PAPER Insulin resistance and abdominal adiposity in young ... - Nature

5 downloads 28 Views 108KB Size Report
men with documented malnutrition during the first year of life. NG Boulé1, A Tremblay1*, J Gonzalez-Barranco3, CA Aguilar-Salinas3, JC Lopez-Alvarenga3,.
International Journal of Obesity (2003) 27, 598–604 & 2003 Nature Publishing Group All rights reserved 0307-0565/03 $25.00 www.nature.com/ijo

PAPER Insulin resistance and abdominal adiposity in young men with documented malnutrition during the first year of life NG Boule´1, A Tremblay1*, J Gonzalez-Barranco3, CA Aguilar-Salinas3, JC Lopez-Alvarenga3, JP Despre´s2, C Bouchard4, FJ Gomez-Perez3, L Castillo-Martinez3 and JM Rios-Torres3 1 Division of Kinesiology, Laval University, Ste-Foy, Que´bec, Canada; 2Department of Food Sciences and Nutrition, Laval University, Ste-Foy, Que´bec, Canada; 3National Institute of Medical Sciences and Nutrition Salvador Zubiran, Mexico City, Mexico; and 4Pennington Biomedical Research Center, Baton Rouge, USA

OBJECTIVE: The main objective of the study was to examine the effect of early life malnutrition on the relation between insulin sensitivity and abdominal adiposity in adulthood. It was hypothesised that participants with early life malnutrition would display a more pronounced deterioration of insulin sensitivity in association with a gain in abdominal fat. DESIGN: As a first attempt to investigate this issue, we studied the effect of body fat gains in a cross-sectional context. SUBJECTS: A total of 26 young adult men with evidence of malnutrition during the first year of life and 27 control subjects were recruited for this study. Malnutrition status was determined from medical files of paediatric hospitals in the Mexico City metropolitan area. MEASUREMENTS: Insulin sensitivity was measured by hyperinsulinaemic euglycaemic clamp, and body composition was measured by anthropometrics, bioelectrical impedance and computed tomography. RESULTS: There was a negative correlation between total abdominal adipose tissue area and insulin sensitivity in the previously malnourished and control groups (r2 ¼ 0.65 and 0.35, Po0.01, respectively). When matched for low amounts of abdominal fat (114 cm2), participants with and without early life malnutrition had similar insulin sensitivity (9.03 vs 8.88 mg kg1 min1). However, when matched for high amounts of abdominal fat (310 cm2) participants who were malnourished during the first year of life had lower insulin sensitivity (4.74 vs 6.85 mg kg1 min1, Po0.05). CONCLUSION: Higher levels of abdominal adipose tissue are more detrimental to insulin sensitivity in previously malnourished individuals. International Journal of Obesity (2003) 27, 598–604. doi:10.1038/sj.ijo.0802288 Keywords: body mass index; abdominal fat; malnutrition; birth weight; insulin resistance

Introduction It is generally accepted that excess body fat, especially when stored in the abdominal region, is a major determinant of insulin resistance. However, individuals with similar amounts of fat mass may have very different levels of insulin resistance. In recent years, there has been growing interest on the potential role of early life nutrition status and/or birth weight on obesity, insulin resistance and type II diabetes.1–6

*Correspondence: Dr A Tremblay, Division of Kinesiology, PEPS, Laval University, Ste-Foy, Que´bec, Canada G1K 7P4. E-mail: [email protected] Received 5 July 2002; revised 15 November 2002; accepted 9 December 2002

Impaired glycaemic control in adulthood may be mediated by impaired insulin action or impaired insulin secretion. Evidence is available to support defects in both these mechanisms in adults who had low birth weights.3,4,7–10 Although impaired insulin action (insulin resistance) may have little deleterious effects in a context where energy intake is low or adequate, it may become a handicap in an obesegenic environment, which offers an increased amount of energy-dense foods and a reduced need for physical activity.11 This theory could partially explain the high prevalence of diabetes in populations migrating to societies with increased availability of foods and decreased energy expenditure.12,13 While previous studies have suggested that insulin resistance may develop in adults with low birth weight, to

Insulin resistance and adiposity in young men NG Boule et al

599 our knowledge there are no studies that have measured insulin sensitivity with the hyperinsulinaemic euglycaemic clamp (a gold standard) in subjects with evidence of malnutrition during the first year of life. Furthermore, since body fat and especially abdominal fat are strongly associated with insulin resistance,14 it is important to take this phenotype into account when considering the impact of early malnutrition on insulin sensitivity. The main objective of the study was to examine the effect of early life malnutrition on the relation between insulin sensitivity and abdominal adiposity in adulthood. It was hypothesised that a positive energy balance leading to fat gain, particularly in the abdominal area, would be more detrimental to insulin sensitivity in young men with early life malnutrition.

