Visceral adipose tissue and body fat mass: Predictive values for and role of gender in cardiometabolic risk among Turks

Nutrition 26 (2010) 382–389

Applied nutritional investigation

www.nutritionjrnl.com

Visceral adipose tissue and body fat mass: Predictive values for and role of gender in cardiometabolic risk among Turks Altan Onat, M.D.a,b,*, Murat Ug˘ur, M.D.c, Gu¨nay Can, M.D.d, Hu¨sniye Yu¨ksel, M.D.b, and Gu¨lay Hergenc¸, Ph.D.e a Turkish Society of Cardiology, Istanbul, Turkey Department of Cardiology, Cerrahpas xa Medical Faculty, Istanbul University, Istanbul, Turkey c Siyami Ersek Center for Cardiovascular Surgery, Istanbul, Turkey d Department of Public Health, Cerrahpas xa Medical Faculty, Istanbul University, Istanbul, Turkey e Biology Department, Yildiz Technical University, Istanbul, Turkey b

Manuscript received February 16, 2009; accepted May 18, 2009.

Abstract

Objective: We investigated the predictive values of visceral adipose tissue area (VAT) and body fat mass for a composite endpoint consisting of type 2 diabetes and coronary heart disease and for incident metabolic syndrome. Methods: We analyzed at 4-y follow-up 157 middle-aged men and women in whom body composition analyzer and single-scan computerized tomography had been used. Results: Sex- and age-adjusted mean areas of visceral fat were 1.5-fold greater in individuals with than without the composite endpoint (P < 0.001), whereas abdominal subcutaneous fat was similar. Analysis of receiver operating characteristics for the optimal criterion regarding the composite endpoint (in 37 participants) indicated a VAT of 130 cm2 and accuracies of 60% in men and 85% in women. Whereas age-adjusted VAT alone significantly predicted the composite endpoint in men, body fat mass or VAT predicted it in women (with 2.2- to 2.6-fold relative risks for 1-SD increment). Age-adjusted incident metabolic syndrome was significantly predicted by each parameter in men but only by fat mass in women. Conclusion: Visceral adiposity in men and body fat mass in women seem to be of greater relevance in cardiometabolic risk for the prediction of which 130 cm2 of VAT in both sexes and/or 27 kg of fat mass in women are useful cutoffs. Sex differences may reflect the predominating role of visceral adiposity in men and of insulin resistance in women in this risk. Ó 2010 Published by Elsevier Inc.

Keywords:

Coronary heart disease; Diabetes type 2; Metabolic syndrome; Obesity; Insulin resistance; Visceral adiposity

Introduction The metabolic syndrome (MetS) is designated as a cluster of atherothrombotic inflammatory risk factors [1] for which visceral adiposity and insulin resistance form the pathophysiologic substrate. Fat in various adipose tissue subcompartments is strongly linked to atherogenic dyslipidemia,

The Turkish Adult Risk Factor survey 2006–2008 was supported by the Turkish Society of Cardiology and the pharmaceutical companies SanofiAventis, Pfizer, Novartis, and AstraZeneca, Istanbul, Turkey. *Corresponding author. Tel.: þ90-212-351-6217; fax: þ90-212-3514235. E-mail address: [email protected] (A. Onat). 0899-9007/10/$ – see front matter Ó 2010 Published by Elsevier Inc. doi:10.1016/j.nut.2009.05.019

cardiovascular disease, and diabetes [2]. Because visceral adiposity is a major driver of cardiometabolic risk [3], studies involving visceral adipose tissue area (VAT) or mass by computerized tomography (CT) or magnetic resonance imaging are essential. It has also been recognized that important ethnic differences exist in susceptibility to visceral adiposity and related metabolic abnormalities [4]. The issue of the major determinant of cardiometabolic risk is still controversial. Some investigators have proposed that abdominal subcutaneous adipose tissue carries as much or more pathogenic significance for the MetS as visceral obesity [2,5]. Furthermore, although waist circumference (WC) or waist-to-hip ratio is generally considered the most appropriate anthropometric surrogate of VAT, this question is not

