1R and downstream-acting signaling proteins, particularly Akt/PKB [16,17]. Finally, when a fourth patient born SGA in length and weight was evaluated for IGF-1 defects, investigators discovered a homozygous G→A missense mutation in the gene that caused replacement of arginine by a glutamine at position 36 (R36Q) in the C domain of the corresponding IGF-1 protein. This change decreased the binding affinity of IGF-1 for IGF-1R by nearly three-fold, but normal affinity for IGFBP-3 was maintained [18,19]. This study confirmed that IGF-1 mutations that lead to only partial loss of IGF-1 protein activity can cause significant postnatal as well as prenatal growth defects. While high-dose hGH therapy (400 mcg/kg/week) promoted successful catch-up growth, the patient’s treatment modality was recently changed to rhIGF-1 therapy [20]. It should be noted that rhIGF-1 therapy has only been approved by the US FDA to treat patients with severe primary IGF-1 deficiency or patients with GH gene deletions who have developed neutralizing antibodies to GH [21]. The genotype-phenotype correlations and response to hGH therapy for each of these patients expressing IGF-1 mutations is summarized in Table 2[11,13,16-18]. Table 2 Phenotypic characteristics and response to hGH therapy for patients with IGF-1 mutations

Finally, when a fourth patient born SGA in length and weight was evaluated for IGF-1 defects, investigators discovered a homozygous G→A missense mutation in the gene that caused replacement of arginine by a glutamine at position 36 (R36Q) in the C domain of the corresponding IGF-1 protein. This change decreased the binding affinity of IGF-1 for IGF-1R by nearly three-fold, but normal affinity for IGFBP-3 was maintained [18,19]. This study confirmed that IGF-1 mutations that lead to only partial loss of IGF-1 protein activity can cause significant postnatal as well as prenatal growth defects. While high-dose hGH therapy (400 mcg/kg/week) promoted successful catch-up growth, the patient’s treatment modality was recently changed to rhIGF-1 therapy [20]. It should be noted that rhIGF-1 therapy has only been approved by the US FDA to treat patients with severe primary IGF-1 deficiency or patients with GH gene deletions who have developed neutralizing antibodies to GH [21]. The genotype-phenotype correlations and response to hGH therapy for each of these patients expressing IGF-1 mutations is summarized in Table 2[11,13,16-18]. Table 2 Phenotypic characteristics and response to hGH therapy for patients with IGF-1 mutations Genetic mutation Phenotype GH response Ref. Deletion of exons 4 and 5 Birth weight −3.9 SD; birth length −5.4 SD; sensorineural deafness and mental retardation; nearly undetectable IGF-1 levels − Woods, 1996 [11] Truncated version of exon 6 Birth weight −4 SD; birth length −6.5 SD; sensorineural deafness and mental retardation; low serum IGF-1 levels − Bonapace, 2003 [13] V44M Birth weight −3.9 SD score; birth length −4.3 SD score; bilateral hearing loss, microcephaly, severe mental retardation; elevated GH levels and IGF-1 levels but normal IGFBP-3 levels n.a. Walenkamp, 2005 [16] Denley, 2005 [17] R36Q Birth weight −2.5 SD score; birth length −3.7 SD score; mild mental development delay; reduced IGF-1 levels but increased IGFBP-3 levels + Netchine, 2006 [18] hGH, human growth hormone; IGF-1, insulin-like growth factor-1; IGFBP-3, insulin-like growth factor binding protein-3; n.a., not available; SD, standard deviation.

rth weight −2.5 SD score; birth length −3.7 SD score; mild mental development delay; reduced IGF-1 levels but increased IGFBP-3 levels + Netchine, 2006 [18] hGH, human growth hormone; IGF-1, insulin-like growth factor-1; IGFBP-3, insulin-like growth factor binding protein-3; n.a., not available; SD, standard deviation. IGF-1R Various compound heterozygous mutations throughout the coding sequence of IGF-1R have been described for multiple families, with each case exhibiting phenotype variations [22]. Typically, IGF-1R mutations can be classified as point mutations or partial deletions. When one patient born SGA with significantly delayed postnatal growth was evaluated for IGF-1R mutations, investigators determined that two point mutations in exon 2 of IGF-1R caused two single-base pair substitutions in the codons for amino acid 108 (CGG→CAG) and 115 (AAA→AAC) of the corresponding protein. This change resulted in two-thirds-lower binding affinity of IGF-1 to IGF-1R in fibroblasts as compared with controls. When treated with hGH therapy (37.5 mcg/kg/week), the patient’s growth rate was increased to the 75th percentile for her age [23].

ino acid 108 (CGG→CAG) and 115 (AAA→AAC) of the corresponding protein. This change resulted in two-thirds-lower binding affinity of IGF-1 to IGF-1R in fibroblasts as compared with controls. When treated with hGH therapy (37.5 mcg/kg/week), the patient’s growth rate was increased to the 75th percentile for her age [23]. Similarly, a second patient born SGA who suffered from postnatal growth delay, microcephaly, and mild mental retardation was evaluated for IGF-1R mutations. A heterozygous point mutation CGA to TGA (Arg59Ter) in exon 2 of IGF-1R caused early termination of transcription of the IGF-1R protein, leading to reduced receptor expression on the cell surface, as well as decreased autophosphorylation and phosphorylation of signaling proteins [23]. When treated with hGH at 30 mcg/kg/day starting at age 6 years, the patient’s height increased by 1.01 SD after two years of therapy, indicating that hGH therapy can improve quality of life for SGA patients with this mutation [24].

as well as decreased autophosphorylation and phosphorylation of signaling proteins [23]. When treated with hGH at 30 mcg/kg/day starting at age 6 years, the patient’s height increased by 1.01 SD after two years of therapy, indicating that hGH therapy can improve quality of life for SGA patients with this mutation [24]. When 24 children born SGA were evaluated by direct sequencing of IGF-1R to identify causal mutations, two patients were observed to have a heterozygous missense mutation (C→T) of IGF-1R, which altered the cleavage site of the proreceptor of IGF-1R from RLRR to RLQR (R709Q). This mutation inhibited the expression of mature IGF-1R from the IGF-1R precursor protein. Interestingly, the two patients who presented with this mutation had different levels of mental development. While patient 1 displayed mental retardation, patient 2 had normal intellectual development. Thus, no link exists between the heterozygous IGF-1R mutation and intellectual development [25].

1R precursor protein. Interestingly, the two patients who presented with this mutation had different levels of mental development. While patient 1 displayed mental retardation, patient 2 had normal intellectual development. Thus, no link exists between the heterozygous IGF-1R mutation and intellectual development [25]. Similarly, two more patients were evaluated and determined to present with a missense mutation in the intracellular kinase domain of IGF-1R. The older patient, a 35-year-old mother, showed above-average intelligence and no dysmorphic features, but her height (−4.0 SD score) and head circumference (−3.0 SD score) showed growth retardation. Her daughter, patient 2, was born SGA and showed normal mental development but delayed motor development by the age of 15 months. Both patients showed increased IGF-1 levels. Sequence analysis of IGF-1R showed a heterozygous G→A nucleotide substitution, which changed the amino acid sequence of IGF-1R at position 1050 from glutamic acid to lysine. This mutation did not affect expression of IGF-1R protein, but the sequence alteration reduced autophosphorylation of IGF-1R and activation of PKB/Akt [26]. Similarly, a 13.6-year-old girl who displayed short stature (−5.0 SD score) and reduced bone age (9.7 years), as well as elevated IGF-1 levels and no improvement in height following six months of treatment with hGH therapy at a daily dose of 70 mcg/kg/day, was evaluated for IGF-1R mutations. A heterozygous G→A point mutation at position 1577 of IGF-1R resulted in substitution of arginine with glutamine at residue 481 of the corresponding protein (R481Q). This mutation altered the α-subunit of IGF-1R, leading to reduced phosphorylation and cell growth [27]. Recently, a third report has described a similar IGF-1R mutation in which alanine replaced glycine at position 1125 in seven patients from the same family, causing reduced receptor autophosphorylation and phosphorylation of downstream kinases [28].

nit of IGF-1R, leading to reduced phosphorylation and cell growth [27]. Recently, a third report has described a similar IGF-1R mutation in which alanine replaced glycine at position 1125 in seven patients from the same family, causing reduced receptor autophosphorylation and phosphorylation of downstream kinases [28]. Finally, a patient born SGA with high IGF-1 levels who showed only marginal improvement in height following treatment with hGH therapy at the age of 7.4 years (doses ranging from 31 to 36 mcg/kg/day) was evaluated for IGF-1R mutation. Gene sequencing showed heterozygous T→A mutation at position 1886, which resulted in substitution of valine with glutamic acid at position 599 of the protein (V599E). This mutation interfered with the receptor trafficking pathway, diminishing the density of the receptor on the cell surface [29]. The genotype-phenotype correlations and response to hGH therapy for each of these patients expressing IGF-1R point mutations is summarized in Table 3[23,25-29]. Table 3 Phenotypic characteristics and response to hGH therapy for patients with IGF-1R mutations

Finally, a patient born SGA with high IGF-1 levels who showed only marginal improvement in height following treatment with hGH therapy at the age of 7.4 years (doses ranging from 31 to 36 mcg/kg/day) was evaluated for IGF-1R mutation. Gene sequencing showed heterozygous T→A mutation at position 1886, which resulted in substitution of valine with glutamic acid at position 599 of the protein (V599E). This mutation interfered with the receptor trafficking pathway, diminishing the density of the receptor on the cell surface [29]. The genotype-phenotype correlations and response to hGH therapy for each of these patients expressing IGF-1R point mutations is summarized in Table 3[23,25-29]. Table 3 Phenotypic characteristics and response to hGH therapy for patients with IGF-1R mutations Genetic mutation Phenotype GH response References R108Q K115N Birth weight −3.5 SD score; delayed motor skill development; psychiatric anomalies; normal IGF-1 levels, delayed motor development + Abuzzahab, 2003 [23] R59X Birth weight −3.5 SD score; birth length −5.8 SD score; microcephaly, mild retardation, and delayed motor and speech development; + Abuzzahab, 2003 [23] R709Q Birth weight −1.5 SD score; birth length −1.0 SD score; significant mental retardation N/A Kawashima, 2005 [25] E1050K Birth height −0.3 SD score, birth weight −2.1 SD score; height at 35 years −4.0 SD score; head circumference at 35 years −3.0 SD score; no dysmorphic features; high IGF-1 levels N/A Walenkamp, 2006 [26] R481Q Height −4.9 SD score, reduced bone age, elevated IGF-1 levels − Inagaki, 2007 [27] G1125A Birth weight −1.7 SD score; head circumference at birth −3.7 SD score; normal mental development N/A Kruis, 2010 [28] V599E Birth weight −2.3 SD score; birth head circumference <3rd percentile; high IGF-1 levels; mental retardation − Wallborn, 2010 [29] hGH, human growth hormone; IGF-1, insulin-like growth factor-1; N/A, not available; SD, standard deviation.

rence at birth −3.7 SD score; normal mental development N/A Kruis, 2010 [28] V599E Birth weight −2.3 SD score; birth head circumference <3rd percentile; high IGF-1 levels; mental retardation − Wallborn, 2010 [29] hGH, human growth hormone; IGF-1, insulin-like growth factor-1; N/A, not available; SD, standard deviation. In addition to point mutations, distal deletions of the terminal long arm of chromosome 15 have also been linked to patients born SGA, although these mutations are quite rare. Often, these patients present with symptoms resembling Prader-Willi or Angelman syndrome, two diseases resulting from deletions in the 15q11q13 region [30]. One patient born SGA who exhibited continued growth retardation at the age of 4.5 years was evaluated for such distal deletion. It was determined that the patient presented with partial monosomy 15q26.2→15qter, correlating to a deleted critical region of approximately 5.7 Mb [31]. This deletion includes the region 15q26.3, to which the IGF-1R gene has been assigned [32]. A similar deletion was observed for a patient born SGA who displayed a heterozygous 8.58 Mb deletion in the same region [33]. Similarly, a patient born SGA who showed significant growth retardation by the age of 2 years was evaluated for deletions in chromosome 15. Results indicated that the maternally derived chromosome 15 had a 4.7 Mb deleted region, which included 15q26.2 [34]. The smallest deletion of chromosome 15 that has been observed to cause SGA birth involves a mutation in exons 11–21 of the IGF-1R gene (a 0.095 Mb deletion) and was associated with SGA births over three generations in a single family [35]. Typically, patients with partial deletions in this region display mental and psychomotor developmental retardations more often than patients with point IGF-1R mutations [22].

tation in exons 11–21 of the IGF-1R gene (a 0.095 Mb deletion) and was associated with SGA births over three generations in a single family [35]. Typically, patients with partial deletions in this region display mental and psychomotor developmental retardations more often than patients with point IGF-1R mutations [22]. Fortunately, patients with partial deletions of chromosome 15 respond favorably to hGH treatment. A patient born SGA who displayed a heterozygous loss of 15q26.2→15qter began hGH treatment at the age of 5.3 years at a dose of 1 mg/m2/day (approximately 30 mcg/kg/day). Rapid growth catch-up was observed, and by the age of 15 years the patient had nearly reached her target height (−1.6 SD score) [36]. Similarly, two patients displaying deletions in exons 1–21 and exons 3–21 were treated with hGH therapy at a dose of 1 mg/m2/day (approximately 30 mcg/kg/day). For both patients, treatment resulted in moderate increase in height of approximately +1 SD after one year [37].

reached her target height (−1.6 SD score) [36]. Similarly, two patients displaying deletions in exons 1–21 and exons 3–21 were treated with hGH therapy at a dose of 1 mg/m2/day (approximately 30 mcg/kg/day). For both patients, treatment resulted in moderate increase in height of approximately +1 SD after one year [37]. Acid-Labile Subunit (ALS) Deficiency In serum, IGF-1 circulates in complex with IGFBP-3 or IGFBP-5 and an ALS, an 85-kDa glycoprotein that functions to prolong the half-life of the IGF-IGFBP-3/IGFBP-5 binary complex [38]. Sixteen different mutations of the IGFALS gene, located at 16p13.3 on chromosome 16, have been observed in patients who presented with reduced postnatal growth. The type of IGFALS gene mutation varies, including missense, nonsense, deletion, duplication and insertion that cause frameshift and premature stop codons, and in-frame duplication mutations, but nearly all of the mutations show autosomal recessive pattern of inheritance and cause defects in the leucine-rich repeat region of the corresponding ALS protein (Table 4) [39-47]. All of these mutations result in circulating ALS levels that are barely detectable based on enzyme-linked immunoabsorbent assay, radioimmunoassay, or Western immunoblot assays, indicating that the mutations likely inhibit the corresponding protein from being secreted by the liver or cause the protein to degrade rapidly after secretion. The circulating ALS deficiency results in a severe reduction in IGF-1 and IGFBP-3 levels, insulin insensitivity, and pubertal delay. hGH therapy was initiated for some of these patients in an effort to increase growth rate. However, despite the treatments, ranging in duration from six months to more than two years, very little growth response was observed. However, it has been suggested that hGH therapy may be beneficial for heterozygous carriers who still carry one intact IGFALS allele.

hese patients in an effort to increase growth rate. However, despite the treatments, ranging in duration from six months to more than two years, very little growth response was observed. However, it has been suggested that hGH therapy may be beneficial for heterozygous carriers who still carry one intact IGFALS allele. Table 4 Genetic mutations involved in ALS deficiency [[39]-[47]] Genetic mutation Type of mutation Homozygous/Heterozygous E35KfsX87 Frameshift, premature stop codon Homozygous El35GfsX17 Frameshift, premature stop codon Heterozygous C60S Missense Compound heterozygous P73L Missense Homozygous L134Q Missense Homozygous L172F Missense Homozygous A183SfsX149 Frameshift, premature stop codon Compound heterozygous S195_R197dup In-frame insertion of 3 amino acids, SLR Compound heterozygous L241P Missense Compound heterozygous L244F Missense Compound heterozygous N276S Missense Homozygous Q320X Nonsense Homozygous L437_L439dup In-frame insertion of 3 amino acids, LEL Homozygous D440N Missense Homozygous L497FfsX40 Frameshift, premature stop codon Homozygous C540R Missense Compound heterozygous ALS, acid-labile subunit.

Compound heterozygous L244F Missense Compound heterozygous N276S Missense Homozygous Q320X Nonsense Homozygous L437_L439dup In-frame insertion of 3 amino acids, LEL Homozygous D440N Missense Homozygous L497FfsX40 Frameshift, premature stop codon Homozygous C540R Missense Compound heterozygous ALS, acid-labile subunit. Select polymorphisms Obesity and diabetes For many individuals born SGA, health concerns such as obesity, type 2 diabetes, hypertension, and ischemic heart disease are often encountered later in life [48-50]. In one study, DNA samples from 546 patients (227 children born SGA and 319 born AGA) were analyzed for 54 single nucleotide polymorphisms (SNPs) associated with diabetes or obesity. Genetic variations in five of these SNPs (KCNJ11, BDNF, PFKP, PTER, and SEC16B) correlated with SGA size. Therefore, genetic factors that contribute to obesity and type 2 diabetes likely correlate with SGA [51].

9 born AGA) were analyzed for 54 single nucleotide polymorphisms (SNPs) associated with diabetes or obesity. Genetic variations in five of these SNPs (KCNJ11, BDNF, PFKP, PTER, and SEC16B) correlated with SGA size. Therefore, genetic factors that contribute to obesity and type 2 diabetes likely correlate with SGA [51]. Angiotensinogen gene variants Angiotensinogen (AGT) is an α2-globulin precursor to angiotensin II that regulates blood pressure and overall homeostasis [52]. In one study, 174 women and their 162 infants born SGA were compared with 400 women and their 240 infants born AGA. The study evaluated these individuals for a methionine to threonine substitution at codon 235 (235Met >Thr) in the AGT gene, a mutation associated with pregnancy complications such as preeclampsia [53]. The results showed a higher frequency of the 235Thr allele in both mothers (0.60 for SGA versus 0.36 for controls) and infants (0.59 for SGA versus 0.38 for controls) who were associated with SGA births [54]. However, the mechanism by which the 235Met >Thr mutation affects maternal-placental and fetal-placental circulation and, consequently, fetal growth is not understood. Interestingly, a prior study found no correlation between this polymorphism and an increased risk of SGA birth. The differences between the findings of the two investigations were attributed, in part, to variation in ethnic diversity between the two study groups [55].

and, consequently, fetal growth is not understood. Interestingly, a prior study found no correlation between this polymorphism and an increased risk of SGA birth. The differences between the findings of the two investigations were attributed, in part, to variation in ethnic diversity between the two study groups [55]. Deletion of exon 3 growth hormone receptor (d3-GHR) The d3-GHR polymorphism, a 2.7 kB deletion in exon 3 of the GHR gene, is a common genetic defect in individuals with normal height and those born SGA [56]. However, for patients born SGA, the d3-GHR polymorphism has been investigated as a potential mutation that affects hGH therapy due to its role in GH signaling. When response to hGH therapy was compared between children born SGA who had only full-length GHR versus at least one d3-GHR allele, results showed that patients with the d3-GHR polymorphism responded 1.7 to 2 times better to hGH therapy than patients with only the full-length gene [57]. Similarly, SGA patients with either two full-length GHRs (fl/fl) or one (d3/fl) or two (d3/d3) d3-GHR alleles were administered hGH for 12 months at a mean dose of 56 ± 11 mcg/kg/day. At the end of 12 months, carriers of either one or two d3-GHR alleles were observed to respond slightly better to hGH therapy than patients with two full-length alleles, although the difference was not statistically significant. The authors suggested that response to hGH therapy for patients with this mutation depends on the specific causes of short stature, such as IGF-1 insensitivity or IGF-1 deficiency [58]. Consequently, children born SGA with the d3-GHR mutation appear to be prime candidates for hGH therapy, although these results are still controversial.

sted that response to hGH therapy for patients with this mutation depends on the specific causes of short stature, such as IGF-1 insensitivity or IGF-1 deficiency [58]. Consequently, children born SGA with the d3-GHR mutation appear to be prime candidates for hGH therapy, although these results are still controversial. For instance, a comparison was made between the GHR genotype (ie, fl/fl, d3/fl, or d3/d3) of patients with GH deficiency and the individual’s response to hGH treatment. Patients were treated with hGH at a mean dose of 0.2 mg/kg/week for one year and then evaluated for height SD score, height velocity, and height velocity SD score. No statistically significant difference with respect to the measured outcomes could be observed between the patients with the d3-GHR allele and patients who were homozygous for the full-length GHR. Furthermore, this study observed that there was no relationship between an individual’s baseline phenotype and his/her GHR genotype, suggesting that the d3-GHR allele does not affect height in GH deficiency [59]. This lack of correlation between d3-GHR genotype and response to hGH treatment was also confirmed in studies for patients born SGA [60,61].

that there was no relationship between an individual’s baseline phenotype and his/her GHR genotype, suggesting that the d3-GHR allele does not affect height in GH deficiency [59]. This lack of correlation between d3-GHR genotype and response to hGH treatment was also confirmed in studies for patients born SGA [60,61]. Recently, a meta-analysis of 15 studies investigating the effects of d3-GHR genotype and a patient’s first-year response to hGH therapy, including height gain and change in growth velocity, was conducted. The results of this analysis indicated that patients with the d3-GHR allele showed improved growth velocity when treated with hGH therapy, but the treatment outcome was affected by the dose (low doses of hGH showed best response) and age at time of treatment (older patients responded more favorably). It should be noted, however, that this meta-analysis did not discriminate with respect to the cause of short stature [62]. In a recent 3-year review, Doerr et al conclude that the determination of GHR isoforms for deletion of exon 3 is not particularly useful in defining the overall response to GH in short SGA children [63].

uld be noted, however, that this meta-analysis did not discriminate with respect to the cause of short stature [62]. In a recent 3-year review, Doerr et al conclude that the determination of GHR isoforms for deletion of exon 3 is not particularly useful in defining the overall response to GH in short SGA children [63]. Uniparental disomy (UPD) and imprinting effects UPD is a process whereby a person inherits two copies of a gene or chromosome from one parent and no copies from the other parent. In most cases, UPD does not affect fetal development. However, if a UPD gene is also an imprinted gene, there may be adverse effects to the fetus, because UPD of imprinted genes is equivalent to functional nullisomy [64]. The transcriptional regulation of imprinted genes varies from normal genes in that imprinted genes are only active from one parent allele. For instance, a gene may be active only when paternally inherited; the maternal allele of this gene is “switched off.” Conversely, imprinted genes can be maternally expressed and paternally imprinted [65]. Thus, if a patient inherits two versions of an imprinted gene (eg, two copies of a maternal, “switched-off” gene), phenotype abnormalities may result. Studies have indicated that several UPDs can be responsible for short stature in patients born SGA.

nted genes can be maternally expressed and paternally imprinted [65]. Thus, if a patient inherits two versions of an imprinted gene (eg, two copies of a maternal, “switched-off” gene), phenotype abnormalities may result. Studies have indicated that several UPDs can be responsible for short stature in patients born SGA. SRS SRS is a disorder characterized by reduced birth weight, facial features including triangular shape and pointed chin, and body asymmetry [66,67]. Growth restrictions continue through life and often correlate with fasting hypoglycemia [68]. hGH treatment, given daily as subcutaneous injections at a dose of 35 mcg/kg/day for up to three years, is usually suggested for these patients [69]. The genetic causes of SRS vary, with cases of autosomal-dominant, autosomal-recessive, and X-linked inheritance all observed (as reviewed by Hitchins and Abu-Amero) [68,70]. However, the most referenced causal candidates for this disease involve mutations on chromosomes 7 and 11, which both contain groups of genes that undergo genomic imprinting [68]. Since the early 1990s, maternal uniparental disomy 7 (mUPD7), both full mUPD7 and mUPD for the long arm of chromosome 7, were documented to be the cause of SRS in approximately 10% of cases [71]. However, the phenotype of an SRS patient presenting UPD7 cannot be predicted, as the exact etiology of the mutation varies [72]. Polymerase chain reaction with microsatellite repeat markers or Southern blot analysis with variable number of tandem repeats can effectively be used to screen patients for mUPD7 [73].

[71]. However, the phenotype of an SRS patient presenting UPD7 cannot be predicted, as the exact etiology of the mutation varies [72]. Polymerase chain reaction with microsatellite repeat markers or Southern blot analysis with variable number of tandem repeats can effectively be used to screen patients for mUPD7 [73]. In addition to mUPD7, the imprinted region on chromosome 11p15 has been associated with SRS in up to 65% of patients. Specifically, hypomethylation at the imprinting center region 1 (ICR1) was associated with fetal growth retardation in SRS patients (Figure 1) [74,75]. Generally, the ICR1 region regulates the expression of IGF2 and H19, and loss of methylation of this region is associated with approximately 50% of SRS cases [68]. However, an inherited duplication (0.76 – 1 Mb) in the ICR2 domain of 11p15 has also been shown to be involved in the etiology of SRS. The duplicated region included the maternally expressed genes KCNQ1, CDKN1C, TSSC5/SLC22A8 and TSSC3/PHDLA2 and the paternally expressed gene LIT1[76]. It should be noted, however, that the distribution of methylation values among patients with SRS is quite varied, making clinical diagnosis of the disease based on methylation analysis difficult [77]. In general, use of hGH has become a standard treatment regimen for patients with SRS, despite the limited number of evaluations regarding the effectiveness of this treatment [78].

ation values among patients with SRS is quite varied, making clinical diagnosis of the disease based on methylation analysis difficult [77]. In general, use of hGH has become a standard treatment regimen for patients with SRS, despite the limited number of evaluations regarding the effectiveness of this treatment [78]. Figure 1 Quantitative representation of methylation indices for H19-IGF2 ICR1 in individuals with SRS (individuals 1–9 and individual 6’s twin) and an individual with isolated hypermethylation of the H19 promoter (BWS1). Five individuals with SRS displayed partial loss of H19-IGF2 ICR1, indicated by the bar below the shaded area near 50%. SRS, Silver-Russell syndrome. Reprinted from [74] with permission from Macmillan Publishers Ltd; copyright 2005.

6’s twin) and an individual with isolated hypermethylation of the H19 promoter (BWS1). Five individuals with SRS displayed partial loss of H19-IGF2 ICR1, indicated by the bar below the shaded area near 50%. SRS, Silver-Russell syndrome. Reprinted from [74] with permission from Macmillan Publishers Ltd; copyright 2005. mUPD UPD of the long arm of chromosome 14 (UPD14) has been associated with both below-average growth and mental retardation. Initially, it was not known whether the congenital anomalies present in UPD14 patients resulted from an extra copy of an active imprinted gene (ie, two genes that were “switched on”) or the absence of gene expression caused by the presence of two repressed alleles (ie, two genes “switched off”). To determine the likely cause of the phenotype, patients with distal partial trisomy for chromosome 14 (Ts14) were evaluated to determine genotype-phenotype correlations to determine whether the partial trisomy was of maternal or paternal origin. By investigating patients with an extra copy of either maternally inherited or paternally inherited copies of chromosome 14, the investigators hoped to observe more pronounced effects of the disease if it was caused by active imprinted genes. All 13 patients with distal maternal Ts14 (mTs14) were born SGA. Conversely, over half of the patients with paternal Ts14 (pTs14) were born at weights AGA, indicating that an absence of paternal information likely causes growth retardation in patients with UPD14. The minimum trisomic regions 14q31.1-14qter and 14q24.3-14qter were identified as possibly containing the imprinted genes [79]. Overall, the phenotype of patients with mUPD14 can be quite variable. A review of 24 cases of patients displaying mUPD14 attributes the growth retardation of these patients to confined placental mosaicism and imprinted genes that cause early skeletal maturation, although unusual phenotypes may also be caused by autosomal, recessively inherited mutations [80].

