| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Birmingham Children’s Hospital (S.E., N.S., T.B.) and University Hospital Birmingham (N.C., P.C.), Birmingham B4 6NH, United Kingdom
Address all correspondence and requests for reprints to: Dr. Sarah Ehtisham, Department of Diabetes and Endocrinology, Birmingham Children’s Hospital, Steelhouse Lane, Birmingham, B4 6NH, United Kingdom. E-mail: s.ehtisham{at}bham.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The goal was to evaluate sex and ethnic differences in insulin sensitivity and body composition in healthy adolescents.
Design: This was a cross-sectional cohort study.
Setting: This was a community-based study.
Participants: One hundred twenty-nine healthy white European and South Asian 14- to 17-yr-old adolescents participated.
Interventions: Body composition was assessed by anthropometry and dual-energy x-ray absorptiometry, and insulin sensitivity by homeostasis model assessment.
Main outcome measures: The main outcome measures were body fat percentage and insulin sensitivity.
Results: We confirmed that South Asian adolescents were less insulin sensitive than white European adolescents (homeostasis model assessment of insulin sensitivity, 52.4 vs. 58.9%, P < 0.05), with a trend toward lower insulin sensitivity in girls. South Asian adolescents had significantly more body fat than white European adolescents (girls, 30.6 vs. 26.0%, P < 0.005; boys, 20.8 vs. 14.8%, P < 0.001), with more central fat (waist-thigh ratio in girls, 1.36 vs. 1.25, P < 0.001; boys, 1.52 vs. 1.42, P < 0.001). The sex-ethnic differences in insulin sensitivity were no longer seen when body fat was included as a covariate.
Conclusions: Ethnic differences in insulin sensitivity are associated with ethnic differences in body fat. South Asian adolescents are more insulin resistant, with more body fat than white European adolescents, which may contribute to their increased risk of developing type 2 diabetes.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The relationship between insulin resistance and fat distribution in childhood and adolescence is not clear. A strong relationship between insulin resistance and visceral fat has been demonstrated in obese children and adolescents but not in lean adolescents, in whom sc fat may be more important (9, 10).
The EarlyBird study of prepubertal white European children found that 5-yr-old girls were more insulin resistant than boys, but it is unclear what role adiposity plays at this age (11). In adults, an android or male body fat pattern (with greater central and upper body fat) is associated with greater metabolic risk, whereas a gynoid or female fat distribution (with relatively more fat in the hip and thigh regions) is associated with lower metabolic risk (12). The emergence of sex differences in body fat distribution has been demonstrated in prepubertal children but with different characteristics in different ethnicities, highlighting the importance of ethnic-specific studies (13). However, it is unclear whether these ethnic and sex differences in body composition play a role in determining insulin resistance and metabolic risk in childhood.
The Ten Towns Heart Health studies of United Kingdom children of age 8–11 yr found significant ethnic differences in cardiovascular and diabetes risk factors (14). The researchers found that South Asian children were more insulin resistant than white European children but were unable to link the insulin resistance to adiposity in their cohort, concluding that the South Asian children were intrinsically more insulin resistant independent of adiposity.
Studies in other ethnic groups have demonstrated ethnic differences in adiposity and insulin resistance in childhood. African-American children have less visceral fat than Caucasian children but are more insulin resistant, independent of visceral fat accumulation (15). This implies that the effects of adiposity may differ according to ethnicity (16).
Our hypothesis was that sex and ethnic differences in insulin sensitivity are directly related to sex and ethnic differences in body composition in children, i.e. that South Asian children have more body fat that is more centrally distributed than white European children and are therefore more insulin resistant.
