Body fat distribution and cardiovascular risk in normal weight ... - Nature

International Journal of Obesity (2000) 24, 751±757 ß 2000 Macmillan Publishers Ltd All rights reserved 0307±0565/00 $15.00 www.nature.com/ijo

Body fat distribution and cardiovascular risk in normal weight women. Associations with insulin resistance, lipids and plasma leptin ES Tai1*, TN Lau1, SC Ho1, ACK Fok1 and CE Tan1 1

Department of Endocrinology, Singapore General Hospital, Outram Road, Singapore 169608

OBJECTIVE: To systematically examine the correlations between insulin resistance, plasma leptin concentration, obesity and the distribution of fat assessed by anthropometry and magnetic resonance imaging in Asian women. DESIGN: A cross sectional study of non-diabetic, normal weight women. SUBJECTS: Twenty-one healthy women aged 38.8 y (s.d. 11.7) and BMI 22.6 kg=m2 (s.d. 2.3). MEASUREMENTS: Intraperitoneal, retroperitoneal and subcutaneous abdominal fat volume was assessed by magnetic resonance imaging. Anthropometric data were collected. Total fat mass was assessed by bioelectric impedance analysis. Fasting serum lipids, insulin and plasma leptin were assayed. RESULTS: Generalized obesity correlated with subcutaneous abdominal fat mass (r 0.83, P < 0.001), but not with intra-abdominal fat mass. Both intraperitoneal fat mass and retroperitoneal fat mass increased with age (r 0.58, P 0.005 and r 0.612, P 0.003, respectively). Abdominal subcutaneous fat mass was the most important determinant of insulin resistance and plasma leptin. Of the serum lipids, only fasting triglyceride correlated signi®cantly with the waist-to-hip ratio. CONCLUSIONS: It is possible that the large size of the subcutaneous depot compared to the intra-abdominal depot overwhelms any metabolic differences between adipose tissue from these two sites, resulting in the stronger correlation between insulin resistance and subcutaneous abdominal fat mass rather than intra-abdominal fat mass. On the other hand, the distribution of fat between subcutaneous fat depots may be important in the metabolic syndrome given the correlation of fasting triglyceride with waist to hip ratio but not with abdominal fat. However, the study population was small, younger and leaner compared to previous studies and we may not be able to generalize these results to all segments of the population. We con®rm that subcutaneous fat mass is the major determinant of plasma leptin. International Journal of Obesity (2000) 24, 751±757 Keywords: obesity; body fat distribution; magnetic resonance imaging; insulin resistance; leptin; female

Introduction Upper body obesity is associated with increased cardiovascular disease in men1 ± 4 and in women.4,5 Its association with the metabolic syndrome, of which insulin resistance is a major feature, may mediate this increased risk. The use of anthropometric measurements (skinfold thickness, waist circumference and waist-to-hip ratio (WHR) has demonstrated that body fat distribution has important effects on glucose and insulin metabolism.6 ± 8 However, it is unclear what these measurements really represent. For instance, anthropometric measures are not able to delineate intra-abdominal fat from subcutaneous fat. Advances in radiological techniques (computed tomography and magnetic resonance (MR) imaging) have, by allowing us to assess the quantity of intra-

*Correspondence: ES Tai, Department of Endocrinology, Singapore General Hospital, Outram Road, Singapore 169608. E-mail: eshyong@paci®c.net.sg Received 16 August 1999; revised 16 November 1999; accepted 5 January 2000

abdominal and subcutaneous fat, helped to clarify some of these issues. MR imaging has an advantage over computed tomography as it does not involve exposure to ionizing radiation, which is important in women of child-bearing age, and allows us to carry out multiple scans in the same individual. It has been suggested that intra-abdominal adipose tissue may play an important role in the pathogenesis of the metabolic syndrome.9 ± 13 Most of the studies conducted to date have utilized the area of adipose tissue in each compartment in a single slice to quantify the amount of intra-abdominal fat. Compared with techniques that employ multiple contiguous slices through the body for assessment of fat volume, this single slice strategy has been shown to increase the measurement uncertainty for fat mass in each body compartment.14 Using MR imaging employing multiple contiguous slices, Abate et al described a reliable technique, which they validated against dissection of human cadavers.15 To our knowledge, these volumetric techniques have only been applied to men and adolescent girls but not to adult women. Men and women have different patterns of body fat distribution. Men have a greater tendency to