Methodology Subjects In total, 26 young male adults with a history of malnutrition during the first year of life (MALNUTR) and 27 young male adults having never-experienced malnutrition (CONTROL) were recruited for this study. The history of malnutrition was documented by the access to medical files through the assistance of paediatric hospitals located in the Mexico City metropolitan area. Malnutrition was assessed using the diagnostic criteria and classification proposed by Gomez et al,15 which takes into account the weight of the infant and his ‘theoretical weight’. This ‘theoretical weight’ is based on a weight for age relation previously validated among normal Mexican children. All MALNUTR subjects maintained a body weight below 90% of their ‘theoretical weight’ during the first year of life. Potential MALNUTR or CONTROL participants were excluded if their files revealed the existence of pathologies susceptible of interfering with the outcome of the study such as diabetes mellitus. The CONTROL group was identified through advertisement, and none of the participants in this group maintained a body weight below 90% of their ‘theoretical weight’ during the first year of life. Each subject gave written consent to participate in the protocol that received the approval of the Human Studies Committee of the National Institute of Medical Sciences and Nutrition Salvador Zubiran.

Measurements Birth weight and gestational age were obtained from the medical files of participants and were confirmed by interview with the participant or his parents. Body weight, height, skinfold thickness (triceps, biceps, subscapular, suprailiac), as well as waist and hip circumferences were measured in adulthood. The amount of fat mass and fat-free mass was estimated by bioelectrical impedance. Visceral, subcutaneous and total abdominal fat were measured by computed tomography scan. Participants were examined in the supine position and the abdominal scan was obtained between the

fourth and fifth lumbar vertebrae. The attenuation interval used in the quantification of the adipose tissue areas was from 190 to 30 Hounsfield units. The abdominal visceral fat area was defined by drawing a line within the muscle wall surrounding the abdominal cavity. A euglycaemic hyperinsulinaemic clamp was performed the morning after an overnight fast. An intravenous polyethylene catheter was inserted into the antecubital vein for the infusion of glucose and insulin and a second polyethylene catheter was inserted retrogradely on the dorsum of the hand kept at 651C. Samples for basal blood glucose concentrations were obtained at times 30, 20, 10 and 0, and measured by a YSI 2700 analyzer (Yellow Springs Instruments Co. Inc., Yellow Springs, OH, USA) with an analysis time of 60 s. After insulin infusion had begun, samples for blood glucose concentrations were obtained every 5 min and samples for plasma insulin concentrations were obtained every 10 min for the 120 min of the clamp. Plasma insulin was measured by a Microparticle Enzyme Immunoassay with the IMxs Insulin assay (Abbott Laboratories, Diagnostics Division, Dainabot, Tokyo, Japan). The insulin infusate was composed of Human zinc insulin, DNA recombinant origin (Humulin R, Eli Lilly Co., IN, USA) diluted to a concentration of 300 mU ml1. The 10 min priming insulin infusion was followed by a constant infusion at the rate of 40 mU m2 min1 for 110 min using a Harvard multiple syringe infusion pump (Harvard Apparatus Inc., MA, USA). This procedure raised insulin concentrations to approximately 100 mU ml1, a level that was maintained throughout the test. Glucose infusion was initiated 4 min after the insulin infusion had begun (20% dextrose in water). The glucose infusion was adjusted every 5 min to maintain plasma glucose within 10% of its baseline value. Insulin sensitivity is expressed as the whole body glucose infusion rate during the steady-state glucose infusion rate during the last 30 min, according to the following equation: insulin sensitivity ¼ glucose infusion rate  subject body weight1  30 min1. The glucose infusion rate required to maintain euglycaemia corresponds to the sum of the decrease in glucose production and the increase in glucose disposal caused by insulin and can thereby serve as an index of the overall effect of insulin on total-body glucose metabolism. A previous study on the dose–response characteristics of the effects of insulin on glucose production and disposal has suggested that glucose production is totally suppressed at plasma insulin concentration and glucose infusion rates similar to the ones reported in our study.16 The metabolic clearance rate of insulin was calculated as follows: metabolic clearance rate of insulin ¼ insulin infusion rate (40 mU  (m2 min)1  (steady-state (30–120) plasma insulin concentration basal plasma insulin concentration)1.