A. Onat et al. / Nutrition 26 (2010) 382–389

fully settled. Data in diverse ethnic groups are needed that may generate differing relations between anthropometric surrogates and the actual amount of ‘‘dangerous’’ fat [6]. Turks have a high and rapidly increasing prevalence of MetS [7]. We previously reported the covariates of VAT as concerned anthropometric indices and serum lipoproteins in a subsample of the Turkish Adult Risk Factor Study [8] in which CT-assessed variables were performed [9]. However, the role of a body composition variable in the likelihood of cardiometabolic risk was not evaluated; and with the current availability of several years of follow-up, we can contribute better to the prediction of adverse outcomes of diabetes and coronary heart disease (CHD) by measurements of abdominal adipose tissue and body composition. Therefore, using the previous baseline variables assessed by CT and hydrodensitometry, we aimed to evaluate the comparative predictive values of VAT and body fat mass for 1) a composite endpoint consisting of diabetes and CHD and 2) MetS. Materials and methods Subjects The study sample of 157 residents of the city of Istanbul who composed the cohort of the Turkish Adult Risk Factor Study [8] was identical to that in which abdominal fat areas had been assessed by CT at the end of 2002, details of which were previously reported [9]. The study was approved by the ethics committee of the Istanbul University Medical Faculty. Written informed consent was obtained from all participants. Measurement of anthropometry and total body fatness at baseline Waist circumference was measured using a spring scale (O. Haus, Florham Park, NJ, USA) with the subject standing and wearing only underwear, at the end of gentle expiration at the level midway between the lower rib margin and the iliac crest, and that of the hip was measured at the level of the great trochanters. Body mass index (BMI) was calculated as weight (kilograms) divided by height (meters) squared. Body composition was measured by bioelectrical impedance analysis (Tanita TBF 300, Tokyo, Japan). Computed tomography was performed on a Siemens Somatom AR SP 40 (Erlangen, Germany) spiral scanner [9], similarly to that described by Ferland et al. [10]. A single axial tomographic scan was made at the abdominal level, between the fourth and fifth lumbar vertebrae. The sagittal diameter of the abdomen was determined from the abdominal image generated by the computer by connecting the anterior and posterior midline points. Total adipose tissue areas were calculated by delineating the abdomen with a graph pen and then computing the adipose tissue surface using an attenuation range of 190 to 30 HU [10]. Subtracting the abdominal VAT area from the total adipose tissue area produced the abdominal subcutaneous area.

383

Measurement of other risk factors Blood pressure was measured in the sitting position on the right arm using an aneroid sphygmomanometer (Erka, Bad To¨lz, Germany) after 5 min of rest, and the mean of two recordings 3 min apart was computed. Plasma concentrations of cholesterol, high-density lipoprotein cholesterol, fasting triacylglycerols, and glucose were determined by enzymatic dry chemistry method using a Reflotron apparatus (Roche Diagnostics, Mannheim, Germany). Glucose was sampled 2 h postprandially in 11% of participants. Serum concentrations of insulin were determined by the chemiluminescent immunometric method using Diagnostic Products Corporation kits and the autoimmunanalyzer Immulite (Los Angeles, CA, USA). External quality control was performed with a reference laboratory in a random selection of 5–6% of participants. Definitions and outcomes Individuals with diabetes were diagnosed with criteria of the American Diabetes Association [11], namely when plasma fasting glucose was 126 mg/dL (or 2-h postprandial glucose >200 mg/dL) and/or current use of diabetes medication. Individuals with MetS were identified when three of the five criteria of the National Cholesterol Education Program (Adult Treatment Panel III [ATP III]) [12] were met, modified for prediabetes (fasting glucose 100–125 mg/dL) [13], and further for abdominal obesity using as a cutpoint 95 cm in men, as recently assessed in the Turkish Adult Risk Factor study [14]. Missing data on triacylglycerols in 25% of the sample did not preclude the identification of MetS because the availability of no more than three criteria was required, and the MetS status of the subsequent survey was adopted in a few individuals presenting two positive criteria. Homeostasis model assessment (HOMA) was calculated with the following formula [15]: insulin (mIU/L) 3 glucose (mmol/L)/22.5. Diagnosis of nonfatal CHD was based on the presence of angina pectoris, a history of myocardial infarction with or without accompanying Minnesota codes of the electrocardiogram [16], or a history of myocardial revascularization. Typical angina and, in women, age >45 y were prerequisite for a diagnosis when angina was isolated. Electrocardiographic changes of ‘‘ischemic type’’ greater than a minor degree (codes 1.1–2, 4.1–2, 5.1–2, 7.1) were considered as myocardial infarct sequelae or myocardial ischemia, respectively. CHD death consisted of death from heart failure of coronary origin and a fatal coronary event. Follow-up Participants were tracked from autumn 2002 when CT examinations were performed until August 2008 over a maximum of 5.7 y. Because 27 subjects had incomplete follow-up, total follow-up amounted to 630 person-years (mean 4.0 y).