The aims of this study were to evaluate sex and ethnic differences in insulin sensitivity and body composition in a cohort of healthy South Asian and white European adolescents. Two independent methods of assessing body fat, anthropometry and dual-energy x-ray absorptiometry (DXA), were used to investigate body composition, and homeostasis model assessment (HOMA) analysis was used to evaluate insulin sensitivity in 129 healthy postpubertal adolescents.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Healthy adolescents were recruited from schools in Birmingham through school assemblies and personal and social education classes, at which information was given about the project. The recruitment was targeted at 14- to 17-yr-old white European and South Asian (Indian, Pakistani, Bangladeshi, and Sri Lankan) adolescents. Volunteers were given written information packs and were recruited if they returned a reply slip indicating their interest (n = 138). The volunteers and their families were subsequently visited at home to undertake data collection, clinical measurements, and fasting blood tests. Informed written consent was obtained from all participants and a parent or guardian. The study protocol was approved by both the Birmingham Children’s Hospital and South Birmingham Local Research Ethics Committees.
Medical questionnaire
A brief medical questionnaire was completed for each participant, which included medical history, medications and allergies, family history, and pubertal status. Boys were asked at what age their voice had broken, to ascertain stage IV puberty, and were excluded from subsequent analyses if they were less than stage IV puberty (17). Girls were asked their age at menarche, to ascertain stage IV puberty, and were subsequently excluded if they had not attained menarche. Ethnicity was self-reported for three generations, and no children of mixed ethnicity were included in the study. The adolescents were also examined for the presence of acanthosis nigricans, a cutaneous marker of insulin resistance (18, 19).
Analytical procedures
Blood samples were taken after an 8-h fast, for the determination of plasma glucose, lipids and insulin, proinsulin, and C-peptide concentrations. Glucose and lipids were analyzed routinely in the Clinical Chemistry Department at Birmingham Children’s Hospital (Vitros 950 Dry Chemistry Analyser; Ortho-Clinical Diagnostics, High Wycombe, UK). Insulin was determined using the mean of three plasma samples drawn over a 10-min period, in view of the pulsatility of insulin secretion (20, 21). The samples for insulin, proinsulin, and C-peptide were centrifuged immediately and the plasma stored on ice until it was transported to the laboratory and frozen at –20 C before analysis. All samples were frozen within 1 h of collection. Insulin was measured using an immunoenzymometric assay (Medgenix; Biosource Technologies, Inc., Nivelles, Belgium), at the Regional Endocrine Laboratory, University Hospital Birmingham. Total proinsulin and C-peptide were measured using an ELISA (C-peptide; DakoCytomation, Ely, UK) at the Regional Endocrine Laboratory, University Hospital Birmingham. HOMA analysis of insulin sensitivity (HOMA %S) and ß-cell function (HOMA %B) was undertaken on the fasting glucose and insulin concentrations, using the computer program (HOMA and CIGMA model 2.00) kindly supplied by Dr. Jonathan Levy, Diabetes Research Laboratories, Oxford, United Kingdom (22, 23, 24).
Anthropometry
Weight was measured in light indoor clothing, to the nearest 0.1 kg (Tanita Digital scales; Tanita UK Ltd., Middlesex, UK); and height was measured without shoes, to the nearest 0.1 cm, using a portable stadiometer (Leicester Height Measure; Invicta Plastics Ltd., Oadby, UK). SD scores (SDS) were calculated using the United Kingdom 1990 Growth Reference data (25). Body mass index (BMI) was calculated as weight (kilograms) divided by height (meters) squared. Overweight and obesity were defined according to the International Obesity Task Force guidelines, which define overweight as a BMI equivalent to 25 at age 18, and obesity as a BMI equivalent to 30 at age 18 (26). Waist, hip, and thigh circumferences were measured to the nearest 0.1 cm using a flexible nonstretchable tape measure, with the subject in light indoor clothing, exposing the relevant parts as needed to undertake the measurement. Waist circumference was taken at the narrowest part of the waist, midway between the 10th rib and the iliac crest, with the subject standing erect. Hip circumference was taken at the widest part of the hip, with the subject standing erect. Thigh circumference was taken at the widest part of the upper thigh on the nondominant side, with the subject standing erect with legs slightly apart. Waist-hip ratio and waist-thigh ratio were subsequently calculated. Skinfold thicknesses were measured to the nearest 0.1 mm with a Holtain caliper, at the following sites: biceps, triceps, suprailiac, and subscapular. All skinfold measurements were taken in duplicate on the nondominant side, and the average of the two readings was used. All the anthropometric measurements were undertaken by one person, using the same techniques and the same equipment for each participant. Percentage body fat was calculated from the skinfold thickness measurements, using linear regression equations (27, 28).
