Distribution and Cutoff Points of Fasting Insulin in Asian Indian Adolescents and their Association with Metabolic Syndrome

Original Article Distribution and Cutoff Points of Fasting Insulin in Asian Indian Adolescents and their Association with Metabolic Syndrome NK Vikram*, A Misra**, RM Pandey***, Kalpana Luthra****, SP Bhatt* Abstract Aim: To evaluate the levels and appropriate cutoff points of fasting insulin, and their association with the metabolic syndrome (MS) in Asian Indian adolescents. Methods: This cross-sectional, population based study included 948 (527 males & 421 females) adolescent subjects aged 14-19 years selected randomly from New Delhi, India. Cutoff points of fasting insulin were defined using Receiver Operating Characteristics curve analysis against overweight, abdominal obesity and high subscapular skinfold thickness. The MS was defined according to NCEP, ATP III and IDF criteria using age-, gender- and ethnicityspecific cutoff points. Results: Fasting insulin levels peaked at 16 y and reduced subsequently in both genders. The derived cutoff points for fasting insulin (pmol/l) were: 14-15 y- 128.5 and 164.8; 16-17 y- 126.1 and 152.8; 18-19 y- 121.2 and 162.4 in males and females, respectively. Prevalence of fasting hyperinsulinemia (39.1%) and MS (NCEP 2.2%, IDF 1.5%) was highest in age group 16-17 years. Conclusion: The data from this first study describing the distribution and cutoff points of fasting insulin in Asian Indian adolescents may be helpful for detection of and application of primary prevention strategies for fasting hyperinsulinemia and the metabolic syndrome in this population. Introduction Global increase in the prevalence of obesity in children and adolescents is a matter of concern. Obesity is associated with insulin resistance and its related metabolic derangements which in turn increase the risk of type 2 diabetes mellitus (T2DM) and coronary heart disease (CHD) at an early age. 1 Therefore it is important to identify the individuals who are insulin resistant for primary prevention. Clustering of insulin resistance related factors has been defined as the metabolic syndrome. Despite a great interest in this area, there is no optimal universal definition of the metabolic syndrome in children and adolescents. Various investigators have used modifications of the National Cholesterol Education Program, Adult Treatment Panel III (NCEP, ATP III) definition and reported the prevalence of the metabolic syndrome. 2 In a recent study we observed that the NCEP, ATP III definition identified very small number *Department of Medicine; ***Biostatistics, and 4Biochemistry, All India Institute of Medical Sciences, New Delhi-110029, India. **Department of Diabetes and Metabolic Diseases, Fortis Hospitals, New Delhi, 110070, India. Received : 17.7.2008; Accepted : 25.10.2008 of at risk Asian Indian adolescents. Further, inclusion of fasting insulin was associated with higher prevalence of the metabolic syndrome. 3,4 Despite the differences in the overall prevalence, the prevalence of the MS was uniformly found to be higher in overweight individuals. 3 Fasting insulin levels correlate with insulin resistance measured by more accurate methods or gold standard methods and therefore may be used as surrogate marker of insulin resistance. 5 Fasting hyperinsulinemia has been shown to be associated with clustering of cardiovascular risk factors like hypertension, dyslipidemia and endothelial dysfunction. Ethnic differences in the fasting insulin levels have been reported; South Asian children had higher levels of fasting insulin as compared to children of European origin. 6,7 Monitoring of fasting insulin levels may be useful in identifying individuals who are insulin resistant and also for monitoring the trends in fasting insulin levels. However, diagnosis of fasting hyperinsulinemia and consequent preventive steps are hampered by absence of clear consensus regarding cutoff points of fasting insulin. Moreover, universal cutoff points are unlikely to be appropriate due to ethnic differences in the fasting insulin concentrations. There is lack of data on the distribution or cutoff points of fasting insulin in Asian Indian adolescents. JAPI VOL. 56 DECEMBER 2008 www.japi.org 949