Body fat distribution in women ES Tai et al

752

accumulate truncal and intra-abdominal fat. Given these gender differences, it is important to investigate the metabolic correlates of abdominal adiposity in women. The availability of these radiological techniques also allows us to investigate the correlation between body fat distribution and another hormone associated with obesity, leptin. Obesity in humans is associated with increased levels of plasma leptin16,17 and this is thought to represent a state of leptin resistance. It appears that plasma leptin is associated with peripheral rather than central obesity when measured by anthropometry in our population.18 This study allows us to con®rm these ®ndings and determine the relationship between visceral fat and plasma leptin concentration. The objective of the study was to systematically examine the correlations between insulin resistance, plasma leptin concentration, obesity and the distribution of fat assessed by anthropometry and MR imaging in Asian women.

Methods We conducted a cross-sectional study involving 21 women recruited among friends and relatives of hospital staff. All subjects were non-diabetic with fasting glucose below 6.0 mmol=l. Individuals with major intercurrent illness in the 3 months prior to examination, renal failure and liver disease were excluded. No subjects had a signi®cant medical history nor were they taking any medication. Characteristics of the subjects are shown in Table 1. No subject had a family history of diabetes mellitus. All subjects were examined between 8 and 9 a.m. after a 10 h fast. They were advised not to exercise strenuously in the 24 h prior to examination. In conjunction with the requirements for body fat estimation, they were also asked not to consume alcohol, caffeine or excessive amounts of ¯uids in the 24 h leading up to the examination. The investigating physicians performed anthropometric measurements on all subjects. Weight was measured in light clothing and without shoes on an electronic scale (Vogel & Halke, Hamburg, Germany) and height was ascertained using a wall-mounted stadiometer. Waist and hip circumference were measured with a non-elastic tape measure. The waist was de®ned as the narrowest circumference between the costal margin and the iliac crests and the hip as the widest circumference between the waist and the thighs. The WHR and body mass index (BMI) were calculated for each subject. Lean body mass was determined by bioelectric impedance analysis using a SEAC bioimpedance meter (Model SFB3, SEAC, Australia) using the formula described by Lukaski et al.19 Total fat mass International Journal of Obesity

was calculated by subtracting the lean body mass from the body weight. Percentage body fat was calculated from the formula [(weight 7 lean body mass)= weight]100%. In a previous study,18 the coef®cient of variation (CV) for total fat mass determined by repeated measurement in six individuals within 3 days was 4.7%. The formulae for measurement of percentage body fat in Western populations using bioelectric impedance analysis have been shown to apply in Chinese, which make up the majority of our study population.20 Blood for serum insulin was collected in plain tubes. For glucose and leptin, blood was collected in ¯uoride oxalate and sodium EDTA tubes respectively. All specimens were centrifuged within 1 h of collection and lipids and glucose were assayed on fresh specimens. Serum and plasma for insulin and leptin determinations were frozen at 7 70 C prior to assay. Glucose was determined by the glucose oxidase method on a Vitros 700 Chemistry Analyser (Rochester, NY). Total cholesterol (TC) and triglyceride (TG) were measured directly by enzymatic methods using Kodak Ektachem clinical chemistry slides. High density lipoprotein (HDL) cholesterol was separated from low density lipoprotein (LDL) and very low density lipoprotein by precipitation with dextran sulphate and magnesium chloride and then measured by enzymatic methods as for cholesterol measurements. The slides were read on the Vitros 700 Chemistry Analyser. Plasma leptin was determined by radioimmunoassay (linco, St Charles, MO) and insulin was assayed by microparticle enzyme immunoassay (intra-assay coef®cient of variation (CV) 4.1%, interassay CV 2.9%, Abbot Imx, Chicago, IL). Insulin resistance was derived by Homeostasis Model Assessment21 using the formula: insulin resistance (fasting glucosefasting insulin)=22.5. MR imaging was performed with a 1.5 T unit (Siemens Magnetom Vision, Siemens AG, Medizinische Technik, Erlangen, Germany) in all subjects. No speci®c subject preparation was required other than to exclude general contra-indications to MR imaging. Scanning was done using the body coil with the subject supine and the arms placed at the side. All subjects had both T1-weighted and T2-weighted sequences done in the axial plane with imaging from the dome of the diaphragm down to the symphysis pubis. T1-weighted gradient echo images of the abdomen (TR 147 ms, TE 4.8 ms) were obtained, using a 107256 matrix and a 2535 cm ®eld of view. Slice thickness was 8 mm with an inter-slice gap of 2 mm such that consecutive sections were at 1 cm intervals. Seven sections were obtained in each acquisition. The acquisition time for each block was 15 s. Usually ®ve to six blocks were required to cover the entire abdomen. T2-weighted true FISP images (TR 6.32 ms, TE 3 ms) was performed with a 160256 matrix, 2535 cm ®eld of view, slice thickness of 8 mm and inter-slice gap of 2 mm. The images were sent across to a Siemens workstation (Sienet Magic-