UPD14 can be quite variable. A review of 24 cases of patients displaying mUPD14 attributes the growth retardation of these patients to confined placental mosaicism and imprinted genes that cause early skeletal maturation, although unusual phenotypes may also be caused by autosomal, recessively inherited mutations [80]. hGH treatment for SGA Much research has correlated genetic mutations with SGA births, but the ability to predict the effectiveness of hGH therapy for each mutation remains controversial. Table 5 summarizes the various mutations that have been shown to cause SGA and the likelihood that hGH therapy will promote growth for individuals with these mutations [11,13,16-18,23,25-29,39-47,51,53-55,57-62,68,70,74-76,78-80]. Some patients with IGF-1 mutations have shown positive growth catch-up when treated with hGH therapy, while others have shown better response to rhIGF-1 therapy. Alternatively, the response to hGH therapy for patients with IGF-1R mutations appears to correlate with the type of mutation; patients with distal deletions of the IGF-1R gene generally have improved GH-induced catch-up growth as compared with patients who have IGF-1R point mutations. Finally, the ability to predict the effectiveness of hGH treatment depending on the specific disease (eg, children with SRS versus children with UPD14) has not been thoroughly reviewed, possibly because a significant number of patients born SGA who undergo hGH therapy are never genetically diagnosed. However, despite the controversies, clinical studies have successfully elucidated some trends about hGH treatment on growth in children born SGA (as reviewed by Simon et al [81] and Saenger et al [1]).

d, possibly because a significant number of patients born SGA who undergo hGH therapy are never genetically diagnosed. However, despite the controversies, clinical studies have successfully elucidated some trends about hGH treatment on growth in children born SGA (as reviewed by Simon et al [81] and Saenger et al [1]). Table 5 Summary of known genetic causes of SGA and the correlating response to hGH therapy [[11],[13],[16]-[18],[23],[25]-[29],[39]-[47],[51],[53]-[55],[57]-[62],[68],[70],[74]-[76],[78]-[80]] Class of genetic mutation Specific genetic variant Response to hGH therapy IGF-1 Generally not effective IGF-1R Good for partial distal deletions; generally not effective for point mutations GH/IGF-1 axis Point Distal ALS deletions Good outcome for heterozygous carriers Obesity/diabetes-related genes Unclear Select Polymorphisms Angiotensinogen gene Unclear d3-GHR Good outcome, but dose and age matter SRS hGH therapy is commonly used for SRS, but correlation between effectiveness and specific genetic mutation has not been carefully evaluated Full mUPD7 UPD/imprinting effects mUPD7 for long arm of chromosome 7 Hypomethylation at ICR1 on 11p15 Duplication of ICR2 on 11p15 UPD14 Unclear ALS, acid-labile subunit; ICR, imprinting control region; IGF-1, insulin-like growth factor-1; d3-GHR, deletion of exon 3 growth hormone receptor; hGH, human growth hormone; SGA, small for gestational age; SRS, Silver-Russell syndrome; UPD, uniparental disomy.

at ICR1 on 11p15 Duplication of ICR2 on 11p15 UPD14 Unclear ALS, acid-labile subunit; ICR, imprinting control region; IGF-1, insulin-like growth factor-1; d3-GHR, deletion of exon 3 growth hormone receptor; hGH, human growth hormone; SGA, small for gestational age; SRS, Silver-Russell syndrome; UPD, uniparental disomy. The rate of catch-up growth promoted by hGH therapy in patients born SGA correlates with the dose; higher doses typically afford rapid height increase, although a similar response can be achieved using lower doses for a longer time. For instance, a height gain of 2 SD was achieved for patients born SGA using doses of either 67 mcg/kg/day over 2.5 years or 33 mcg/kg/day over 5.5 years (Figure 2A) [82]. However, the low-dose regimen requires three times as many injections and 50% more hGH overall than the high-dose method. The method of administration of hGH therapy can also affect height gain, though less significantly than dose. Patients who received discontinuous high-dose hGH therapy (67 mcg/kg/day for one or two years) have shown slightly increased height gain compared with patients receiving a continuous low-dose regimen (33 mcg/kg/day doses for three or four years), although discontinuation of the treatment typically corresponds with reduction in growth velocity [83]. A similar trend was observed previously by De Zeghers et al, who found that after six years, height SD scores were similar for high-dose hGH course for two years and continuous low-dose hGH treatment for six years (Figure 2B) [82].

tinuation of the treatment typically corresponds with reduction in growth velocity [83]. A similar trend was observed previously by De Zeghers et al, who found that after six years, height SD scores were similar for high-dose hGH course for two years and continuous low-dose hGH treatment for six years (Figure 2B) [82]. Figure 2 (A) Amount of time required to increase height SD score by 2 was 2.5 years for hGH therapy administered at a dose of 67 mcg/kg/day and 5.5 years for hGH therapy administered at a dose of 33 mcg/kg/day in children born SGA. Figure 2A reprinted with permission from [82]. (B) After 6 years, similar height SD scores were achieved using 2 years of high-dose (100 mcg/kg/day) hGH therapy and 6 years of low-dose (33 mcg/kg/day) hGH therapy for children born SGA. hGH, human growth hormone; SD, standard deviation; SGA, small for gestational age. Figure 2B reprinted with permission from [82].

. (B) After 6 years, similar height SD scores were achieved using 2 years of high-dose (100 mcg/kg/day) hGH therapy and 6 years of low-dose (33 mcg/kg/day) hGH therapy for children born SGA. hGH, human growth hormone; SD, standard deviation; SGA, small for gestational age. Figure 2B reprinted with permission from [82]. In addition to the dose and method of administration, the age of initiation of hGH therapy significantly affects the outcome. Patients treated before the onset of puberty achieve optimal results. A recent study showed that children treated for one year with hGH therapy before the age of 4 years achieved greater height gain (1.7 SD score, 12.5 cm) than those treated after 4 years of age (1.2 SD score) [84]. Even among older patients, this trend persists. Patients receiving hGH therapy more than two years before puberty showed increased height gain (1.7 SD, ~12 cm) compared with patients treated fewer than two years before puberty (0.9 SD gain, 6 cm). However, nearly 90% of these patients achieved adult height within the normal range [85]. Conversely, patients treated during puberty achieved height gain of only 0.6 SD score, and fewer than 50% of these patients achieved normal adult height [86].

tients treated fewer than two years before puberty (0.9 SD gain, 6 cm). However, nearly 90% of these patients achieved adult height within the normal range [85]. Conversely, patients treated during puberty achieved height gain of only 0.6 SD score, and fewer than 50% of these patients achieved normal adult height [86]. In 2003, Ranke et al developed a model that essentially summarized the trends that we have described and that could be used by physicians to individualize hGH treatment for SGA patients. Using a pharmacoepidemiological survey of 613 children, various trends were elucidated. In fact, the model could be used to explain approximately 50% of the variability associated with hGH therapy response during the first and second years of treatment. Nearly 35% of the variability could be attributed to the dose, followed by the patient’s age at the start of treatment. Subsequent growth during the second year of treatment could be predicted based on a successful first year of treatment [87].

d with hGH therapy response during the first and second years of treatment. Nearly 35% of the variability could be attributed to the dose, followed by the patient’s age at the start of treatment. Subsequent growth during the second year of treatment could be predicted based on a successful first year of treatment [87]. It must be mentioned, though not stressed, that some controversy regarding the use of hGH arose in 2011 due to results from a study conducted in France, the Santé Adulte GH Enfant (SAGhE) study. The results from this study indicated that long-term use of hGH in children with short stature could increase a patient’s risk of death [88]. The SAGhE study reported that hGH therapy, when administered to patients at doses above 50 mcg/kg/day, increased the risk of death by 30% as compared to the general population in France. This effect was attributed to an increased likelihood of bone-tumor formation, cardiovascular disease, and cerebrovascular events. These results concerned patients born SGA, as the normal recommended dose of hGH therapy can be approximately 70 mcg/kg/day (Table 6) [1]. However, recent publications and an FDA report have noted flaws associated with the SAGhE study design [88-90]. In many other long-term evaluations of large groups of patients undergoing hGH therapy, the overall safety profile is favorable [91-95]. No increased risk of death due to leukemia, cancer, or cardiovascular disorders was observed. Table 6 Use of hGH therapy in SGA children in the United States and Europe

It must be mentioned, though not stressed, that some controversy regarding the use of hGH arose in 2011 due to results from a study conducted in France, the Santé Adulte GH Enfant (SAGhE) study. The results from this study indicated that long-term use of hGH in children with short stature could increase a patient’s risk of death [88]. The SAGhE study reported that hGH therapy, when administered to patients at doses above 50 mcg/kg/day, increased the risk of death by 30% as compared to the general population in France. This effect was attributed to an increased likelihood of bone-tumor formation, cardiovascular disease, and cerebrovascular events. These results concerned patients born SGA, as the normal recommended dose of hGH therapy can be approximately 70 mcg/kg/day (Table 6) [1]. However, recent publications and an FDA report have noted flaws associated with the SAGhE study design [88-90]. In many other long-term evaluations of large groups of patients undergoing hGH therapy, the overall safety profile is favorable [91-95]. No increased risk of death due to leukemia, cancer, or cardiovascular disorders was observed. Table 6 Use of hGH therapy in SGA children in the United States and Europe FDA-approved indication in 2001 EMEA-approved indication in 2003 Age at start of treatment (year) 2 4 Height SDS at start Not stated −2.5 SD Growth velocity before treatment No catch-up growth Less than 0 SD for age Reference to midparental height Not stated Height SDS > 1 SD below midparental height SDS Dose (mcg/kg/day) 70 35 FDA, United States Food and Drug Administration; EMEA, European Agency for the Evaluation of Medicinal Products; hGH, human growth hormone; SDS, standard deviation score; SGA, small for gestational age.

age Reference to midparental height Not stated Height SDS > 1 SD below midparental height SDS Dose (mcg/kg/day) 70 35 FDA, United States Food and Drug Administration; EMEA, European Agency for the Evaluation of Medicinal Products; hGH, human growth hormone; SDS, standard deviation score; SGA, small for gestational age. Reprinted with permission from [1]. Conclusions Based on results from more than 20 years of research, numerous genetic causes for SGA births have been realized. Genetic defects in either IGF-1 or IGF-1R that result in SGA size typically correlate with phenotypical features such as microcephaly and mental retardation. The most predictive factors for IGF-1R deletion include small birth size, head size, and stature, as well as high IGF-1 levels, developmental delay, and micrognathia. hGH therapy in patients with mutations in IGF-1 has shown moderate success. Furthermore, for patients with IGF-1R mutations, hGH treatment has been shown to be especially promising, particularly for those with distal deletions of the terminal long arm of chromosome 15. Overall, in studies in which the genotype of SGA patients was not known and hGH therapy was conducted, improvements were observed for most of the patient population, particularly if therapy was begun at a young age.

ially promising, particularly for those with distal deletions of the terminal long arm of chromosome 15. Overall, in studies in which the genotype of SGA patients was not known and hGH therapy was conducted, improvements were observed for most of the patient population, particularly if therapy was begun at a young age. However, despite these positive results, a number of questions regarding the effectiveness of the treatment remain. For instance, hGH therapy for children with SRS has shown positive results, but overall the improvements are often not statistically significant. Furthermore, the differences in SGA patient response to hGH therapy are still only slightly understood. While much of the diversity in response rates to hGH therapy for SGA patients correlates with the type of genetic mutation, the role of additional factors, such as ethnicity, on this treatment still requires significant research. Abbreviations (AGA): Appropriate for gestational age; (AGT): Angiotensinogen; (ALS): Acid-labile subunit; (FDA): Food and Drug Administration; (GH): Growth hormone; (ICR1): Imprinting center region 1; (IGFBP-3): Insulin-like growth factor binding protein-3; (IGF): Insulin-like growth factor; (IGF-1R): Insulin-like growth factor-1 receptor; (mUPD7): Maternal uniparental disomy 7; (hGH): Recombinant human GH; (rhIGF-1): Recombinant human IGF-1; (SAGhE): Santé Adulte GH Enfant study; (SGA): Small for gestational age; (SD): Standard deviation; (SRS): Silver-Russell syndrome; (Ts14): Trisomy for chromosome 14; (UPD): Uniparental disomy; (US): United States.

; (mUPD7): Maternal uniparental disomy 7; (hGH): Recombinant human GH; (rhIGF-1): Recombinant human IGF-1; (SAGhE): Santé Adulte GH Enfant study; (SGA): Small for gestational age; (SD): Standard deviation; (SRS): Silver-Russell syndrome; (Ts14): Trisomy for chromosome 14; (UPD): Uniparental disomy; (US): United States. Competing interests Dr. Saenger reports that he receives grant support from Novo Nordisk Inc., and that he is a consultant for LG and Biopartners. Dr. Reiter reports that he has received payment from Novo Nordisk Inc. for board membership; from Abbott Pharmaceuticals as a consultant; from Quintiles for development of educational presentations, and from various pharmaceutical companies for lectures, including service on speakers bureaus. Authors’ contributions The authors contributed equally to this work and were involved in the development of its concept, outline, and narrative. At all stages, the authors discussed the data presented and commented on the manuscript. Both authors read and approved the final manuscript. Acknowledgments The authors would like to thank Meredith A. Mintzer, PhD, and Emma Hitt, PhD, of MedVal Scientific Information Services, LLC, for providing medical writing and editorial assistance. This manuscript was prepared according to the International Society for Medical Publication Professionals’ Good Publication Practice for Communicating Company-Sponsored Medical Research: The GPP2 Guidelines. Funding to support the preparation of this manuscript was provided by Novo Nordisk Inc.

Introduction POLYCYSTIC OVARY SYNDROME (PCOS) is an endocrine-metabolic disorder, which is highly prevalent (5–10%) in reproductive-age women [1-3]. PCOS is characterized by hyperandrogenism [4,5]. The phenotypic characterization is heterogeneous; PCOS can manifest in the prepubertal years as premature pubarche [6]; hirsutism, acne and anovulatory cycles may remain clinically silent until late adolescence. Peripheral insulin resistance plays a key role in the pathogenesis of this syndrome. It has been suggested that insulin excess facilitates ovarian and/or adrenal hyperandrogenism [7-10]. Insulin resistance in PCOS women has long-term health consequences, predisposing to type 2 diabetes mellitus, cardiovascular disease and pregnancy-associated disorders like infertility, miscarriage, premature labor and gestational diabetes [11-15]. Female first-degree relatives (FDRs) of PCOS-affected women are at higher risk for developing PCOS symptoms [16]. FDRs are also at higher risk of developing the endocrine and metabolic co-morbidities of PCOS, such as obesity, insulin resistance and impaired insulin sensitivity, hyperlipidemia and metabolic syndrome[17-21]. An abundant literature supports familial clustering of hyperandrogenemia in PCOS women, consistent with a genetic contribution to the disease [16,22]. Recent studies have shown an increased prevalence of hyperandrogenism and insulin resistance in adult FDRs of PCOS [16].

y, hyperlipidemia and metabolic syndrome[17-21]. An abundant literature supports familial clustering of hyperandrogenemia in PCOS women, consistent with a genetic contribution to the disease [16,22]. Recent studies have shown an increased prevalence of hyperandrogenism and insulin resistance in adult FDRs of PCOS [16]. Despite the alarming increase in the prevalence of type 2 diabetes in children, and efforts to identify risk factors for the development of this disease in children studies of glucose homeostasis in pediatric FDRs of PCOS, have not been performed. In the present study we assessed the insulin secretion (β-cell function), insulin sensitivity, adrenal and ovarian steroid levels in peri-pubertal daughters and sisters of women diagnosed with PCOS. Our purpose was to determine the presence of early biochemical changes in females at risk of PCOS before clinical manifestations occurred and to evaluate whether insulin resistance or hyperandrogenemia occurs before in FDRs of PCOS women. We performed IVGTT to more comprehensively study the insulin dynamics in this cohort, and compared the results with those of an age- and weight-matched group of controls who were daughters or sisters of women without PCOS. The short version of IVGTT provides enough testing points to calculate indexes to assess insulin sensitivity and β-cell function. We also compared the steroid levels in both groups to evaluate the existence of the hormonal imbalance in FDRs. We hypothesized that being an FDR of a PCOS conveys an independent risk for the development of type 2 diabetes independent of other biochemical or clinical evidence of PCOS.

n sensitivity and β-cell function. We also compared the steroid levels in both groups to evaluate the existence of the hormonal imbalance in FDRs. We hypothesized that being an FDR of a PCOS conveys an independent risk for the development of type 2 diabetes independent of other biochemical or clinical evidence of PCOS. Subjects and methods Study subjects We studied 18 healthy premenarchal girls (mean age of 11.6 ± 1.1 years, range 8–14) whose mothers or sisters had been diagnosed with PCOS and were followed in adult and pediatric endocrinology clinics at Maimonides Medical Center, University of Sherbooke, University of Chile, Vanderbilt University and Cedars-Sinai Medical Center. All had been diagnosed with PCOS based upon the National Institute of Child Health and Human Development criteria for PCOS [23], including history of documented chronic oligomenorrhea or amenorrhea and hyperandrogenism, with the exclusion of secondary causes. The anthropometric, insulin dynamics and steroid data of FDR subjects was determined for NIH K23 HD040325, “Insulin Resistance in Adolescents at High Risk for Polycystic Ovary Syndrome".

], including history of documented chronic oligomenorrhea or amenorrhea and hyperandrogenism, with the exclusion of secondary causes. The anthropometric, insulin dynamics and steroid data of FDR subjects was determined for NIH K23 HD040325, “Insulin Resistance in Adolescents at High Risk for Polycystic Ovary Syndrome". The control group consisted of 21 healthy premenarchal girls (mean age of 12.1 ± 0.4 years, range 8–14) whose mothers had no history of irregular menstrual cycles or hirsutism. Children were recruited from a school-based study that is part of Reduce Obesity and Diabetes (ROAD) Project, a collaborative project between Maimonides Medical Center and Columbia University Medical Center, Cohen Children’s Medical Center, Mount Sinai Medical Center, Winthrop University Hospital and New York City public schools. This ongoing project is supported by Academy for Medical Development and Collaboration (AMDeC). These children had no personal history of diabetes or family history of irregular menstruation, diabetes or hirsutism.

edical Center, Mount Sinai Medical Center, Winthrop University Hospital and New York City public schools. This ongoing project is supported by Academy for Medical Development and Collaboration (AMDeC). These children had no personal history of diabetes or family history of irregular menstruation, diabetes or hirsutism. First-degree relatives of PCOS women were recruited and tested following approval from the Institutional Review Boards at Cedars-Sinai Medical Center (Los Angeles), Sherbrooke University (Sherbrooke, Canada), University of Chile (Santiago, Chile), Vanderbilt University (Nashville, TN), and Maimonides Medical Center (Brooklyn, New York). The ROAD Project study was approved by the Institutional Review Board for each participating hospital, school boards and the Department of Health and Education. Written informed consent was obtained from all parents and assent from the peri-adolescent study subjects.

and Maimonides Medical Center (Brooklyn, New York). The ROAD Project study was approved by the Institutional Review Board for each participating hospital, school boards and the Department of Health and Education. Written informed consent was obtained from all parents and assent from the peri-adolescent study subjects. Study protocol Assessment All study subjects were either tested at Maimonides Medical Center Clinic (MMCC), ROAD school based study or at Cedars-Sinai Medical Center, Vanderbilt University, University of Sherbrooke or University of Chile. Enrollees presented after an overnight fast. A comprehensive medical assessment and medical history were obtained, including a personal and family history of type 2 diabetes, metabolic syndrome or heart disease. Anthropometric measurements included weight, height, body mass index (BMI), and waist circumference (waist: midway between the lower rib margin and the iliac crest). Age- and sex-specific BMI z-scores were calculated by using National Center for Health Statistics (NCHS) data [24]. Body fat composition was determined by bioelectrical impedance (Body Fat Analyzer, Model HBF-306, Omron, Gays Mills, WI). A Tanner stage was assigned to each study subject based on levels of both DHEAS and E2. No participants in the study were taking medications known to affect either sex steroids or carbohydrate metabolism.

at composition was determined by bioelectrical impedance (Body Fat Analyzer, Model HBF-306, Omron, Gays Mills, WI). A Tanner stage was assigned to each study subject based on levels of both DHEAS and E2. No participants in the study were taking medications known to affect either sex steroids or carbohydrate metabolism. IVGTT Rapid frequently sampled intravenous glucose tolerance testing (IVGTT) was performed on all FDRs [25]. Each subject was given IV 25% Dextrose at 2 ml/Kg (max 25gm of Dextrose) delivered over 1 min, and blood was drawn through the same butterfly needle for measurements of serum glucose and insulin at 2, 3, 4 and 5 min after glucose administration. Short version of IVGTT (5 point over 5 minutes versus 3 hr classic IVGTT) had been chosen in control group because it was less time consuming in the school-based study. Hormonal assay Additional baseline blood samples were obtained for determination of DHEAS and Estradiol (E2). Serum glucose was determined by the glucose hexokinase procedure from Raichem. The intra- and interassay coefficient of variation of this method was less than 3%. Serum insulin was determined by a solid phase sandwich immunoassay developed by Wallace/Perkin-Elmer. The intra-assay coefficient of variation of this method is 3-6%. Serum assays for DHEAS and E2 were performed by Labcorp Institute (Burlington, NC). Assay sensitivities for DHEAS and E2 were 1.7 ng/dl, 10 pg/ml, respectively. Intra- and interassay coefficients of variation were 7.9-9.8% for DHEAS and 97% for E2.

Elmer. The intra-assay coefficient of variation of this method is 3-6%. Serum assays for DHEAS and E2 were performed by Labcorp Institute (Burlington, NC). Assay sensitivities for DHEAS and E2 were 1.7 ng/dl, 10 pg/ml, respectively. Intra- and interassay coefficients of variation were 7.9-9.8% for DHEAS and 97% for E2. Calculations Insulin resistance was estimated by the Homeostasis Model Assessment for Insulin Resistance (HOMA-IR), calculated as [fasting insulin (μU/ml) x fasting glucose (mg/dl)]/405] [26] and insulin sensitivity was estimated by Quantitative Insulin Sensitivity Check Index (QUICKI), calculated as [1/(log fasting insulin μU/ml) + log (fasting glucose mg/dl)] [27]. The acute insulin response (AIRg) was calculated as mean incremental rise in plasma insulin at 3 and 5 min after IV glucose. Pancreatic β-cell function was assessed by calculating the Glucose Disposal index as [log10 (AIRg x fasting glucose concentration/fasting insulin concentration)] [25,28]. Statistical methods and analysis Data are expressed as mean ± SD. Freidman’s repeated measures ANOVA was used to compare variables (age, BMI Z-score, HOMA, QUICKI, GDI, DHEAS, etc.) within the same group. Comparisons of means between the PCOS affected first-degree relatives group (PCOS-FDR) and the control group were performed using the Unequal Variance, Unequal Sample Size t-test. Regression analysis and Spearman correlations were used to evaluate the relationship between the variables of interest. Statistical analysis was performed using SPSS Statistics 17.0.

cted first-degree relatives group (PCOS-FDR) and the control group were performed using the Unequal Variance, Unequal Sample Size t-test. Regression analysis and Spearman correlations were used to evaluate the relationship between the variables of interest. Statistical analysis was performed using SPSS Statistics 17.0. Results Anthropometry and hormonal assay There were no significant differences between the PCOS-FDR and the control groups with respect to age, BMI, BMI Z score, waist circumference or percent of body fat (Table 1). There were no significant differences in serum DHEA-S between groups. E2 levels were significantly higher in PCOS-FDR group comparing to the control group. Both groups had similar Tanner stage distributions, as a function of DHEA-S and E2 levels. Table 1 Anthropometric characteristic of PCOS FDR and Control Daughters (Cd) FDR-PCOS (n = 18) Cd group (n = 21) Age (yr) 11.6 ± 1.4 12. ± 0.8 BMI (kg/m2) 21.5 ± 3.5 21.5 ± 3.2 BMI Z Score 1.01 ±0.85 0.93 ± 0.9 Waist Circumference (cm) 72 ± 10.9 76 ±8.7 Body Fat % 25 ± 2.5 25.76 ± 3.2 Data are presented as the mean ± SD. *P value <0.05. IVGTT The HOMA-IR ratio was significantly higher, and both the QUICKI and GDI parameters were significantly lower in the PCOS-FDR group compared to the control group (Table 2). The HOMA-IR ratio findings were unchanged even after segregating the PCOS-FDR and control groups according to Tanner stage. AIR was not significantly different between groups. Table 2 Biochemical characteristics in FDR-PCOS and Control Daughters (Cd) groups

IVGTT The HOMA-IR ratio was significantly higher, and both the QUICKI and GDI parameters were significantly lower in the PCOS-FDR group compared to the control group (Table 2). The HOMA-IR ratio findings were unchanged even after segregating the PCOS-FDR and control groups according to Tanner stage. AIR was not significantly different between groups. Table 2 Biochemical characteristics in FDR-PCOS and Control Daughters (Cd) groups FDR-PCOS Group (n = 18) Cd Group (n = 21) DHEA-S 63 (65) 99.5 (71) AIRg 80 (54) 97 (43) QUICKI 0.32 (0.03) * 0.35(0.02) HOMA-IR 3.45 (1.7)* 2.04 (1.6) GDI 2.6 (0.46) * 2.98 (0.27) Data are presented as the mean ± SD. *P value <0.05. Discussion The major findings of this study are that both decreased insulin sensitivity and beta-cell function are evident in premenarachal peripubertal female FDR’s of PCOS without clinical or biochemical evidence of PCOS. These data suggest that having a first-degree relative with PCOS may be an independent risk factor for the development of type 2 diabetes in childhood. The implication is that FDR’s of PCOS should potentially be screened more aggressively for pre-diabetic risk factors, including obesity, hypertriglyceridemia, and a pro-inflammatory state and should be considered an at-risk group in terms of efforts to prevent the development of other risk factors. Finally, the detection of impaired glucose homeostasis prior to the onset of hyperandrogenism is in agreement with the hypothesis that hyperinsulinism is a cause, rather than the result, of PCOS.

matory state and should be considered an at-risk group in terms of efforts to prevent the development of other risk factors. Finally, the detection of impaired glucose homeostasis prior to the onset of hyperandrogenism is in agreement with the hypothesis that hyperinsulinism is a cause, rather than the result, of PCOS. PCOS is likely the cumulative product of a number of genetic, epigenetic, environmental factors and/or familial habits [29]. PCOS may be inherited in an autosomal dominant or X-linked dominant pattern [30-32]. Genome-wide genetic and linkage studies have found associations with PCOS for many genes including fibrillin-3, PPAR-γ and IL-6, though replication has proven elusive [33-36] and development of the characteristic syndrome may occur in the absence of known mutation. In complex, heterogeneous conditions with variable presentation such as PCOS, studies of first-degree relatives of affected females may help to separate the biochemical contributions from genetic and habitual influence. Hyperandrogenism is the consistent finding in prior studies of the adult relatives of PCOS women [22,37-39] but the data on hyperandrogenism in peripubertal FDR is scant. One study observed higher androgen levels in the later stages of puberty (Tanner 4 and 5) in PCOS-FDR subjects compared to control daughters [40], likely an expression of the normal maturation of the hypothalamic-pituitary gonadal axis.

COS women [22,37-39] but the data on hyperandrogenism in peripubertal FDR is scant. One study observed higher androgen levels in the later stages of puberty (Tanner 4 and 5) in PCOS-FDR subjects compared to control daughters [40], likely an expression of the normal maturation of the hypothalamic-pituitary gonadal axis. Many have proposed that hyperinsulinism is the fundamental pathophysiological event leading to ovarian and/or adrenal production of excessive androgen [41]. Hyperandrogenism is the primary feature in the emergence of PCOS [22,42]. In our study, significant differences in the androgen precursor DHEA-S between the two groups were not detected. Conversely, beta-cell function was impaired not only in affected PCOS probands but also in their first-degree relatives, regardless of whether PCOS or other metabolic abnormalities were yet manifest. Prior reports of insulin resistance and glucose insensitivity in first-degree relatives utilized manipulation of the less informative oral glucose tolerance test. A recent study demonstrated hyperinsulinemia and increased ovarian volumes present in PCOS daughters even prior to the onset of puberty, the hyperinsulinemia persisting throughout pubertal development. Other biochemical abnormalities of PCOS emerge only in later puberty [40], suggesting that metabolic disturbances are fundamental to establishment of permanent states of androgen excess.

olumes present in PCOS daughters even prior to the onset of puberty, the hyperinsulinemia persisting throughout pubertal development. Other biochemical abnormalities of PCOS emerge only in later puberty [40], suggesting that metabolic disturbances are fundamental to establishment of permanent states of androgen excess. Frequently-sampled IVGTT is a well-validated method of estimating insulin sensitivity and considered superior to OGTT-derived measures of insulin dynamics [43-45] as it allows determination of the glucose disposal index (GDI), a highly sensitive reflection of the capacity of pancreatic islets to compensate for lower insulin sensitivity [46]. We utilized IVGTT to assess both glucose tolerance and insulin resistance as well as the acute insulin response and glucose disposition indices that better define the beta-cell function.

sal index (GDI), a highly sensitive reflection of the capacity of pancreatic islets to compensate for lower insulin sensitivity [46]. We utilized IVGTT to assess both glucose tolerance and insulin resistance as well as the acute insulin response and glucose disposition indices that better define the beta-cell function. Our data provides evidence of early development of insulin resistance in the peripubertal first-degree female relatives of women with PCOS. All three measures, QUICKI, HOMA-IR and GDI, demonstrated lower insulin sensitivity among PCOS first-degree relatives versus weight, Tanner, age-matched controls without family history of PCOS, diabetes mellitus and hypertension. Perturbed beta-cell dysfunction and the resulting inadequate compensation for deteriorating insulin sensitivity has been demonstrated in the daughters of PCOS-affected women prior to puberty and independently of body weight. Our findings suggest that peripubertal insulin resistance (IR) even prior to biochemical or clinically apparent androgen excess may also be an early hallmark of risk for PCOS in the genetically vulnerable peri-adolescent population as well. This emphasizes the necessity of early and ongoing biochemical monitoring of relatives of women with PCOS, affording the opportunity for both earlier diagnosis and therapeutic intervention to prevent the long-term morbidity inherent in this disorder.