In this study we evaluated the distribution of fasting insulin levels and attempted to define appropriate cutoff points for Asian Indian adolescents. In addition we also evaluated the relationship of these cutoffs with the metabolic syndrome. Methodology Data from a population based study, the Epidemiological Study of Adolescents and Young adults (ESAY study), details of which are mentioned elsewhere, 3,4 were included for the present analysis. Briefly, the ESAY study included adolescents and young adults aged 14-25 years selected from various schools and colleges of southwest New Delhi. Subjects were recruited using a multi-stage cluster sampling methodology based on the modified World Health Organization Expanded Program of Immunization Sampling Plan as described previously. 3.4 The study was approved by the Director of Education, Ministry of Education, Government of New Delhi and institutional ethics committee. A written informed consent was obtained from subjects who were 18 years or older and from the parents/guardians of subjects less than 18 years of age. Data from 948 subjects (527 males and 421 females), selected randomly from a total population of 1447 subjects (768 males and 679 females) aged 14-19 years were included for the present analysis. The anthropometric and biochemical characteristics of subjects aged 14-19 years not included in the present analysis (n=499) did not differ significantly from those included in the present analysis. Clinical Profile and Measurements Demographic details were recorded and a detailed clinical examination was performed. Anthropometric measurements included; height, weight, waist circumference (WC), body mass index (BMI), waist-to-hip circumference ratio (W HR) and skinfold thickness at four sites (biceps, triceps, subscapular and suprailiac). The measurements were made as described previously. 8 Anthropometric measurements were made by two observers, one each for males and females. These observers were trained prior to initiation of the study. The inter-observer variation in the measurements was less than 5%, which was acceptable. Percentage body fat was measured using a four-point bioelectrical impedance apparatus (Tanita TBF 300, TANITA Corp., Tokyo, Japan), which has been validated for Asian children and adolescents. 9 Subjects stood on the apparatus with both the feet in firm contact with the surface of the apparatus and hands not touching any surface. Blood pressure was measured by a single physician using a standard mercury sphygmomanometer (Industrial Electronic and Allied Products, Pune, India) which was periodically validated against a Hawksley Random Zero Sphygmomanometer (Hawksley, Lancing, Sussex, UK). Measurements were made using appropriate cuff size, after the subject had rested for 5 minutes in the sitting position and phase 5 Korotkoff sounds were taken for the categorization of diastolic blood pressure. Metabolic Parameters and Serum Insulin Venous blood samples were obtained after an overnight fast. Fasting blood glucose (FBG), total cholesterol (TC), serum triglycerides (TG), and high-density lipoprotein cholesterol (HDL-c) concentrations estimations were done using commercially available reagent kits (Randox Laboratory, San Francisco, CA, USA) as mentioned elsewhere. 4 The value of low-density lipoprotein cholesterol (LDL-c) was calculated according to Friedewald s equation if fasting serum triglycerides levels were <400 mg/dl. 10 Serum insulin levels were determined using a radioimmunoassay kit (Medicorp, Montreal, Canada) as described previously. 3 The intra-assay and inter-assay percentage coefficient variables were 2.5% and 3%, respectively. The quality control check on insulin assays were rigorously maintained by a biochemist (K.L.). Definitions Since no cutoff points are available for Asian Indian adolescents for anthropometric and metabolic parameters, the percentile data of subjects aged 14-19 years from the ESAY study cohort was used as reference for deducing the cutoff points (Table 1). Overweight was defined as value of BMI 85 th percentile for age and gender. Values 90 th percentile for age and gender were used to define high values of WC and subscapular skinfold (SS) thickness (Table 1). Hypertriglyceridemia was defined as a value 90 th percentile ( 128 mg/dl) and for low HDL-c was defined as levels <10 th percentile (<40 mg/dl). 2 Elevated blood pressure was defined as presence of either systolic or diastolic blood pressure to 90 th percentile for age, gender and height, 2,11 and those on antihypertensive treatment. Impaired fasting glucose (IFG, FBG 100 mg/ dl and <126 mg/dl) and diabetes (FBG 126 mg/dl) were defined according to the American Diabetic Association criteria. 12 Homeostasis model assessment was computed by using the formula: fasting glucose (mmol/l) X fasting insulin (µu/ml)/22.5. 