DXA
To validate the estimations of percentage body fat derived from skinfold thickness measurements, participants were invited to attend for a DXA scan within 4 months of the home visit, to give measures of total body composition (Lunar DPX-L pencil beam scanner; GE Systems, Madison, WI). Trunk thickness and body weight were used to ensure that each adolescent was scanned in the most appropriate acquisition mode. Scan acquisition was performed by trained personnel and analysis performed by a single trained operator (N.C.). Body composition was calculated from the whole-body scan, to provide measures of lean and fat body mass.
Statistical analysis
All statistical analysis was undertaken using Microsoft Excel 2000 and SPSS software version 11.0.1, with statistical advice from Dr. Paul Davies, Statistician, Birmingham Children’s Hospital. Data are expressed as means ± SE. Any variables that were not normally distributed were log-transformed before data analysis. Pearson correlation coefficients were used to quantify relationships between variables. t tests were used to compare means between sex and ethnic groups. Two-way ANOVA was used to explore the effect of sex and ethnicity on the continuous variables and to evaluate the effects of covariates, with P < 0.05 taken to indicate a statistically significant difference. Sex and ethnic differences in categorical variables were evaluated with binomial logistic regression. Multiple linear regression analysis was undertaken to examine the contribution of variables to insulin sensitivity. A power calculation based on published data in young adults indicated that recruiting 30 adolescents per sex-ethnic group would give over 95% power for detecting differences between groups of 10.8 mg/dl (0.6 mmol/liter) in fasting plasma glucose values and of 0.06 in waist-hip ratio in this cohort (6).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The study cohort was equally divided by sex and ethnic group, with 66 girls (33 white European, 33 South Asian) and 63 boys (31 white European, 32 South Asian). The majority of the South Asian adolescents, 51, were Pakistani; 10 were Indian; 3 were Bangladeshi; and 1 was Sri Lankan.
Characteristics of the cohort
The clinical characteristics of the study population are given in Table 1
.
|
Sex and ethnic anthropometric differences
The South Asian girls were significantly shorter than their white European peers, but there were no significant sex-ethnic differences in weight or BMI SDS between the groups. In view of the higher prevalence of overweight and obesity in the South Asian adolescents, subsequent statistical analyses on anthropometric data were undertaken using ANOVA with BMI SDS as a covariate.
The South Asian adolescents had larger skinfold thickness measurements compared with the white European adolescents; these differences were statistically significant even after adjusting for BMI SDS. Body fat calculated from skinfold thickness correlated well with that measured on DXA scan, in both South Asian and white European adolescents, implying that the skinfold thickness equations are valid for use in South Asian adolescents (South Asian, r = 0.932, P < 0.001; white European, r = 0.946, P < 0.001). The Bland-Altman plot in Fig. 1 shows that the DXA estimates of body fat are higher than calculated body fat and that the difference in the result of the two methods increases with increasing percent body fat (30). Because the two measurements correlated well, body fat calculated from skinfolds was used in subsequent analyses because it was available for all adolescents.
|
). These ethnic differences remained statistically significant even after adjusting for BMI SDS (mean relative increase of 10% in the girls and 34% in the boys, P < 0.001 on ANOVA). The DXA data also showed that the South Asian adolescents had significantly lower percentage lean body mass independent of BMI SDS.
|
).
|
One of the South Asian boys was found to have impaired fasting glycemia as defined by the World Health Organization Criteria, with a fasting plasma glucose of 117.1 mg/dl (6.5 mmol/liter); he was referred for further investigation but included in the statistical analyses (31). All the other adolescents had fasting plasma glucose values within the normal range. After exclusion of the boy with impaired fasting glycemia, the South Asian girls had higher fasting plasma glucose values than all the other groups (mean glucose ± SE, 82.6 ± 1.22 mg/dl (4.58 mmol/liter) vs. 79.6 ± 0.62 ng/dl (4.42 mmol/liter), P < 0.05 on t test).