Statistics A one-way ANOVA was used to compare insulin sensitivity or body composition between the MALNUTR and CONTROL International Journal of Obesity

Insulin resistance and adiposity in young men NG Boule et al

600 groups. When appropriate, covariables or the interaction between a covariable and groups were also entered in the general linear model. In a secondary analysis, participants were matched for high and low amounts of abdominal adipose tissue and a 2  2 factorial ANOVA was performed (MALNUTR vs CONTROL  low vs high abdominal adiposity). The Levene test was performed to determine unequal variances. Insulin sensitivity had a skewed distribution and was normalised by logarithmic transformation. Insulin sensitivity is considered to be the reciprocal of the definition of insulin resistance. Stepwise multiple regression analysis was performed to determine which body composition variable best predicted insulin resistance. Values are expressed as means (standard deviations). The analyses were conducted with JMP software 3.2.2 from the SAS Institute Inc. (Cary, NC, USA).

Results The main characteristics of the MALNUTR and CONTROL participants are described in Table 1. Although gestational age was similar in both groups, birth weight tended to be lower in the MALNUTR group (P ¼ 0.07). The MALNUTR group was also slightly younger than the CONTROL (22.0 vs 26.5 y, Po0.01). Furthermore, all participants were Mexican and although the CONTROL group had a higher level of education (Po0.01, results not shown) there was no statistical difference in income level between the two groups. In adulthood, the previous malnourished participants were smaller, as indicated by a significantly lower body weight, height, body mass index (BMI), and waist and hip circumference. Computed tomography and bioelectric impedance measurements revealed lower amounts of abdominal fat and percent body fat in the MALNUTR group

Table 1

Participant characteristics and body composition

Birth weight (kg) Gestational age (week) Age (y) Weight (kg) Height (cm) BMI (kg m2) Sum of skinfolds (mm) Waist circumference (cm) Hip circumference (cm) Visceral abdominal tissue (cm2) Subcutaneous abdominal tissue (cm2) Body fat (%)

N

MALNUTR

CONTROL

P

26/23 26/22 26/27 26/27 26/27 26/27 25/26 25/26 25/26 26/24

2.94 (0.66) 39.7(1.6) 22.0 (3.6) 59.8 (9.6) 165.6 (7.0) 21.8 (2.8) 47.5 (17.7) 78.9 (9.7) 91.1 (7.1) 46.2 (40.0)

3.25 (0.52) 39.2 (1.1) 26.5 (2.1) 71.0 (9.3) 169.8 (5.7) 24.6 (2.6) 57.8 (15.1) 85.6 (6.5) 97.9 (6.9) 73.5 (36.4)

0.07 0.19 o0.01 o0.01 0.02 o0.01 0.03 o0.01 o0.01 0.02

26/24

105.9 (76.1)

152.1 (72.4)

26/27

15.1 (6.9)

20.2 (5.3)

0.03 o0.01

Values expressed as Mean (s.d.), n ¼ number of participants in MALNUTR/ CONTROL. MALNUTR: group with a history of early life malnutrition, CONTROL: group without a history of early life malnutrition.