13 Subjects were categorized into three age groups; 14-15 y, 16-17 y and 18-19 y to determine appropriate cutoffs of fasting insulin. This was done because of differences in the maturation between the age groups and also to have an adequate sample size in each group. In our earlier studies we have demonstrated that BMI, WC and subscapular skinfold thickness correlate and are useful in identification of fasting hyperinsulinemia. 4,14 Therefore cutoff points of fasting insulin were determined against overweight, high WC and high subscapular skinfold thickness as independent variables. The metabolic syndrome was defined according to the NCEP, ATP III 15 and the International Diabetes Federation (IDF) criteria 16 using percentile based cutoff points specific for Asian Indian adolescents for various components as mentioned earlier. Statistical Methods Mean and standard deviation were used to summarize the data. Differences in anthropometric and metabolic parameters between males and females were compared using the Z-test. The relationships of BMI and fasting insulin levels with other components of the metabolic syndrome 950 www.japi.org JAPI VOL. 56 DECEMBER 2008

were evaluated by Spearman s correlation method. For each age and gender group Receiver Operating Characteristics (ROC) curve analysis was used to determine the appropriate cutoff values of fasting insulin that corresponds separately to overweight, high WC and high SS. The mean of the three cutoff values of fasting insulin as obtained above for each age and gender group was defined as the final cutoff point for analysis. STATA 9.0, Intercooled version statistical software (STATA Corporation, College Station, TX, USA) was used for the statistical analysis. In this study statistical significance was considered at a P value <0.05. Results Clinical, Anthropometric and Biochemical profiles Female subjects were slightly older (16.3±0.8 y) than male subjects (16.5±0.6 y, p<0.05). Mean systolic and diastolic blood pressures were higher in males (114.8±9.9 and 74.4±7.4 mmhg) as compared to females (110.5±8.9 and 73.0±6.9 mm Hg, p<0.01). The mean BMI was similar in males (19.6±3.3 kg/m 2 ) and females (20.0±3.3 kg/m 2 ) whereas males had significantly higher WC (69.9±8.7 vs. 66.8±7.6 cm, p<0.001) and WHR (0.81±0.05 vs. 0.75±0.06, p<0.001) as compared to females. Waist-to-height ratio was similar in both the genders (males: 0.42±0.05, females: 0.43±0.05, p=ns). All the skinfold thicknesses were higher in females as compared to males (p<0.001). None of the subjects was detected to be diabetic. Fasting blood glucose levels were comparable in both the genders (males: 90.0 ± 9.3, females: 89.5±8.5 mg/dl, p=ns). Impaired fasting glucose was present in 11.2% subjects (12.7% males and 9.3% females). Mean values of all the lipid parameters were higher in females as compared to males (p<0.01). Fasting insulin levels The distribution of mean fasting insulin levels at ages 14-19 is presented in Figure 1. Serum levels of fasting insulin were higher in females than males at all ages. The fasting insulin levels showed an increase from age of 14 y, were highest at 16 y of age and subsequently showed a decline with increasing age (Fig. 1a). After adjusting for percentage body fat, fasting insulin concentrations were higher in females than males and were observed to increase till age of 17 y with a subsequent decline which was much more pronounced in males as compared to females (Fig. 1b). The distribution of HOMA values followed a pattern similar to that of fasting insulin across all ages. Fasting insulin levels were significantly higher in overweight subjects (161.6±53.7 pmol/l) as compared to normal weight subjects (124.3±46.7 pmol/l, p<0.001). Similarly subjects with high WC and high subscapular skinfold thickness had higher levels of fasting insulin as compared to subjects with normal values of these parameters (p<0.001 for all). To determine appropriate cutoff of fasting insulin in each age group, ROC analysis was performed with overweight, high WC and high subscapular skinfold thickness as independent predictors (Table 2). The mean value of the cutoff points of fasting insulin generated after ROC analysis Fig. 1: Distribution of mean fasting insulin concentrations in adolescents, (a) unadjusted values and (b) values adjusted for percentage body fat against each predictor variable was defined as the final cutoff. In males the cutoff points of fasting insulin were: for age group 14-15 y 128.5 pmol/l, for 16-17 y 126.1 pmol/l and for 18-19 y 121.2 