Body fat distribution in women ES Tai et al Table 1 Age, obesity, insulin resistance and plasma leptin of subjects n Age (y) Body mass index (kg=m2) Lean body mass (kg)* Fat mass (kg) Percentage body fat (%) Waist circumference (cm)* Hip circumference (cm) Waist ± hip ratio* Intraperitoneal fat mass (kg)* Retroperitoneal fat mass (kg)* Abdominal subcutaneous fat mass (kg)* Systolic blood pressure (mmHg)* Diastolic blood pressure (mmHg) HOMA insulin resistance* Plasma leptin (ng=ml)*

21 38.8 (11.8) 22.6 (2.3) 37.2 (34.4 ± 44.9) 17.5 (4.9) 31.0 (5.5) 70.0 (66.0 ± 88.0) 94.6 (5.7) 0.76 (0.68 ± 0.86) 0.60 (0.15 ± 1.71) 0.50 (0.26 ± 1.31) 2.05 (1.16 ± 4.77) 110.(90 ± 150) 68.(8) 1.42 (0.56 ± 3.07) 15.54 (6.18 ± 36.30)

Figures represent mean (standard deviation) except ones marked*, which are geometric mean (minimum ± maximum).

View 1000, Siemens AG, Medixinische Technik, Erlangen, Germany) where the areas of fat were mapped out using a mouse. All analyses were performed by the same radiologist (TNL). Fat appears as areas of higher signal intensity on both the T1- and T2weighted images and was easily distinguished from adjacent structures. Abdominal body fat was divided into subcutaneous, intraperitoneal and retroperitoneal compartments. Anatomical landmarks such as the ascending and descending colon, aorta, inferior vena cava, pancreas and kidneys were used to demarcate the intra-peritoneal and retro-peritoneal compartments. Whilst we measured the intra-abdominal components of body fat from the diaphragm down to the symphysis pubis, subcutaneous fat was only measured from sections between the level of the con¯uence of splenic and superior mesenteric veins and the iliac crests. Because each contiguous section is 1 cm thick, the volume of fat in each slice (in cm3) is obtained by multiplying the areas measured by a factor of 1. Fat volumes of each compartment were obtained by summating the respective volumes of fat from all the contiguous sections. Assuming the density of fat

to be 0.9196 kg=l,15 the mass of adipose tissue (in kg) was then computed by multiplying the fat volumes of each compartment by 0.0009196. Statistical analysis was performed using SPSS for Windows version 7.5 (SPSS Inc., Chicago, IL). Normally distributed data was represented as arithmetric mean and standard deviation whereas non-parametric data was represented by geometric mean and range (Table 1). Pearson product ± moment correlation and stepwise multiple linear regression were used to assess the association between age, insulin resistance, plasma leptin, serum lipids and various measures of generalised or regional obesity (Tables 2 and 3). Fasting TG, HDL cholesterol, plasma leptin and insulin resistance were loge transformed to reduce skewness.