384

A. Onat et al. / Nutrition 26 (2010) 382–389

Table 1 Sex- and age-adjusted (48.8 y) characteristics of study sample (n ¼ 157) at baseline by presence of VAT* VAT 130y (n ¼ 71) Sagittal diameter (cm) Visceral abdominal tissue (cm2)z Subcutaneous abdominal tissue (cm2) Total body fat (%) Fat mass (kg) Waist circumference (cm) Body mass index (kg/m2) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) HOMA indexz (n ¼ 119) Fasting insulin (mIU/L)z (n ¼ 120) Total cholesterol (mmol/L) HDL cholesterol (mmol/L) LDL cholesterol (mmol/L) (n ¼ 115) Fasting triacylglycerol (mmol/L)z (n ¼ 115) Fasting glucose (mmol/L) (n ¼ 139)

{

24.7 6 0.3 175.8{ 6 1.03 315{ 6 11 35.5{ 6 0.75 29.5{ 6 1.0 100.0{ 6 1.17 30.3{ 6 0.49 133.6 6 2.7 85.1x 6 1.6 2.32{ 6 1.08 9.60x 6 1.08 5.10 6 0.13 0.98{ 6 0.03 3.15 6 0.12 1.60x 6 0.01 5.56x 6 0.28

VAT <130 (n ¼ 86) 20.6 6 0.3 86.3 6 1.04 247 6 10 29.1 6 0.68 20.9 6 0.9 87.6 6 1.07 26.1 6 0.44 128.2 6 2.4 80 6 1.4 1.51 6 1.10 7.46 6 1.07 5.0 6 0.11 1.11 6 0.03 3.15 6 0.11 1.30 6 0.01 4.7 6 0.26

HDL, high-density lipoprotein; HOMA, homeostasis model assessment; LDL, low-density lipoprotein; VAT, visceral adipose tissue area x P < 0.05, { P ¼ 0.001. * Values are means 6 SEs. y Log-transformed values. z Geometric means.

Data analysis Descriptive parameters are shown as age-adjusted mean estimate 6 standard error or in percentages. Log-transformed values were used for insulin, HOMA, and VAT due to their skewed distribution. Two-sided t tests and Pearson’s chisquare tests served to analyze differences in means and proportions between groups. Estimates (95% confidence intervals [CIs]) for relative risk (RR) of a dependent variable were obtained by use of multiple logistic regression analyses in models that controlled for potential confounders. An RR for incident MetS was obtained after exclusion of participants with MetS at baseline examination. Among men and women, 1 standard deviation (SD) of BMI corresponded to 4 and 5.6 kg/m2 and WC to 11 and 12.4 cm, respectively. P < 0.05 on the two-sided test was considered statistically significant. Statistical analyses were performed using SPSS 10 for Windows (SPSS Inc., Chicago, IL, USA). Results The study sample consisted of 79 men and 78 women. Mean age 6 SD of the sample was 49 6 9 y at baseline examination. Compared with the entire cohort of the Turkish Adult Risk Factor Study, this study sample at baseline excluded subjects 70 y of age and, hence, was 4 y younger (P < 0.001), had a narrower WC (0.6 cm, P < 0.03), and lower BMIs (0.5 kg/m2, P < 0.036) and women were more frequent current smokers (35% versus 17.4%, P < 0.001). Sex prevalence, blood pressures, values of glucose, lipids, and lipoproteins were similar in this sample. The composite endpoint occurred in 37 participants; five had type 2 diabetes combined with CHD, one of whom died in follow-up, nine

had CHD, two of whom subsequently succumbed, and six had diabetes alone; diabetes developed newly in nine and CHD in eight participants during follow-up. Age-adjusted geometric mean VATs were 132.4 6 1.05 cm2 in men and 107.2 6 1.06 cm2 in women (P ¼ 0.003). Age-adjusted means for abdominal subcutaneous adipose tissue were 220.3 6 10.7 and 335.7 6 10.7 cm2 (P < 0.001), for sagittal abdominal diameter 22.4 6 0.37 cm (similar in men and women), and for fat mass 22.7 6 1.0 and 26.8 6 1.01 kg (P < 0.01). Table 1 shows that at baseline sex- and age-adjusted means of individuals with a VAT 130 compared with <130 cm2 differed significantly in WC, HOMA, high-density lipoprotein cholesterol (P < 0.001 in all), fasting glucose, triacylglycerols, insulin, and diastolic blood pressure (P < 0.05 in all). Distribution of adiposity compartments and composite endpoint Table 2 demonstrates that in subjects with the composite endpoint compared with the remaining subjects, sex- and age-adjusted mean values of body fat content and fat compartments were significantly higher (by 53% for VAT, 11% for sagittal diameter, 27% for body fat mass, and 14% for body fat percentage). Only abdominal subcutaneous fat area was similar in the two groups. Receiver operating characteristics analysis of abdominal fat compartments Analysis of receiver operating characteristics curves of three adiposity surrogates for optimal criteria of predicting the composite endpoint are presented in Table 3, by gender.