pmol/l. In females, the cutoffs were; for 14-15 y 164.8 pmol/l, for 16-17 y 152.6 pmol/l and for 18-19 y 162.4 pmol/l. Based on the generated cutoff points, fasting hyperinsulinemia was present in 35.4% subjects (14-15 y- 32.6%, 16-17 y-39.1% and 18-19 y- 32.7%). The prevalence of fasting hyperinsulinemia was 67.3%, 67.4% and 67.7% in those who were overweight, had high WC and high subscapular skinfold thickness, respectively (p<0.001 for all), than those who had normal values of BMI (29.3%), WC (32.1%) and subscapular skinfold thickness (29.1%). The Metabolic syndrome and Fasting Hyperinsulinemia Body mass index showed positive correlation with WC (males: p (rho)=0.87, females p (rho)=0.76, p<0.001 for both), TG (males: p (rho)=0.20, females p (rho)=0.14, p<0.01 for both), subscapular skinfold thickness (males: p (rho)=0.79, females p (rho)=0.79, p<0.001 for both) and fasting insulin (males: p (rho)=0.36, females p (rho)=0.35, p<0.001 for both). Fasting insulin levels showed significant correlation with WC (males: p (rho)=0.36, females p (rho)=0.23, p<0.001 for both), TG (males: p (rho)=0.13, p<0.01 in males only), and subscapular skinfold thickness (males: p (rho)=0.39, females p (rho)=0.37, p<0.001 for both). The metabolic syndrome was defined according to the NCEP and IDF definitions with modified percentile-based cutoff points for Asian Indian adolescents. The prevalence of the metabolic syndrome was recorded to be 1.7 % according to the NCEP definition JAPI VOL. 56 DECEMBER 2008 www.japi.org 951

Table 1 : Age- and gender-specific cutoff points to define abnormal values of anthropometric parameters Age (years) Percentile based cutoff points Body mass index Waist circumference Subscapular skinfold 85 th centile (kg/m 2 ) 90 th centile (cm) thickness 90 th centile (mm) Males Females Males Females Males Females 14 21.5 21.6 80.8 74.8 25.5 30.1 15 21.9 22.8 81.4 75.5 23.1 31.0 16 22.7 23.7 87.0 78.2 30.1 34.2 17 22.8 23.9 82.0 75.0 25.3 33.1 18 23.2 23.9 82.6 76.9 28.0 28.2 19 22.9 22.8 81.8 79.0 20.9 29.4 Table 2 : Cutoff values of fasting insulin derived after ROC analysis Age group Overweight High waist circumference High subscapular skinf Mean (years) old thickness cutoff Insulin cutoff Sensitivity Specificity Insulin cutoff Sensitivity Specificity Insulin cutoff Sensitivity Specificity value (pmol/l) (%) (%) (pmol/l) (%) (%) (pmol/l) (%) (%) (pmol/l) Males 14-15 122.5 76.2 72.7 135.2 75.0 74.6 127.8 83.3 72.0 128.5 16-17 126.0 76.7 73.5 126.1 71.4 68.9 126.3 74.1 70.6 126.1 18-19 112.6 63.2 62.6 127.4 80.0 78.4 123.6 70.0 69.6 121.2 Females 14-15 164.0 73.7 73.5 160.3 72.7 62.6 170.3 83.3 72.9 164.8 16-17 159.0 60.9 60.6 145.8 47.4 46.0 153.6 63.6 53.0 152.8 18-19 153.8 75.0 70.6 163.4 81.2 74.0 170.1 80.0 78.0 162.4 and 1.3% according to the IDF definition. With inclusion of fasting hyperinsulinemia in the definition of the metabolic syndrome, the prevalence increased to 7.0% and 3.8% according to the NCEP and IDF definitions, respectively. Comparison of Fasting Insulin levels with the NHANES data (Table 3) The data from the present study were compared with the recent the data on insulin levels among individuals aged 12-19 years from the third National Health and Nutrition Examination Survey, U.S.A., 17 which included Caucasians, African-Americans and Mexican-Americans, but not Asians. The comparison of fasting insulin levels of across similar age groups is presented in Table 3. Fasting insulin levels were substantially higher in Asian Indian adolescents in all age groups as compared to adolescents from USA. Discussion This is the first population-based study on the distribution of fasting insulin levels in urban Asian Indian youth age 14-19 y. The levels of fasting insulin were higher in females across all ages than males even after adjusting for percentage body fat. Unadjusted values of fasting insulin levels were highest at 16 y of age. After adjusting for percentage body fat, fasting insulin concentrations were highest at 17 y of age with subsequent decline with increasing age, the change being more prominent in males than females. In the ESAY study, percentage body fat and subscapular skinfold thickness were observed to increase till 16 y of age with subsequent decline in adolescents aged 14-19 y (unpublished data). Fasting insulin levels to some extent may reflect this change in percentage body fat and subscapular skinfold thickness with age. Since fasting insulin levels are influenced by adiposity, the proposed cutoff points of fasting insulin were generated for different