753

Results Table 2 shows the Pearson correlation coef®cients between the fat mass in the various compartments and the anthropometric data as well as the total body fat mass. It can be seen that the measures of general adiposity (body mass index, total fat mass, percentage body fat) correlated with each other and also with the waist and hip circumferences but not the WHR. With regards the abdominal fat depots studied, all three markers of general adiposity correlated strongly with the abdominal subcutaneous fat mass but not with the intraperitoneal fat mass or retroperitoneal fat mass. The relationship between total fat mass and subcutaneous abdominal fat mass, intraperitoneal fat mass and retroperitoneal fat mass is depicted in Figure 1. Abdominal subcutaneous fat mass accounts for 69% of the variation of total fat mass. Intraperitoneal fat mass showed a tendency to increase with increasing total fat mass but this did not reach statistical signi®cance. No correlation was found between retroperitoneal fat mass and total fat mass. Apart from a

Table 2 Pearson correlation coef®cients between measures of obesity and body fat distribution Body mass index Body mass index Waist Hip Waist ± hip ratio Total fat mass Percentage body fat I=P fat mass R=P fat mass S=C fat mass

Ð 0.86** P < 0.001 0.789** P < 0.001 0.33 P 0.145 0.86** P < 0.001 0.70** P < 0.001 0.42 P 0.059 0.22 P 0.335 0.86** P < 0.001

Waist

Hip

Waist ^ hip ratio Total fat mass Percentage body fat I=P fat mass R=P mass

Ð 0.631** P 0.002 0.67** P 0.001 0.73** P < 0.001 0.53** P 0.014 0.54* P 0.011 0.36 P 0.335 0.81** P < 0.001

Ð 7 0.151 P 0.512 0.843** P < 0.001 0.774** P < 0.001 0.429 P 0.052 0.265 P 0.264 0.662** P 0.001

Ð 0.13 P 0.578 7 0.06 P 0.798 0.27 P 0.241 0.19 P 0.406 0.38 P 0.086

Ð 0.95** P < 0.001 0.41 P 0.068 0.28 P 0.215 0.83** P < 0.001

Ð 0.33 P 0.147 0.26 P 0.252 0.74** P < 0.001

Ð 0.88** Ð P < 0.001 0.58** 0.40 P 0.006 P 0.075

*P < 0.05. **P < 0.01. International Journal of Obesity


754

Table 3 Pearson correlation coef®cients between measure of obesity and age, fasting lipids, insulin resistance and plasma leptin

Body mass index Waist Hip Waist ± hip ratio Total fat mass Percentage body fat I=P fat mass R=P fat mass S=C fat mass

Age

Total cholesterol

Triglyceride

HDL-C

LDL-C

Insulin resistance

Leptin

7 0.184 P 0.425 7 0.124 P 0.592 7 0.42 P 0.857 7 0.151 P 0.513 7 0.149 P 0.519 7 0.086 P 0.712 0.584** P 0.005 0.612** P 0.003 7 0.018 P 0.937

7 0.052 P 0.823 7 0.14 P 0.546 7 0.21 P 0.361 0.015 P 0.947 7 0.103 P 0.658 7 0.088 P 0.703 0.101 P 0.664 0.001 P 0.998 7 0.013 P 0.955

0.201 P 0.382 0.273 P 0.231 7 0.1 P 0.667 0.476* P 0.029 0.143 P 0.537 0.051 P 0.826 0.022 P 0.926 7 0.021 P 0.929 0.113 P 0.627

7 0.31 P 0.168 7 0.393 P 0.078 7 0.169 P 0.465 7 0.371 P 0.097 7 0.315 P 0.164 7 0.214 P 0.352 7 0.057 P 0.808 7 0.058 P 0.804 7 0.219 P 0.34

0.098 P 0.672 0.004 P 0.987 7 0.0785 P 0.714 0.09 P 0.697 0.07 P 0.765 0.05 P 0.83 0.14 P 0.546 0.038 P 0.87 0.108 P 0.641

0.528* P 0.014 0.475* P 0.03 0.667** P 0.001 7 0.024 P 0.918 0.71** P < 0.001 0.688** P 0.001 0.333 P 0.141 0.344 P 0.126 0.572** P 0.007

0.724** P < 0.001 0.502* P 0.02 0.694** P < 0.001 7 0.005 P 0.981 0.661** P 0.001 0.558** P 0.009 0.266 P 0.243 0.086 P 0.711 0.564** P 0.008

* P < 0.05. ** P < 0.01.