A. Onat et al. / Nutrition 26 (2010) 382–389

385

Table 2 Sex- and age-adjusted distribution of indicators of body fat and abdominal adiposity in subjects with and without the composite endpoint

Sagittal diameter (cm) Visceral abdominal tissue (cm2)y Abdominal subcutaneous tissue (cm2) Body fat (%) Body fat mass (kg) Body mass index (kg/m2) Waist circumference (cm)

No endpoint* (n ¼ 120)

With endpoint* (n ¼ 37)

P

21.8 6 0.3 107.6 6 1.04 270.4 6 8.8 31.0 6 0.62 23.3 6 0.8 27.4 6 0.4 90.5 6 0.95

24.2 6 0.56 164.8 6 1.08 301.3 6 16.9 35.4 6 1.18 29.6 6 1.5 30.1 6 0.8 96.1 6 1.8

<0.001 <0.001 0.12 0.002 0.001 0.003 0.009

* Values are means 6 SEs. y Geometric means.

A VAT of 130 cm2 had 88% sensitivity and 52% specificity for the composite endpoint in men and 86% sensitivity and 84% specificity in women. Area under the curve of total fat mass (Table 3 and Fig. 1) failed to reach significance in men but was significant in women (and greater than the percentage of fat). A sagittal diameter of 24 cm had a lower sensitivity, although a higher specificity, than VAT for the composite endpoint in both sexes. Regression analyses for composite endpoint For better delineating the specific association between the risk of composite endpoint and each adiposity component or obesity measurement, age-adjusted logistic regression analyses were performed using VAT, body fat mass, WC, and BMI as independent variables, separately, and in various dual combinations (Table 4). In terms of 1-SD increment (50 cm2), in the basic sex- and age-adjusted model, VAT revealed an RR of 2.56 in predicting the endpoint, in men an RR of 2.02 (95% CI 1.15–3.6), and in women an RR of 3.44 (95% CI 1.64–7.1). A 1-SD increment (9 kg) of sexand age-adjusted fat mass predicted the endpoint significantly with an RR of 2.24 (95% CI 1.39–3.66), in women with an RR of 2.46 (95% CI 1.26–4.7), and at borderline significance in men with an RR of 2.00 (95% CI 0.94–4.24). When VAT and fat mass, dichotomized by criteria of 130 cm2 and 27/22 kg, respectively, were entered in the analysis, they largely attenuated each other; VAT retained signif-

icance in women and attenuated to borderline significance in men. Although WC was significant for the endpoint of cardiometabolic risk only in men, BMI was so in both sexes. When VAT equal to 130 cm2 was analyzed in conjunction with either obesity measurement, WC and BMI lost significance, but a non-significant residual risk seemed to remain for WC in men. Collectively, these findings indicate that WC, at variance to men, does not contribute in women independently of BMI to cardiometabolic risk, although fat mass and VAT do. Furthermore, WC does not appear as good a marker of VAT as in men. For 1 kg of total body fat mass, age-adjusted men and women had a geometric mean of log-transformed VAT/fat mass ratio (6.28 versus 4.27 cm2/kg, P < 0.001), indicating a 1.46-fold higher susceptibility of men to visceral fat accumulation than women. The age-adjusted geometric mean VAT/fat mass ratio was significantly higher in individuals with than in those without the composite endpoint (P ¼ 0.02). By logistic regression, the composite endpoint was predicted by the stated ratio in women (for a doubling of the ratio, RR 3.11, 95% CI 1.20–8.02) and both genders combined (RR 2.62, 95% CI 1.30–5.28), but not in men (P ¼ 0.17). Prediction of incident MetS After exclusion of the prevalent cases of MetS, the syndrome newly developed in 13 men and 9 women of the remaining 100 participants. Age-adjusted MetS was