International Journal of Obesity

compared to the CONTROL group. The variance in the body composition measures was higher in the MALNUTR group, but the differences were not significant according to the Levene test. Insulin sensitivity as estimated by glucose infusion during the last 30 min of the euglycaemic hyperinsulinaemic clamp was not significantly different between the two groups, see Table 2. Fasting insulin, fasting glucose and the metabolic clearance rate of insulin were not significantly different between the MALNUTR and CONTROL groups. After correcting for birth weight, age and percent body fat, insulin sensitivity was lower in the MALNUTR group but the difference was not statistically significant (see Table 2). However, after the same adjustments, fasting glucose and fasting insulin levels were higher in the MALNUTR group (Po0.05). Within the entire sample as well as within the CONTROL and MALNUTR groups separately, birth weight was not associated with insulin sensitivity or abdominal fat surface areas (all r2o0.03, all P40.5). In the CONTROL group, the strongest predictor of insulin sensitivity was visceral fat (r 2 ¼ 0.35, Po0.01). In this group, subcutaneous abdominal fat was also associated with insulin sensitivity (r 2 ¼ 0.27, Po0.01). However, in the MALNUTR group, subcutaneous abdominal fat was the strongest predictor of insulin sensitivity (r 2 ¼ 0.66, Po0.01), although visceral fat was also a strong predictor (r 2 ¼ 0.50, Po0.01). Fat mass was associated with insulin sensitivity in the MALNUTR group (r 2 ¼ 0.52, Po0.01) and in the CONTROL group (r 2 ¼ 0.17, P ¼ 0.03). Stepwise regressions were performed to determine which indicators of body composition (BMI, fat mass, percent body fat, subcutaneous abdominal fat, visceral abdominal fat and total abdominal fat) best predicted insulin sensitivity. Within the entire sample, total abdominal fat explained 49% of the variance in insulin sensitivity (Po0.01). The next variable to be entered in the model was BMI, which explained an additional 2% of the variance in insulin sensitivity, however it was not statistically significant (P ¼ 0.11). The correlations between BMI or total abdominal fat and glucose infusion rate during the clamp are illustrated separately for the CONTROL and MALNUTR groups in Figure 1. BMI was not significantly associated with insulin sensitivity in CONTROL subjects (P ¼ 0.12), while in the MALNUTR subjects, BMI accounted for 41% of the variance in insulin sensitivity. Total abdominal adipose tissue area was associated with insulin sensitivity in both groups, accounting for 35 and 65% of the variance in insulin sensitivity in the CONTROL and MALNUTR groups, respectively. An ANCOVA was performed to compare insulin sensitivity between the MALNUTR and CONTROL groups, while considering total abdominal fat as a covariable (since it was the most powerful independent predictor of insulin sensitivity in both the groups combined). There was a significant

Insulin resistance and adiposity in young men NG Boule et al

601 Table 2

Results from the hyperinsulinaemic euglycaemic clamp

Insulin sensitivity (mg kg1 min1) Insulin sensitivity (log mg kg1 min1) Insulin sensitivity (log mg FFM1 min1) Mean plasma insulin 90–120 min (pmol/l) MCRI (ml kg1 min1) Fasting insulin (pmol l1) Fasting C-peptide (nmol l1) Fasting glucose (mmol l1)

N

MALNUTR

CONTROL

P

P*

26/26 26/26 26/26 26/27 26/27 26/27 26/25 26/27

8.5 (3.6) 0.89 (0.19) 0.97 (0.16) 616 (151) 0.57 (0.12) 48.6 (33.6) 0.20 (0.10) 4.5 (0.3)

7.9 (1.9) 0.89 (0.10) 0.98 (0.09) 606 (76) 0.56 (0.08) 36.6 (23.4) 0.18 (0.16) 4.4 (0.3)

0.43 0.88 0.62 0.77 0.55 0.15 0.48 0.35

0.62 0.21 0.23 0.19 0.83 0.01 0.25 0.04

Values expressed as mean (s.d.), n ¼ number of participants in MALNUTR/CONTROL. MALNUTR: group with a history of early life malnutrition, CONTROL: group without a history of early life malnutrition. P*: P values adjusted for age, birth weight and percent body fat as covariables. Insulin sensitivity was calculated as whole body glucose infusion from 90 to 120 min of the euglycaemic hyperinsulinaemic clamp. FFM ¼ kg of fat-free mass. MCRI ¼ metabolic clearance rate of insulin.

Figure 1 Correlations between insulin sensitivity (log transformed) as measured by glucose infusion during the euglycaemic hyperinsulinaemic clamp and body mass index (BMI) or abdominal fat (Abd Fat) in MALNUTR and CONTROL. **Po0.01. MALNUTR: group with a history of early life malnutrition, CONTROL: group without a history of early life malnutrition

interaction between the effects of the groups (MALNUTR and CONTROL) and total abdominal fat on insulin sensitivity (P ¼ 0.01), indicating that the slopes of the regressions between insulin sensitivity and total abdominal fat were significantly different between groups. This significant interaction remained after adjusting insulin sensitivity and

total abdominal fat for birth weight and age (P ¼ 0.02). Similar results were found when visceral abdominal fat, subcutaneous abdominal fat, fat mass or body mass index were considered separately as covariables instead of total abdominal fat. A significant interaction between total adipose tissue area and early life malnutrition status was International Journal of Obesity