COS in the genetically vulnerable peri-adolescent population as well. This emphasizes the necessity of early and ongoing biochemical monitoring of relatives of women with PCOS, affording the opportunity for both earlier diagnosis and therapeutic intervention to prevent the long-term morbidity inherent in this disorder. Abbreviations PCOS: Polycystic ovarian syndrome; AIR: Acute insulin response; GDI: Glucose disposal index; QUICKI: Quantitative insulin sensitivity check index; HOMA: Homeostatic model assessment; BMI: Body mass index; IVGTT: IV glucose tolerance test; OGTT: Oral glucose tolerance test; DHEAS: Dehydroepiandrosterone sulfate. Competing interests NR, AK, RP, SS, SB, AB, SM, RM, J-PB, MIH, MR, ST and DG have no competing interests. Authors’ contributions DG participated in the planning and execution of the protocols performed on study subjects, as well as the preparation of the manuscript. NR, AK, RP, SS, SB, AB, SM, RM, J-PB, MIH, MR, ST participated in the execution of the protocols performed on study subjects, as well as the preparation of the manuscript. All authors read and approved the final manuscript.

tion of the protocols performed on study subjects, as well as the preparation of the manuscript. NR, AK, RP, SS, SB, AB, SM, RM, J-PB, MIH, MR, ST participated in the execution of the protocols performed on study subjects, as well as the preparation of the manuscript. All authors read and approved the final manuscript. Acknowledgments We acknowledge the Reduce Obesity and Diabetes (ROAD) project under aegis of Academy for Medical Development and Collaboration (AMDeC). We would like to gratefully acknowledge the invaluable participation of all the students, teachers, and school administrators, as well as the NYC Board of Health and Department of Education. Funding was provided through AMDeC (Academy for Medical Development and Collaboration) and NIH grant # 1 UL1 RR024156-01. The study was supported by funding from K23 HD40325 “Insulin Resistance in Adolescents at High Risk for Polycystic Ovary Syndrome” (PI: David Geller) and M01-RR000425 (Cedars-Sinai General Clinical Research Center Grant from the NCRR).

Correction Three errors have been noted by the author in an article published in this journal in 2010 [1], all three in paragraphs three and four under the heading: “4. Effects of psychotropic drugs on prolactin secretion”: 1. Reference 19 in the original article [1] (Frazier JA, et al. Risperidone treatment for juvenile bipolar disorder: a retrospective chart review. J Am Acad Child Adolesc Psychiatry. 1999;38:960–5) was quoted as indicating that one of the 9 of 11 outpatients aged 4–17 years treated with risperidone who developed hyperprolactinemia had amenorrhea and one had gynecomastia. In fact, the only hyperprolactinemia associated complication reported was the one case of amenorrhea; there were no instances of gynecomastia. 2. Reference 23 in the original article [1] (Bunker MT, Marken PA, Schneiderhan ME, Ruehter VL. Attenuation of antipsychotic induced hyperprolactinemia with clozapine. J Am Acad Child Adolesc Psychiatry 1997; 7:65–9) was mistakenly described as reporting a single patient treated with risperidone who developed galactorrhea and had resolution of the problem when switched to clozapine. In fact, the subject had been treated with thioridazine (haloperidol).

erprolactinemia with clozapine. J Am Acad Child Adolesc Psychiatry 1997; 7:65–9) was mistakenly described as reporting a single patient treated with risperidone who developed galactorrhea and had resolution of the problem when switched to clozapine. In fact, the subject had been treated with thioridazine (haloperidol). 3. The author described reference 24 of the original article [1] (Findling et al. Prolactin levels during long-term risperidone treatment in children and adolescents. J Clin Psychiatry 2003; 64:1362–9), as a retrospective study of prolactin levels and hyperprolactinemia attributable side effects from 5 clinical trials involving 592 children and adolescents aged 5–15 years that yielded a 2.2% frequency of side effects of gynecomastia, amenorrhea, or galactorrhea, quoting the published report. However, the frequency noted in the article was based on Findling et al. erroneously using the denominator for the entire 5-15-year-old population (n = 592) rather than the subpopulation that excluded boys over 10 years of age that they separately analyzed, in order to avoid confounding by adolescent gynecomastia (n = 360). It was within this smaller population that the 13 adverse events occurred. The correct frequency for hyperprolactinemia attributable side effects in these studies, therefore, should be 3.6% for this subpopulation and for the entire 5-15-year-old population in which 30 hyperprolactinemia attributable side effects occurred, 5%.

within this smaller population that the 13 adverse events occurred. The correct frequency for hyperprolactinemia attributable side effects in these studies, therefore, should be 3.6% for this subpopulation and for the entire 5-15-year-old population in which 30 hyperprolactinemia attributable side effects occurred, 5%. The first error, though regrettable, removes a single case that does not affect the conclusion that, “First-generation antipsychotics, particularly haloperidol, and the second-generation antipsychotic drugs, most prominently risperidone, appear to be associated with greatest risk for hyperprolactinemia: some treated individuals developing hyperprolactinemia will have galactorrhea, amenorrhea, or gynecomastia. [1]” The second error, in which haloperidol rather than risperidone was implicated, remains consistent with the conclusion. Correction of the more substantial and significant third error serves to reinforce this conclusion.

Background The primary goal of diabetes care for children and adolescents is to achieve an optimal metabolic control to prevent or to minimize the risk of acute (e.g. hypoglycemia) and long-term complications such as retinopathy, nephropathy and neuropathy [1,2]. The recommended everyday treatment regimen for a patient with Type 1 Diabetes Mellitus (T1DM) is complex and demanding. Parents or other adult care takers initially play a key role in this intensive care system. But as children grow older responsibility for taking care of their chronic condition is placed upon them. During adolescence deteriorations in diabetes management and control are common [3]. These deteriorations raise the risk of acute or long-term complications and are also associated with higher health care costs. It is known that an optimal self-care behavior, independently of age, impacts positively on glycemic control [4]. Therefore, professionals aim to help adolescent patients and their families to become experts in self-management of their disease. A recently published systematic review investigated demographic and inter- or intrapersonal factors associated with metabolic control and self-care in adolescent patients with T1DM [4]. This revealed that adolescence is associated with both, decreased self-care and deterioration in metabolic control. Factors like a lower socioeconomic status, lower parental responsibility for, and involvement in diabetes-focused daily tasks, higher peer orientation or also intrapersonal characteristics like low conscientiousness and low emotional stability were associated with lower self-care and higher HbA1c values.

bolic control. Factors like a lower socioeconomic status, lower parental responsibility for, and involvement in diabetes-focused daily tasks, higher peer orientation or also intrapersonal characteristics like low conscientiousness and low emotional stability were associated with lower self-care and higher HbA1c values. Self-care of diabetes in daily routine involves insulin administration, decisions around food – choices and intake, physical activity, timing of glucose measurements and analysis as well as response to the results. This calls for well-organized treatment instructions and continuous coaching by a multidisciplinary team but also for patients cognitive and executive skills. In recent years cognitive and executive functioning in T1DM gained attention in the literature [5,6]. These studies essentially showed only mild differences between the neurocognitive performance of children and adolescents with T1DM when compared to controls. In a meta-analysis of the literature in 2008 only a mildly reduced intellectual quotient was found in children with diabetes [5]. The largest effects, but still within a very small range, were on visuospatial ability, motor speed and writing, and on sustained attention and reading. Most of these investigations focused on cognition.

lysis of the literature in 2008 only a mildly reduced intellectual quotient was found in children with diabetes [5]. The largest effects, but still within a very small range, were on visuospatial ability, motor speed and writing, and on sustained attention and reading. Most of these investigations focused on cognition. Overall, there is a body of knowledge about cognitive and executive functioning in T1DM and also of factors associated with self-care, adherence to therapy and metabolic control. By contrast, there is very limited knowledge about the T1DM patients’ capacity of self-assessment which obviously is a prerequisite for good self-care. Characteristics of self-assessment for example are self-perception of HbA1c value, patient’s memory of the HbA1c value, knowledge on target HbA1c or patients’ suggestions on how to improve metabolic control. We found only limited literature concerning the role that recall plays in diabetic management. Only very recently a study investigated the prospective recall and glycemic control in children with T1DM [7]. No clear association between glycemic control and memory was found. Similarly, no literature is available for the difference between patients’ and professionals’ assessment of metabolic control. Our daily experience suggests that patients’ self-assessment of the actual glycemic control depends primarily on the perception of their own diabetes management at home, including daily blood glucose self-monitoring, insulin applications and diet, whereas professionals’ assessment depends mainly on measured HbA1c levels and blood glucose measurements from home devices.

ssessment of the actual glycemic control depends primarily on the perception of their own diabetes management at home, including daily blood glucose self-monitoring, insulin applications and diet, whereas professionals’ assessment depends mainly on measured HbA1c levels and blood glucose measurements from home devices. Therefore, the aim of our study was to test the hypothesis if there was a difference between patients’ self-perception and an objective assessment (HbA1c) of metabolic control in T1DM children and adolescents; and to investigate factors that may be involved. Methods Patients and study design We performed a cross-sectional, retrospective and prospective cohort study. We recruited patients with T1DM, seen at the outpatient clinic of the University Children’s Hospital in Bern between April and September 2011. Inclusion criteria were an age between 8 – 18 years, diagnosis of T1DM for ≥ 12 months, at least 3 regular consultations in our department during the past 12 months and informed consent. Exclusion criteria were a change in the modality of insulin therapy in the past 12 months, less than 3 regular consultations in our department over the past 12 months, other chronic illnesses influencing the metabolic control of T1DM (such as malignancy or neuromuscular disease) and other types of diabetes. The study fulfilled the criteria of the Declaration of Helsinki and was approved by the cantonal ethics committee of Bern, Switzerland. Participating patients and caregivers were informed about the study and gave their written consent.

l of T1DM (such as malignancy or neuromuscular disease) and other types of diabetes. The study fulfilled the criteria of the Declaration of Helsinki and was approved by the cantonal ethics committee of Bern, Switzerland. Participating patients and caregivers were informed about the study and gave their written consent. A total of 91 children (53 boys and 38 girls) were included in the study. 39 T1DM patients between 8 – 18 years did not participate for the following reasons: 3 refused to participate, 33 did not fulfill the inclusion criteria and 3 did not provide full information on the questionnaires. Details on patient characteristics are summarized in Table 1. Table 1 Patient characteristics

A total of 91 children (53 boys and 38 girls) were included in the study. 39 T1DM patients between 8 – 18 years did not participate for the following reasons: 3 refused to participate, 33 did not fulfill the inclusion criteria and 3 did not provide full information on the questionnaires. Details on patient characteristics are summarized in Table 1. Table 1 Patient characteristics Number of patients (n) All 91 Male 53 Female 38 Mean Range HbA1c (%) All 8.03 6.1 - 10.9 Male 7.99 6.3 - 10.5 Female 8.09 6.1 - 10.9 Age (years)   13.22 8.23 - 17.81 Duration of T1DM (years) All 6.13 1.05 - 15.77 Body mass index (SDS) All 0.06 -2.61 - 1.98 Male -0.10 -2.61 - 1.93 Female 0.28 -1.65 - 1.98 n % Modality of therapy Conventional insulin therapy 9 9.9 Functional insulin therapy 59 64.8 Insulin pump 23 25.3 Parental socioeconomic level Low 8 8.8 Moderate 64 70.3 High 17 18.7 Not determined 2 2.2 Data collection Clinical and demographic data such as age, duration of disease, modality of insulin therapy and HbA1c of the last consultation were collected from patients’ clinical records. Height and weight were measured during the visit at the outpatient clinic. Standard deviation score of the Body Mass Index (BMI) was calculated according to the LMS model taking the Kromeyer-Hauschild percentiles as a reference [8]. Data concerning self-monitored blood glucose levels were taken from memory functions of personal glucometers. Average values per day were calculated over the past 2–4 weeks.

Number of patients (n) All 91 Male 53 Female 38 Mean Range HbA1c (%) All 8.03 6.1 - 10.9 Male 7.99 6.3 - 10.5 Female 8.09 6.1 - 10.9 Age (years)   13.22 8.23 - 17.81 Duration of T1DM (years) All 6.13 1.05 - 15.77 Body mass index (SDS) All 0.06 -2.61 - 1.98 Male -0.10 -2.61 - 1.93 Female 0.28 -1.65 - 1.98 n % Modality of therapy Conventional insulin therapy 9 9.9 Functional insulin therapy 59 64.8 Insulin pump 23 25.3 Parental socioeconomic level Low 8 8.8 Moderate 64 70.3 High 17 18.7 Not determined 2 2.2 Data collection Clinical and demographic data such as age, duration of disease, modality of insulin therapy and HbA1c of the last consultation were collected from patients’ clinical records. Height and weight were measured during the visit at the outpatient clinic. Standard deviation score of the Body Mass Index (BMI) was calculated according to the LMS model taking the Kromeyer-Hauschild percentiles as a reference [8]. Data concerning self-monitored blood glucose levels were taken from memory functions of personal glucometers. Average values per day were calculated over the past 2–4 weeks. All other information was collected with the help of three specific questionnaires: One for the professional, one for the care taker and one for the patient. Patients were requested to fill in the questionnaires without the help of their care takers.

Data concerning self-monitored blood glucose levels were taken from memory functions of personal glucometers. Average values per day were calculated over the past 2–4 weeks. All other information was collected with the help of three specific questionnaires: One for the professional, one for the care taker and one for the patient. Patients were requested to fill in the questionnaires without the help of their care takers. A classification of the socioeconomic level was performed based on the self-declared educational level and occupational status of both parents as published elsewhere [9]. In brief, the classification “low” consisted of public school without professional training; the “intermediate” level included secondary school with completed professional training, and a “high” level was defined as having completed academic studies at a university. HbA1c was determined by the Latex-Immunagglutination method (DCA 2000 Analyzer, Bayer Corporation, Elkart, IN 46514 USA). For this assay, reference values for healthy, non-diabetic individuals range between 4.0 - 5.6%.

A classification of the socioeconomic level was performed based on the self-declared educational level and occupational status of both parents as published elsewhere [9]. In brief, the classification “low” consisted of public school without professional training; the “intermediate” level included secondary school with completed professional training, and a “high” level was defined as having completed academic studies at a university. HbA1c was determined by the Latex-Immunagglutination method (DCA 2000 Analyzer, Bayer Corporation, Elkart, IN 46514 USA). For this assay, reference values for healthy, non-diabetic individuals range between 4.0 - 5.6%. Self assessment score (SAS) We created a questionnaire and a scoring system to evaluate the quality of the self-assessment of patients’ metabolic control. Patients were asked by questionnaire whether they felt that the actual HbA1c might be better, equal or worse than the HbA1c measured 3 months ago. Better or worse were defined as a difference in HbA1c ≥ +/- 0.5%. Data were analyzed and categorized as follows. SAS 0 meant, that patient’s perception overlapped with the objective result. SAS +1 or +2 meant, that the measured HbA1c value showed an improvement which the patient did not perceive (e.g. the patient meant that the actual HbA1c was worse than the last HbA1c, but in fact it was equal (+1) or better (+2). SAS -1 or -2 meant, that the measured HbA1c value showed a worsening of the metabolic control which the patient did not realize (e.g. the patient meant that the actual HbA1c was equal or better than the last HbA1c, but in fact it was worse (-1 to -2).

e than the last HbA1c, but in fact it was equal (+1) or better (+2). SAS -1 or -2 meant, that the measured HbA1c value showed a worsening of the metabolic control which the patient did not realize (e.g. the patient meant that the actual HbA1c was equal or better than the last HbA1c, but in fact it was worse (-1 to -2). Data analysis Data were analyzed using SPSS 19.0 (IBM® SPSS® Statistics 19). For group comparison the Kruskal-Wallis test was used. A p-value < 0.05 was considered to indicate statistical significance. Most data are shown as boxplots with the top of the box representing the 75th percentile, the bottom of the box representing the 25th percentile, and the line in the middle representing the 50th percentile. The whiskers represent the highest and lowest values, that were not outliers or extreme values. Results Patient characteristics and metabolic control in the study cohort Mean HbA1c of the 91 studied T1DM patients was 8.03% (range: 6.1 – 10.9%) (Table 1). In boys the mean HbA1c was 7.99%, in girls 8.09%. Mean duration of T1DM (time since the initial diagnosis) was about 6 years. Two thirds of the patients were treated with a functional insulin therapy using multiple daily injections, 25% of the patients with an insulin pump, and 10% were on a conventional 2–3 insulin injection regimen with fixed meals. Two thirds of the patients had care givers classified as having a moderate level of socioeconomic status, 17% had a high and 8% a low level.

Results Patient characteristics and metabolic control in the study cohort Mean HbA1c of the 91 studied T1DM patients was 8.03% (range: 6.1 – 10.9%) (Table 1). In boys the mean HbA1c was 7.99%, in girls 8.09%. Mean duration of T1DM (time since the initial diagnosis) was about 6 years. Two thirds of the patients were treated with a functional insulin therapy using multiple daily injections, 25% of the patients with an insulin pump, and 10% were on a conventional 2–3 insulin injection regimen with fixed meals. Two thirds of the patients had care givers classified as having a moderate level of socioeconomic status, 17% had a high and 8% a low level. Figure 1 shows the HbA1c values in the study cohort in relation to age, duration of diabetes, glucose self-monitoring and socioeconomic level. We found significant correlations between HbA1c values and the duration of diabetes, with higher HbA1c values in patients with diabetes for > 2 years. Similarly, HbA1c values were significantly higher in the lowest socioeconomic group as compared to the moderate and high socioeconomic group (Figure 1D). Finally, we observed a trend towards higher HbA1c values in older patients (p = 0.065).

of diabetes, with higher HbA1c values in patients with diabetes for > 2 years. Similarly, HbA1c values were significantly higher in the lowest socioeconomic group as compared to the moderate and high socioeconomic group (Figure 1D). Finally, we observed a trend towards higher HbA1c values in older patients (p = 0.065). Figure 1 HbA1c in relation to (A) age, (B) duration of diabetes, (C) glucose self-monitoring and (D) socioeconomic level. There is a tendency towards higher HbA1c values with age (p = 0.065). HbA1c values correlate with the duration of diabetes (p = 0.025). HbA1c does not correlate with the number of blood glucose self-measurements (p = 0.173) but correlates with the socioeconomic level (p = 0.032). Data are given as boxplots and were statistically analyzed by Kruskal-Wallis tests with a significance level of p ≤ 0.05.

correlate with the duration of diabetes (p = 0.025). HbA1c does not correlate with the number of blood glucose self-measurements (p = 0.173) but correlates with the socioeconomic level (p = 0.032). Data are given as boxplots and were statistically analyzed by Kruskal-Wallis tests with a significance level of p ≤ 0.05. Memory of the HbA1c measured at the last consultation To investigate the impact of regular consultations with diabetes professionals at our center, the memorized HbA1c of the last visit was studied. Recollection of the HbA1c measured during the former visit 3–4 months ago was assessed by questionnaire and compared with the HbA1c value from the laboratory. The difference between the recalled and the measured HbA1c were then compared to the actual HbA1c, age, frequency of blood glucose self-monitoring, duration of diabetes and socioeconomic level. We found that patients with HbA1c values > 8.5% tended to have a poorer recollection of their last HbA1c than better controlled subjects (p = 0.069) (Figure 2). By contrast no relationship was found between the anamnestic HbA1c and age, frequency of glucose self-monitoring, duration of diabetes and socioeconomic level (data not shown).

ients with HbA1c values > 8.5% tended to have a poorer recollection of their last HbA1c than better controlled subjects (p = 0.069) (Figure 2). By contrast no relationship was found between the anamnestic HbA1c and age, frequency of glucose self-monitoring, duration of diabetes and socioeconomic level (data not shown). Figure 2 Memory of last HbA1c. The recollection of the last measured HbA1c values was assessed by comparing the objective HbA1c values 3 months ago with the patient’s recollection of this HbA1c. The gap between the last measured and remembered HbA1c value was then compared to the actual HbA1c. Data are shown as boxplots with the actual HbA1c in categorized form on the x-axis. Note that there is a tendency towards wrong positive memory of the last HbA1c in patients having an HbA1c > 8.5% (p = 0.069). Data were analyzed by the Kruskal-Wallis test with a significance level of p ≤ 0.05.

ared to the actual HbA1c. Data are shown as boxplots with the actual HbA1c in categorized form on the x-axis. Note that there is a tendency towards wrong positive memory of the last HbA1c in patients having an HbA1c > 8.5% (p = 0.069). Data were analyzed by the Kruskal-Wallis test with a significance level of p ≤ 0.05. Knowledge of target HbA1c Quality of metabolic control in diabetic patients is followed by regular HbA1c measurements. Internationally a target HbA1c of < 7.5% is recommended for all age groups [10]. This basic information on diabetes is conveyed to our patients and parents/caregivers by our team during initial instructions and is part of the communication during every follow-up visit. Therefore, we asked our patients for their target HbA1c and then correlated this value with their measured HbA1c value at the time of the visit, age, blood glucose self-monitoring, duration of diabetes and socioeconomic level (Figure 3). Overall, we found no relationship between knowledge of target HbA1c and measured HbA1c (Figure 3A). By contrast, older patients indicated higher target values of HbA1c than younger patients (p = 0.017) (Figure 3B). No relationship was found between target HbA1c and the frequency of glucose self-monitoring, duration of diabetes and socioeconomic level (data not shown).

e of target HbA1c and measured HbA1c (Figure 3A). By contrast, older patients indicated higher target values of HbA1c than younger patients (p = 0.017) (Figure 3B). No relationship was found between target HbA1c and the frequency of glucose self-monitoring, duration of diabetes and socioeconomic level (data not shown). Figure 3 Knowledge of target HbA1c. All patients were asked for the currently recommended HbA1c level for good glycemic control (y-axis). A) These data were then compared to the actual HbA1c of each patient (x-axis). No significant difference was found (p = 0.154). B) Data were also correlated with the age finding significantly higher target HbA1c levels in older patients (p = 0.017). Data were analyzed by the Kruskal-Wallis test with a significance level of p ≤ 0.05. Self-perception of metabolic control in T1DM To assess our patients’ self-perception of their metabolic control, we invited them to predict whether the current measured HbA1c would be better, same or worse than the HbA1c assessed during the prior visit. Data were scored (SAS) and related to the actual HbA1c, age, frequency of blood glucose self-monitoring, duration of diabetes and socioeconomic level. For details concerning the SAS see the Methods section. Generally, patients with a SAS of 0 had a perfect fit between their prediction and the actual HbA1c measurement, while patients with a SAS of +/-2 had the biggest difference between their prediction and the objective measurement.

diabetes and socioeconomic level. For details concerning the SAS see the Methods section. Generally, patients with a SAS of 0 had a perfect fit between their prediction and the actual HbA1c measurement, while patients with a SAS of +/-2 had the biggest difference between their prediction and the objective measurement. We found that nearly half of the patients with a HbA1c value < 7.6% had a perfect fit showing a SAS of 0, whereas only 36% of the patients with a HbA1c value > 8.5% had a SAS of 0 (Figure 4). However, this effect was not significant (p = 0.99). There was a trend, that patients with a longer duration of diabetes overestimated their actual HbA1c false-positively (p = 0.095) (data not shown). Figure 4 Self-perception of metabolic control in T1DM. HbA1c levels were put in relation to a self assessment score (SAS). Patients were asked to predict their HbA1c qualitatively. Data were collected with questionnaires and categorized from -2 to +2. A SAS 0 meant that patient’s perception overlapped with the objective result. A SAS of +1 or +2 meant that the measured HbA1c value was better than the last one but this improvement was not perceived by the patient. A SAS -1 or -2 meant that the actual HbA1c value was worse than the last one but predicted otherwise by the patient. No significant correlation was found between the SAS and the actual HbA1c level (p = 0.99). Data are shown as bar graphs and were analyzed by the Kruskal-Wallis test.

nt was not perceived by the patient. A SAS -1 or -2 meant that the actual HbA1c value was worse than the last one but predicted otherwise by the patient. No significant correlation was found between the SAS and the actual HbA1c level (p = 0.99). Data are shown as bar graphs and were analyzed by the Kruskal-Wallis test. Interestingly, the largest proportion of patients predicted their metabolic control correctly irrespective of their actual HbA1c (36-45%) while only few made a grossly wrong prediction (Figure 4). Suggestions for improving metabolic control Professionals and patients were invited to make suggestions on how to improve or maintain metabolic control. A list of items was given. Data were analyzed descriptively and results are shown as percentage (Figure 5). Professionals often suggested a Change of the treatment regimen or No change. By contrast, patients more often suggested a change in their daily routine like Intensified glucose monitoring, Modification of nutrition or More elaborate self-protocol of therapy. Figure 5 Comparison between professionals’ and patients’ suggestions to improve metabolic control in T1DM. Professionals and patients were invited to make suggestions to improve or maintain metabolic control. A list of items was given. Professionals and patients could choose one or more of the listed items. Only professionals had the possibility to choose the item Change of treatment regimen while only patients could choose the answer Don’t know. Data were analyzed descriptively and are shown as % of all.

metabolic control. A list of items was given. Professionals and patients could choose one or more of the listed items. Only professionals had the possibility to choose the item Change of treatment regimen while only patients could choose the answer Don’t know. Data were analyzed descriptively and are shown as % of all. Additional analysis revealed a relationship between the number of daily measurements of blood glucose and the age, with a higher number of daily measurements of blood glucose in younger patients (p = 0.012), who also tended to have lower HbA1c levels. On average patients in the age category of 8 – 10 years (n = 12) performed 5.3 glucose self-measurements daily, patients in the age category of 10 – 13 years (n = 16) 5.7, patients in the age category of 13 – 16 years (n = 41) 4.6 and patients in the age category of 16 – 18 years (n = 9) 3.4 only.

e lower HbA1c levels. On average patients in the age category of 8 – 10 years (n = 12) performed 5.3 glucose self-measurements daily, patients in the age category of 10 – 13 years (n = 16) 5.7, patients in the age category of 13 – 16 years (n = 41) 4.6 and patients in the age category of 16 – 18 years (n = 9) 3.4 only. Our study questionnaire also included a query concerning the most annoying thing in the patients daily diabetes care: “If you could skip something in your daily diabetes care, what would it be?” We suggested the following items: Insulin injections, Glucose measurements, Self-protocol of the therapy in a booklet or electronic device, Diet issues or Other. From a total of 87 answers, 39% (n = 34) chose the answer Insulin injections, 37.9% (n = 33) chose Self-protocol of the therapy in a booklet or electronic device, 10.9% (n = 9) answered with Glucose measurements, 6.9% (n = 6) were annoyed with Diet issues and 5.7% (n = 5) chose Other issues including regular change of catheters of insulin pump or drawing venous blood for recommended laboratory control once a year. When we related these answers to the age of the patients, we observed, that older patients were especially annoyed at having to self-protocol the therapy in a booklet or electronic device and at glucose self-measurements, while younger patients would rather skip insulin injections or diet issues. Discussion This study in a small cohort of a single center shows that self-perception of metabolic control is good in children and adolescents with T1DM irrespective if well or poorly controlled.

Our study questionnaire also included a query concerning the most annoying thing in the patients daily diabetes care: “If you could skip something in your daily diabetes care, what would it be?” We suggested the following items: Insulin injections, Glucose measurements, Self-protocol of the therapy in a booklet or electronic device, Diet issues or Other. From a total of 87 answers, 39% (n = 34) chose the answer Insulin injections, 37.9% (n = 33) chose Self-protocol of the therapy in a booklet or electronic device, 10.9% (n = 9) answered with Glucose measurements, 6.9% (n = 6) were annoyed with Diet issues and 5.7% (n = 5) chose Other issues including regular change of catheters of insulin pump or drawing venous blood for recommended laboratory control once a year. When we related these answers to the age of the patients, we observed, that older patients were especially annoyed at having to self-protocol the therapy in a booklet or electronic device and at glucose self-measurements, while younger patients would rather skip insulin injections or diet issues. Discussion This study in a small cohort of a single center shows that self-perception of metabolic control is good in children and adolescents with T1DM irrespective if well or poorly controlled. Little is known about T1DM patients’ capacity to self-assess therapy. This includes self-perception of HbA1c value, patient’s recall of the HbA1c value, knowledge of target HbA1c level as well as patients’ suggestions on how to improve metabolic control. This is in contrast to good knowledge on neurocognitive functioning in T1DM patients and of factors associated with self-care, adherence to therapy and metabolic control.