Table 3 Area under curve of receiver operating characteristics analysis and predictive values of cutoffs of visceral fat, total fat mass, and sagittal diameter for composite endpoint Men

Women

VAT Selected thresholds Area under curve P Sensitivity (%) Specificity (%) PPV NPV Accuracy

130 cm 0.77 0.001 88 52 32 94 60

Fat mass 2

22 kg 0.62 0.15 68.8 54 27.5 87 57

Sagittal diameter 24 cm 0.705 0.012 63 71 36 88 70

VAT 2

130 cm 0.90 <0.001 86 84 67 94 85

NPV, negative predictive value; PPV, positive predictive value; VAT, visceral adipose tissue area

Fat mass

Sagittal diameter

27 kg 0.72 0.003 71.4 70 46.9 87 70.5

24 cm 0.745 0.001 52 77 46 81.5 70.5

A. Onat et al. / Nutrition 26 (2010) 382–389

386

Male

1,00

,75

Sensitivity

,75

Sensitivity

Female

1,00

,50

,25

,50

,25 Visceral A.

0,00 0,00

Visceral A.

Fat mass ,25

,50

,75

0,00 0,00

1,00

1 - Specificity

Fat mass ,25

,50

,75

1,00

1 - Specificity

Fig. 1. Receiver operating characteristics curves of visceral adipose tissue area and body fat mass for the composite endpoint in women (areas under the curve 0.90 and 0.72) and men (areas under the curve 0.77 and 0.62).

significantly predicted in men by either body fat component equally well (RR 2.69 with each; Table 5), in women only by fat mass, although in a weaker way than in men (for 1-SD increment, RR 1.92, 95% CI 1.01–3.66). VAT did not predict MetS significantly (P ¼ 0.33), nor did sex- and age-adjusted abdominal adipose tissue area in either gender (P ¼ 0.15).

women in contrast to VAT alone in men. Evidence was provided that WC, not as good a marker of VAT in women as in men, and, at variance to men, does not contribute independently of BMI to the cardiometabolic risk, whereas fat mass and VAT tended to predict risk independently in women, again at variance to men. MetS was significantly predicted by either age-adjusted measurement in men, but only by fat mass in women. Hence, in this population sample, fat mass appeared equally relevant in predicting cardiometabolic risk in women, although VAT was greater in men.

Discussion In a sample of residents of the city of Istanbul, Turkey, predictive values of VAT and body fat mass, assessed by CT and hydrodensitometry, respectively, were examined for a composite endpoint consisting of prevalent and incident cases of diabetes, fatal and nonfatal CHD, or death. Receiver operating characteristics analysis showed highest areas under the curve for VAT in both sexes; total body fat mass was additionally significant only in women. Both age-adjusted parameters predicted the composite endpoint significantly in

Visceral fat accumulation Deposition of fat into ectopic areas depends on total body fat mass and an individual’s basic susceptibility that in turn is related to age, gender, and ethnicity. Gender and ethnicity were found to influence group differences in visceral adiposity; compared with white women, men had significantly

Table 4 Age-adjusted prediction of composite endpoint by body fat measurements*

Female sex Age (y) VAT (cm2) Body fat mass (kg) Dual obesity markersy 1 VAT, >130 versus <130 cm2 Body fat mass, >27/22 versus <27/22 kg 2 Body fat mass, >27/22 versus <27/22 kg Body mass index (kg/m2) 3 Waist circumference (cm) Body mass index (kg/m2) 4 Waist circumference (cm) VAT, >130 versus <130 cm2 5 Body mass index (kg/m2) VAT, >130 versus <130 cm2

Total (n ¼ 157)

Men (n ¼ 79)

RR

95% CI

RR

95% CI

2.52 1.127 1.019 1.094

0.95–6.67 1.06–1.19 1.01–1.028 1.037–1.155

1.126 1.014 1.08

1.04–1.22 1.003–1.026 0.993–1.174

1.122 1.025 1.105

1.02–1.23 1.01–1.04 1.028–1.188

10.2 2.79 3.65 1.063 1.06 1.16 1.01 12.7 1.048 11.6

3.1–33.7 0.99–7.84 1.15–11.6 0.94–1.21 1.015–1.11 1.05–1.29 0.96–1.06 3.6–44.4 0.93–1.18 3.4–39.4

4.75 2.13 2.05 1.16 1.094 1.225 1.055 3.83 1.11 4.26

0.78–28.8 0.47–9.7 0.38–11.2 0.91–1.47 1.01–1.19 1.01–1.49 0.96–1.16 0.55–26.7 0.88–1.40 0.62–29.2

16.8 4.23 5.64 1.02 1.044 1.134 0.993 23.3 1.033 19.1

3.67–77.2 0.97–18.5 1.14–27.9 0.88–1.19 0.993–1.10 1.005–1.28 0.93–1.06 4.5–119 0.90–1.19 4.0–90.3

CI, confidence interval; RR, relative risk; VAT, visceral adipose tissue area * Included were 16 men (20.3%) and 21 women (26.9%) with the composite endpoint. y Dual independent variables were jointly analyzed in sex- and age-adjusted models.