Insulin resistance and adiposity in young men NG Boule et al

602

Figure 2 Insulin sensitivity (log transformed mean, s.d.) as measured by glucose infusion during the euglycaemic hyperinsulinaemic clamp in CONTROL and MALNUTR groups matched for high and low amounts of abdominal fat (Abd Fat). Factorial ANOVA (2 groups  2 Abd Fat levels): effect of groups (P ¼ 0.08), effect of Abd Fat level (Po0.0001), interaction group  Abd Fat level (P ¼ 0.05), n ¼ 28 (seven per group). MALNUTR: group with a history of early life malnutrition, CONTROL: group without a history of early life malnutrition.

also observed when insulin sensitivity was not logarithmically transformed (P ¼ 0.04). The impact of abdominal adipose tissue on insulin sensitivity may be better illustrated by matching MALNUTR and CONTROL subjects according to high and low amounts of abdominal adipose tissue area. A total of 28 participants were matched. The mean (s.d.) for abdominal adipose tissue area were 114.1 (60.8) cm2 and 113.6 (59.5) cm2 for the low abdominal adipose tissue CONTROL and MALNUTR groups, respectively. Abdominal adipose tissue was almost three times more elevated in the high adipose tissue CONTROL and MALNUTR groups: 306.7 (53.5) cm2 and 314.2 (55.2) cm2, respectively. The 2  2 factorial ANOVA (group vs abdominal fat level) indicated that there was a significant interaction between the two factors (P ¼ 0.05). As illustrated in Figure 2, insulin sensitivity was similar in CONTROL and MALNUTR when matched for a low abdominal adipose tissue area. However, in the groups matched for high amounts of abdominal adipose tissue the MALNUTR subjects were less insulin sensitive.

Discussion The results of this study demonstrated that insulin sensitivity is not altered by previous malnutrition in individuals maintaining a low level of body fatness and abdominal adiposity. On the other hand, the subjects of the MALNUTR group displaying high levels of abdominal fat displayed a significantly lower insulin sensitivity level than their control subgroup matched for abdominal fat. Despite its crosssectional nature, this study tends to suggest that when one has a personal history of early life malnutrition, he/she would be more vulnerable to an environment promoting a International Journal of Obesity

positive energy balance and fat gain. To some extent, these results are concordant with those reported by Ravussin et al,17 who demonstrated that Pima Indians maintaining a traditional lifestyle preventing obesity are characterised by a healthy metabolic profile. However, in Pima Indians, having experienced an environment promoting obesity in Arizona, the prevalence of insulin resistance and type II diabetes has increased to a level that is one of the highest in the world.18 This finding is important and justifies the relevance to investigate this issue in a longitudinal context. The significant statistical interaction observed between early life malnutrition status and the amount abdominal fat in adulthood implies that the direct comparison of insulin sensitivity between groups having different early life nutritional status is dependent on the level of abdominal fat in the studied population. In this regard, even if the overall glucose infusion rate during the hyperinsulinaemic euglycaemic clamp was similar in both groups in the present study, it would be false to conclude that insulin sensitivity is not affected by early life malnutrition. Previous studies have suggested an interaction between low birth weight and obesity on insulin resistance. A study in Indian children had suggested that the inverse relation between birth weight and insulin resistance, as measured by HOMA, was strongest in the children who were heavier at age 8.19 Phillips et al4 have studied insulin sensitivity in older adults (mean age ¼ 52 y) with the short insulin tolerance test and demonstrated that individuals with low birth weight and high BMI had the highest level of insulin resistance. In addition to extending these previous observations to young adults with a history of malnutrition during the first year of life, the major strength of the present investigation was the use of direct measures for the main outcome variables: insulin sensitivity (hyperinsulinaemic euglycaemic clamp) and abdominal obesity (computed tomography). The results suggest that the amount of abdominal adipose tissue accounted for 65% of the variance in insulin sensitivity in previously malnourished adults. Interestingly, when participants were matched for a low abdominal adipose tissue cross-sectional area (mean ¼ 114 cm2), the previously malnourished had similar insulin sensitivity compared to controls. However, when matched for higher amounts of abdominal fat (mean ¼ 310 cm2), the previously malnourished subjects had lower insulin sensitivity compared to the controls. Although this cross-sectional study cannot support a causeto-effect relation, it does suggest that similar gains in abdominal fat are more detrimental to insulin sensitivity in people with a history of early life malnutrition. Unlike previous studies on the effect of low birth weight, in the present study early life malnutrition alone was not associated with decreased insulin sensitivity. However, the finding that larger decreases in insulin sensitivity in subjects with previous malnutrition occur with increasing body fat has important practical implications for the prevention of insulin resistance.