Figure 1 The relationship between total body fat mass and A subcutaneous abdominal fat mass, B intraperitoneal fat mass and C retroperitoneal fat mass.

Figure 2 Relationship between age and A intra-abdominal fat mass, B intraperitoneal fat mass and C retroperitoneal fat mass.

correlation between intraperitoneal fat mass and waist circumference, there was no signi®cant correlation between either intraperitoneal fat mass or retroperitoneal fat mass and any other anthropometric measurements. Table 3 shows the correlation coef®cients between age, fasting serum lipids, insulin resistance, plasma leptin and measures of obesity and body fat distribution. There was no correlation seen between age and

any of the anthropometric measures or total fat mass. However, there was a strong positive correlation between intra-abdominal fat mass, intraperitoneal fat mass and retroperitoneal fat mass with age (Figure 2). Both insulin resistance and plasma leptin were highly correlated with generalized obesity (Table 3). However, when it came to abdominal adiposity, only the abdominal subcutaneous fat mass was correlated with either insulin resistance or plasma leptin. None

International Journal of Obesity


of the measures of intra-abdominal fat (intraperitoneal, retroperitoneal or the total of these two parameters) correlated with insulin resistance. With regards the serum lipids, no signi®cant correlation was seen except for a positive correlation between WHR and serum TG.

Discussion The use of MR imaging with multiple contiguous slices through the body (as opposed to a single slice sample or dual energy X-ray absorptiometry) allowed us to delineate the relationship between body fat distribution and insulin resistance more precisely than in previous studies. While this is a strength of our study, there were some limitations. One of these was the small number studied (n 21), which was mandated by the high cost of the imaging and the very labour-intensive nature of the image analysis. However, it can be seen from the subject characteristics (Table 1) that they represent a wide range of obesity, insulin resistance, plasma leptin and serum lipids. In addition, while we appreciate that homeostasis model assessment is not the gold standard for measuring insulin resistance, it has been shown to be a good predictor of the development of diabetes mellitus,22 which is an important problem in our country.23 The various anthropometric measures of obesity and body fat distribution were highly correlated (Table 2). Our data concurred with previous data showing that the waist circumference was a better indicator of both subcutaneous and intraperitoneal fat than the WHR.24 Our study highlights an important difference in patterns of fat distribution between men and women. In men, fat mass in all three compartments is highly correlated with the total fat mass.25 In contrast, our data (Figure 1) shows that, with increasing obesity, women deposit fat in the subcutaneous compartment rather than in either intra-abdominal compartment, as described in Swedish women using similar methodology to ours.26 The subcutaneous abdominal and the gluteofemoral compartments seem to be the preferential areas for fat distribution. We carried out stepwise multiple linear regression analysis with total fat mass as the dependant variable and waist circumference, hip circumference, intraperitoneal fat mass, retroperitoneal fat mass and abdominal subcutaneous fat mass as independent variables. Only hip circumference and abdominal subcutaneous fat mass remained signi®cantly correlated with total fat mass. Together they explained 84.5% of the variation in total fat mass, thus reinforcing the importance of these two fat depots in female obesity. Omental adipocytes are smaller and show greater lipolysis in response to stimulation by catecholamines.27 It has been suggested that the resulting free fatty acid release into the hepatic portal system