age groups against measures of overall, abdominal and truncal adiposity in the current study. These cutoff points may be useful for identification of adolescents with fasting hyperinsulinemia in Asian Indian adolescents. In the study by Ford et al. 17 the insulin levels were observed to increase from 12 to 14 y before declining and were highest at 14 y of age in males and at 13 y of age in females. In our study children below 14 y of age were not included. However, fasting insulin levels were observed to be highest at 16-17 y of age in the present study. No other major data are available regarding fasting insulin level cutoffs in any other ethnic groups. Comparison of the fasting insulin levels of the present study with those from the NHANES data 17 across similar age groups reveals that insulin levels are substantially higher in Asian Indian adolescents in all age groups as compared to adolescents from USA. Though there is lack of uniformity in the methods of estimation of insulin in the two studies, it further emphasizes the fact that Asian Indians have higher insulin levels as compared to the other ethnic groups. It would however be interesting to compare the degree of adiposity among these two datasets. Differences in body composition in the form of excess body fat, in particular truncal subcutaneous adipose tissue in Asian Indians may be an important determinant of fasting insulin levels. 4 We recently reported that Asian Indian children had thicker truncal skinfolds as compared to White and Black children. 4 In a recent study, South Asian adolescents were 952 www.japi.org JAPI VOL. 56 DECEMBER 2008

Table 3 : Comparison of fasting insulin values of Asian Indian adolescents in the present study with those of data from NHANES III* Age groups Percentiles of serum insulin (pmol/l) (years) 25 th 50 th 75 th NHANES Asian NHANES Asian NHANES Asian III Indians III Indians III Indians Males 14-15 48.66 85.40 63.00 104.60 92.88 145.30 NHANES III: n=199; Asian Indians: n=131 16-17 47.16 84.30 59.58 110.55 78.54 136.20 NHANES III: n=259; Asian Indians: n=262 18-19 41.34 84.10 56.58 101.80 84.72 131.60 NHANES III: n=224; Asian Indians: n=135 Females 14-15 49.80 110.60 65.34 139.85 90.00 179.10 NHANES III: n=214; Asian Indians: n=102 16-17 49.26 122.30 64.56 150.40 82.56 180.30 NHANES III: n=209; Asian Indians: n=145 18-19 51.42 107.50 63.00 138.25 86.46 175.10 NHANES III: n=184; Asian Indians: n=174 NHANES: National Health and Nutrition Examination Survey III. *: Data from Ford et al. 17 observed to have higher amount of total and central fat, and lower insulin sensitivity as compared to White European adolescents, however, this difference in insulin sensitivity was no longer significant when body fat was matched. 6 There is no consensus regarding the appropriate definition of the metabolic syndrome in children and adolescents. Different investigators have used modifications of the existing definitions of the metabolic syndrome. 2 The current definitions of the metabolic syndrome may not be appropriate for Asian Indians since they identified only a small proportion of subjects at risk. 3 Inclusion of fasting insulin in the definition identified more number of at risk subjects by either of two, NCEP and IDF definitions. A major limitation of this study is the lack of data on the sexual maturity status of the subjects. The stage of sexual maturation has significant influence on insulin sensitivity. 18 In a recent study involving children and adolescents across different Tanner stages, insulin resistance increased significantly from T2, remained constant from T2 to T4 and subsequently returned almost to T1 levels by stage T5. 19 These changes in insulin resistance did not correlate with changes in BMI, suggesting that changes in the body composition may not be responsible for insulin resistance of puberty. Changes in the growth hormone/ insulin like growth factor-1 axis mirror the changes in insulin resistance, and may be responsible for the pubertal insulin resistance. 19 In a study involving urban north Indian children, it was observed that the mean age of attainment of G-4 stage of genital development in boys was 14.6 years, and about 75% of boys between 13-16 years were in G-4 stage. 20 In girls the mean age of attainment of menarche was observed to be 12.6 years and the mean age for B-4 was 13.5 years. 