A1c value, patient’s recall of the HbA1c value, knowledge of target HbA1c level as well as patients’ suggestions on how to improve metabolic control. This is in contrast to good knowledge on neurocognitive functioning in T1DM patients and of factors associated with self-care, adherence to therapy and metabolic control. We found that the self-perception of actual metabolic control is similar in well or poorly controlled T1DM patients. This raises the question what factors influence metabolic control, and what factors influence the ability of self-assessment. It is well known, that for example the frequency of blood-glucose self-monitoring, the age of the patients, the duration of disease or the socioeconomic background influence metabolic control [2,11]. Therefore, we wondered whether these same factors were also associated with the ability of self-assessment of metabolic control.

ll known, that for example the frequency of blood-glucose self-monitoring, the age of the patients, the duration of disease or the socioeconomic background influence metabolic control [2,11]. Therefore, we wondered whether these same factors were also associated with the ability of self-assessment of metabolic control. In general, our patients with T1DM have a satisfactory metabolic control with a mean HbA1c of 8.03%. This compares to a cross-sectional study from our center in 2008 with a mean HbA1c of 7.6% [12]. The difference in HbA1c may be explained by the fact that in the study in 2008 all T1DM patients aged 0 – 20 years were enrolled without further limitations. In this study a large proportion (69%) of the patients had a short diabetes duration of 0 – 24 months with presumed residual activity. In line with the actual study the subgroup of diabetic adolescents also had a mean HbA1c of 8.1%. Compared to a large, international multicentre study which reported a mean HbA1c of 8.2% [13], our results are slightly better. Similar to other studies [2,14], we show that metabolic control is better with shorter duration of diabetes and higher socioeconomic level.

diabetic adolescents also had a mean HbA1c of 8.1%. Compared to a large, international multicentre study which reported a mean HbA1c of 8.2% [13], our results are slightly better. Similar to other studies [2,14], we show that metabolic control is better with shorter duration of diabetes and higher socioeconomic level. We found, that poorly controlled patients (HbA1c > 8.5%) have a worse recollection of their last HbA1c compared to better controlled subjects. Only one recent study investigated the prospective memory in correlation with glycemic control in children with T1DM [7]. Prospective memory was defined as the memory which is required to carry out intended actions. This study employed PROMS, an innovative prospective memory screen and a series of cognitive tests. Overall, this was a largely negative study which found no association between total PROMS score and glycemic control. Most studies investigating neurocognitive functioning in pediatric T1DM patients conclude that severely low blood glucose levels increase the risk of learning difficulties and a range of cognitive deficits and memory function [5,6]. As we found that poorly controlled patients have a worse recollection of their last HbA1c, we assumed that they overestimated their metabolic control in personal favor. In fact, false-positive recollection of metabolic control can harm the diabetic patient because no actions will be taken to achieve euglycemia (including insulin dose adjustments, intensified glucose monitoring, and diet control). Therefore, regularly measured HbA1c and discussions with professionals are strongly recommended to prevent wrong self-assessment. Factors like age, frequency of glucose self-monitoring, duration of diabetes and socioeconomic level alone don’t seem to correlate with the ability of memorizing the personal HbA1c level.

). Therefore, regularly measured HbA1c and discussions with professionals are strongly recommended to prevent wrong self-assessment. Factors like age, frequency of glucose self-monitoring, duration of diabetes and socioeconomic level alone don’t seem to correlate with the ability of memorizing the personal HbA1c level. In regards to the knowledge about target HbA1c, no correlation was found with metabolic control. By contrast the personal target HbA1c level correlated with age, with higher personal target levels in older patients. This is inconsistent with the findings of the Hvidoere Childhood Diabetes Study 2005 [13], where reported target HbA1c levels were associated with the actual metabolic control, but not associated with age. The fact that target levels in our study did not correlate with metabolic control, is probably due to the small number of patients in our study. The observation that our older T1DM patients have a higher target HbA1c in mind remains unexplained. It has been reported that, if members of the diabetes care team are consistent in their advice on target HbA1c, adolescents’ HbA1c correlates with those targets [13]. So we presume that it is an important teaching point in diabetes care, that patients are aware of the internationally recommended target HbA1c, which is < 7.5% for all age groups [10].

embers of the diabetes care team are consistent in their advice on target HbA1c, adolescents’ HbA1c correlates with those targets [13]. So we presume that it is an important teaching point in diabetes care, that patients are aware of the internationally recommended target HbA1c, which is < 7.5% for all age groups [10]. Furthermore, it is discussed in the literature that lower HbA1c levels and longer duration of diabetes might be factors that increase the risk for hypoglycemia in children and adolescents with diabetes [13,15,16]. Therefore, it is conceivable that higher HbA1c levels or a higher personal target HbA1c level might result out of fear of hypoglycemic episodes, especially in patients with hypoglycemia unawareness or recurrent severe hypoglycemia. However, the question whether frequent and/or severe hypoglycemic episodes affect T1DM patients’ self-perception of metabolic control is not solved in the literature and remains unsolved as we did not record hypoglycemic episodes for analysis in our study.

with hypoglycemia unawareness or recurrent severe hypoglycemia. However, the question whether frequent and/or severe hypoglycemic episodes affect T1DM patients’ self-perception of metabolic control is not solved in the literature and remains unsolved as we did not record hypoglycemic episodes for analysis in our study. We found no correlation between the self-assessment score (SAS) and the actual measured HbA1c or other parameters. Interestingly, the largest proportion of patients in our study predicted their metabolic control correctly irrespective of their actual HbA1c. This may result from the therapeutic approach of our diabetes team to discuss the actual metabolic control with patients and parents and try to support patients in their efforts to improve metabolic control with personal advice. Currently there is no literature to compare these findings. When we assessed professionals’ and patients’ suggestions to improve actual metabolic control, we found that professionals often suggested a change of treatment regimen or no change, while patients rather suggested changes in their daily routine at home, like improving glucose monitoring or self-protocol or adapting nutrition. This reflects the different perspectives on diabetes management between professionals and patients well. While professionals are primarily preoccupied with values of HbA1c, glucose and insulin doses, patients deal with blood glucose self-monitoring, insulin applications and their diabetes diet regimen and know about their personal compliance. The ideal professional diabetes care has to integrate these two perspectives to reach consensus on what needs to be done to achieve good metabolic control. This goal may only be achieved with a multidisciplinary specialist team consisting of psychologists, social workers, dieticians, diabetes nurse instructors and pediatric diabetologists. Partners of the team may also be pediatricians, teachers or day-care professionals.

eeds to be done to achieve good metabolic control. This goal may only be achieved with a multidisciplinary specialist team consisting of psychologists, social workers, dieticians, diabetes nurse instructors and pediatric diabetologists. Partners of the team may also be pediatricians, teachers or day-care professionals. Interestingly, when focusing on the answers of the patients concerning our question of the most annoying thing in their daily diabetes care, we found that our patients are just as annoyed by insulin injection as to having self-protocol the therapy in a booklet or electronic device. Especially the older patients were annoyed at the continuous task of keeping a diary. There is hope, that further development of electronic devices will facilitate and simplify patients’ self-protocol of therapy in the future. Conclusion Self-perception of metabolic control in children and adolescents with T1DM treated according to international standards is good, even if the objective metabolic control does not meet the target. Patients with poor metabolic control are less attentive to their actual HbA1c. In theory, T1DM patients know the target HbA1c levels for excellent metabolic control. Overall, current diabetes care strategies seem to achieve the goal to make T1DM patients experts of their own diabetes. Abbreviations T1DM: Type 1 Diabetes Mellitus; IDF: International Diabetes Federation; ISPAD: International Society for Pediatric and Adolescent Diabetes; BMI: Body mass index; LMS: Least mean square; SAS: Self assessment score.

Conclusion Self-perception of metabolic control in children and adolescents with T1DM treated according to international standards is good, even if the objective metabolic control does not meet the target. Patients with poor metabolic control are less attentive to their actual HbA1c. In theory, T1DM patients know the target HbA1c levels for excellent metabolic control. Overall, current diabetes care strategies seem to achieve the goal to make T1DM patients experts of their own diabetes. Abbreviations T1DM: Type 1 Diabetes Mellitus; IDF: International Diabetes Federation; ISPAD: International Society for Pediatric and Adolescent Diabetes; BMI: Body mass index; LMS: Least mean square; SAS: Self assessment score. Competing interests The authors declare that they have no competing interests. Authors’ contribution The authors’ contribution to the paper is as follow: AB and CEF: study concepts and design, data analysis and interpretation, statistical analysis, critical revision of the manuscript for important intellectual content and manuscript preparation; MOM, CCM, KPS, PEM: acquisition of data, critical revision of the manuscript for important intellectual content: MJ: study concepts and statistical analysis. All authors read and approved the final manuscript. Acknowledgments We thank all patients and families for participating in this study. We also thank Siemens Healthcare Diagnostics AG, Zürich, Switzerland for their financial support in paying the publication fees.

Background Congenital hypothyroidism (CH) is defined as thyroid hormone deficiency present at birth and is classified into permanent and transient CH [1]. An exact cause for the vast majority of cases of CH (or thyroid dysgenesis) remains unknown [1]. CH is a relatively common congenital disorder, occurring in about 1 of 1,500 to 4,000 live births [2-4]. The incidence rate of CH has been reported to be increasing [1,5,6]. The reasons for the increased incidence rate are not clear, but several possibilities have been pointed out, including a change in the screening test cutoff point [2], enhanced detection [5], the increase of preterm infants [1], and the misclassification of some cases of transient CH as permanent CH [7].

asing [1,5,6]. The reasons for the increased incidence rate are not clear, but several possibilities have been pointed out, including a change in the screening test cutoff point [2], enhanced detection [5], the increase of preterm infants [1], and the misclassification of some cases of transient CH as permanent CH [7]. Several risk factors of CH have been presented in epidemiologic studies. Females are consistently reported to more frequently suffer from CH than males [1,8]. The association between birth weight and CH was reported to show a U-shaped curve [9]. According to a population-based case–control study in Italy by Medda et al. [10], the statistically significant risk factors for permanent CH were twins, additional birth defects, female gender, and gestational age >40 weeks. Using the same CH registry, Olivieri et al. [11] showed that the incidence rate of CH in multiples was more than 3-fold higher compared to singletons. Another population-based study in the U.S. also showed that the incidence rates of CH in twins doubled as compared to that in singletons, and the incidence rate was even higher with triplets/+ [6]. Moreover, many studies reported a higher incidence rate of extrathyroidal additional birth defects among neonates with CH compared with the general population [8,12-24], especially congenital heart disease [8,14,17,21-25].

ed as compared to that in singletons, and the incidence rate was even higher with triplets/+ [6]. Moreover, many studies reported a higher incidence rate of extrathyroidal additional birth defects among neonates with CH compared with the general population [8,12-24], especially congenital heart disease [8,14,17,21-25]. Many case studies showed that CH occurs sporadically [26], although dyshormonogenic cases are often recessively inherited, and recent cohort analyses estimated that approximately 2% of cases with thyroid dysgenesis are familial [27,28]. Olivieri et al., in the only population-based twin study published to date [11], noted that the pairwise concordance rate of twin pairs for permanent CH was low (4.3%). Given these circumstances, assisted reproductive technology (ART) data present a unique opportunity for twin study [29]. The percentage of multiple births among the ART population is much higher than that in the general population. Most twins after ART are dizygotic (DZ) siblings that develop together in the same womb. Thus, the aim of the present study was to compare the incidence rate of CH in multiples with singletons, examine additional birth defects with CH, and analyze the familial aggregation of CH using nationwide data on ART in Japan.

n. Most twins after ART are dizygotic (DZ) siblings that develop together in the same womb. Thus, the aim of the present study was to compare the incidence rate of CH in multiples with singletons, examine additional birth defects with CH, and analyze the familial aggregation of CH using nationwide data on ART in Japan. Materials and methods Outline of Japanese birth defects data after ART The method for collecting data is described in detail elsewhere [29]. Almost all medical institutions performing ART, which does not include simple ovulation stimulation/enhancement, are registered with the Japan Society of Obstetrics and Gynecology (JSOG). The JSOG administers questionnaire surveys for these medical institutions. Some of the survey data are presented in simple annual reports of aggregate, not individual, data. The individual list of all ART pregnancies resulting in congenital diseases, not all ART pregnancies, is presented every year in the JSOG annual reports on ART (in Japanese). The presented items are method of treatment, blastocyst transfer, maternal age, perinatal outcome and the gestational week, plurality, sex, early neonatal infant death up to day 6, and name of congenital disease. The author used case reports on birth defects or congenital anomaly data from 2005–2009 as the initial information for the present secondary data analyses. The mean response rate throughout the 5 years was 99.0% (3,026/3,056), meaning that almost a complete database reflecting the current situation of ART in Japan could be constructed.

ase reports on birth defects or congenital anomaly data from 2005–2009 as the initial information for the present secondary data analyses. The mean response rate throughout the 5 years was 99.0% (3,026/3,056), meaning that almost a complete database reflecting the current situation of ART in Japan could be constructed. All live births after ART were analyzed in the present study. The types of defects were reclassified according to the International Classification of Diseases, tenth edition (ICD-10, 2003 version). Diseases that were classified in the category of ICD-10 code Q00-Q99 (i.e., congenital malformations, deformations, and chromosomal abnormalities) were regarded as birth defects in the present study. CH was classified as part of E00 (congenital iodine-deficiency syndrome) or E03 (other hypothyroidism). The present data did not allow the author to distinguish between permanent CH and transient CH. Thus, all types of CH were treated as CH in the present study. The total number of ART live births, singletons and multiples, were available from 2007 to 2009. The author estimated the number of ART singletons and multiples between 2005 and 2006 using approximation formulae [29]. Statistical analyses All patients with CH were listed with their obstetric data and neonatal outcome to further examine the features of CH patients. The frequency of additional birth defects in neonates with CH was calculated. Next, the relative risk (RR: the incidence rate of multiples/singletons) with the corresponding 95% confidence interval (CI) was calculated with singletons as the reference group.

onatal outcome to further examine the features of CH patients. The frequency of additional birth defects in neonates with CH was calculated. Next, the relative risk (RR: the incidence rate of multiples/singletons) with the corresponding 95% confidence interval (CI) was calculated with singletons as the reference group. Finally, the familial aggregation of CH was analyzed. The probandwise concordance rate [30] was calculated. The probandwise concordance rate is in a restricted sense the probability that a twin is affected given that his/her co-twin is affected. These rates can be compared directly to risk rates reported for other familial pairings and to population incidence rate figures [30]. Probandwise concordance rates were calculated as 2×C/(2×C+D), assuming complete and double ascertainment, where C denotes the number of affected concordant pairs and D denotes the number of discordant pairs [30]. For the triplets, the neonates were counted as concordant only when all neonates have CH in the present study. A widely used measure of familial aggregation is the sibling recurrence risk ratio (RRR), defined as the ratio of the risk of disease manifestation, given that one’s sibling is affected, compared with the disease incidence rate in the general population [31,32]. The probandwise concordance rate of multiple births was divided by the incidence rate of the CH in the total ART population or the Japanese general population in this study. Statistical analysis was conducted using Microsoft Excel 2010 and SAS for Windows ver. 9.2.

A widely used measure of familial aggregation is the sibling recurrence risk ratio (RRR), defined as the ratio of the risk of disease manifestation, given that one’s sibling is affected, compared with the disease incidence rate in the general population [31,32]. The probandwise concordance rate of multiple births was divided by the incidence rate of the CH in the total ART population or the Japanese general population in this study. Statistical analysis was conducted using Microsoft Excel 2010 and SAS for Windows ver. 9.2. Results Demographic and perinatal outcome data of all neonates with CH are presented in Table 1. There were 18 patients total, consisting of 12 singletons and 6 multiples. The percentages of preterm delivery in singletons and multiples were 9% (=1/11) and 100% (=5/5), respectively. The sex ratio was 1 (9 males and 9 females). Additional birth defects were present in three patients with CH (17%=3/18). Multiples were more frequently affected by other birth defects (33%=2/6) than singletons (8%=1/12). The authors of previous studies noted that patent ductus arteriosus (PDA) is related to prematurity and is consequently more prevalent in twins [33,34]. When the frequency was calculated by excluding PDA, the result was 11% (=2/18). Multiples were also more frequently affected by other birth defects (17%=1/6) than singletons (8%=1/12). Table 1 Demographic and perinatal outcome data of all neonates with CH

Results Demographic and perinatal outcome data of all neonates with CH are presented in Table 1. There were 18 patients total, consisting of 12 singletons and 6 multiples. The percentages of preterm delivery in singletons and multiples were 9% (=1/11) and 100% (=5/5), respectively. The sex ratio was 1 (9 males and 9 females). Additional birth defects were present in three patients with CH (17%=3/18). Multiples were more frequently affected by other birth defects (33%=2/6) than singletons (8%=1/12). The authors of previous studies noted that patent ductus arteriosus (PDA) is related to prematurity and is consequently more prevalent in twins [33,34]. When the frequency was calculated by excluding PDA, the result was 11% (=2/18). Multiples were also more frequently affected by other birth defects (17%=1/6) than singletons (8%=1/12). Table 1 Demographic and perinatal outcome data of all neonates with CH ID Maternal age Method of ART Blastocyst transfer Gestational weeks Plurality Concordance/ Discordance in multiples Sex Early neonatal death Other birth defects 1 24 FET no 37 singleton   male no 2 39 IVF no 25 triplet discordance female no 3 33 IVF yes 41 singleton   male unknown 4 39 ICSI no unknown singleton   male unknown 5 31 ICSI yes 35 twin discordance male no hypospadias 6 31 FET no 27 twin concordance male no 7 31 FET no 27 twin concordance male no patent ductus arteriosus 8 30 IVF no 39 singleton   male unknown 9 33 ICSI yes 40 singleton   female no 10 37 FET yes 40 singleton   female no 11 28 ICSI no 39 singleton   female no cleft lip, congenital genu recurvatum 12 40 FET yes 41 singleton   female no 13 36 IVF no 36 twin discordance male no 14 29 ICSI no 34 twin discordance female no 15 39 IVF, ICSI yes 36 singleton   female no 16 38 ICSI no 40 singleton   female no 17 41 FET yes 37 singleton   male no 18 33 FET yes 37 singleton   female no CH: congenital hypothyroidism, ART: assisted reproductive technology, FET: frozen embryo transfer, IVF: in-vitro fertilization, ICSI: intracytoplasmic sperm injection.

VF, ICSI yes 36 singleton   female no 16 38 ICSI no 40 singleton   female no 17 41 FET yes 37 singleton   male no 18 33 FET yes 37 singleton   female no CH: congenital hypothyroidism, ART: assisted reproductive technology, FET: frozen embryo transfer, IVF: in-vitro fertilization, ICSI: intracytoplasmic sperm injection. The incidence rate of CH per 1,000 live births was 0.17 (=18/106,678). The rate of CH was more than twofold higher in multiple births (0.31 ‰=6/19,533) than singleton births (0.14 ‰=12/87,145), but the difference was not statistically significant (RR=2.2, 95% CI 0.8–5.9). The proportion of multiple births observed in the CH patients (33%=6/18) was twofold higher than that estimated in the total ART population (18%=19,533/106,678) in the same period (2005–2009). The calculated concordance rate and RRR are presented in Table 2. Six multiple-births patients were derived from one concordant twin pair, which consisted of two twin patients; three discordant twin pairs, which consisted of three twin patients; and one discordant triplets set, which consisted of one triplet patient. Thus, the probandwise concordance rate was 33.3% (=(2×1)/(2×1+4)). The estimated RRR was 1976 for the total ART population or 609 for the general Japanese population. Table 2 Concordance rate and RRR of CH, 2005-2009

The calculated concordance rate and RRR are presented in Table 2. Six multiple-births patients were derived from one concordant twin pair, which consisted of two twin patients; three discordant twin pairs, which consisted of three twin patients; and one discordant triplets set, which consisted of one triplet patient. Thus, the probandwise concordance rate was 33.3% (=(2×1)/(2×1+4)). The estimated RRR was 1976 for the total ART population or 609 for the general Japanese population. Table 2 Concordance rate and RRR of CH, 2005-2009 Multiples Total ART population RRR(=A/X) Japanese general population RRR (=A/Y) Concordant Discordant Probandwise concordance rate (A)(%) N Incidence rate ( ‰) (X) N Incidence rate( ‰) (Y) pair(N) pair or set(N) 1 4 33.3 18 0.17 1976 2956 0.55 609 RRR: recurrence risk ratio, CH: congenital hypothyroidism, ART: assisted reproductive technology. RRR was calculated probandwise concordance rate divided by the incidence rate in the general population. The incidence rate in the Japanese general population was calculated using the data presented by the Japan Child and Family Research Institute (http://www.boshiaiikukai.jp/img/milk/kensajokyoH22.pdf (In Japanese). Accessed December 23, 2012). The prevalence throughout 2005–2009 was 0.55 ‰ (=2,956/5,398,934). Probandwise concordance rate was calculated as 33.3%=2/(2+4). RRR for total ART population was calculated as 1976=(2/6)/(18/106,678). RRR for Japanese general population was calculated as 609=(2/6)/(2,956/5,398,934).

The incidence rate in the Japanese general population was calculated using the data presented by the Japan Child and Family Research Institute (http://www.boshiaiikukai.jp/img/milk/kensajokyoH22.pdf (In Japanese). Accessed December 23, 2012). The prevalence throughout 2005–2009 was 0.55 ‰ (=2,956/5,398,934). Probandwise concordance rate was calculated as 33.3%=2/(2+4). RRR for total ART population was calculated as 1976=(2/6)/(18/106,678). RRR for Japanese general population was calculated as 609=(2/6)/(2,956/5,398,934). Discussion Incidence rate and RR To my knowledge, there has been no published study in which the authors directly analyzed the effect of ART on CH. CH incidence rate in developed countries is usually estimated by the data from neonatal screening tests, which are performed after birth. The incidence rate of CH noted in the present study is lower than that noted in many other studies published recently. A neonatal screening test for CH has been mandatory in Japan since 1979. According to the Japanese data on CH, the mean incidence rate of CH was 1/3,600 between 1979 and 2004, and 1/2,000 in 2005, and it then increased slightly to 1/1,800 in 2009 (presented by the Japan Child and Family Research Institute, http://www.boshiaiikukai.jp/img/milk/kensajokyoH22.pdf (In Japanese). Accessed December 23, 2012). The lower incidence rate of CH in the present study might be attributed to the incomplete reporting. The follow-up period of the present data did not necessarily reflect the result of the mass screening test. Moreover, some obstetricians might not regard CH as a birth defect in the narrowest sense (namely, based on ICD-10). Nevertheless, the objective of this study was to evaluate the incidence rate of CH in multiple births compared to singletons, and not to compare the CH incidence rate across different populations. Therefore, the comparison of CH in multiple births and singletons maybe biased only if there is differential reporting according to plurality, which is not likely to have occurred.

the incidence rate of CH in multiple births compared to singletons, and not to compare the CH incidence rate across different populations. Therefore, the comparison of CH in multiple births and singletons maybe biased only if there is differential reporting according to plurality, which is not likely to have occurred. In the present study, the incidence rate of CH in multiples was about twofold higher than singletons, and even though the difference was not statistically significant, the finding suggests that multiple birth is one of the risk factors of CH. According to the recent population-based case control study by Medda et al. [10], an increased risk for permanent CH was detected in twins by a multivariate analysis (RR=12.2, 95% CI 2.4–62.3). According to the Harris and Pass study in New York [6], the incidence of CH was nearly double in twin births (1:876) as compared to singletons (1:1765) and even higher with triplets/+ (1:575) in 2002–2003. According to the population-based study by Olivieri et al. [11], a more than 3-fold higher frequency of multiples was found in the CH population (10.1 in 10,000) than in the general population (3.2 in 10,000 live births). As is well known, multiple births occur far more often in ART than in cases of spontaneous conception in almost all developed countries [29]. The multiple-birth rate (per 1,000 live births) was increased nearly doubled (12.4 in 1986 and 22.7 in 2005) in Japan, mainly due to iatrogenic multiple births of advanced-age mothers [29]. Thus, the widespread use of fertility treatment, including ART, might in part have contributed to the rise in the total incidence rate of CH in the general population, by indirectly increasing multiple births. Unfortunately, the present results did not reflect this hypothesis.

tiple births of advanced-age mothers [29]. Thus, the widespread use of fertility treatment, including ART, might in part have contributed to the rise in the total incidence rate of CH in the general population, by indirectly increasing multiple births. Unfortunately, the present results did not reflect this hypothesis. Additional birth defects Many studies [8,12-23] reported that the frequencies of additional birth defects with CH were between 8% and 20%, except that of a recent study by Reddy (59%) [24]. The present result (17%) was well within this range. The present results also suggested that this tendency is more obvious in multiples than in singletons. The mechanism of a higher frequency of additional birth defects in patients with CH was unclear. Olivieri et al. [11] proposed the hypothesis that genes involved in the development of the thyroid and other organs may be affected during the early stages of embryogenesis. However, it should be noted with caution that permanent CH showed a lower frequency of additional birth defects than did transient CH [18].

clear. Olivieri et al. [11] proposed the hypothesis that genes involved in the development of the thyroid and other organs may be affected during the early stages of embryogenesis. However, it should be noted with caution that permanent CH showed a lower frequency of additional birth defects than did transient CH [18]. It has been well established that congenital heart disease is the most frequently occurring additional birth defect in patients with CH [8,14,17,21-25]. In the present study, PDA was found in one twin. A recent study by Reddy et al. [24] reported that two out of ten patients had PDA in addition to CH. Other birth defects with CH in the present patients were hypospadias, cleft lip, and congenital genu recurvatum. Olivieri et al. [22] performed a population-based study of the frequency of additional birth defects in patients with CH. Cleft lip, urological malformation, and musculoskeletal anomalies were reported as additional birth defects in patients with CH, although the numbers were small [22]. The present results might be supported by these findings. Familial aggregation There have been few genetic epidemiologic studies on CH. Family study is a useful tool to show familial aggregation. Intensive family studies on CH were performed by a French group [27,28,35] who found that approximately 2% of CH cases with thyroid dysgenesis were familial. Although familial cases represent a minority of cases of CH caused by thyroid disgenesis, such cases were observed in more than 15-fold higher proportion than would be expected from chance alone, suggesting genetic factors.

up [27,28,35] who found that approximately 2% of CH cases with thyroid dysgenesis were familial. Although familial cases represent a minority of cases of CH caused by thyroid disgenesis, such cases were observed in more than 15-fold higher proportion than would be expected from chance alone, suggesting genetic factors. Apart from case reports, the only population-based twin study was performed by Olivieri et al. [11]. According to their study, the pairwise concordance rate of twin pairs with unknown zygosity for permanent CH was low (4.3%=3/70) and was due to there being three pairs. They suggested that the sporadic occurrence of CH was likely due to noninheritable postzygotic events that may have included epigenetic modifications and early somatic mutation. We view their results with caution because their calculation was based on a pairwise, not probandwise method. The probandwise concordance rate of their data was recalculated as 8.2% (=6/73). This value would be compared with the incidence rate in the general population, or would be calculated according to the zygosity of twin pairs. Using the probandwise concordance rate and their incidence rate in the general population (0.032% in singletons), the RRR of twins was 256, which might be high enough to suggest familial aggregation. Olivieri et al. [11] also found a high recurrence risk (35.4) for CH in siblings of affected babies and indicated that environmental risk factors may act as a trigger in persons with a susceptible genetic background. There was one concordant pair in the present study, which produced a relatively high probandwise concordance rate (33.3%) and RRR. Although the present concordance rates and RRR were influenced by chance factors, familial aggregation of CH was suggested.

ct as a trigger in persons with a susceptible genetic background. There was one concordant pair in the present study, which produced a relatively high probandwise concordance rate (33.3%) and RRR. Although the present concordance rates and RRR were influenced by chance factors, familial aggregation of CH was suggested. Population-based twin study using univariate/multivariate genetic analyses based on the structural equation modeling is the most powerful tool to clarify the genetic/environmental contribution to CH and comorbidity with other birth defects [36]. Record linkage between a mass screening registry of CH and twin registry would make this possible.

udy using univariate/multivariate genetic analyses based on the structural equation modeling is the most powerful tool to clarify the genetic/environmental contribution to CH and comorbidity with other birth defects [36]. Record linkage between a mass screening registry of CH and twin registry would make this possible. Limitations This study has the following limitations, most of which could be attributed to the dataset, based on the fact that individual information was obtained only from the subjects with birth defects and/or CH after ART, not the total ART live births. The first limitation is that the author could not check the reliability of the data directly. This is the essential limitation of secondary data analyses. Second, although the present dataset was from a multi-year nationwide survey, it still did not have sufficiently high statistical power to detect statistical significance. Third, the author could not control for confounding factors that can affect ART and/or CH. Fourth, follow-up after birth was limited to the neonatal period at the latest, and was incomplete [29]. Both CH and some birth defects are not obvious within a few days after birth. Fifth, the CH could not be distinguished between permanent and transient type. From the view of disease prevention, all types of CH should be properly followed up. This problem should also be discussed from the viewpoint of medical economics for mass screening of CH.

fects are not obvious within a few days after birth. Fifth, the CH could not be distinguished between permanent and transient type. From the view of disease prevention, all types of CH should be properly followed up. This problem should also be discussed from the viewpoint of medical economics for mass screening of CH. Conclusions CH was more frequent in multiples compared to singletons. A higher percentage of additional birth defects was also observed in multiples compared to singletons. The familial aggregation of CH was suggested. Abbreviations CH: Congenital hypothyroidism; ART: Assisted reproductive technology; RR: Relative risk; CI: Confidence interval; RRR: Recurrence risk ratio; PDA: Patent ductus arteriosus; FET: Frozen embryo transfer; IVF: In-vitro fertilization; ICSI: Intracytoplasmic sperm injection. Competing interests The author declare that I have no competing interests. Author’ contributions SO carried out data gathering, analyses, and writing of manuscript. Acknowledgements I would like to thank Toshimi Ooma for assistance with data analysis. This work was supported in part by a Grant-in-Aid for Challenging Exploratory Research (Grant Number 23659356) and Grant-in-Aid for Scientific Research (B) (Grant Number 24390167) from the Japan Society for the Promotion of Science.

chromosomes and the parental origin (monoallelic expression of the imprinted genes) may play a part in the broad and variable clinical features spectrum seen in TS. Therefore, TS could be a model for understanding the role of genomic imprinting on the clinical features and putative imprinted genes on the X-chromosome. A continuing controversy remains regarding the impact of the retained X-chromosome on the clinical features of TS according to their parental origin, in particular with regards to height, deafness, cardiovascular and metabolic abnormalities and to the growth response to rhGH. In the context of this controversy, the current study was performed to evaluate the effect of the parental origin of the X-chromosome on the clinical features, auxological data, associated complications, lipid metabolism and response to rhGH in a group of patients with TS and a non-mosaic 45,X karyotype.