results in insulin resistance. These metabolic differences between the adipocytes in the various fat depots have been used to explain previous ®ndings that intraabdominal fat mass was a major determinant of insulin resistance. We did not see this relationship in our study. We have a hypothesis that may partially explain our ®ndings. In our study, intraperitoneal and retroperitoneal fat make up only a small proportion of the total body fat mass (3.2 and 3.8%, respectively), which is similar to that reported in other populations of similar obesity.14 It was not surprising that these fat depots, even if more metabolically active, did not have a major effect on the insulin resistance in the entire body. We postulate that the preferential deposition of fat in the subcutaneous compartment results in a mass effect. The large size of the subcutaneous fat depot compared to the intra-abdominal fat depot may result in this becoming the major determinant of insulin resistance (Table 3). The same association between subcutaneous fat and insulin resistance has been found in men.25 We feel it is important to point out some important differences between our study and previous studies, which showed that intra-abdominal fat correlated signi®cantly with insulin resistance. Firstly, our study population was not selected on the basis of obesity or body fat distribution because we wanted to study a cohort representative of the majority of our population. The body mass index of our study population (mean body mass index 22.57 kg=m2) was only slightly higher than that of the entire population of women in Singapore (21.29 kg=m2).23 Compared with other studies which included signi®cant numbers of women10 ± 12,28 our study population was relatively lean. There is some thought that generalized obesity may have a permissive effect such that increased total fat mass is required before the deleterious effects of intra-abdominal fat are observed. Landin et al reported that intra-abdominal fat was an important determinant of insulin resistance in obese but not in lean subjects.29,30 In adolescent girls, visceral fat was a signi®cant correlate of insulin-mediated glucose disposal in obese but not in lean subjects.31 Secondly, this population was younger than that studied by Cefalu et al.13 They studied a population in whom half the women were aged 60 ± 80, a much higher proportion than in our study. They also reported that intra-abdominal fat mass increases with advancing age, a ®nding that is con®rmed by our data (Figure 2). It is conceivable that, with advancing age, the intra-abdominal fat mass could increase to a level where it becomes an important in¯uence on insulin resistance. It is therefore important that we do not give the impression that intra-abdominal fat has no impact on insulin resistance. Our study was not capable of making this conclusion. We are merely suggesting that subcutaneous fat mass correlates better with insulin resistance than intra-abdominal fat mass in a group of subjects which are representative of our general population.

755



756

The level of physical activity may play a role in explaining some of our ®ndings. Lean individuals could be more physically active and that this could result in lower insulin resistance in these individuals. Furthermore, the level of physical activity may contribute to the increase of intra-abdominal fat with age. It has been found that exercise, while decreasing the size of all fat depots, has a preferential effect on visceral fat.32 We also looked at the correlation between fasting lipids and body fat distribution. The dyslipidaemia associated with the metabolic syndrome manifests as a raised serum triglyceride and low HDL cholesterol. Hypercholesterolaemia is not a feature and it was not surprising that no correlation was found between total or LDL cholesterol and obesity (Table 3). However, the WHR was signi®cantly correlated with fasting serum TG. It also showed a correlation with HDL cholesterol but this was not statistically signi®cant. We have previously reported that the WHR was the best anthropometric predictor of both triglyceride is not mediated by variation in the intra-abdominal fat mass. Instead, our data seems to suggest that differential deposition of adipose tissue between the abdominal subcutaneous compartment and the glutefemoral fat depot is associated with variations in the fasting triglyceride and possibly HDL cholesterol as well. Thus heterogeneity of adipose tissue metabolism exists not only between visceral and subcutaneous tissue, but also between subcutaneous compartments at different sites. In support of this hypothesis is previous data showing that the subcutaneous fat distribution rather than the level of intra-abdominal fat correlated with plasma lipoprotein levels in lean women.34 The mechanism may be related to fat cell hypertrophy and greater basal lipolytic activity in the subcutaneous adipose tissue of the abdomen in women with upper segment obesity compared to adipose tissue from a similar site in lower segment obese women.35 The correlation of plasma leptin with subcutaneous but not intraperitoneal or retroperitoneal fat mass is more easily explained. As for insulin resistance, the proportion of total fat mass in the intra-abdominal compartment is small and thus a mass effect exists. Furthermore, it has been shown that omental adipocytes contain less leptin mRNA and also secrete less leptin than subcutaneous adipocytes.36 Our data thus con®rms the data of other investigators37 ± 39 that abdominal fat, whether intra- or retro-peritoneal is not a determinant of peripheral plasma leptin. In conclusion, our data suggests that obesity in women, unlike men, is associated with fat deposition in subcutaneous rather than intraabdominal compartments. Although advancing age is associated with accumulation of intraabdominal fat, in a younger non-obese population, there may need to be a shift of focus away from intra-abdominal fat to subcutaneous fat when dealing with the features of the metabolic syndrome and plasma leptin.