20 Therefore, even though data on the precise pubertal stage were not available, it is likely that the majority of the subjects were post-pubertal in the present study. Other limitations of the present study include its cross-sectional design and the lack of use of more precise methods of determination of body composition and insulin sensitivity. The usefulness of the cutoff points of fasting insulin generated in predicting future metabolic complications needs to be evaluated in a longitudinal study. In conclusion, the data from the present study may serve as reference for future research studies and may be useful in identifying Asian Indian adolescents with fasting hyperinsulinemia so that appropriate primary prevention strategies can be instituted early. Acknowledgements This study was partially funded by research grant from the Department of Science and Technology, Ministry of Science and Technology, Government of India. The cooperation of the children who took part in the study, and the help extended by the principals, teachers, and staff of the various schools and colleges where the study was conducted is greatly appreciated Competing Interests None to declare. References 1. Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics 1999;103:1175-82. 2. Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of JAPI VOL. 56 DECEMBER 2008 www.japi.org 953

a metabolic syndrome phenotype in adolescents: findings from the third National Health and Nutrition Examination Survey, 1988-1994. Arch Pediatr Adolesc Med 2003;157:821-27. 3. Vikram NK, Misra A, Pandey RM, Luthra K, Wasir JS, Dhingra V. Heterogeneous phenotypes of insulin resistance and its implications for defining metabolic syndrome in Asian Indian adolescents. Atherosclerosis 2006;186:193-99. 4. Misra A, Vikram NK, Arya S, Pandey RM, Dhingra V, Chatterjee A, et al. High prevalence of insulin resistance in postpubertal Asian Indian children is associated with adverse truncal body fat patterning, abdominal adiposity and excess body fat. Int J Obes Relat Metab Disord 2004;28:1217-26. 5. Gungor N, Saad R, Janosky J, Arslanian S. Validation of surrogate estimates of insulin sensitivity and insulin secretion in children and adolescents. J Pediatr 2004;144: 47-55. 6. Ehtisham S, Crabtree N, Clark P, Shaw N, Barrett T. Ethnic differences in insulin resistance and body composition in United Kingdom adolescents. J Clin Endocrinol Metab 2005;90: 3963-69. 7. Whincup PH, Gilg JA, Papacosta O, Seymour C, Miller GJ, Alberti KG, et al. Early evidence of ethnic differences in cardiovascular risk: cross sectional comparison of British South Asian and white children. BMJ 2002;324:635. 8. Dudeja V, Misra A, Pandey RM, Devina G, Kumar G, Vikram NK. BMI does not accurately predict overweight in Asian Indians in northern India. Br J Nutr 2001;86:105-12. 9. Sung RY, Lau P, Yu CW, Lam PK, Nelson EA. Measurement of body fat using leg to leg bioimpedance. Arch Dis Child 2001;85:263 67. 10. Friedewald WT, Levy RI, Fredrikson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem 1972;18:499-502. 11. Madhavan M, Pandey RM, Misra A, Vikram NK, Dhingra V, Luthra K, et al. Centile values for serum lipids and blood pressure for Asian Indian adolescents. Lipids Health Dis 2005;4:20. 12. American Diabetic Association. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. In: Diabetes Care 2003; S5-S20. 13. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28: 412-19. 14. Misra A, Madhavan M, Vikram NK, Pandey RM, Dhingra V, Luthra K. Simple anthropometric measures identify fasting hyperinsulinemia and clustering of cardiovascular risk factors in Asian Indian adolescents. Metabolism 2006;55:1569-73. 15. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285: 2486-97. 16. The IDF consensus worldwide definition of the metabolic syndrome. Avaialable at http://www.idf.org/webdata/docs/idf_metasyndrome_ definition.pdf (accessed September, 2005). 2005. 17. Ford ES, Li C, Imperatore G, Cook S. Age, sex, and ethnic variations in serum insulin concentrations among U.S. youth: findings from the National Health and Nutrition Examination Survey 1999-2002. Diabetes. Care 2006; 29 : 2605-11. 18. Moran A, Jacobs DR Jr, Steinberger J, Hong CP, Prineas R, Luepker R et al. Insulin resistance during puberty: results from clamp studies in 357 children. Diabetes 1999; 48: 2039-44. 19. Cook JS, Hoffman RP, Stene MA, Hansen JR. Effects of maturational stage on insulin sensitivity during puberty. J Clin Endocrinol Metab 1993; 77: 725 730. 20. Agarwal DK, Agarwal KN, Upadhyay SK, Mittal R, Prakash R, Rai S. Physical and sexual growth pattern of affluent Indian children from 5 to 18 years of age. Indian Pediatr 1992;29:1203-82. 954 www.japi.org JAPI VOL. 56 DECEMBER 2008