Background In 2010, the Boston Children’s Hospital Program for Patient Safety and Quality convened a Task Force to identify the frequency of fragility or insufficiency fractures in hospitalized infants, children and adolescents and to reduce the risk of these fractures. As low bone density has been identified as a risk factor for fractures in children [1,2] a component of this bone health safety initiative was to identify factors associated with low bone mineral density in patients seen in our institution. Specific pediatric populations are known to be at high risk for a low bone density, including children and adolescents with cerebral palsy and other non-ambulatory states [1-4], chronic renal failure [5], malnutrition and malabsorptive states [6-8], cystic fibrosis [9-11], pubertal delay [12], and 25-hydroxyvitamin D (25OHD) deficiency [13,14]. In addition, medical therapies such as treatment with anticonvulsants [15], glucocorticoids [15], and chemotherapy [16,17] are associated with a compromise of bone density. As fracture risk in children is inversely related to bone density in some reports [18-22], understanding the risk factors associated with low bone mineral density may provide greater opportunities for early identification and intervention for those at risk for skeletal fragility. To our knowledge, no study has examined the relative risk of these conditions and treatments on low bone density in a pediatric population.

ding the risk factors associated with low bone mineral density may provide greater opportunities for early identification and intervention for those at risk for skeletal fragility. To our knowledge, no study has examined the relative risk of these conditions and treatments on low bone density in a pediatric population. Dual-energy x-ray absorptiometry (DXA) is a common methodology used to quantify bone mineral density (BMD), and provides a measure of bone mineral (g) per projected area scanned (cm2). In children, a Z-score is used to compare a child’s BMD with an age- and gender-matched norm, with appropriate adjustments for bone age or pubertal status often needed [20]. Guidelines established by the International Society for Clinical Densitometry define a low bone density as a BMD Z-score of less than or equal to -2.0 SD [23]. The objectives of this study were to determine risk factors for low bone density (BMD Z-score ≤ -2) in a pediatric population, and to identify which risk factors have the strongest correlation with BMD. A secondary aim was to examine whether fracture history was correlated with BMD Z-scores in children and adolescents who are within designated risk groups for low bone density.

for low bone density (BMD Z-score ≤ -2) in a pediatric population, and to identify which risk factors have the strongest correlation with BMD. A secondary aim was to examine whether fracture history was correlated with BMD Z-scores in children and adolescents who are within designated risk groups for low bone density. Methods Subjects We identified all patients between the ages of 4-21 years who were referred to the Bone Health Program at Children’s Hospital Boston for BMD measurements by DXA as part of routine clinical care from October 2008 to September 2009. Patients who had the DXA performed as part of a research study were excluded. Subjects were categorized based on their lowest BMD Z-score and divided into three groups: BMD Z-score > -1.0 SD, between -1 to -1.9 SD, or ≤ -2.0 SD. Based on the results of a power analysis (described in ‘Statistical Analysis’), we reviewed charts consecutively by scanning date until we identified 100 subjects in each of the three subgroups (532 charts reviewed). Once 100 subjects were identified in a group, additional subjects in that group were not included. The Boston Children’s Hospital Committee on Clinical Investigation approved this protocol.

iewed charts consecutively by scanning date until we identified 100 subjects in each of the three subgroups (532 charts reviewed). Once 100 subjects were identified in a group, additional subjects in that group were not included. The Boston Children’s Hospital Committee on Clinical Investigation approved this protocol. Densitometry measurements Scans were performed on a single densitometer. Areal (two-dimensional) bone density was quantified by DXA using a Hologic Discovery A scanner [Hologic Inc, Bedford, MA]. Bone mineral density (BMD, g/cm2) measurements were obtained at the left total hip and lumbar spine (L1-L4), and in some cases, at the whole body. Pediatric normative data were used to calculate BMD Z-scores at each skeletal site [24] using pediatric software, to allow for comparison with age- and gender-matched controls [24]. With this instrument in our DXA Center, the average in vivo precision for aBMD (expressed as percent coefficient of variation) was 0.62% at the spine and 0.72% at the total hip in children and adolescents.

h skeletal site [24] using pediatric software, to allow for comparison with age- and gender-matched controls [24]. With this instrument in our DXA Center, the average in vivo precision for aBMD (expressed as percent coefficient of variation) was 0.62% at the spine and 0.72% at the total hip in children and adolescents. Data collection Height and weight were obtained using a calibrated stadiometer (Kalamazoo, MO) and scale. Body mass index (BMI) was expressed as body weight in kilograms divided by the square of height in meters (kg/m2) as a weight-for-height index and was converted to percentiles and corresponding Z-scores by using age- and gender-specific normative values for US children [25]. We used the normative values for maximal age (20 years) [25] to calculate BMI for older subjects. Underweight was defined as BMI < 5th percentile and overweight was defined as BMI > 85th percentile for age and gender. Demographic and medical history data and DXA reports were obtained through a retrospective chart review of the Boston Children’s Hospital medical record. Data were collected from outpatient clinic notes, radiology reports, and DXA reports. Data included ethnicity, gender, fracture history, age at menarche, history of 25OHD insufficiency (defined as 25OHD level < 30 ng/mL, the lower limit of the normal range for our clinical laboratory), family history of osteoporosis, and history of prematurity. We recorded 25OHD values for subjects with 25OHD insufficiency. Information about specific medical conditions and treatments was collected from the medical record.

ned as 25OHD level < 30 ng/mL, the lower limit of the normal range for our clinical laboratory), family history of osteoporosis, and history of prematurity. We recorded 25OHD values for subjects with 25OHD insufficiency. Information about specific medical conditions and treatments was collected from the medical record. Data recorded from each subject’s DXA report included the absolute BMD and corresponding Z-scores at the left total proximal hip and lumbar spine (L1-L4). Only selected patients had total body DXA measurements, as the DXA measurements for many patients within this sample were performed prior to published recommendations by pediatric experts regarding preferred skeletal sites [23]. When available, DXA measurements adjusted for bone age were recorded when bone age differed significantly from chronological age. Additional data collected from the DXA report included the date of the study, the subject’s age, height, and weight at time of the BMD measurements, the indication(s) for the scan as provided by the referring physician, and the total number of prior DXA scans the subject had undergone.

ificantly from chronological age. Additional data collected from the DXA report included the date of the study, the subject’s age, height, and weight at time of the BMD measurements, the indication(s) for the scan as provided by the referring physician, and the total number of prior DXA scans the subject had undergone. Statistical analysis A two proportion power analysis was used to determine the minimum number of cases and controls necessary to detect a 15 percentage point difference in risk factors associated with low BMD Z-score (power = 0.8, alpha = 0.05). We determined that a minimum of 100 cases (BMD Z-score ≤ -2) and 200 controls (BMD Z-score > -2) would give our study the power necessary to detect this difference. For the descriptive analysis, patients were stratified into three groups based on the patient’s lowest BMD Z-score (for multiple DXA readings): > -1.0 SD, between -1 to -1.9 SD, or ≤ -2.0 SD. Patient demographics among the BMD Z-score groups were summarized using means and standard deviations for continuous variables and proportions for categorical variables. Statistical differences across the three groups were analyzed using Pearson’s chi-square or Fisher’s Exact test for categorical variables and one way analysis of variance (ANOVA) for continuous variables. Additionally, we assessed differences across DXA indications by gender using Pearson’s chi-square or Fisher’s Exact test for categorical variables.

he three groups were analyzed using Pearson’s chi-square or Fisher’s Exact test for categorical variables and one way analysis of variance (ANOVA) for continuous variables. Additionally, we assessed differences across DXA indications by gender using Pearson’s chi-square or Fisher’s Exact test for categorical variables. For the univariate analysis, we dichotomized the BMD Z-scores into two groups: ≤ -2 and > -2 and assessed individual factors that may be associated with low BMD Z-score. Pearson’s chi-square and Fisher’s Exact test were used for categorical variables and ANOVA was used for continuous variables. Variables in the univariate model with a p-value ≤ 0.05 were considered for inclusion in a multivariate logistic regression model. A gender specific sub-analysis looking at fracture history and low BMD was performed using the Pearson’s chi-square test and ANOVA to assess differences between BMD groups. All analyses were performed using SAS software version 9.2 (SAS Institute Inc, Cary, NC), and a 2-sided p value ≤ 0.05 was considered indicative of statistical significance.

ysis looking at fracture history and low BMD was performed using the Pearson’s chi-square test and ANOVA to assess differences between BMD groups. All analyses were performed using SAS software version 9.2 (SAS Institute Inc, Cary, NC), and a 2-sided p value ≤ 0.05 was considered indicative of statistical significance. Results We obtained information on 304 children and young adults between the ages of 4-21 years old who underwent DXA scans at Boston Children’s Hospital. Height and weight measurements were available for 282 subjects. Subject characteristics at the time of the DXA scans are presented in Table 1, classified by BMD Z-score. Of note, there were differences among the mean age in each group, with a rise in mean age correlating with increasing BMD Z-scores (p <0.01). We did not identify differences between BMD Z-scores and race or ethnicity, with more than 80% of subjects in our sample self-identifying as white (p = 0.40). BMI was significantly different between the groups, with the lowest BMD group having the lowest BMI Z-scores (p < 0.001). Fifteen percent of underweight children and young adults, defined as those with a BMI < 5th percentile for age and gender, had a BMD Z-score ≤ -2, while only 3% of those with BMD Z-score > -1 were underweight (p < 0.01). We also found significant differences in height Z-scores between the groups, with the lowest BMD group having the lowest height Z-scores (p < 0.001). As expected based upon the study design, the mean hip and spine Z-scores were significantly different between the groups (p < 0.001). In our sample population, we also identified that male gender was associated with a BMD Z-score ≤ -2 (p = 0.01).

with the lowest BMD group having the lowest height Z-scores (p < 0.001). As expected based upon the study design, the mean hip and spine Z-scores were significantly different between the groups (p < 0.001). In our sample population, we also identified that male gender was associated with a BMD Z-score ≤ -2 (p = 0.01). Table 1 Characteristics of study subjects at time of DXA

with the lowest BMD group having the lowest height Z-scores (p < 0.001). As expected based upon the study design, the mean hip and spine Z-scores were significantly different between the groups (p < 0.001). In our sample population, we also identified that male gender was associated with a BMD Z-score ≤ -2 (p = 0.01). Table 1 Characteristics of study subjects at time of DXA Characteristic BMD Z-score ≤ - 2 - 2 < BMD Z-score < -1 BMD Z-score ≥ -1 p N =102 N = 101 N = 101 Female 56 (55%) 63 (62%) 76 (75%) 0.009 Age (years) Mean ± SD 13.6 ± 4.1 14.5 ± 4.3 15.5 ± 3.3 0.003 Race/Ethnicity 0.397 White 88 (86%) 79 (78%) 83 (82%) Black 1 (1%) 2 (2%) 4 (4%) Hispanic 1 (1%) 4 (4%) 3 (3%) Asian 2 (2%) 4 (4%) 3 (3%) Other 6 (6%) 2 (2%) 3 (3%) Not Documented 4 (4%) 10 (10%) 5 (5%) BMI Mean ± SD 18.8 ± 3.6 20.2 ± 4.4 23.1 ± 5.2 <0.001 Underweight 13 (15%) 9 (9%) 3 (3%) 0.003 Healthy weight 62 (72%) 75 (76%) 65 (67%) Overweight 9 (11%) 6 (6%) 16 (17%) Obese 2 (2%) 9 (9%) 13 (13%) BMI Z-score Mean ± SD -0.43 + 1.0 -0.11 + 1.2 0.52 + 1.0 <0.001 Height Z-score Mean ± SD -1.2 ± 1.5 -0.7 ± 1.3 -0.1 + 1.1 <0.001 Hip Z-score Mean ± SD -2.18 ± 1.1 -1.0 ± 0.6 0.45 ± 0.8 <0.001 Spine Z-score Mean ± SD -2.4 ± 0.9 -1.1 ± 0.6 0.30 ± 0.9 <0.001 Indications for DXA, as documented on the DXA report form from the referring provider, are presented in Table 2. Providers had the option of selecting multiple indications if applicable. We looked at differences in indication by gender, and found that a slightly higher percentage of males were referred for a history of fracture, compared to females (p = 0.05). Additionally, a greater percentage of males reported a history of gastrointestinal disease (p < 0.01), while a greater percentage of females had a history of hypogonadism (p < 0.001). Furthermore, as detailed in the univariate analysis (Table 3), report of a history of a chronic disease associated with bone loss was predictive of a BMD Z-score ≤ -2 (p < 0.001).

reported a history of gastrointestinal disease (p < 0.01), while a greater percentage of females had a history of hypogonadism (p < 0.001). Furthermore, as detailed in the univariate analysis (Table 3), report of a history of a chronic disease associated with bone loss was predictive of a BMD Z-score ≤ -2 (p < 0.001). Table 2 Differences in indication for DXA by gender Indication Male Female p N = 109 N= 195 Medical History of Fracture 47 (43.1%) 62 (31.8%) 0.05 History of gastrointestinal disease* 35 (32.1%) 32 (16.4%) <0.01 Osteopenia noted on a prior x-ray 9 (8.3%) 21 (10.8%) 0.48 Hypogonadism or delayed puberty 6 (5.5%) 77 (39.5%) <0.001 Chronic disease associated with bone loss 41 (37.6%) 56 (28.7%) 0.11 *Includes inflammatory bowel disease, celiac disease, and malabsorption. Table 3 Univariate analysis: Factors associated with BMD Z-score ≤ - 2

Indication Male Female p N = 109 N= 195 Medical History of Fracture 47 (43.1%) 62 (31.8%) 0.05 History of gastrointestinal disease* 35 (32.1%) 32 (16.4%) <0.01 Osteopenia noted on a prior x-ray 9 (8.3%) 21 (10.8%) 0.48 Hypogonadism or delayed puberty 6 (5.5%) 77 (39.5%) <0.001 Chronic disease associated with bone loss 41 (37.6%) 56 (28.7%) 0.11 *Includes inflammatory bowel disease, celiac disease, and malabsorption. Table 3 Univariate analysis: Factors associated with BMD Z-score ≤ - 2 Factor ≤ - 2 BMD Z-score BMD Z-score > - 2 p N = 102 N= 202 Indication for DXA (reported on referral form) Gastrointestinal disease* 14 (13.7%) 53 (26.2%) 0.013 Osteopenia noted on a prior x-ray 13 (12.8%) 17 (8.4%) 0.232 Hypogonadism or delayed puberty 23 (22.6%) 60 (29.7%) 0.186 Chronic disease associated with bone loss 51 (50.0%) 46 (22.8%) <0.001 Risk Factors 25OHD insufficiency (< 30 ng/mL) 68 (66.7%) 91 (45.1%) <0.001 Family history of osteoporosis 23 (22.6%) 26 (12.9%) 0.009 Has reached menarche (females only) 26 (47.3%) 95 (68.4%) 0.012 Amenorrhea (females only) 12 (35.3%) 53 (51.5%) 0.257 Delayed puberty (males only) 5 (10.9%) 5 (7.8%) 0.649 Fracture History** 42 (42.4%) 62 (31.0%) 0.051 Eating disorder 8 (7.8%) 26 (12.9%) 0.189 Cerebral palsy and/or non-ambulatory 22 (21.6%) 4 (2.0%) <0.001 Cystic fibrosis 6 (5.9%) 7 (3.5%) 0.371 Malnutrition § 17 (16.7%) 26 (12.9%) 0.370 Malabsorption or IBD 17 (16.7%) 64 (31.7%) 0.005 Osteogenisis imperfecta 3 (2.9%) 2 (0.99%) 0.339 Recipient of an organ transplant 5 (4.9%) 2 (1.0%) 0.032 Prior treatment with radiation and/or chemotherapy 8 (7.8%) 11 (5.5%) 0.415 Recipient of a bone marrow transplant 7 (6.9%) 2 (1.0%) 0.008 Glucocorticoid use ( > 2 weeks) 39 (38.2%) 74 (36.6%) 0.785 Anticonvulsant use 19 (18.6%) 13 (6.4%) 0.001 History of prematurity ‡ 9 (8.8%) 10 (4.9%) 0.022 Height z score (Mean + SD) -1.15 + 1.5 -0.37 + 1.2 <0.001 BMI z score (Mean + SD) -0.4 + 1.1 0.2 + 1.1 <0.001 *Includes inflammatory bowel disease, celiac disease, and malabsorption.

( > 2 weeks) 39 (38.2%) 74 (36.6%) 0.785 Anticonvulsant use 19 (18.6%) 13 (6.4%) 0.001 History of prematurity ‡ 9 (8.8%) 10 (4.9%) 0.022 Height z score (Mean + SD) -1.15 + 1.5 -0.37 + 1.2 <0.001 BMI z score (Mean + SD) -0.4 + 1.1 0.2 + 1.1 <0.001 *Includes inflammatory bowel disease, celiac disease, and malabsorption. **Excludes patients with Osteogenisis Imperfecta (N=5). § Based upon clinical assessment by a nutritionist or physician, as documented in medical record. ‡ Gestational age < 36 weeks. In the univariate analysis of factors associated with BMD Z-score ≤ -2 (Table 3), we found that correlates of low BMD included low BMI Z-score for age and gender, a history of fracture, a history of 25OHD insufficiency and a family history of osteoporosis. Additionally, we found that female subjects who had not reached menarche at the time of the DXA assessment were at increased risk of low BMD, while amenorrhea was not identified as a risk factor. For young women who had reached menarche, there was no difference in average age of menarche between the two groups (13.1 ± 1.6 years v. 12.8 ± 1.8, p = 0.646).

ubjects who had not reached menarche at the time of the DXA assessment were at increased risk of low BMD, while amenorrhea was not identified as a risk factor. For young women who had reached menarche, there was no difference in average age of menarche between the two groups (13.1 ± 1.6 years v. 12.8 ± 1.8, p = 0.646). Medical diagnoses associated with low BMD included cerebral palsy or non-ambulatory state, a history of organ or bone marrow transplant (BMT), treatment with anticonvulsants, and a history of prematurity, defined herein as gestational age < 36 weeks (all at p < 0.05). We did not find a difference (p = 0.27) in duration of anticonvulsant use between subjects with BMD Z-score ≤ -2 (67.9 ± 53.5 months) and those with Z-score > -2 (45.7 ± 45.1 months). Our sample size was too small to analyze the associations between specific anticonvulsants and low BMD. Unexpectedly, both report of a history of a gastrointestinal disease as an indication for DXA (p < 0.01) and history of malabsorption or inflammatory bowel disease (p < 0.01) were associated with a normal BMD. For all subjects with a fracture history, we did not find a difference (p = 0.63) in mean fracture number between those with BMD Z-score ≤ -2 (3.0 ± 5.0) and those with Z-scores > - 2 (2.7 ± 2.9). We also examined whether the relation between low BMD and fracture history was modified by gender, and found that a history of fracture was associated with an increased risk of a low BMD for males (p < 0.01), but not for females (p = 0.95) (data not shown).

re ≤ -2 (3.0 ± 5.0) and those with Z-scores > - 2 (2.7 ± 2.9). We also examined whether the relation between low BMD and fracture history was modified by gender, and found that a history of fracture was associated with an increased risk of a low BMD for males (p < 0.01), but not for females (p = 0.95) (data not shown). All of the patients with osteogenesis imperfecta (OI) in our cohort were actively treated with bisphosphonates. As a result of bisphosphonate treatment, some subjects with OI had BMD Z-scores that exceeded -2 SD. Additionally, the 5 patients with OI included in this study all had sustained multiple fractures. Since bisphosphonate treatment improves bone density [26-28], we excluded the 5 patients with OI and repeated our univariate analysis to assess the impact of these patients on our results. Without the children with OI, fracture history still remained a risk factor for BMD Z-score ≤ -2 SD only for boys (p = 0.01). As fracture risk may increase as the BMD Z-score falls below -1 [18], we sought to identify factors associated with this BMD threshold as well. Similar to our initial univariate analysis, BMI < 5th percentile remained a predictor of low BMD. Additionally, a history of cerebral palsy or non-ambulatory state and history of BMT were associated with BMD Z-score < -1.

ore falls below -1 [18], we sought to identify factors associated with this BMD threshold as well. Similar to our initial univariate analysis, BMI < 5th percentile remained a predictor of low BMD. Additionally, a history of cerebral palsy or non-ambulatory state and history of BMT were associated with BMD Z-score < -1. As there is uncertainty in characterizing (25OHD) values between 20-30 ng/mL, we considered both of these values in assessing for vitamin D insufficiency. For subjects with 25OHD < 30 ng/mL, we did not find a difference (p = 0.24) between mean 25OHD measurements in those with BMD Z-score ≤ -2 (19.3 ± 6.9 ng/mL, n = 67) and those with Z-score > -2 (21.6 ± 14.6 ng/mL, n = 90). 44.8% of subjects with BMD Z-score ≤ -2 and 41.1% of those with Z-score > -2 had a 25OHD measurement < 20 mg/mL, while 66.7% of subjects with BMD Z-score ≤ -2 and 45.1% of those with Z-score > -2 had a 25OHD measurement < 30 mg/mL.

≤ -2 (19.3 ± 6.9 ng/mL, n = 67) and those with Z-score > -2 (21.6 ± 14.6 ng/mL, n = 90). 44.8% of subjects with BMD Z-score ≤ -2 and 41.1% of those with Z-score > -2 had a 25OHD measurement < 20 mg/mL, while 66.7% of subjects with BMD Z-score ≤ -2 and 45.1% of those with Z-score > -2 had a 25OHD measurement < 30 mg/mL. A multivariable logistic regression model was constructed to examine factors associated with a BMD Z-score ≤ -2 (Table 4). Significant predictors include low BMI Z-score, low height Z-score, a 25OHD measurement < 30 ng/mL, and history of a BMT. BMI Z-score was inversely related to the risk of low BMD, with an odds ratio (OR) of 0.52 (95% confidence interval (CI) 0.39 – 0.69). Height Z-score was also inversely related to the risk of low BMD (OR 0.71, CI 0.57, 0.88). We found a notable clinical association between history of 25OHD insufficiency and low BMD, with an almost 4-fold increased risk of low BMD (OR 3.97, CI 2.08, 7.59). In comparison, a history of BMT led to a nearly 6-fold increased risk of low BMD (OR 5.78, CI 1.00, 33.45). Similar to our findings in the univariate analysis, malabsorption or IBD was found to be associated with a normal BMD Z-score in the multivariate analysis. There were no interactions among predictors in the final model. The c-statistic, used to assess the predictive validity of the model, was 0.789. Table 4 Multivariate analysis: Risk factors associated with BMD Z-score ≤ -2

A multivariable logistic regression model was constructed to examine factors associated with a BMD Z-score ≤ -2 (Table 4). Significant predictors include low BMI Z-score, low height Z-score, a 25OHD measurement < 30 ng/mL, and history of a BMT. BMI Z-score was inversely related to the risk of low BMD, with an odds ratio (OR) of 0.52 (95% confidence interval (CI) 0.39 – 0.69). Height Z-score was also inversely related to the risk of low BMD (OR 0.71, CI 0.57, 0.88). We found a notable clinical association between history of 25OHD insufficiency and low BMD, with an almost 4-fold increased risk of low BMD (OR 3.97, CI 2.08, 7.59). In comparison, a history of BMT led to a nearly 6-fold increased risk of low BMD (OR 5.78, CI 1.00, 33.45). Similar to our findings in the univariate analysis, malabsorption or IBD was found to be associated with a normal BMD Z-score in the multivariate analysis. There were no interactions among predictors in the final model. The c-statistic, used to assess the predictive validity of the model, was 0.789. Table 4 Multivariate analysis: Risk factors associated with BMD Z-score ≤ -2 Factor Odds ratio (95% CI) p BMI z-score   <0.001 1 unit increase 0.52 (0.39, 0.69) Height z-score   0.002 1 unit increase 0.71 (0.57, 0.88) Vitamin D Insufficiency   <0.001 Yes 3.97 (2.08, 7.59) Not Documented/No 1.00 Malabsorption or IBD   0.001 Yes 0.29 (0.14, 0.60) No 1.00 Recipient of a bone marrow transplant   0.050 Yes 5.78 (1.00, 33.45) No 1.00 BMD Z-score ≤ -2, N = 86; BMD Z-score > -2, N = 196. c statistic for model = 0.789.