Insulin resistance and adiposity in young men NG Boule et al

603 Low birth weight and a history of early life malnutrition are not identical selection criteria.20 Small birth weight may reflect the intrauterine environment (including nutrition), maternal factors and the fetal genotype whereas early life malnutrition specifically referred to the nutritional status during the first year of life. Although birth weight tended to be different in the two groups in our study, birth weight was not associated with insulin sensitivity or body composition. Furthermore, taking into account the differences in birth weight between the two groups did not alter principal findings of our study. These findings do not mean that birth weight does not have a role in the aetiology of insulin resistance as determined by hyperinsulinaemic euglycaemic clamp. Jaquet et al9 have demonstrated that subjects born with intrauterine growth retardation had a higher percent body fat and had lower peripheral glucose uptake during a euglycaemic hyperinsulinaemic clamp. However, in this study the mean birth weight in subjects born with intrauterine growth retardation was 2.4 kg. In contrast, only four of the 26 subjects in our study who were affected by early life malnutrition had a birth weight below 2.5 kg. This may explain why birth weight was not associated with insulin sensitivity in our analyses. A third group of participants that had a similar birth weight as to the malnourished group, but was nutritionally rehabilitated after birth, would have provided an interesting comparison. The significant interaction between malnutrition status and abdominal fat found in our study has important implications. Abdominal fat explained approximately twice as much of the variance in insulin sensitivity in the previously malnourished group compared to controls. It is important to note that the reduced insulin sensitivity in these subjects is only evident at higher levels of abdominal fat. Epidemiological data about the prevalence of diabetes in Mexico reinforce the idea that previous nutritional restriction decreases the ability to perform adequate metabolic regulation in a context promoting a positive energy balance. In the desertic area of San Luis Potosi, the prevalence of diabetes is less than 1% in the rural areas whereas it exceeded 20% in the city of this Mexican state.21 An earlier animal study by Holness and Sugden22 was consistent with our findings. In this study, rats exposed to protein restriction during the first 20 weeks of life had lower insulin-stimulated glucose disposal compared to the control group after both groups had been transferred to a high-fat diet for 8 weeks.22 The mechanisms by which early life malnutrition reduces insulin sensitivity are not well understood. Animal studies have suggested that low protein intake during fetal nutrition and lactation can have long-term effects on glucose transport and the insulin signalling pathway in the adipocytes.23 It has also been suggested that there is a sensitive period for the development of adipose tissue in humans, which extends approximately between the third trimesters of gestation to the first year of life.24 Energy restriction during this critical period could have a long-term impact on the number and size of adipocytes. The size of the

adult abdominal adipocytes is also independently associated with insulin resistance.25 In summary, higher levels of body fat and especially abdominal fat seem to be more detrimental to insulin sensitivity in previously malnourished individuals. In recent years, Mexico and other developing countries have experienced rapid increases in the prevalence of type II diabetes. The findings from the present study support the hypothesis that the negative consequences of early life malnutrition on insulin resistance may only become apparent in a context that promotes an increase in body fat.

Acknowledgements This study was supported by a grant from the Nestle´ Foundation. Mr Boule´ is supported by NSERC of Canada postgraduate scholarship. We acknowledge the contribution of the following institutions in Mexico City for the recruitment of participants: Instituto Nacional de Pediatra, Hospital General, Centro Infantil de Rehabilitacion Nutricional, and Hospital Infantil Frederico Gomez.