Acknowledgements

This work was supported by a research grant from the Singapore General Hospital Research Fund. The authors would like to thank Dr Su Chi Lim for his assistance in the preparation of this manuscript. References

1 Donahue RP, Abbott RD, Bloom E, Reed DM, Yano K. Central obesity and coronary heart disease in men. Lancet 1987; 1: 821 ± 824. 2 Ducimetiere P, Richard J, Cambien F. The pattern of subcutaneous fat distribution in middle-aged men and the risk of coronary heart disease: the Paris Prospective Study. Int J Obes 1986; 10: 229 ± 240. 3 Larsson B, Svardsudd K, Welin L, Wilhelmsen L, Bjorntorp P, Tibblin G. Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: 13 year follow up of participants in the study of men born in 1913. Br Med J (Clin Res Ed) 1984; 288: 1401 ± 1404. 4 Kannel WB, Cupples LARegional Adiposity and Risk of Cardiovascular Disease; the Framingham Study. J Clin Epidemiol 1991; 44: 183 ± 190. 5 Lapidus L, Bengtsson C, Larsson B, Pennert K, Rybo E, Sjostrom L. Distribution of adipose tissue and risk of cardiovascular disease and death: a 12 year follow up of participants in the population study of women in Gothenburg, Sweden. Br Med J (Clin Res Ed) 1984; 289: 1257 ± 1261. 6 Kohrt WM, Kirwan JP, Staten MA, Bourey RE, King DS, Holloszy JO. Insulin resistance in aging is related to abdominal obesity. Diabetes 1993; 42: 273 ± 281. 7 McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet 1991; 337: 382 ± 386. 8 Haffner SM, Karhapaa P, Mykkanen L, Laakso M. Insulin resistance, body fat distribution, and sex hormones in men. Diabetes 1994; 43: 212 ± 219. 9 Sparrow D, Borkan GA, Gerzof SG, Wisniewski C, Silbert CK. Relationship of fat distribution to glucose tolerance. Results of computed tomography in male participants of the Normative Aging Study. Diabetes 1986; 35: 411 ± 415. 10 Fujioka S, Matsuzawa Y, Tokunaga K, Tarui S. Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism in human obesity. Metabolism 1987; 36: 54 ± 59. 11 Peiris AN, Soth mann MS, Hennes MI, Lee MB, Wilson CR, Gustafson AB, Kissebah AH. Relative contribution of obesity and body fat distribution to alterations in glucose insulin homeostasis: predictive values of selected indices in premenopausal women. Am J Clin Nutr 1989; 49: 758 ± 764. 12 Zamboni M, Armellini F, Turcato E, de Pergola G, Todesco T, Bissoli L, Bergamo Andreis IA, Bosello O. Relationship between visceral fat, steroid hormones and insulin sensitivity in premenopausal obese women. J Intern Med 1994; 236: 521 ± 527. 13 Cefalu WT, Wang ZQ, Werbel S, Bell-Farrow A, Crouse JR III, Hinson WH, Terry JG, Anderson R. Contribution of visceral fat mass to the insulin resistance of aging. Metabolism 1995; 44: 954 ± 959. 14 Thomas EL, Saeed N, Hajnal JV, Brynes A, Goldstone AP, Frost G, Bell JD. Magnetic resonance imaging of total body fat. J Appl Physiol 1998; 85: 1778 ± 1785. 15 Abate N, Burns D, Peshock R, Garg A, Grundy S. Estimation of adipose tissue mass by magnetic resonance imaging: validation against dissection in human cadavers. J Lipid Res 1994; 35: 1490 ± 1496. 16 Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 1996; 334: 292 ± 295.