South Asian adolescents had significantly higher fasting insulin levels and were less insulin sensitive than white European adolescents (mean HOMA %S ± SE, 52.4% ± 2.3 vs. 58.9% ± 2.3, P < 0.05 on ANOVA). Girls were less insulin sensitive than boys, although this did not reach statistical significance (mean HOMA %S ± SE, 52.7 ± 2.3% vs. 58.6% ± 2.4%, P = 0.08 on ANOVA). South Asian adolescents also had higher total proinsulin and C-peptide levels, although these differences were not statistically significant. There was evidence of ethnic differences in lipid profile, with South Asian adolescents having a more adverse profile with significantly higher total cholesterol and lower high-density lipoprotein (HDL) cholesterol levels. Girls had significantly higher HDL cholesterol levels than boys. There was a trend toward South Asian adolescents having higher HOMA %B, although this did not reach statistical significance (P = 0.09 on ANOVA).
HOMA %S was linearly inversely correlated with percent body fat (Pearson correlation coefficient, –0.492, P < 0.001), and the relationship between insulin sensitivity and degree of adiposity was the same in the white European and South Asian adolescents (Fig. 4). When total body fat was included as a covariate in ANOVA, the ethnic differences in insulin sensitivity were no longer seen (mean HOMA %S, 55.0 vs. 56.0%, P = 0.569).
|
). Essentially identical results were obtained when waist circumference was substituted for percentage body fat in the model. This indicates that the main factors contributing to variation in insulin sensitivity are body fat and gender, implying that the ethnic differences demonstrated in insulin sensitivity are a function of ethnic differences in body composition.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
South Asian adolescents had a higher frequency of a positive family history of diabetes, with a strikingly high prevalence of having an affected parent. Up to a quarter of South Asian adolescents had acanthosis nigricans, which was not seen in the white European adolescents. In keeping with this, the South Asian adolescents were more insulin resistant, with evidence of higher insulin levels and a more adverse lipid profile. They also had higher body fat levels, with increased central fat deposition. South Asian girls also had higher fasting plasma glucose levels than any other group. Ethnic differences in insulin sensitivity were no longer seen when body fat was included as a covariate, implying a major role for body fat in this ethnic variation in insulin sensitivity.
HOMA %S has previously been shown to correlate strongly with insulin sensitivity measured by both hyperinsulinemic-euglycemic clamp and frequently sampled iv glucose tolerance test (24, 32). HOMA has been validated for use in both prepubertal and pubertal, and obese and nonobese children, and it correlates similarly with clamp data in both African-American and white children (33, 34). HOMA has the advantage of being simpler, quicker, less expensive and time-consuming, and more acceptable to children, making it ideal for larger studies.
The gold standard for measurement of body composition is a four-compartment (4C) model, which uses independent measurements of bone mineral content, total body water, and body density to formulate a body fat measurement. However, the complexity and cost of 4C analysis limit its clinical use. DXA has been evaluated against 4C models in children, and a number of studies have shown that it does overestimate body fat by 1–4%, depending on age, sex, and degree of obesity (35, 36, 37). The good correlation between DXA fat measurements and 4C measurements, with such a small magnitude of error, make it a convenient and suitable method for studying children (38). We used two independent methods to assess body fat levels, using the DXA data to validate our use of skinfold-derived estimates of body fat. Body fat calculated from skinfolds correlated well with that measured on DXA scan, and the difference between the two measures was unaffected by sex or ethnicity, implying that the skinfold thickness equations used are valid in both white European and South Asian adolescents.
Sex and ethnic differences in body composition have previously been reported in prepubertal children, but South Asian children were not one of the groups studied (13). In the Ten Towns Heart Health studies of prepubertal children, the South Asian children were less insulin sensitive than their white European peers, but ethnic differences in adiposity were not detected (14). The authors suggested that hyperinsulinemia precedes the body composition changes recognized in adults, and that South Asian children may be intrinsically more insulin resistant at a given level of adiposity. The measures of adiposity used in that study were ponderal index, waist, and hip circumferences; and these alone may not have been sensitive enough to detect subtle ethnic differences. Our study, however, involved more detailed assessment of body composition and found clear ethnic differences in level of adiposity at a given BMI SDS level but no ethnic difference in the relationship of insulin sensitivity with adiposity, suggesting that the ethnic differences in insulin sensitivity relate to ethnic differences in adiposity.