Factor Odds ratio (95% CI) p BMI z-score   <0.001 1 unit increase 0.52 (0.39, 0.69) Height z-score   0.002 1 unit increase 0.71 (0.57, 0.88) Vitamin D Insufficiency   <0.001 Yes 3.97 (2.08, 7.59) Not Documented/No 1.00 Malabsorption or IBD   0.001 Yes 0.29 (0.14, 0.60) No 1.00 Recipient of a bone marrow transplant   0.050 Yes 5.78 (1.00, 33.45) No 1.00 BMD Z-score ≤ -2, N = 86; BMD Z-score > -2, N = 196. c statistic for model = 0.789. Discussion Multiple underlying health problems and associated treatments can impair bone mineral accrual during childhood and adolescence. We sought to identify the risk factors associated with low bone density in subjects referred for DXA measurements of BMD at a tertiary care pediatric hospital to help inform a bone health patient safety initiative. Using a multivariable logistic regression model, we found that low BMI Z-score, history of BMT, and 25OHD insufficiency were the medical conditions most predictive of a BMD Z-score ≤ -2, a threshold deemed by expert consensus to represent a significant low bone density for age [23].

ne health patient safety initiative. Using a multivariable logistic regression model, we found that low BMI Z-score, history of BMT, and 25OHD insufficiency were the medical conditions most predictive of a BMD Z-score ≤ -2, a threshold deemed by expert consensus to represent a significant low bone density for age [23]. We identified that history of BMT significantly increases the risk of low BMD. Multiple mechanisms have been postulated to mediate the low BMD seen after BMT, including glucocorticoid use, chemotherapy, irradiation, hypogonadism, decreased activity levels, and chronic graft versus host disease [29]. Prior studies support our findings, demonstrating that BMT during childhood increases the risk of low BMD and fracture [29,30], and that bone formation is decreased and bone resorption is increased in the months immediately following BMT [31].

dism, decreased activity levels, and chronic graft versus host disease [29]. Prior studies support our findings, demonstrating that BMT during childhood increases the risk of low BMD and fracture [29,30], and that bone formation is decreased and bone resorption is increased in the months immediately following BMT [31]. Our multivariate analysis demonstrated that 25OHD insufficiency (< 30 ng/mL) is associated with a nearly 4-fold increased risk of low BMD. This finding is consistent with the observation of Cheng et al. demonstrating that low 25OHD levels correlate with lower cortical BMD in pre-pubertal and pubertal girls [32]. Unlike many of the underlying medical conditions included in our analysis, 25OHD insufficiency may be improved with 25OHD supplementation and thus represents an area amenable to intervention for improved bone health in children and adolescents. Our selection of values less than 30 ng/mL for defining 25OHD deficiency was based upon the lower limit of the normal range of our hospital laboratory. After we designed and initiated our study, the Institute of Medicine reported on dietary reference intakes for calcium and 25OHD, and concluded that an adequate 25OHD level is ≥ 20 ng/mL [33]. However, controversy regarding adequate serum 25OHD levels remains (e.g., above 20 or 30 ng/mL or higher) [34] and additional research is needed to identify the optimal therapeutic target.

edicine reported on dietary reference intakes for calcium and 25OHD, and concluded that an adequate 25OHD level is ≥ 20 ng/mL [33]. However, controversy regarding adequate serum 25OHD levels remains (e.g., above 20 or 30 ng/mL or higher) [34] and additional research is needed to identify the optimal therapeutic target. We found that low BMI is a significant risk factor for low BMD, yet despite the robust relationship between underweight and low BMD, we did not find a relationship between malnutrition and low bone density. Prior studies have demonstrated a strong relationship between malnutrition in young women with anorexia nervosa, a model of malnutrition-induced bone loss, and low BMD [35-37], which we did not observe in this sample. Bone density in these patients can vary depending on weight loss or gain, and severity of disease. The patients with anorexia nervosa included in this study could have represented a skewed subgroup, with less severe disease and bone loss. A prospective study of larger sample size would likely be powered to reveal the expected relationship between malnutrition and low BMD.

eight loss or gain, and severity of disease. The patients with anorexia nervosa included in this study could have represented a skewed subgroup, with less severe disease and bone loss. A prospective study of larger sample size would likely be powered to reveal the expected relationship between malnutrition and low BMD. The multivariate analysis demonstrated that low height Z-score is a risk factor for low BMD. Any child with an underlying health problem may have short stature, confounding bone density measurements obtained by DXA [38]. It is important to recognize that DXA measurements may be confounded by bone size, an important issue to consider when studying small underweight adolescents with anorexia nervosa or other pediatric chronic diseases [23]. Zemel et al. recently proposed a method for height adjustment in calculating DXA measurements of bone mass in growing children [39] which will be useful to account for the effect of stature on BMD measurement. Despite these limitations in smaller children, our findings suggest that children who are underweight may be at risk for decreased BMD.

posed a method for height adjustment in calculating DXA measurements of bone mass in growing children [39] which will be useful to account for the effect of stature on BMD measurement. Despite these limitations in smaller children, our findings suggest that children who are underweight may be at risk for decreased BMD. Consistent with prior reports of gender differences in fracture risk [40,41], we found that fractures occurred in a higher percentage of boys than girls in a population referred for DXA measurement. This gender difference has been attributed to higher rates of sports-related injuries in boys [40,41]. Surprisingly, in our univariate analysis of factors associated with BMD Z-score ≤ - 2, we found that male gender was a risk factor for low BMD. This finding is questionable, as there is a known association between male gender and higher BMD [42], and as gender is no longer predictive in the multivariate model, there may be other confounders influencing this finding. As our subjects were selected from a retrospective sample of patients referred for DXA measurements, this observation may also represent selection bias and merits further study.

and higher BMD [42], and as gender is no longer predictive in the multivariate model, there may be other confounders influencing this finding. As our subjects were selected from a retrospective sample of patients referred for DXA measurements, this observation may also represent selection bias and merits further study. Prior studies have demonstrated that IBD is both associated with bone loss [43-45], and increased fracture risk in adults [46]. Thus, the finding that a history of malabsorption or IBD was associated with decreased risk of low BMD in the univariate and multivariate analyses was unexpected. The retrospective design of our study may have introduced bias, or referral bias may have influenced the selection of subjects with IBD referred for DXA. Additionally, the timing of the BMD measurements in relation to the IBD diagnosis and associated malabsorption and therapy may have contributed to this finding.

cted. The retrospective design of our study may have introduced bias, or referral bias may have influenced the selection of subjects with IBD referred for DXA. Additionally, the timing of the BMD measurements in relation to the IBD diagnosis and associated malabsorption and therapy may have contributed to this finding. Limitations of this study should be acknowledged. The study design was a retrospective cohort study; therefore, the findings represent associations and causality cannot be inferred. Our results are limited by information obtained solely from subjects referred for DXA measurements at a tertiary care medical center. However, DXA is the assessment tool used most commonly in clinical practice around the world for bone density measurements in children and adults. Thus, we are hopeful that the information gleaned will be useful for clinicians, and in particular, experts in bone health. Measurements of volumetric BMD of the peripheral skeleton were not available and would be valuable to obtain, especially in a cohort of young patients, although this tool is only infrequently used at the present time for clinical purposes. By design, our study was not able to include or identify subjects who were not able to have DXA measurements performed. The retrospective nature of our study required us to rely upon diagnoses and referrals made by providers and the samples sizes in many diagnostic categories were small, limiting our ability to detect differences in BMD. Additionally, due to the reliance on the medical record, we were not able to include assessments of physical activity. We acknowledge that physical activity data would have been useful, as exercise may positively impact BMD [18] and chronic disease often limits activity for a child or adolescent. Obtaining an accurate family history of osteoporosis was also limited by the retrospective nature of this study, as some charts did not include documentation of this information. However, despite these limitations, our results identified the relative risks of a variety of medical conditions known to be associated with low BMD in a sample of patients referred for DXA at a tertiary care center. The information gained from our study highlights pediatric populations who may be at highest risk for a low BMD and may be candidates for bone density screening.

the relative risks of a variety of medical conditions known to be associated with low BMD in a sample of patients referred for DXA at a tertiary care center. The information gained from our study highlights pediatric populations who may be at highest risk for a low BMD and may be candidates for bone density screening. Conclusions We used a novel approach to identify factors which lead to the highest risk of low BMD in a pediatric cohort referred for DXA. Prior studies have focused on BMD in specific populations, such as children and young adults with a particular diagnosis or those receiving a certain therapy. Our findings suggest that for all children referred for DXA, a low BMI Z-score, a history of BMT, and vitamin D insufficiency are significant risk factors for a low BMD. Identification of the medical conditions and therapies that lead to the greatest risk of low BMD in a pediatric population is especially important, given that adequate bone mineral accretion during childhood may impact the peak bone mass that is achieved during adulthood, especially for children with chronic medical conditions or therapies. Additionally, we hope that this information can be used to help decrease the frequency of fragility fractures in hospitalized children. Earlier identification of risk factors could potentially lead to the initiation of strategies that maximize bone health during childhood and adolescence and ultimately reduce the future risk for developing osteoporosis.

ormation can be used to help decrease the frequency of fragility fractures in hospitalized children. Earlier identification of risk factors could potentially lead to the initiation of strategies that maximize bone health during childhood and adolescence and ultimately reduce the future risk for developing osteoporosis. Abbreviations BMD: Bone mineral density; DXA: Dual-energy x-ray absorptiometry; BMI: Body mass index; OR: Odds ratio; CI: Confidence interval; SD: Standard deviation; IBD: Inflammatory bowel disease; BMT: Bone marrow transplant. Competing interests None of the authors have any competing interest relevant to the topic of this manuscript. Catherine Gordon reports that she was the Co-Director of the Clinical Investigator Training Program through Harvard/MIT with support of Pfizer and Merck, through June 2012. Authors’ contributions All authors meet criteria for authorship, have participated in the writing of the manuscript, and have seen and approved the final version. LST and C. Giancaterino wrote the first draft of the manuscript and were responsible for data acquisition. LST and C.Gordon developed the concept and design of the study. PM contributed to study design, writing of the manuscript, and was responsible for statistical analysis. C.Gordon had senior responsibility for overall data interpretation, analysis, and writing. All authors provided critical review and revision for important content. All authors read and approved the final manuscript.

PM contributed to study design, writing of the manuscript, and was responsible for statistical analysis. C.Gordon had senior responsibility for overall data interpretation, analysis, and writing. All authors provided critical review and revision for important content. All authors read and approved the final manuscript. Acknowledgements We thank Anisa Djermoun for her assistance with data entry and Dionne Graham for helpful biostatistical advice regarding the design of this study.

Introduction A recent analysis of the Healthcare Cost and Utilization Project National Inpatient Sample (HCUP-NIS) database evaluated outcomes in children undergoing thyroid and parathyroid surgery [1]. Pediatric patients were observed to have higher complication rates than those reported for adults [1,2]. Pediatric complication rates were lower when the surgery was performed by a high volume endocrine surgeons, defined as performing >30 cases per year [3-5], but remained higher than those reported for adults [1,3,6]. We hypothesized that an endocrine surgical team consisting of a high volume endocrine surgeon, a pediatric surgeon, a pediatric endocrinologist and an experienced endocrine surgery nurse could attenuate the outcome gap between pediatric and adult surgical patients. To evaluate the outcome of this approach, we compared the outcomes of 100 consecutive pediatric and adult patients who underwent thyroidectomy for Graves’ disease (GD) at a single institution by high volume adult and pediatric endocrine surgery team.

enuate the outcome gap between pediatric and adult surgical patients. To evaluate the outcome of this approach, we compared the outcomes of 100 consecutive pediatric and adult patients who underwent thyroidectomy for Graves’ disease (GD) at a single institution by high volume adult and pediatric endocrine surgery team. Methods The Yale Pediatric Thyroid Center is a multidisciplinary team composed of a high volume endocrine surgeon (greater than 200 thyroidectomies/year), a pediatric surgeon, pediatric endocrinologists, pediatric anesthesiologists and experienced nursing and supportive staff. The Yale Pediatric Thyroid Center sees about 300 pediatric patients per year with thyroid disorders and approximately 30 pediatric patients undergo thyroid surgery annually. All pediatric patients who underwent total thyroidectomy for GD as part of the Yale Pediatric Thyroid Center between June 2002 and November 2010 were entered into a database. Adult patients operated on over the same period were identified, as well from CPT codes (60240, 60252). All pediatric procedures were performed by both attending surgeons. All but one patient were treated pre-operatively with supersaturated potassium iodide (SSKI) for at least 5 days prior to surgery. The one pediatric patient not treated fully with SSKI developed an allergic reaction after two doses and the medication was discontinued.

ocedures were performed by both attending surgeons. All but one patient were treated pre-operatively with supersaturated potassium iodide (SSKI) for at least 5 days prior to surgery. The one pediatric patient not treated fully with SSKI developed an allergic reaction after two doses and the medication was discontinued. Twenty pediatric patients were treated with 0.5 mcg of calcitriol twice daily for 3 days before surgery. Post-operatively the calcitriol was weaned over 15 days (0.5 mcg bid × 5 days; then 0.5 mcg qd × 5 days; then 0.5 mcg qod for 5 days). In all cases either a total or bilateral near-total thyroidectomy was performed. A total thyroidectomy was defined as a complete resection of the thyroid gland via an extracapsular dissection [7,8]. If it is was determined intraoperatively that a complete extracapsular dissection would likely result in irreversible damage to either the recurrent laryngeal nerve (RLN) or parathyroid gland(s), the capsule was entered and a miniscule amount of thyroid tissue was left in situ to avoid injury to either the RLNs or parathyroid glands, a procedure referred to as a near-total thyroidectomy [7,8].

ction would likely result in irreversible damage to either the recurrent laryngeal nerve (RLN) or parathyroid gland(s), the capsule was entered and a miniscule amount of thyroid tissue was left in situ to avoid injury to either the RLNs or parathyroid glands, a procedure referred to as a near-total thyroidectomy [7,8]. All patients were admitted post-operatively. Serum calcium levels were measured every 4 to 8 hrs after surgery in the pediatric patients and in all patients on the morning of post-operative day one. Intravenous calcium gluconate therapy was administered if serum calcium levels (corrected for total protein) measured <7.5 mg/dL without symptoms of hypocalcemia or measured <8.0 mg/dL in the presence of symptomatic hypocalcaemia (e.g. hand parasthesias, perioral numbness, muscle cramps, the presence of a Chvostek’s sign). Pre-operative demographics, comorbidities and medications were recorded. Operative details including detailed contemporaneous operative illustrations, post-operative course, serum calcium levels and operative pathology were recorded. Permanent recurrent laryngeal nerve (RLN) injury, hematoma and permanent hypoparathyroidism were classified as major complications. Transient hypocalcemia requiring intravenous calcium infusion and transient RLN neuropraxia were considered to be minor complications. Statistical analysis was performed using Students’t-test and Fisher’s exact test as appropriate. Statistical significance was determined when p < 0.05. Values shown are mean ± standard deviation. The Yale Human Investigations Committee approved this study.

Pre-operative demographics, comorbidities and medications were recorded. Operative details including detailed contemporaneous operative illustrations, post-operative course, serum calcium levels and operative pathology were recorded. Permanent recurrent laryngeal nerve (RLN) injury, hematoma and permanent hypoparathyroidism were classified as major complications. Transient hypocalcemia requiring intravenous calcium infusion and transient RLN neuropraxia were considered to be minor complications. Statistical analysis was performed using Students’t-test and Fisher’s exact test as appropriate. Statistical significance was determined when p < 0.05. Values shown are mean ± standard deviation. The Yale Human Investigations Committee approved this study. Results Thirty two children and 68 adults underwent total or near-total thyroidectomy. The clinical characteristics of the patients are described in Table 1. The mean age at time of surgery was 9.7 years (range 3.4-17.9 yrs) for children and 44.9 years (range 18.4-84.2 yrs) for adults. Females constituted 81.3% of the children and 86.7% of the adult patients. Table 1 Clinical characteristics of adult and pediatric patients Adult Pediatric p-value Patients 68 32 Patients requiring IV Calcium 1 6 0.004* RLN Injuries 0 1 0.32* Hematomas 2 0 0.46* Incidental Cancers 12 1 0.01* Thyroid mass (g)/Body mass (kg) 0.67 0.94 0.05† Operative Time (hour:min) 1:37 2:09 0.001† Inpatient Post Operative Days 1.03 1.41 0.004† * - Fisher's Exact test. † - Student's t-test.

Adult Pediatric p-value Patients 68 32 Patients requiring IV Calcium 1 6 0.004* RLN Injuries 0 1 0.32* Hematomas 2 0 0.46* Incidental Cancers 12 1 0.01* Thyroid mass (g)/Body mass (kg) 0.67 0.94 0.05† Operative Time (hour:min) 1:37 2:09 0.001† Inpatient Post Operative Days 1.03 1.41 0.004† * - Fisher's Exact test. † - Student's t-test. Pediatric procedures required longer operative time than those of adults (1.18 ± 0.08 hrs vs. 2.09 ± 0.03 hrs, p = 0.003). Pediatric thyroid specimens averaged 38.6.0 ± 8.9 gm (range 9–293 gm) (p = 0.34) and adult thyroid specimens averaged 48.0 ± 6.4 gm (range 6.6-203 gm) (p = 0.34). When the thyroidectomy specimen weight was considered in relation to that of the patient, the pediatric patients had larger glands per kilogram of body mass as compared to the adults (0.94 ± 0.11 gm/kg vs. 0.66 ± 0.07 gm/kg, p = 0.05) Pathological analysis of the operative specimens revealed occult malignancy in one of the 32 pediatric patients and in 12 of the 68 adult patients (p = 0.03). The one pediatric patient had a papillary microcarcinoma of 0.4 cm in greatest diameter confined to the thyroid capsule. In the adults, 10 of 12 malignancies represented papillary microcarcinomas that were less than 0.3 cm in greatest diameter. One case was a multifocal papillary carcinoma that included a papillary carcinoma that was 2.5 cm in greatest diameter. There was one Hürthle cell adenoma in an adult patient.

the thyroid capsule. In the adults, 10 of 12 malignancies represented papillary microcarcinomas that were less than 0.3 cm in greatest diameter. One case was a multifocal papillary carcinoma that included a papillary carcinoma that was 2.5 cm in greatest diameter. There was one Hürthle cell adenoma in an adult patient. Pediatric patients were more likely to require postoperative intravenous calcium infusion than adults (18.0% vs. 1.4%, p = 0.004). Pediatric patients required a longer length of stay than adults (1.41 ± 0.12 vs. 1.03 ±0.03 days, p = 0.004) reflecting the greater need for calcium infusions. When preoperative calcitriol was not given 50% of pediatric patients required calcium infusion with a mean duration of infusion of 1.5 ±0.3 days. When calcitriol was given, 16% of patients required calcium infusion with a mean duration of infusion of 1.25 days (p 0.02 vs. no caclitriol). There was one case of transient RLN neuropraxia in a child that recovered within 6 months after surgery. One child sustained a RLN transaction during surgery that was recognized and repaired during the initial procedure. Long term follow-up revealed that RLN function was restored as evidenced by bilateral vocal cord mobility 12 months after surgery. There were no RLN injuries in the adults.

within 6 months after surgery. One child sustained a RLN transaction during surgery that was recognized and repaired during the initial procedure. Long term follow-up revealed that RLN function was restored as evidenced by bilateral vocal cord mobility 12 months after surgery. There were no RLN injuries in the adults. There were 2 post-operative hematomas in the adult group requiring operative exploration and none in the pediatric group (p = 0.46). There was no operative or perioperative mortality, and no cases of permanent hypoparathyroidism. Follow up data indicated no recurrences of hyperthyroidism in either children or adults over up to a maximal follow-up period of 8 years. The overall rates of major complications were indistinguishable comparing children and adults. Discussion Graves’ disease is rare in children with a prevalence of 1 in 10,000 [9,10]. Because only a minority of pediatric patients achieve remission [11-14], the majority of patients will require either radioactive iodine or surgery [4,9]. Our observations support the notion that in the hands of high volume thyroid surgeons, thyroidectomy is an appropriate treatment option for pediatric patients.

,10]. Because only a minority of pediatric patients achieve remission [11-14], the majority of patients will require either radioactive iodine or surgery [4,9]. Our observations support the notion that in the hands of high volume thyroid surgeons, thyroidectomy is an appropriate treatment option for pediatric patients. To date the largest series detailing the surgical treatment of GD in children is from the Mayo Clinic [15]. Seventy-eight pediatric patients, ages 3.1 to 17.9 years, with GD underwent surgical treatment [15]. Sixty patients (77%) underwent total or near-total thyroidectomy; 18 patients (23%) underwent bilateral subtotal thyroidectomy [15]. When a sub-total thyroidectomy is performed, a small amount of thyroid tissue is left behind in the region of the RLN or parathyroid glands in the hope of minimizing risks of RLN injury or hypoparathyroidism [8]. In the Mayo Clinic series, no RLN injuries, hematomas, nor instances of permanent hypoparathyroidism were reported [15]. One patient (1%) developed transient RLN neuropraxia, and 5 patients (6%) had transient hypoparathyroidism [15]. Five microcarcinomas (6%) were identified in the pathological specimens [15]. Of note, 4 of the 18 patients (22%) who underwent bilateral subtotal thyroidectomy experienced recurrence of their hyperthyroidism [15].

ient (1%) developed transient RLN neuropraxia, and 5 patients (6%) had transient hypoparathyroidism [15]. Five microcarcinomas (6%) were identified in the pathological specimens [15]. Of note, 4 of the 18 patients (22%) who underwent bilateral subtotal thyroidectomy experienced recurrence of their hyperthyroidism [15]. Recognizing the high recurrence rate associated with subtotal thyroidectomy, we adopted an aggressive surgical approach for the management of GD in children and adults by performing total or near-total thyroidectomies in all patients. Supporting the utility of our approach in curing the disease, no patient in our series experienced relapse of their GD. HCUP-NIS data show that pediatric patients suffer complications more frequently following thyroidectomy than adults (9.3% vs. 6.1%, p < 0.01) [1,2]. In our cohort, we observed transient hypocalcemia, presumable due to transient hypoparathyroidism, as the most common minor complication in pediatric patients. However, with preoperative calcitriol therapy, the need for post-operative calcium infusions was reduced from 50% to 16% and the duration of intravenous calcium infusion was shortened by more than 50%. RLN injury was observed in two of the children in our cohort. In one child, a soft voice was observed shortly after surgery, and direct laryngoscopy performed two months after surgery revealed unilateral vocal cord paresis. At six months after surgery, there was normal bilateral vocal cord function and the voice was normal.

njury was observed in two of the children in our cohort. In one child, a soft voice was observed shortly after surgery, and direct laryngoscopy performed two months after surgery revealed unilateral vocal cord paresis. At six months after surgery, there was normal bilateral vocal cord function and the voice was normal. One child, 4 years of age, had transection of one RLN during surgery that was immediately recognized and repaired. At 6 months after surgery, unilateral vocal cord paresis was observed by direct laryngoscopy. At 18 months after surgery, direct laryngoscopy revealed bilateral vocal cord movement and the voice was normal. Of note, the RLN injury occurred in the lone patient who could not be treated preoperatively with SSKI due to an anaphylactic reaction to the medication. It is interesting that occult malignancy was seen in 12 of 68 adult patients and in 1 of 32 pediatric patients. Ten of the 12 malignancies seen in adults were unifocal papillary microcarcinomas, and one case was a multifocal papillary carcinoma that was 2.5 cm in the greatest diameter. A papillary microcarcinoma of 0.4 cm confined to the thyroid capsule was seen in one pediatric patient. These observations highlight the need for clinicians to obtain sonographic imaging of the thyroid gland if asymmetry or gland size change is noted in the setting of GD. Should nodules be observed, they should be evaluated per recent guidelines [16].

cm confined to the thyroid capsule was seen in one pediatric patient. These observations highlight the need for clinicians to obtain sonographic imaging of the thyroid gland if asymmetry or gland size change is noted in the setting of GD. Should nodules be observed, they should be evaluated per recent guidelines [16]. The observation of microcarcinomas in the setting of GD, which will be difficult to identify by preoperative ultrasound, should not be construed as an argument to in favor of surgery over radioactive iodine in the definitive treatment of GD. If GD patients with microcarcinomas are treated with recommended activities of radioactive iodine that are intended to ablate thyroid tissue [4,17,18], it is anticipated that microcarcinomas will be destroyed along with normal thyroid tissue. Support for this notion come from data from the Thyrotoxicosis Study Group follow-up data showing that rates of differentiated thyroid cancer in patients with GD are substantially lower in those treated with radioactive iodine or surgery than in individuals treated with antithyroid medications alone who underwent spontaneous remission [19].

n come from data from the Thyrotoxicosis Study Group follow-up data showing that rates of differentiated thyroid cancer in patients with GD are substantially lower in those treated with radioactive iodine or surgery than in individuals treated with antithyroid medications alone who underwent spontaneous remission [19]. Overall, we demonstrate that total or near-total pediatric thyroidectomy can be performed safely in experienced hands and is not associated with GD recurrence. A multidisciplinary team consisting of a high volume endocrine surgeon, a pediatric surgeon, pediatric endocrinologists, pediatric anesthesiologists, and experienced endocrine nurses can achieve excellent outcomes with very low complication rates. These observations support recent recommendations that children and adults requiring surgery for GD should be operated on by high volume endocrine surgical teams [4,5]. Competing interests The authors declare that they have no competing interests. Authors’ contributions SR, CB conceived and participated in the design of the study. RU, CB and SR performed the surgeries. DS and PD participated in the sequence and coordination of the study. SR and RU helped draft the manuscript. All authors read and approved the final manuscript.

Background The Working Party on Disorders of Sex Development (DSD) met in Annecy, France from March 15 to 17, 2012 to explore objective data sets to better describe the presentation, treatment and long-term outcome of patients with DSD. One goal was to identify objective data that could be validated and then applied internationally. Ideal outcomes would include normal adult phenotypes, functionality and quality of life. This conference aimed to identify the important issues concerning the evaluation of DSD patients, by assessing needs in three primary areas: diagnosis, treatment and outcome. This workshop was not a consensus meeting and was not designed to develop clinical guidelines, but rather sought to develop a status update on DSD, to be achieved by presenting and discussing relevant data. The Journal of Pediatric Urology [1] has published a supplement issue including referenced information from these three pertinent areas.

consensus meeting and was not designed to develop clinical guidelines, but rather sought to develop a status update on DSD, to be achieved by presenting and discussing relevant data. The Journal of Pediatric Urology [1] has published a supplement issue including referenced information from these three pertinent areas. The Chicago consensus conference [2] was frequently referenced regarding diagnosis, medical/surgical treatment and outcome information in DSD. Over the past decade and particularly since the consensus conference, there continues to be a major shift in gender assignment of 46,XY DSD patients to a male gender assignment when there is evidence of testicular function and in utero androgen exposure [3,4]. Conversely, however, the failure of the consensus statement to address gender assignment for 46XX individuals with markedly masculinized external genitalia has largely led to the female assignment because of the interpretation of the default statement regarding sex of rearing for 46XX CAH patients with Prader 4 or 5 genitalia. Recent outcome information has provided a strong rational for consideration of a male assignment among such individuals [5,6]. Hence, this provides an example where the consensus conference notation of inadequate information is being supplemented by additional -- but as yet incomplete -- data.

Prader 4 or 5 genitalia. Recent outcome information has provided a strong rational for consideration of a male assignment among such individuals [5,6]. Hence, this provides an example where the consensus conference notation of inadequate information is being supplemented by additional -- but as yet incomplete -- data. The purpose of this publication is to discuss four main topics from this conference that may help to provide a broader perspective to practicing pediatric endocrinologists. These include; 1) complexities of hypospadias, 2) problems related primarily to the surgical treatment of virilized genitalia of 46,XX DSD individuals because of anatomic variability and difficulties of visualing and defining the internal anatomy, 3) advances in phalloplasty and 4) updated information on psychological, social and sexual outcomes of DSD individuals. Each section contains comments and suggestions relayed from the conference regarding next steps in improving this knowledge and treatment, particularly surgical treatment for children with DSD.

atomy, 3) advances in phalloplasty and 4) updated information on psychological, social and sexual outcomes of DSD individuals. Each section contains comments and suggestions relayed from the conference regarding next steps in improving this knowledge and treatment, particularly surgical treatment for children with DSD. Hypospadias The Masculinization Programming Window (MPW) based on evidence from rodents is a period early in fetal life that determines phallic development and growth potential. Diminished masculinization may result in hypospadias and may be associated with a diminished anogenital distance (AGD), the distance between the anus and the edge of scrotum, in the male [7]. However, the pragmatic application of the impact of the MPW is unclear, including those with hypospadias since care is directed toward producing as functional and normal sized phallus as possible.

ciated with a diminished anogenital distance (AGD), the distance between the anus and the edge of scrotum, in the male [7]. However, the pragmatic application of the impact of the MPW is unclear, including those with hypospadias since care is directed toward producing as functional and normal sized phallus as possible. It is important to recognize that ‘distal hypospadias’ as defined by the urologist is not synonymous with what a pediatric endocrinologist would call ‘distal’ or ‘coronal; or ‘glanular’ hypospadias’. To the pediatric urologist, distal hypospadias indicates a phallus with a urethral meatus on the distal half of the penis. It is unclear which patients with distal hypospadias (as defined by urologists) will benefit from surgery not necessarily because of the programming of the MPW but because the status of underlying tissues cannot be determined until exploration at surgery. Hence, surgical approach and outcomes cannot be predicted. Current indications for surgical repair are presence of abnormal urinary stream and/or significant chordee. Once exploration has begun, if primary tubulization can be used, a low complication rate can be expected. While surgical techniques [8,9], are continually being refined, anatomic defects remain difficult to ascertain until visualization during surgery.

are presence of abnormal urinary stream and/or significant chordee. Once exploration has begun, if primary tubulization can be used, a low complication rate can be expected. While surgical techniques [8,9], are continually being refined, anatomic defects remain difficult to ascertain until visualization during surgery. This defect involving hypospadias of the distal half of the penis is associated with a narrow urethral plate that results from the lack of fusion of the corpus spongiosum overlying the urethra (Figure 1). Since the location of the urethral meatus is an inadequate surrogate of the severity of hypospadias, the underlying anatomy, and therefore the extent of repair required, cannot be inferred by visual or manual inspection. While the extent of this defect may be proportional to the chordee ranging from mild to severe, the extent of corpus spongiosum fusion can only be determined after degloving of the phallus at surgery. The point of fusion of the spongiosum, which often falls short of the point of urethral fusion at the meatus, can then be determined and the surgical approach planned. This area, the urethral plate, consists of a ventral triangular defect with its base at the existing urethral opening and its apex at the point where the corpus spongiosum prematurely divides in an area of fusion with surrounding dysplastic tissue; the urethral plate exists between the base and the apex of the triangular area (Figure 1). Beyond lack of fusion, development is considered good if the spongiosum is developed along the urethral plate with the urethra, but poor if there is a narrow, diminished spongiosum. The extent of fusion of the spongiosum and the overall quality of the urethral plate is used to determine the use of adjacent tissues for repair. Surrounding ventral tissues should be avoided if the urethral plate is dysplastic tissue of poor quality. If possible, dorsal tissue should be used. Post-operative growth of the hypospadiac penis suggests that postnatal growth of the penis is greater proximally than distally since the distal urethra grows at a slower pace after reconstruction of distal hypospadias.

f the urethral plate is dysplastic tissue of poor quality. If possible, dorsal tissue should be used. Post-operative growth of the hypospadiac penis suggests that postnatal growth of the penis is greater proximally than distally since the distal urethra grows at a slower pace after reconstruction of distal hypospadias. Figure 1 Scheme of distal hypospadias with blue triangle outlining the defect of the urethral plate illustrating the more distal urethral meatus at a more distgal point than the point of fusion of the corpus spongiosum. This later is not apparent on physical examination. Figure modified from several presented at meeting in Annecy. Improved descriptions of this anatomy are not only necessary for optimal surgical repair, communication between healthcare providers may also allow for a more meaningful comparison between outcome studies. The extent of urethral fusion and underlying fusion categorized as proximal or distal division of the corpus spongiosum [10] together with a description of the surgery could provide a basis for both surgical guidelines and outcome assessment. Such information is needed for prospective long-term studies utilizing multivariate analyses [11] to permit comparison of the initial anatomy and surgical correction with anatomic, functional and psychosocial outcome considering the perspectives of surgeons, patients, family members and partners.