References 1 Garnett SP, Cowell CT, Baur LA, Fay RA, Lee J, Coakley J, Peat JK, Boulton TJ. Abdominal fat and birth size in healthy prepubertal children. Int J Obes Relat Metab Disord 2001; 25: 1667–1673. 2 Ravelli GP, Stein ZA, Susser MW. Obesity in young men after famine exposure in utero and early infancy. N Engl J Med 1976; 295: 349–353. 3 Flanagan DE, Moore VM, Godsland IF, Cockington RA, Robinson JS, Phillips DI. Fetal growth and the physiological control of glucose tolerance in adults: a minimal model analysis. Am J Physiol 2000; 278: E700–E706. 4 Phillips DI, Barker DJ, Hales CN, Hirst S, Osmond C. Thinness at birth and insulin resistance in adult life. Diabetologia 1994; 37: 150–154. 5 Hales CN, Barker DJ, Clark PM, Cox LJ, Fall C, Osmond C, Winter PD. Fetal and infant growth and impaired glucose tolerance at age 64. BMJ 1991; 303: 1019–1022. 6 Barker DJ, Hales CN, Fall CH, Osmond C, Phipps K, Clark PM. Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 1993; 36: 62–67. 7 Robinson S, Walton RJ, Clark PM, Barker DJ, Hales CN, Osmond C. The relation of fetal growth to plasma glucose in young men. Diabetologia 1992; 35: 444–446. 8 Cook JT, Levy JC, Page RC, Shaw JA, Hattersley AT, Turner RC. Association of low birth weight with beta cell function in the adult first degree relatives of non-insulin dependent diabetic subjects. BMJ 1993; 306: 302–306. 9 Jaquet D, Gaboriau A, Czernichow P, Levy-Marchal C. Insulin resistance early in adulthood in subjects born with intrauterine growth retardation. J Clin Endocrinol Metab 2000; 85: 1401–1406. 10 Leger J, Levy-Marchal C, Bloch J, Pinet A, Chevenne D, Porquet D, Collin D, Czernichow P. Reduced final height and indications for insulin resistance in 20 year olds born small for gestational age: regional cohort study. BMJ 1997; 315: 341–347. 11 Hales CN, Barker DJ. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia 1992; 35: 595–601.

International Journal of Obesity

Insulin resistance and adiposity in young men NG Boule et al

604 12 Zimmet PZ, McCarty DJ, de Courten MP. The global epidemiology of non-insulin-dependent diabetes mellitus and the metabolic syndrome. J Diabetes Complications 1997; 11: 60–68. 13 Fujimoto WY. Overview of non-insulin-dependent diabetes mellitus (NIDDM) in different population groups. Diabet Med 1996; 13: S7–S10. 14 Despre´s JP. Abdominal obesity as important component of insulin-resistance syndrome. Nutrition 1993; 9: 452–459. 15 Gomez F, Ramos GR, Frenk S, Cravioto MJ, Chavez R, Vazquez J. Mortality in second and third degree malnutrition. J Trop Pediatr 1956; 2: 77–83. 16 Rizza RA, Mandarino LJ, Gerich JE. Dose–response characteristics for effects of insulin on production and utilization of glucose in man. Am J Physiol 1981; 240: E630–E639. 17 Ravussin E, Valencia ME, Esparza J, Bennett PH, Schulz LO. Effects of a traditional lifestyle on obesity in Pima Indians. Diabetes Care 1994; 17: 1067–1074. 18 Bennett PH. Type 2 diabetes among the Pima Indians of Arizona: an epidemic attributable to environmental change? Nutr Rev 1999; 57: S51–S54. 19 Bavdekar A, Yajnik CS, Fall CH, Bapat S, Pandit AN, Deshpande V, Bhave S, Kellingray SD, Joglekar C. Insulin resistance syndrome in

International Journal of Obesity

20

21

22

23

24 25

8-year-old Indian children: small at birth, big at 8 years, or both? Diabetes 1999; 48: 2422–2429. Godfrey KM, Barker DJ. Maternal nutrition in relation to fetal and placental growth. Eur J Obstet Gynecol Reprod Biol 1995; 61: 15–22. Quibrera IR, Herna´ndez RHG, Aradillas GT, Gonza´lez RC, CallesEscando´n J. Prevalencias de diabetes, intolerancia a la glucosa, hiperlipemia y factores de riesgo en funcio´n de nivel socioecono´mico. Rev Invest Clin 1994; 46: 25–36. Holness MJ, Sugden MC. Antecedent protein restriction exacerbates development of impaired insulin action after high-fat feeding. Am J Physiol 1999; 276: E85–E93. Ozanne SE, Nave BT, Wang CL, Shepherd PR, Prins J, Smith GD. Poor fetal nutrition causes long-term changes in expression of insulin signaling components in adipocytes. Am J Physiol 1997; 273: E46–E51. Brook CG. Evidence for a sensitive period in adipose-cell replication in man. Lancet 1972; 2: 624–627. Weyer C, Foley JE, Bogardus C, Tataranni PA, Pratley RE. Enlarged subcutaneous abdominal adipocyte size, but not obesity itself, predicts type II diabetes independent of insulin resistance. Diabetologia 2000; 43: 1498–1506.