Our data show that the adult android and gynoid body proportions are well established by 14–16 yr and there are clear ethnic differences, with South Asian adolescents having larger waist circumferences and waist-thigh ratios, with more truncal fat than the white European adolescents. South Asian adolescents also have higher levels of body fat, compared with white European adolescents, across the range of BMI.
This data confirms that the increased adiposity and central fat distribution seen in South Asian adults is also seen in South Asian adolescents. This implies that interventions aimed at primary prevention of the metabolic and cardiovascular consequences of adiposity in high-risk populations should be targeted at children. The ethnic and sex differences described emphasize the importance of interpreting data on body composition and insulin sensitivity in childhood in the context of sex and ethnicity and of interpreting BMI in the context of body composition. Further studies are needed to investigate how early these sex and ethnic differences are apparent. The increased adiposity seen in South Asian adolescents within the normal range of BMI implies that the lowering of the action-point BMI for obesity-related interventions recommended in Asian adults may also be relevant during adolescence, and that we may need to define overweight and obesity at lower BMI values in different ethnicities in childhood (39, 40).
This study shows that South Asian adolescents with BMI SDS well within the normal range are more insulin resistant than white European adolescents. A South Asian child with a BMI SDS within the normal range is more likely to have higher levels of body fat that is more centrally distributed and associated with adverse metabolic risk. This study confirms that the ethnic differences seen in adulthood have their origins in childhood. The ethnic differences in both insulin sensitivity and body composition seen in healthy adolescents may account for the observed sex and ethnic differences in type 2 diabetes prevalence in childhood.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online April 19, 2005
Abbreviations: BMI, Body mass index; 4C, four-compartment; DXA, dual-energy x-ray absorptiometry; HDL, high-density lipoprotein; HOMA, homeostasis model assessment; HOMA %B, HOMA of ß-cell function; HOMA %S, HOMA of insulin sensitivity; SDS, SD score(s).
Received October 11, 2004.
Accepted April 13, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. F. Krebs, J. H. Himes, D. Jacobson, T. A. Nicklas, P. Guilday, and D. Styne Assessment of Child and Adolescent Overweight and Obesity Pediatrics, December 1, 2007; 120(Supplement_4): S193 - S228. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. M. Shields, B. Knight, H. Hopper, A. Hill, R. J. Powell, A. T. Hattersley, and P. M. Clark Measurement of Cord Insulin and Insulin-Related Peptides Suggests That Girls Are More Insulin Resistant Than Boys at Birth Diabetes Care, October 1, 2007; 30(10): 2661 - 2666. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. J Shaw, N. J Crabtree, M. S Kibirige, and J. N Fordham Ethnic and gender differences in body fat in British schoolchildren as measured by DXA Arch. Dis. Child., October 1, 2007; 92(10): 872 - 875. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ramachandran, C. Snehalatha, A. Yamuna, N. Murugesan, and K.M. V. Narayan Insulin Resistance and Clustering of Cardiometabolic Risk Factors in Urban Teenagers in Southern India Diabetes Care, July 1, 2007; 30(7): 1828 - 1833. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. S. Ford, C. Li, G. Imperatore, and S. Cook 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, December 1, 2006; 29(12): 2605 - 2611. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. L. Minton, K. R. Owen, C. J. Ricketts, N. Crabtree, G. Shaikh, S. Ehtisham, J. R. Porter, C. Carey, D. Hodge, R. Paisey, et al. Syndromic Obesity and Diabetes: Changes in Body Composition with Age and Mutation Analysis of ALMS1 in 12 United Kingdom Kindreds with Alstrom Syndrome J. Clin. Endocrinol. Metab., August 1, 2006; 91(8): 3110 - 3116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Wilkin Testing the Accelerator Hypothesis: Body Size, {beta}-Cell Function, and Age at Onset of Type 1 (Autoimmune) Diabetes: Response to Dabelea et al. Diabetes Care, June 1, 2006; 29(6): 1462 - 1463. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