nd outcome assessment. Such information is needed for prospective long-term studies utilizing multivariate analyses [11] to permit comparison of the initial anatomy and surgical correction with anatomic, functional and psychosocial outcome considering the perspectives of surgeons, patients, family members and partners. Interestingly, available outcome information does not support the concept that post-surgical normalization of genital anatomy leads to a satisfactory outcome. In fact, some with poor outcomes indicate satisfaction, implying that future studies should consider all aspects of normalcy from the patients’ perspective with comparison to other adult controls. Normal variation of the adult penis needs to be considered as the degree of penile curvature following hypospadias repair is similar to control men. In addition, studies need to account for the relatively large portion of the control population who do not have a urethral meatus at tip of the glans. A validated tool involves scoring (from 1 to 4) for 8 criteria: flaccid and erect penis size, penis thickness and meatal position, glans size and shape, penis appearance in general and testis/scrotum [12,13]. General dissatisfaction among repaired men who had hypospadias compared with normal men appears not to be related to anatomy as much as to sexuality. Poorer satisfaction among men who had hypospadias seems to be related to having fewer nocturnal emissions, fewer daytime sexual fantasies, less masturbation, and less foreplay and decreased frequency of sexual activity with fewer partners.

ed with normal men appears not to be related to anatomy as much as to sexuality. Poorer satisfaction among men who had hypospadias seems to be related to having fewer nocturnal emissions, fewer daytime sexual fantasies, less masturbation, and less foreplay and decreased frequency of sexual activity with fewer partners. Urinary outcome reports also indicate general overall dissatisfaction relating to increased incidence of urinary tract infections, urinary stream problems and poor voiding control. Long-term outcome studies indicate more frequent complication rates than indicated by short term assessments and may be as high as 33% [14]. Complications including fistulas, urethral strictures and meatal stenosis become more prevalent the longer the duration of follow-up. Further, overall outcome including appearance, urinary and sexual function cannot be determined until after puberty.

n indicated by short term assessments and may be as high as 33% [14]. Complications including fistulas, urethral strictures and meatal stenosis become more prevalent the longer the duration of follow-up. Further, overall outcome including appearance, urinary and sexual function cannot be determined until after puberty. The need for long-term, evidence- based outcome studies with uniform descriptions of defects at presentation and at surgery, the surgery itself and a multi-faceted assessment of adult anatomy, function, and quality of life is clear. A goal of this conference and the publication therefrom was to provide the basis for such studies1. It was suggested that standardized criteria be required for journal publication involving pre-operative anatomic description, any additional intraoperative revelations, detailed operative procedures, followed by a formalized outcome description. Outcome criteria would document cosmetic appearance compared with pre-surgical status, urinary flow and penile perception scores by the patient (and as age-appropriate-by the parents) based on questionnaire. Anatomic outcome details would include the straightness of erection, site of the meatus, quality of skin, any rotation of skin, glans configuration and peno-scrotal relationship and scrotal description.

nary flow and penile perception scores by the patient (and as age-appropriate-by the parents) based on questionnaire. Anatomic outcome details would include the straightness of erection, site of the meatus, quality of skin, any rotation of skin, glans configuration and peno-scrotal relationship and scrotal description. The issues regarding hypospadias reviewed point out the key problem that the anatomic development of the complex development of the penile urethra cannot be determined before exposure at the time of sugery. Hence, the need for further research currently include a details description of the actual anatomy, involving primarily the development of the urethral triangle underlying the surface anatomy since the urethral meatus is often not indicative of the underlying development. Further, research needs to involve the description of new or previous surgical techniques with modifications, together with the various aspects of long term outcome regarding appearance and function.

triangle underlying the surface anatomy since the urethral meatus is often not indicative of the underlying development. Further, research needs to involve the description of new or previous surgical techniques with modifications, together with the various aspects of long term outcome regarding appearance and function. Virilized genitalia in 46,XX DSD individuals: anatomic description Determination of anatomy of genital ambiguity While sonography of normal genitalia at 12 weeks gestation can differentiate male from female based on the angle of the phallus, even though phallic size differences between male and female are not present until late gestation. However, sufficient visualization of the internal reproductive anatomy in those with ambiguous genitalia is difficult and can be misleading at any age making preoperative planning difficult. The urinary outlet and vaginal introitus may each opening separately into the urogenital sinus or may create a urogenital confluence very high in the pelvis with the vagina opening into the urethra just below the bladder (Figure 2). Generally the more masculine the urethral formation, the more proximal and smaller the vagina but consistency between virilization as suggested by Prader staging and internal anatomy relationships cannot be assumed. Also, larger phallic size does not predict the extent of fusion of the labial folds. Hence, a detailed description strategy beyond Prader staging is needed.

the more proximal and smaller the vagina but consistency between virilization as suggested by Prader staging and internal anatomy relationships cannot be assumed. Also, larger phallic size does not predict the extent of fusion of the labial folds. Hence, a detailed description strategy beyond Prader staging is needed. Figure 2 Urogenital sinus anatomy: high (left) and low (right) confluence of vagina and urethra which exit as a common channel. Unrestricted use from open access journal: Leslie JA, Cain MP, Rink RC. Feminizing genital reconstruction in congenital adrenal hyperplasia, Indian J Urol 2009; 25 (1):17–26.

the more proximal and smaller the vagina but consistency between virilization as suggested by Prader staging and internal anatomy relationships cannot be assumed. Also, larger phallic size does not predict the extent of fusion of the labial folds. Hence, a detailed description strategy beyond Prader staging is needed. Figure 2 Urogenital sinus anatomy: high (left) and low (right) confluence of vagina and urethra which exit as a common channel. Unrestricted use from open access journal: Leslie JA, Cain MP, Rink RC. Feminizing genital reconstruction in congenital adrenal hyperplasia, Indian J Urol 2009; 25 (1):17–26. Need for complete descriptions In terms of both surgical decision making and generation of outcome study datasets it is critical that a complete description of the pre-surgical and post-surgical genitalia be documented including phallic size, the presence of 1 or 2 orifices, the configuration and extent of fusion, symmetry, and relative tissue quantity of both pairs of labial folds. Likewise, a complete description of internal anatomy includes visualization at onset of surgery documenting the presence and position of gonads, uterine size, vaginal length and location in relation to the bladder neck, urethra and urogenital sinus. Anogenital distance is an indication of extent of androgen exposure during early fetal life. The degree of masculinization of the bony pelvis should be documented since the vaginal angle may be altered enough to cause problems with intercourse. As described for hypospadias above, a record of all information preoperatively and during surgery provides the basis for the development of standardized protocols for treatment including surgery and key factors to be considered in outcome studies.

vaginal angle may be altered enough to cause problems with intercourse. As described for hypospadias above, a record of all information preoperatively and during surgery provides the basis for the development of standardized protocols for treatment including surgery and key factors to be considered in outcome studies. MRI or endoscopy may be useful in identification of confluence of the urethra and vagina, but genitography was not felt to be generally helpful. Ultrasound studies or MRI can be used for distance measurements for descriptions of the internal phenotype. Length and width of the urogenital sinus and visualized internal structures such as the uterus should be included. However, the location of the urethral sphincter, an important structure to be avoided at surgery, usually cannot be determined.

be used for distance measurements for descriptions of the internal phenotype. Length and width of the urogenital sinus and visualized internal structures such as the uterus should be included. However, the location of the urethral sphincter, an important structure to be avoided at surgery, usually cannot be determined. Use of standardized data sheets could provide uniform recording of information including prenatal assessment, initial post-natal and subsequent pre-operative evaluation, findings and techniques at surgery and at specified time points for years or even decades after surgery would provide the basis for recommended clinical approach and outcome studies. Outcome data should include whether excess clitoral tissue used for concomitant vaginoplasty, the age of vaginoplasty, any repeat surgeries required, and the use of vaginal dilation without surgery or after vaginal reconstruction; vaginal dilation was considered inappropriate until after puberty. The role of estrogen upon healing after surgery; short-term estrogen therapy for the prepubertal child or in the estrogenised pubertal female has been suggested [15], the use of administered estrogen or the clinical evidence of estrogenization at surgery should be recorded.

considered inappropriate until after puberty. The role of estrogen upon healing after surgery; short-term estrogen therapy for the prepubertal child or in the estrogenised pubertal female has been suggested [15], the use of administered estrogen or the clinical evidence of estrogenization at surgery should be recorded. An aim of prospective outcome studies is to clarify how complications are related to pre-surgical anatomy and surgical techniques. Surgical techniques vary from flap vaginoplasties which have been recommended for those with low junction of the vagina and urethra in which the urogenital sinus may be left intact; to vaginal pull-through procedures which are difficult when using a perineal approach, and may involve partial mobilization [16] or total mobilization [17,18], which commonly results in a low bladder neck; and an anterior surgical-transrectal approach [19] for the high urogenital sinus. After clitoroplasty, information concerning the preservation of Buchs fascia with the neurovascular bundle and histologic description of nerves within removed erectile tissue are pertinent. Since the goal is normal appearing genitalia with a vagina adequate to admit 2 fingers and greater than 6 cm. depth at the end of therapy, this information should be included in outcome datasets. Surgical outcome evaluation for vaginoplasty includes satisfaction with sexual function, and comparison of the utility of skin grafts versus intestinal mucosal replacement. The former has been suggested to be problematic because of lubrication difficulties and hair growth when hair-bearing skin is used while excessive mucus secretion is a complaint when intestinal mucosal lining is employed. At present, outcome studies suggest that urinary complaints are more common following vaginoplasty than occurs in the general population [20], it is possible that future outcome studies will help to refine surgical techniques and postoperative treatments to improve vaginoplasty outcomes.

nal mucosal lining is employed. At present, outcome studies suggest that urinary complaints are more common following vaginoplasty than occurs in the general population [20], it is possible that future outcome studies will help to refine surgical techniques and postoperative treatments to improve vaginoplasty outcomes. Decisions regarding age of surgery The decision regarding surgery and the age of surgery are informed by internal anatomic relationships. If both orifices open into the urogenital sinus, surgery may be deferred until a much older age, depending upon other factors. For the patient with CAH with high confluence, it is generally, although not universally, accepted that early surgery in infancy should be considered because the distance to the perineum is less, there is likely to be adjacent available tissues to use and it is generally technically easier. There was no agreement regarding the criteria for early or late genital surgery, particularly vaginoplasty. Suggested considerations include a potential for better wound healing after surgery during infancy, the use of urogenital tissue with early clitoroplasty [21] particularly if the urogenital sinus is long, the perception that early surgery is technically easier with better outcome for those with high vaginal insertion into the urethra and higher urogenital sinuses [22]. The likelihood of the need for repeat surgery is greater after early surgery even though such surgery may be relatively minor. If surgery is not done as an infant, it is generally considered that it should not be done until adolescence. Currently, a delay of surgery is recommended until adolescence for an absent or foreshortened vagina as present in the patient with Rokitansky or Complete Androgen Insensitivity Syndromes. For the cloacal extrophy group, vaginal surgery should be done at the time of ano/cloacal rectal surgery.

e done until adolescence. Currently, a delay of surgery is recommended until adolescence for an absent or foreshortened vagina as present in the patient with Rokitansky or Complete Androgen Insensitivity Syndromes. For the cloacal extrophy group, vaginal surgery should be done at the time of ano/cloacal rectal surgery. Goal of improved outcomes It has been accepted that in the past some genital surgeries done in childhood have resulted in poor cosmesis and reduced genital sensitivity leading to adult sexual dysfunction. While current nerve sparring feminizing surgery is expected to enjoy better outcomes regarding appearance and genital sensitivity [23], long-term outcomes supporting this have not yet been documented. Based on follow-up data in DSD patients who underwent genital surgery in childhood, common complications include scarring and stenosis; in those with CAH who did not have ideal adrenal suppression, scrotalization is common and, if skin was used for vaginoplasty, hair in the vagina. If surgery resulted in reduced sensation, this negatively impacts sexual function [24]. Currently, surgical techniques involve corporal sparing clitoroplasty with, if necessary, glans reduction using techniques such as a subtotal de-epithelization-wedge. The approach considers external genital proportions [25], before surgery with appropriate proportions created or maintained after surgery.

function [24]. Currently, surgical techniques involve corporal sparing clitoroplasty with, if necessary, glans reduction using techniques such as a subtotal de-epithelization-wedge. The approach considers external genital proportions [25], before surgery with appropriate proportions created or maintained after surgery. It is anticipated that outcome of those DSD individuals having surgery currently will have better outcomes when compared to previous surgical techniques because of improved surgical techniques and a better appreciation of the female sexual response. Such outcome needs to be documented using a multifactorial approach as discussed earlier. Future research needs involve careful documentation of decision to proceed with or delay surgery, together with descriptions of surgery, together with details outcome information. As with hypospadias above, much of this information must involve the surgeon and hence this will be the primary group involved in these studies. Phalloplasty Historically, a major limitation in sex assignment of 46,XY DSD patients with phallic deficiency was the problem of surgically creating a functional penis. This limitation often resulted in a female assignment since surgery to create a vagina was considered likely to be more successful. Current surgical advances when care is taken to preserve innervation and sensitive tissues have resulted in the ability to create a more functional phallus.

ly creating a functional penis. This limitation often resulted in a female assignment since surgery to create a vagina was considered likely to be more successful. Current surgical advances when care is taken to preserve innervation and sensitive tissues have resulted in the ability to create a more functional phallus. Most of these significant surgical advances have been made due to the demand for phalloplasty [26,27], from female to male transsexuals. Prostheses have erectile capabilities and attempts to connect nerves for erogenous sensitivity and proprioception. Donor sites include arm and thigh tissues, with an option of the use of separate donor sites for the urethra from the hairless, ventral side of the forearm. Candidates for surgery include those with penile agenesis, dysfunctional or “cripple” exstrophy, “cripple” hypospadias and epispadias, circumcision accidents, micropenis, small penis with DSD diagnoses and “shriveled” penis (<6 cm erect length). Figure 3A shows the genitalia of an exstrophy patient who lost his penis during reconstructive surgery and who was offered a phalloplasty at age 16 years with implantation of erectile device 1 year later. Currently, he is a 20 year old student who his extremely happy with his phallus, has a girlfriend and lives like any other adolescent of his age. Figures 3B and 3C show pre- and post-operative genitalia of three brothers with Partial Androgen Insensitivity Syndrome (PAIS) who underwent phalloplasty and subsequent erectile implant. Before surgery, multiple hypospadias repairs resulted in a insufficient penis.

friend and lives like any other adolescent of his age. Figures 3B and 3C show pre- and post-operative genitalia of three brothers with Partial Androgen Insensitivity Syndrome (PAIS) who underwent phalloplasty and subsequent erectile implant. Before surgery, multiple hypospadias repairs resulted in a insufficient penis. Figure 3 Results of phalloplasty for 2 46,XY boys. Genitalia of: A. Exstrophy patient who lost his penis during reconstructive surgery, B. Preoperatively of Partial Androgen Insensitivity Syndrome (PAIS) patient with an insufficient penis after multiple hypospadias repairs. C. Post-operative PAIS after phalloplasty and subsequent implant one year later. Kindly provided by Piet Hoebeke, Department of Urology, University Hospital, Gent, Belgium. Evidence suggests that outcomes are improved when existing penile structures or other sensitive genital tissues can be incorporated into the prosthesis. Penises smaller than 6 cm can be incorporated into the prosthesis while that > 6 cm cannot. Outcome is less favorable in extrophy when the poorly developed penile structure cannot be incorporated. The basic approach involves keeping and using any tissue that is erectile plus all tissue that is sensitive. Preoperative assessment includes determination of whether tissues, such as any corpora, can be lengthened. Surgery after puberty is best since sexual sensitivity can be assessed and the patient can be fully involved in all decision making.

keeping and using any tissue that is erectile plus all tissue that is sensitive. Preoperative assessment includes determination of whether tissues, such as any corpora, can be lengthened. Surgery after puberty is best since sexual sensitivity can be assessed and the patient can be fully involved in all decision making. Outcome research for those having phalloplasty, as for the other post-surgical issues described here must involve the initial genital anatomy, description of which tissue were used for construction of the phallus as well as the reconstruction procedure, together with long term functional and cosmetic satisfaction. Psychological, social and sexual outcomes Previous evaluations of DSD psychological and gender issues have often involved case histories rather than case–control studies. These have shown that important information concerning their condition may be unknown to the patients, such as a married 46XY female who believed that her previous feminizing surgery had caused her infertility. It is important to note that this may occur even when attempts toward full disclosure and psychological support are made. Children develop their own ideas of what is normal. Future studies need to assess not only an individual’s basic knowledge of his/her condition but also the patient’s perception of if and how they are different and how they perceive why they are different.

mpts toward full disclosure and psychological support are made. Children develop their own ideas of what is normal. Future studies need to assess not only an individual’s basic knowledge of his/her condition but also the patient’s perception of if and how they are different and how they perceive why they are different. Particularly among those who had sex assignment in infancy, it is important to consider apparent gender identity among the young as provisional while preparing parents for potential, but unlikely, gender change. Gender change, a consequence of both pre- and post-natal biological factors and social factors, must always be patient-initiated. When gender issues are unclear, as for example in a pubertal-aged DSD person being raised female and a potential need for gonadectomy; one should precede cautiously before considering irreversible surgery. A temporary option is GnRHa suppression, as used for similarly aged transsexual persons. Patients at this age or older need to ponder their perception of gender assignment since it appears that fewer 46,XY DSD individual have gender questions over time, with the incidence decreasing to about 12%. It has long been noted that among 46,XX CAH females, while there may be significant overlap in feminine-masculine gender role behaviors that gender identity is usually female indicating that many of the social cues commonly used to indicate gender, such as gender role behaviors, are not closely aligned with gender identity outcomes.

en noted that among 46,XX CAH females, while there may be significant overlap in feminine-masculine gender role behaviors that gender identity is usually female indicating that many of the social cues commonly used to indicate gender, such as gender role behaviors, are not closely aligned with gender identity outcomes. There is more psychological, social and sexual outcome information among 46XX CAH females that other condition. Differences in play behavior and toy preference during childhood are well documented [28]; as is involvement in male dominated professions as adults [29,30]. These differences are apparently influenced by both intrinsic and environmental factors. Outcome information also indicates that adult 46XX CAH females often live alone, have fewer social contacts and are less likely to be employed than age-matched peers [31]. An increased incidence of sexual dysfunction is also present and appears to be related to genital insensitivity, difficulty/inability to orgasm, atypical genital appearance and decreased fertility [32]. Table 1 summarizes some findings suggesting that dysfunction is related to both the severity of the underlying condition (based on genotype) as well as surgery. It is of interest that the ability to achieve orgasm does not differ from control women and does not appear to correlate with sensitivity and that as a group sexual life satisfaction does not differ between CAH women and controls nor does it correlate with ability to orgasm. In addition, this report also provides preliminary evidence of better outcome with the currently used nerve sparing clitoroplasty when compared to older techniques.

with sensitivity and that as a group sexual life satisfaction does not differ between CAH women and controls nor does it correlate with ability to orgasm. In addition, this report also provides preliminary evidence of better outcome with the currently used nerve sparing clitoroplasty when compared to older techniques. Table 1 Sexual Function among CAH Women: Relationships with genotype* and surgery Sexual function Based on genotype and surgery Clitoral sensitivity Highest for nerve sparing and clitoral placement No difference between no surgery and partial resection Poorest with subcutaneous clitoral placement/partial resection Ability to achieve orgasm Did not correlate with sensitivity Not different between genotype groups or from controls Sexual function score Score correlated with severity of genotype (9 aspects of sexual life) Lowest in null phenotype group Lower in null and 12-splice phenotype than controls Satisfaction with sexual life No difference between CAH women and controls CAH group with null phenotype (who had more surgical complications) differed from other phenotypes Satisfaction did not correlate with ability to orgasm *CYP21A2 allele mutations: null, 12-splice, I172N and V281L.

-splice phenotype than controls Satisfaction with sexual life No difference between CAH women and controls CAH group with null phenotype (who had more surgical complications) differed from other phenotypes Satisfaction did not correlate with ability to orgasm *CYP21A2 allele mutations: null, 12-splice, I172N and V281L. The psycho-sexual impact of genital surgery involves both an individual’s perception of his/her genitalia before surgery and the impact of surgery on genital sensitivity and the presence of actual or perceived change in genital appearance. Most of these are factors that are impacted by current and previous positive or negative social and environmental influences. Sexual function can be hindered by actual or perceived genital differences and anatomic and sensual physical limitations. Stigmas including fear, guilt, and shame may lead to avoidance of situations that could lead to sexual activity. A recent review of ninety eight 46,XX DSD CAH patients suggests that psychological problems result in negative outcome in 68% most of whom blame prenatal androgen exposure [33]. This group was also found to have less sexual experience, more sexual anxiety, more behavior avoidance, and more passivity compared with type 1 diabetes mellitus patients [34]. Also among XY women who had vaginoplasty, a poor understanding of their development and lack of a sense of entitlement regarding sex was found to be associated with limited social interaction including even a lack of engagement conversations simply for pleasure.

ity compared with type 1 diabetes mellitus patients [34]. Also among XY women who had vaginoplasty, a poor understanding of their development and lack of a sense of entitlement regarding sex was found to be associated with limited social interaction including even a lack of engagement conversations simply for pleasure. With the latest medical and surgical care, disclosure and counseling are crucial for good outcomes. Such must focus upon the underlying conditions and expected gonad function and will differ for those with hormonal (cAIS, CAH, 5alpha, etc.) versus non-hormonal DSD (bladder/ cloacal exstrophy, epispadias, aphallia, and vaginal agenesis). Learning disabilities and a higher rate of internalizing psychopathologic disorders with risks for suicide and behavior aberrations have been identified in DSD patients and if suspected, appropriate evaluation including neuropsychologic testing should be considered. Before proceeding to full disclosure, careful consideration is needed and agreement from the individual is pertinent, since full medical disclosure is not always positive at all times. Regarding sexual issues, a desire for sexual satisfaction should always be assumed. If surgery is planned, its impact on current or future mates should be discussed. Learning how to be verbally intimate is important before approaching physical intimacies.

nce full medical disclosure is not always positive at all times. Regarding sexual issues, a desire for sexual satisfaction should always be assumed. If surgery is planned, its impact on current or future mates should be discussed. Learning how to be verbally intimate is important before approaching physical intimacies. Psychosocial wellness involves being self-determined, resourceful, and knowledgeable, feeling good enough for intimacy and finding actual experiences pleasurable, leading to enjoyment. Predictors of good outcome involve lack of stigma related to body image and genital perceptions, knowledge and understanding of condition, attitudes and aspirations, general physical and mental health, understanding and outcome of surgical interventions, coupled with social interactions and impact of education, culture and religion. Assessment with Quality of Life indicators based on judgment of patients satisfaction and achievement and sensitivity to life situations including sexual experiences are crucial. Such may reveal individual’s unawareness or suppression of negative factors (WHOQOL) [35,36]. Future research needs to involve prospective observational studies with repeated measure design within bio-psycho-social models, be hypothesis-driven and will likely require non-standardized questionnaires.

iences are crucial. Such may reveal individual’s unawareness or suppression of negative factors (WHOQOL) [35,36]. Future research needs to involve prospective observational studies with repeated measure design within bio-psycho-social models, be hypothesis-driven and will likely require non-standardized questionnaires. Long-term psychosocial and psychosexual outcomes are highly complex and while details studies must involve assessment of qualify to life outcome information, multifaceted analyses are needed to correlate outcome with initial development issues, psychological and social support and surgical aspects, if involved. Conclusion A primary goal of this conference was to provide a format by which multi-centered outcome studies could be conducted. This report focuses upon four areas: distal hypospadias, genital assessment and repair for virilized 46, XX CAH patients, potential for phalloplasty and psychological, social and sexual outcomes study perspectives. Each continues to be problematic regarding assessment and therapy and careful outcome studies are needed for each, to assess past therapy and improve it in the future. Consent Written informed consent was obtained from the patient for the publication of this report and any accompanying images. Competing interests The authors declare that they have no competing interests. Authors’ contributions Both authors took extensive notes of the sessions of the working conference on disorders of sex development and both incorporated these notes into this meeting report. Hence, both have written, read and approve the final manuscript.

Background Short stature homeobox-containing gene (SHOX) is a growth regulating gene present on the pseudoautosomal region 1 (PAR1) on the distal end of the X and Y chromosomes. SHOX haploinsufficiency is considered in the differential diagnosis for short stature in children and is currently an FDA approved indication for growth hormone therapy. Leri-Weill syndrome (LWS) is a genetic disorder caused by deletions or mutations in the SHOX gene or by deletions downstream of the gene. It is classically characterized by short stature, mesomelic shortening of forearms and legs, and Madelung deformity [1]. We present the rare case of a male infant diagnosed with Leri-Weill syndrome in which an originally X-located SHOX deletion from father was transmitted to son’s Y chromosome by crossover during meiosis. We also review the literature to date and discuss the future implications of such findings.

d Madelung deformity [1]. We present the rare case of a male infant diagnosed with Leri-Weill syndrome in which an originally X-located SHOX deletion from father was transmitted to son’s Y chromosome by crossover during meiosis. We also review the literature to date and discuss the future implications of such findings. Case presentation The male index patient (IV: 1, Figure 1) was the first child of Caucasian non-consanguinous parents. He was born at 39½ weeks gestation with a normal birth weight of 2.95 kg, (25th centile), and length 50 cm (50th centile). On prenatal ultrasound, fetal long bones were found to be about three weeks delayed in growth during the second trimester. The infant underwent genetic evaluation at 23 days of life since the father had been previously diagnosed with Leri-Weill syndrome with a documented SHOX deletion on his X chromosome del (X)(p22.33p22.33) confirmed by FISH using a SHOX gene probe. The father’s deletion was confirmed with a FISH probe specific for the SHOX gene. On examination, the patient measured 51.5 cm (25-50th centile) and weighed 3.45 kg (10-25th centile) with head circumference of 36 cm (25th centile). Although physical examination was unremarkable, a DNA microarray study was performed to rule out the presumed extremely small possibility of the deletion crossing over to the Y chromosome of the index patient (IV: 1, Figure 1). Figure 1 Pedigree of the family.