17 Ostlund RE, Jr., Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996; 81: 3909 ± 3913. 18 Ho SC, Tai ES, Eng PHK, Ramli A, Tan CE, Fok ACK. A study in the relationships between leptin, insulin, and body fat in Asian subjects. Int J Obes 1999; 23: 246 ± 252. 19 Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using biolelectric impedance measurement of the human body. Am J Clin Nutr 1985; 41: 410 ± 414. 20 Wang J, Deurenberg P. The validity of predicted body composition in Chinese adults from anthropometry and bioelectrical impedance in comparison with densitometry. Br J Nutr 1996; 76: 175 ± 182. 21 Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and B-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412 ± 419. 22 Haffner S, Gonzales C, Miettinen H, Kenndy E, Stern M. A prospective analysis of the HOMA model. The Mexico City Diabetes Study. Diabetes Care 1996; 19: 1138 ± 1141. 23 Tan CE, Emmanuel SC, Tan BY, Jacob E. Prevalence of diabetes mellitus and ethnic differences in cardiovascular risk factors: the 1992 Singapore National Health Survey. Diabetes Care 1999; 22: 241 ± 247. 24 Pouliot M, Despres J, Lemieux S, Moorjani S, Bouchard C, Tremblay A, Nadeau A, Lupien P. Waist circumference and abdominal saggital diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 1994; 73: 460 ± 468. 25 Abate N, Garg A, Peshock RM, Stray-Gundersen J, Grundy SM. Relationships of generalized and regional adiposity to insulin sensitivity in men. J Clin Invest 1995; 96: 88 ± 98. 26 Sohlstrom A, Wahlund LO, Forsum E. Adipose tissue distribution as assessed by magnetic resonance imaging and total body fat by magnetic resonance imaging, underwater weighing, and body-water dilution in healthy women. Am J Clin Nutr 1993; 58: 830 ± 838. 27 Ostman J, Arner P, Engfeldt P, Kager L. Regional differences in the control of lipolysis in human adipose tissue. Metabolism 1979; 28: 1198 ± 1205.

28 Despres JP, Nadeau A, Tremblay A, Ferland M, Moorjani S, Lupien PJ, Theriault G, Pinault S, Bouchard C. Role of deep abdominal fat in the association between regional adipose tissue distribution and glucose tolerance in obese women. Diabetes 1989; 38: 304 ± 309. 29 Landin K, Krotkiewski M, Smith U. Importance of obesity for the metabolic abnormalities associated with an abdominal fat distribution. Metabolism 1989; 38: 572 ± 576. 30 Landin K, Lonnroth P, Krotkiewski M, Holm G, Smith U. Increased insulin resistance and fat cell lipolysis in obese but not lean women with a high waist=hip ratio. Eur J Clin Invest 1990; 20: 530 ± 535. 31 Caprio S, Hyman LD, Limb C, McCarthy S, Lange R, Sherwin RS, Shulman G, Tamborlane WV. Central adiposity and its metabolic correlates in obese adolescent girls. Am J Physiol 1995; 269: E118 ± 126. 32 Smith SR, Zachwieja JJ. Visceral adipose tissue: a critical review of intervention strategies. Int J Obes Relat Metab Disord 1999; 23: 329 ± 335. 33 Tai ES, Ho SC, Fok ACK, Tan CE. Measurement of obesity by anthropometry and bioelectric impedance analysis: correlation fasting lipids and insulin resistance in an Asian population. Ann Acad Med Singapore 1999; 28: 445 ± 450. 34 Despres JP. Lipoprotein metabolism in visceral obesity. Int J Obes 1991; 15(Suppl 2): 45 ± 52. 35 Kissebah AH, Vydelingum N, Murray R, Evans DJ, Hartz AJ, Kalkhoff RK, Adams PW. Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab 1982; 54: 254 ± 260. 36 Van Harmelen V, Reynisdottir S, Eriksson P, Thorne A, Hoffstedt J, Lonnqvist F, Arner P. Leptin secretion from subcutaneous and visceral adipose tissue in women. Diabetes 1998; 47: 913 ± 917. 37 Takahashi M, Funahashi T, Shimomura I, Miyaoka K, Matsuzawa Y. Plasma leptin levels and body fat distribution. Horm Metab Res 1996; 28: 751 ± 752. 38 Dua A, Hennes MI, Hoffmann RG, Maas DL, Krakower GR, Sonnenberg GE, Kissebah AH. Leptin: a signi®cant indicator of total body fat but not of visceral fat and insulin insensitivity in African-American women. Diabetes 1996; 45: 1635 ± 1637. 39 Ramachandran A, Sbnehalatha C, Vijay V, Satyavani K, Latha E, Haffner SM. Plasma leptin in non-diabetic Asian Indians: association with abdominal adiposity. Diabet Med 1997; 14: 937 ± 941.

757