Case presentation The male index patient (IV: 1, Figure 1) was the first child of Caucasian non-consanguinous parents. He was born at 39½ weeks gestation with a normal birth weight of 2.95 kg, (25th centile), and length 50 cm (50th centile). On prenatal ultrasound, fetal long bones were found to be about three weeks delayed in growth during the second trimester. The infant underwent genetic evaluation at 23 days of life since the father had been previously diagnosed with Leri-Weill syndrome with a documented SHOX deletion on his X chromosome del (X)(p22.33p22.33) confirmed by FISH using a SHOX gene probe. The father’s deletion was confirmed with a FISH probe specific for the SHOX gene. On examination, the patient measured 51.5 cm (25-50th centile) and weighed 3.45 kg (10-25th centile) with head circumference of 36 cm (25th centile). Although physical examination was unremarkable, a DNA microarray study was performed to rule out the presumed extremely small possibility of the deletion crossing over to the Y chromosome of the index patient (IV: 1, Figure 1). Figure 1 Pedigree of the family. The mother (III:1) was of normal height and stature at 154.9 cm tall. The father (III:2) was of short stature at 157.5 cm tall. The father had been diagnosed with Leri-Weill syndrome at 12 years of age, and had not been treated with growth hormone. His past medical history was also notable for an insulinoma diagnosed at 8 years of age with a partial pancreatectomy at 18 years. Three paternal uncles, two paternal aunts, one paternal cousin and patient’s paternal grandmother, in addition to patient’s paternal great aunt, and paternal great-grandmother, were also noted to be affected with LWS (see Figure 1).

or an insulinoma diagnosed at 8 years of age with a partial pancreatectomy at 18 years. Three paternal uncles, two paternal aunts, one paternal cousin and patient’s paternal grandmother, in addition to patient’s paternal great aunt, and paternal great-grandmother, were also noted to be affected with LWS (see Figure 1). A SHOX deletion of 262 Kb on Yp11.32 was identified using single nucleotide polymorphism microarray (SNP) analysis in the patient and confirmed by FISH using a SHOX gene probe. Our infant patient was diagnosed with Leri-Weill syndrome resulting from an unusual documented inheritance between father and son due to crossover of the SHOX deletion between X and Y chromosomes during paternal meiosis. We report the youngest patient in literature documented by FISH analysis to have an X to Y chromosome transfer of an originally X-located SHOX deletion. The mechanisms resulting in SHOX deficiency include gene mutations and whole gene deletions of the pseudoautosomal region1 (PAR1) of differing sizes [2,3]. Approximately two-thirds of individuals with SHOX mutations have large scale SHOX deletions that vary in size between 90 kb and 2.5 Mb or more. Point mutations comprise the remaining one-third of SHOX mutations causing SHOX-related haploinsufficiency [4]. Since SHOX is located in the pseudoautosomal regions 1 (PAR1) present on both the X and Y chromosomes, mutations within these genes may segregate independent of sex and be inherited in an autosomal fashion, termed pseudo-autosomal dominant inheritance [2].

of SHOX mutations causing SHOX-related haploinsufficiency [4]. Since SHOX is located in the pseudoautosomal regions 1 (PAR1) present on both the X and Y chromosomes, mutations within these genes may segregate independent of sex and be inherited in an autosomal fashion, termed pseudo-autosomal dominant inheritance [2]. Transfer of the deleted or mutated SHOX gene to the alternate sex chromosome due to crossover during meiosis has been described. However, most reports discussed recombination events in the PAR1 in the context of a Y-located SHOX deletion transmitted from father to daughter or included pedigrees that did not differentiate between Y- and X- chromosomal SHOX mutations [5-8]. Indeed, to our knowledge, the transfer of an originally X-located SHOX deletion to the Y chromosome after transmission from father to son has only been documented by FISH in two patients in the literature to date, and not in an apparently phenotypically normal male child [9,10].

and X- chromosomal SHOX mutations [5-8]. Indeed, to our knowledge, the transfer of an originally X-located SHOX deletion to the Y chromosome after transmission from father to son has only been documented by FISH in two patients in the literature to date, and not in an apparently phenotypically normal male child [9,10]. In a study by Ross et al., subjects were referred for short stature or Madelung deformity and 17 unrelated families with LWD with complete gene deletion in 33 subjects were identified. An X to Y chromosome transfer was mentioned in an 8.3 year old male from his father, with the child’s height noted to be at −2.5 SDS [9]. Musebeck et al. analyzed the frequency of SHOX deletions in short stature children and found 5 patients with deletions, one of whom was an 11.75 year old male with a height of 139 (−1.4 SDS) found to have a transfer of the X-located SHOX deletion from his father to his Y chromosome [10]. The 7 month old infant described in our case is the youngest patient in literature documented by FISH analysis to have an X to Y chromosome transfer and is the first patient of these three cases presenting prior to onset of short stature or Madelung deformity.

ted SHOX deletion from his father to his Y chromosome [10]. The 7 month old infant described in our case is the youngest patient in literature documented by FISH analysis to have an X to Y chromosome transfer and is the first patient of these three cases presenting prior to onset of short stature or Madelung deformity. This crossover of SHOX between sex chromosomes is particularly important in the context of genetic counseling. Since 2006, growth hormone (GH) therapy has been an FDA approved indication for treatment of short stature for patients with SHOX deficiency. Leri-Weill syndrome caused by deletions or mutations of the SHOX gene is known to be clinically highly variable [11]. Correct identification of SHOX deficiency in children with growth problems is vital to the implementation of GH therapy in a timely manner.

treatment of short stature for patients with SHOX deficiency. Leri-Weill syndrome caused by deletions or mutations of the SHOX gene is known to be clinically highly variable [11]. Correct identification of SHOX deficiency in children with growth problems is vital to the implementation of GH therapy in a timely manner. Conclusion Our case highlights the importance of advising affected SHOX patients of risks to future offspring and screening off-spring of parents carrying SHOX abnormalities regardless of sex. Patients should be informed of the possibility that a father carrying a SHOX mutation on the X chromosome can transmit this mutation not only to a daughter but to a son as well due to crossing over between the pseudo-autosomal regions of the X and Y chromosomes during paternal meiosis, albeit a rare occurrence. This finding can also occur in mothers with SHOX mutations on X chromosomes transmitting mutations to the Y chromosomes of their sons. Our case contributes to the knowledge regarding meiotic crossover of the SHOX gene region between the X and Y chromosomes and brings attention to an unusual documented inheritance between father and son. Our patient was identified prior to growth failure and can now be monitored for growth abnormalities with the ability to implement growth augmentation therapy without delay. Consent Written informed consent was obtained from the patient’s father for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Conclusion Our case highlights the importance of advising affected SHOX patients of risks to future offspring and screening off-spring of parents carrying SHOX abnormalities regardless of sex. Patients should be informed of the possibility that a father carrying a SHOX mutation on the X chromosome can transmit this mutation not only to a daughter but to a son as well due to crossing over between the pseudo-autosomal regions of the X and Y chromosomes during paternal meiosis, albeit a rare occurrence. This finding can also occur in mothers with SHOX mutations on X chromosomes transmitting mutations to the Y chromosomes of their sons. Our case contributes to the knowledge regarding meiotic crossover of the SHOX gene region between the X and Y chromosomes and brings attention to an unusual documented inheritance between father and son. Our patient was identified prior to growth failure and can now be monitored for growth abnormalities with the ability to implement growth augmentation therapy without delay. Consent Written informed consent was obtained from the patient’s father for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Abbreviations SHOX: Short stature homeobox-containing gene; LWS: Leri-Weill syndrome; PAR1: Pseudoautosomal region 1. Competing interests The authors declare that they have no competing interests.

Consent Written informed consent was obtained from the patient’s father for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Abbreviations SHOX: Short stature homeobox-containing gene; LWS: Leri-Weill syndrome; PAR1: Pseudoautosomal region 1. Competing interests The authors declare that they have no competing interests. Authors’ contributions All authors have made significant intellectual contributions to the manuscript. All authors have contributed to the concept and design of the case report. SEO, KA, RW, ES, EG, and MC have diagnosed and/or treated the patient and/or mother. MC, SEO and KA drafted the manuscript, and all authors have critically revised the manuscript for important intellectual content. All authors have read and given approval of the final manuscript version to be published. Acknowledgements This work was supported by an NIH NIDDK 5 T32 DK 06552–07 in Pediatric Endocrinology (PI SE Oberfield). We thank Dr. Brynn Levy of Columbia University Medical Center for his assistance in the genetic testing of our patient and Dr. Lakshmi Mehta of the Mt. Sinai Medical Center for discussions regarding the father’s prior genetic consultation.

Introduction Turner syndrome (TS) is defined as the combination of short stature, gonadal dysgenesis, typical visible somatic dysmorphic stigmata, and urinary, cardiovascular, metabolic and skeletal abnormalities associated with a missing or structurally abnormal second sexual chromosome [1]. This disorder is one of the most common human chromosomal anomalies, occurring in approximately 1:2500 live female births. More than 50% of TS patients have an apparent non-mosaic 45,X karyotype and a wide and variable spectrum of clinical features and associated complications of TS are present even in the individuals with non-mosaic pure 45,X karyotypes. The absence of the second X-chromosome in TS leads to a loss of expression of either paternally or maternally derived genes on the X-chromosome, depending of the X chromosome retained [2]. The haploinsufficiency of genes that are normally expressed from two sexual chromosomes is the most plausible mechanism to explain the TS phenotype. However, the absence of the second sexual chromosome has led to the speculation that there may be genes present on the X chromosome which are expressed differently depending upon whether they are maternally or paternally derived [3]. Thus, both the haploinsufficiency of genes on the two sexual chromosomes and the parental origin (monoallelic expression of the imprinted genes) may play a part in the broad and variable clinical features spectrum seen in TS. Therefore, TS could be a model for understanding the role of genomic imprinting on the clinical features and putative imprinted genes on the X-chromosome.

ponse to rhGH. In the context of this controversy, the current study was performed to evaluate the effect of the parental origin of the X-chromosome on the clinical features, auxological data, associated complications, lipid metabolism and response to rhGH in a group of patients with TS and a non-mosaic 45,X karyotype. Subjects and methods This is a cross-sectional multicenter correlational study carried out in 6 Latin-American university hospitals over the past three years. The study protocol was approved by the medical ethics institutional review board of all participating hospitals (Unidad de Genetica Medica, Universidad del Zulia; Unidad de Endocrinologia Pediatrica, Hospital de Clínicas Caracas; Departamento de Endocrinología y Laboratorio, Hospital de Niños de la Santísima Trinidad; Unidad de Endocrinología, Hospital T Álvarez; Instituto de Investigaciones Materno Infantil, Facultad de Medicina, Universidad de Chile; División de Endocrinología, Hospital Domingo Luciani; and Departamento de Pediatría, Hospital de Especialidades Pediátricas). Initially, unrelated prepubertal patients with TS (n =  100) examined at our hospitals were enrolled in this study along with their mothers. However, the study group was reduced to 93 patients after a hidden mosaicism in the second X-chromosome was detected by molecular analysis in duplicate blood samples in seven subjects. Patients included in this study were from the following countries: Venezuela (n =  55) Argentina (n =  29), and Chile (n =  9). The diagnosis of TS was based on typical clinical features and on cytogenetic analysis. Only patients with a non-mosaic 45,X karyotype were included and the average number of analyzed metaphases was 50. A subgroup (n =  34) of the patients that fitted eligibility criteria for rhGH treatment was analyzed both longitudinally and prospectively. We have used the following eligibility criteria for rhGh treatment: 1) age over 2 years old; 2) girls with a height below the third percentile in normal height curves, decreased growth velocity and without significant spontaneous catch-up growth; and 3) no previous administration of growth-stimulating therapy or of estrogen replacement. Compliance was determined by the clinical response to rhGH therapy (significant increase in their linear growth and in their growth velocity following rhGH administration). In some of the girls, compliance was also determined by significant increases in serial IGF-1 measurements during rhGH therapy.

trogen replacement. Compliance was determined by the clinical response to rhGH therapy (significant increase in their linear growth and in their growth velocity following rhGH administration). In some of the girls, compliance was also determined by significant increases in serial IGF-1 measurements during rhGH therapy. With respect to parental age, chronological age, height-SDS, weight-SDS, BMI-SDS, bone age delay, and parental origin of retained X-chromosome, no statistical difference was found between patients of different countries at the beginning of the study. Thus, these country subgroups were pooled as one group. Written informed consents for DNA sampling and for participation in the study were obtained from parents or adult patients before beginning the study.

osome, no statistical difference was found between patients of different countries at the beginning of the study. Thus, these country subgroups were pooled as one group. Written informed consents for DNA sampling and for participation in the study were obtained from parents or adult patients before beginning the study. Each patient was examined by two clinicians in each center. Clinical data were collected from the patients using a previously designed standardized checklist for TS which included more than 100 variables and recorded. The standardized checklist will be sent upon request. Clinical features encompassed facial dysmorphic stigmata and ophthalmologic, auditive, thoracic, skeletal and cutaneous signs. In addition, cardiovascular, renal metabolic and endocrinological abnormalities were prospectively evaluated. Serum fasting levels of glucose, total-, LDL-, VLDL- and HDL-cholesterol and triglyceride concentrations were measured in all subjects by the enzymatic method. Measurements were performed directly in each of the 6 reference hospitals. When a patient was receiving HRT, this was discontinued 3 months prior to assessment. Imaging studies performed included renal and cardiac ultrasonography. All clinicians were blinded to the results of the molecular analysis.

e enzymatic method. Measurements were performed directly in each of the 6 reference hospitals. When a patient was receiving HRT, this was discontinued 3 months prior to assessment. Imaging studies performed included renal and cardiac ultrasonography. All clinicians were blinded to the results of the molecular analysis. The mean age for the whole group was 18.3 ± 8.5 years (range 0.4-46 years ). Auxological evaluation was performed by trained physicians using standard techniques and included measurements of weight (expressed in kg and in standard deviation scores (SDS)), parental height, standing height expressed in cm and transformed into SDS according to national growth chart references for each country (patient value minus mean age- and sex-matched control values (cm) divided by the corresponding SD according to age) in order to evaluate all patients from different countries. The mid-parental height was calculated as paternal height minus 12.5 cm plus maternal height divided by 2, then converted to SDS using national normative height data for females and males at 19 years of age. To analyze the growth response to rhGH, anthropometric data were evaluated at the start of rhGH therapy and every 3 months thereafter for 2 years . Growth velocities, total height gain and differences in the increment were calculated [4]. Recombinant human growth hormone (rhGH) was administered at a mean dose of 37 ± 5 μg/kg/day seven days a week by subcutaneous injection for a two year period. Body mass index (BMI) was expressed as SDS. Left hand and wrist radiographs were assessed prior to starting therapy and then yearly for bone age determination using the methods of Greulich and Pyle.

H) was administered at a mean dose of 37 ± 5 μg/kg/day seven days a week by subcutaneous injection for a two year period. Body mass index (BMI) was expressed as SDS. Left hand and wrist radiographs were assessed prior to starting therapy and then yearly for bone age determination using the methods of Greulich and Pyle. Molecular analysis Isolation, quantification and purification of genomic DNA, selection of Short Tandem Repeats on the X-chromosome (X-STRs), PCR amplification, and fragments analysis are described in Additional file 1. Briefly, DNA samples were obtained from a few drops of venous blood placed on the filter paper from patients and their mothers. Genomic DNA was extracted and quantitated using the DNA IQ™ System kit (Promega, Madison, WI, USA). PCR conditions were optimized in our laboratory using the QIAGEN Multiplex PCR kit according Gusmao et al. and Edelman et al. [5,6]. The X-STRs were selected due to 1) their location on both Xp and Xq; 2) their high degree of heterozygosity (between 66 to 82%); and 3) their high discrimination power obtained in Latin-American populations, both in males (≥1 in 5 × 105) and females (≥1 in 3 × 109), as well as high mean exclusion chance in father/daughter duos (≥99.953%) and in father/mother/daughter trios (≥99.999%) [5,6]. Separation and detection of the different amplicons generated during the PCR was performed in an ABI PRISM 310 Genetic Analyzer (Applied Biosystems). The parental origin of the single X chromosome was assigned as maternal when the patient and her mother shared an allele at each of the ten loci tested using the X-STRs decaplex. Paternal DNA was excluded to avoid potential ethical problems.

e PCR was performed in an ABI PRISM 310 Genetic Analyzer (Applied Biosystems). The parental origin of the single X chromosome was assigned as maternal when the patient and her mother shared an allele at each of the ten loci tested using the X-STRs decaplex. Paternal DNA was excluded to avoid potential ethical problems. Stastical analysis Data analysis was performed using GraphPad Prism for Windows XP, version 4.0 (GraphPad Software, San Diego, CA, USA). Quantitative variables are shown as mean ± SD or median (range). The Mann–Whitney U-test was performed to examine the possible relationship between age, birth height and weight, height, weight, BMI, maternal, paternal and mid-parental heights and the parental origin of the X chromosome. The Kolmogorov-Smirnov test was used to test the normal distribution of these variables and Pearson correlation coefficients were used to assess the correlations among the various variables. To compare the frequencies of categorical variables between the Xp and Xm groups we performed the Fisher’s exact test. The Wilcoxon signed ranks test was used to relate the pre- and the on rhGH treatment patient variables. To estimate the parental origin effect on the response to rhGH treatment, multiple linear regression analysis was performed with growth response to rhGh treatment parameters as the dependent variables and parental origin of the retained X-chromosome as potentially influencing variable. The t-test was used to compare biochemical data between Xp and Xm patients. ANCOVA was used to adjust for age and BMI, as covariates followed by the Fisher protected least-significant-difference tests. Two-sided p values were calculated for age, BMI, total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides and fasting glucose. Statistical significant differences were assumed if p < 0.05.

VA was used to adjust for age and BMI, as covariates followed by the Fisher protected least-significant-difference tests. Two-sided p values were calculated for age, BMI, total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides and fasting glucose. Statistical significant differences were assumed if p < 0.05. Results In Table 1 we can see the patient and parental data according to origin of the retained X-chromosome. No significant difference between the origin of the retained X-chromosome and height-SDS of patients was found (p < 0.31). A stronger, but non significant correlation (p ≤ 0.07) between final height of subjects carrying the Xm chromosome and maternal height was found (Table 2). There was also a trend towards significance between maternal height and patients height in the ≤ 7 years of age group (p ≤ 0.09) who retained the Xm. However, among the group of subjects who retained the Xp no correlation was demonstrated at any age group. The correlation observed between the patients’ height and mid-parental height was similar to the correlation seen between patients’ height and maternal height, but the statistical significance was less at each age group. We found no correlation at any age group between paternal and patients height in the Xp group. Table 1 Patient and parental data according to origin of the retained X-chromosome

Results In Table 1 we can see the patient and parental data according to origin of the retained X-chromosome. No significant difference between the origin of the retained X-chromosome and height-SDS of patients was found (p < 0.31). A stronger, but non significant correlation (p ≤ 0.07) between final height of subjects carrying the Xm chromosome and maternal height was found (Table 2). There was also a trend towards significance between maternal height and patients height in the ≤ 7 years of age group (p ≤ 0.09) who retained the Xm. However, among the group of subjects who retained the Xp no correlation was demonstrated at any age group. The correlation observed between the patients’ height and mid-parental height was similar to the correlation seen between patients’ height and maternal height, but the statistical significance was less at each age group. We found no correlation at any age group between paternal and patients height in the Xp group. Table 1 Patient and parental data according to origin of the retained X-chromosome Xm (n =  67) Xp (n =  26) p Patients’ height-SDS −3.8 ±1.32 −3.7 ±1.5 0.31 Maternal height SDS −1.1 ±0.9 −0.8 ±0.7 0.14 Paternal height SDS 0.8 ±0.6 0.9 ±0.8 0.75 Mid-parental height −0.9 ±0.9 −0.9 ±0.6 0.23 Values are expressed as group means ± SD, sample sizes as n =  . Table 2 Correlation of patient’s height-SDS with parental height according to the retained X-chromosome at different age groups

Xm (n =  67) Xp (n =  26) p Patients’ height-SDS −3.8 ±1.32 −3.7 ±1.5 0.31 Maternal height SDS −1.1 ±0.9 −0.8 ±0.7 0.14 Paternal height SDS 0.8 ±0.6 0.9 ±0.8 0.75 Mid-parental height −0.9 ±0.9 −0.9 ±0.6 0.23 Values are expressed as group means ± SD, sample sizes as n =  . Table 2 Correlation of patient’s height-SDS with parental height according to the retained X-chromosome at different age groups Age groups Maternal height Paternal height Mid-parental height Xm =  67 Xp =  26 Xm =  67 Xp =  26 Xm =  67 Xp =  26 Neonatal r =  0.341 r =  0.282 r =  0.301 r =  0.282 r =  0.263 r =  0.287 p =  0.24 p =  0.45 p =  0.62 p =  0.39 p =  0.44 p =  0.43 ≤ 7 years old r =  0.58 r =  0.405 r =  0.381 r =  0.241 r =  0.489 r =  0.375 p =  0.09 p =  0.11 p =  0.23 p =  0.43 p =  0.12 p =  0.21 Final Height r =  0.601 r =  0.598 r =  0.362 r =  0.213 r =  0.521 r =  0.421 p =  0.07 p =  0.09 p =  0.32 p =  0.52 p =  0.09 p =  0.17 Sample sizes as Xm =  or Xp=; and number of measures (in brackets). Kolmogorov-Smirnov test showed no deviation from linearity (p < 0.01). Values are correlations of patient and parental height (Pearson’s correlation coefficient) p ≤ 0.05.

=  0.521 r =  0.421 p =  0.07 p =  0.09 p =  0.32 p =  0.52 p =  0.09 p =  0.17 Sample sizes as Xm =  or Xp=; and number of measures (in brackets). Kolmogorov-Smirnov test showed no deviation from linearity (p < 0.01). Values are correlations of patient and parental height (Pearson’s correlation coefficient) p ≤ 0.05. No significant differences in the incidence of any of the visible somatic dysmorphic stigmata, or in urinary, cardiovascular, metabolic and skeletal abnormalities were observed among the patients from the different countries. Additionally, no significant differences between the subjects who retained the Xm or Xp chromosome were found in relation to the frequency of somatic stigmata or associated abnormalities (Table 3). The incidence of cardiovascular anomalies was similar between the two different genotype groups (18/67: 26.9% vs 8/26: 30.8%). Renal anomalies were present in 16/67 (23.9%) of the patients who retained the Xm versus 6/26 (23.1%) in those with retained Xp. There was no statistically significant difference among the genotypes described above (p =  0.9). Similar results were found for other anomalies, including ear and thyroid abnormalities (Table 3). Table 3 Associated disorders and anomalies according to the parental origin of the retained X chromosome

No significant differences in the incidence of any of the visible somatic dysmorphic stigmata, or in urinary, cardiovascular, metabolic and skeletal abnormalities were observed among the patients from the different countries. Additionally, no significant differences between the subjects who retained the Xm or Xp chromosome were found in relation to the frequency of somatic stigmata or associated abnormalities (Table 3). The incidence of cardiovascular anomalies was similar between the two different genotype groups (18/67: 26.9% vs 8/26: 30.8%). Renal anomalies were present in 16/67 (23.9%) of the patients who retained the Xm versus 6/26 (23.1%) in those with retained Xp. There was no statistically significant difference among the genotypes described above (p =  0.9). Similar results were found for other anomalies, including ear and thyroid abnormalities (Table 3). Table 3 Associated disorders and anomalies according to the parental origin of the retained X chromosome Type Cardiac anomalies Renal anomalies Webbed neck Ear disorders Thyroid disorders Parental origin Xm Xp Xm Xp Xm Xp Xm Xp Xm Xp A/T: (%); 18/67a (26.9) 8/26a (30.8) 16/67b (23.9) 6/26b (23.1) 23/67b (34.3) 9/26b (34.6) 31/67c (46.2) 12/26c (46.1) 14/67d (21) 6/26d (23.1) A/T: affected/total (%); Fisher’s exact test: ap =  0.7980; bp =  1.0, cp =  0.1547; dp =  0.7862.

Ear disorders Thyroid disorders Parental origin Xm Xp Xm Xp Xm Xp Xm Xp Xm Xp A/T: (%); 18/67a (26.9) 8/26a (30.8) 16/67b (23.9) 6/26b (23.1) 23/67b (34.3) 9/26b (34.6) 31/67c (46.2) 12/26c (46.1) 14/67d (21) 6/26d (23.1) A/T: affected/total (%); Fisher’s exact test: ap =  0.7980; bp =  1.0, cp =  0.1547; dp =  0.7862. Thirty-four of our TS patients received rhGH therapy, 24 of them retained the Xm, and 10 retained the Xp. At the beginning of therapy, there were no statistically significant differences in baseline auxological features between the two genotype groups. The chronological age was similar in both groups (Xm =  10.8 ± 2.8 years vs Xp =  9.9 ± 3.1 years ). The height-SDS of patients was not significantly different between those who retained the Xm (−2.81 ± 0.71) and those who retained the Xp (−2.66 ± 0.63). Growth hormone therapy significantly (p < 0.0001) increased growth velocity (cm/yr) and standing height (SDS and cm) during the first (7.56 ± 1.35 cm/yr; -2.47 ± 0.65) and second years (5.66 ± 1.51 cm/yr; -2.35 ± 0.71) in the total group (Xm + Xp). However, these changes were not significantly different among the patients who retained the Xm and those who retained the Xp (p value =0.51; 0.49; 0.26; 0.54; respectively) (Table 4). Additionally, no differences in the total two-year height gain or in the increment of height-SDS among patients with different retained X chromosome were found (Table 4). Table 4 Clinical and auxological data of the growth response to rhGH according to the parental origin of the retained X chromosome

Thirty-four of our TS patients received rhGH therapy, 24 of them retained the Xm, and 10 retained the Xp. At the beginning of therapy, there were no statistically significant differences in baseline auxological features between the two genotype groups. The chronological age was similar in both groups (Xm =  10.8 ± 2.8 years vs Xp =  9.9 ± 3.1 years ). The height-SDS of patients was not significantly different between those who retained the Xm (−2.81 ± 0.71) and those who retained the Xp (−2.66 ± 0.63). Growth hormone therapy significantly (p < 0.0001) increased growth velocity (cm/yr) and standing height (SDS and cm) during the first (7.56 ± 1.35 cm/yr; -2.47 ± 0.65) and second years (5.66 ± 1.51 cm/yr; -2.35 ± 0.71) in the total group (Xm + Xp). However, these changes were not significantly different among the patients who retained the Xm and those who retained the Xp (p value =0.51; 0.49; 0.26; 0.54; respectively) (Table 4). Additionally, no differences in the total two-year height gain or in the increment of height-SDS among patients with different retained X chromosome were found (Table 4). Table 4 Clinical and auxological data of the growth response to rhGH according to the parental origin of the retained X chromosome Xm (n =  24) Xp (n =  10) P Age at start yr 10.8 ± 2.8 9.9 ± 3.1 0.45 rhGH dose (mg/kg/day) 0,37 ±0.08 0,37 ± 0.02 0.87 Pretreatment height-SDS −2.81 ± 0.71 −2.66 ± 0.63 0.12 HV during first yr (cm/y) 7.65 ± 1.73 7.87 ± 1.1 0.51 HV during second yr (cm/y) 5.74 ± 1.57 5.59 ± 1.77 0.49 Height-SDS at first yr −2.21 ± 0.64 −2.25 ± 0.79 0.26 Δ height-SDS 0.55 ± 0.43 0.56 ± 0.41 0.54 According to SDS in relation to national growth chart references for each country. Multiple linear regression analysis was performed with growth response to rhGH treatment parameters as the dependent variables and parental origin of the retained X-chromosome as potentially influencing variable.

± 0.43 0.56 ± 0.41 0.54 According to SDS in relation to national growth chart references for each country. Multiple linear regression analysis was performed with growth response to rhGH treatment parameters as the dependent variables and parental origin of the retained X-chromosome as potentially influencing variable. Pooled data from fasting serum levels of glucose, total-, LDL-, VLDL- and HDL-cholesterol were indistinguishable between the two genotype groups. Triglycerides, total-, HDL- and LDL-cholesterol and fasting glucose levels were similar in the Xm and Xp groups in subjects <20 years . However, serum levels of triglycerides (p < 0.001), total-(p < 0.03), and LDL-cholesterol (p < 0.04) were significantly higher in the subjects ≥20 years who retained the Xm chromosome, when compared to those in the the Xp group (Table 5). Table 5 Serum lipid and carbohydrate data according to the parental origin of the retained X chromosome

Pooled data from fasting serum levels of glucose, total-, LDL-, VLDL- and HDL-cholesterol were indistinguishable between the two genotype groups. Triglycerides, total-, HDL- and LDL-cholesterol and fasting glucose levels were similar in the Xm and Xp groups in subjects <20 years . However, serum levels of triglycerides (p < 0.001), total-(p < 0.03), and LDL-cholesterol (p < 0.04) were significantly higher in the subjects ≥20 years who retained the Xm chromosome, when compared to those in the the Xp group (Table 5). Table 5 Serum lipid and carbohydrate data according to the parental origin of the retained X chromosome < 20 years (n =  67) ≥ 20 years (n =  48) Xm (n =  48) Xp (n =  19) P Xm (n =  32) Xp (n =  16) p Total cholesterol (mmol/l) 175 ± 29 183 ± 35 0.21 199 ± 38 181 ± 41 0.03 HDL-cholesterol (mmol/l) 61 ± 15 62 ± 13 0.76 55 ± 11 61 ± 15 0.13 LDL-cholesterol (mmol/l) 98 ± 21 101 ± 15 0.65 105 ± 17 95 ± 21 0.04 Triglycerides (mmol/l) 95 ± 52 97 ± 45 0.61 127 ± 55 99 ± 52 0.001 Fasting glucose (mg/dl) 81 ± 9 85 ± 9 0.18 91 ± 8 88 ± 10 0.38 BMI 19.9 ± 1.34 20.7 ± 2.13 0.54 25.1 ± 7.3 24.7 ± 6.1 0.81 The means of Xm and Xp were compared by one-way ANCOVA analysis to age and BMI used as covariates, followed by Tukey’s significant-difference test. Two-sided p values were calculated for age, BMI, total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides and fasting glucose, but were found to be significantly different only for total cholesterol (p ≤ 0.001), and for LDL cholesterol (p ≤ 0.037) in the ≥ 20-year-group.

y Tukey’s significant-difference test. Two-sided p values were calculated for age, BMI, total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides and fasting glucose, but were found to be significantly different only for total cholesterol (p ≤ 0.001), and for LDL cholesterol (p ≤ 0.037) in the ≥ 20-year-group. Discussion As reported in previous studies [7-14], but with an appropriate sample size and with more sensitive/specific molecular analysis, we found that 2 out of 3 patients with monosomy for 45,X retained the Xm. To explain the higher frequency of maternal X-chromosome several hypotheses have been proposed [15-19]. However, and we agree with the hypothesis, given the nonviability of 45,Y cells, the reason for this ratio is that women have two chances of contributing to a monosomic offspring, while man has only one. Thus, the 2:1 ratio is not consistent with a protective effect of the maternal X-chromosome [8,9,20-22].