Women (n ¼ 78) RR

95% CI

A. Onat et al. / Nutrition 26 (2010) 382–389

387

Table 5 Sex- and age-adjusted prediction of new metabolic syndrome by VAT or body fat mass* Total (n ¼ 100) 2

VAT (cm ) Body fat mass (kg)

1.011 1.083

Men (n ¼ 53) 1.002–1.02 1.024–1.147

1.020 1.116

Women (n ¼ 47) 1.004–1.035 1.012–1.23

1.006 1.075

0.994–1.018 1.001–1.155

VAT, visceral adipose tissue area * Included were 13 men (24.5%) and 9 women (19%) who developed metabolic syndrome in follow-up.

greater total VAT volume, and their maximum VAT occurred higher in the abdomen, whereas this was less conspicuous in African-American adults [17], yet a gender difference has not been a consistent finding [3,18,19]. Data of this sample confirmed that, compared with women, men showed a higher VAT per kilogram of body fat mass.

Expandability of the fat compartment The higher thresholds of total fat mass associated with a VAT of 130 cm2 in women than in men reflect the greater expandability of the overall fat compartment and lower susceptibility to accumulate visceral fat. Nonetheless, a significant prediction of cardiometabolic endpoint by sex- and age-adjusted VAT/fat mass ratio was revealed. This implies that the ATP III criterion of waist girth for MetS is too small in Turkish women to signify a high likelihood of visceral adiposity, and that other factors such as impaired insulin sensitivity originating from overall fatness might be involved for the high cardiometabolic risk of generally identified cases of MetS. Histology and metabolic activity of adipose tissue exhibits differences. Only fat cell hypertrophy has been observed in omental fat, whereas hypertrophy and hyperplasia were present in the subcutaneous compartment in 40 obese women [20]. In another study, interleukin-6 levels were significantly higher in 189 asymptomatic men with high versus low VAT and were found to contribute independently to the variation of fasting insulin, whereas tumor necrosis factor-a was associated rather with BMI as an index of total body fatness [21]. Body fat mass was the best covariate of circulating leptin in 95 obese non-diabetic patients [22]. Total fat mass approached VAT in women in predicting diabetes and CHD and surpassed it in the prediction of incident MetS. This is not necessarily at variance with the findings of Goodpaster et al. [23] to the effect that subcutaneous thigh adipose tissue was inversely associated with MetS in obese elderly adults, and with the Hoorn Study [24] that, compared with trunk fat, accumulation of fat in the legs seemed to be oppositely protective against a disturbed glucose metabolism, particularly in women, because the measurement of total body fat mass is more comprehensive. In contrast, visceral fat tissue, not total fat mass, was an independent predictor of incident myocardial infarction in elderly women, whereas both were not associated with the event in elderly men in the Health, Aging and Body Composition study [25].

Possible factors for fat mass predicting risk in women Which features of Turkish women make fat mass, an index of overall fatness, as good a predictor of diabetes and CHD? One factor might be its sheer mass (being >10 times greater than that of visceral adiposity [6]) contributing through an abnormal regulation of lipolysis the majority of systemic free fatty acid delivery [5]. Second, we have evidence that, as distinct from Turkish men, the correlation of insulin resistance with fat mass in women is stronger than with VAT (observations to be published), i.e., more discordance exists between visceral adiposity and insulin resistance in women; in fact, 21.8% of Turkish women versus 17.2% of men presented with MetS in the absence of HOMA for insulin resistance [26]. Third, cigarette smoking may constitute a lesser confounder for fat mass than for VAT by reducing visceral fat accumulation (to an extent beyond reducing fat mass) in women [27]. Clinical implications of relevance of overall fatness to cardiometabolic risk in women only A first consequence of this has been some difficulty in selecting an appropriate cutoff in women [9,28] for the identification of MetS with ATP III [12] (or International Diabetes Federation) criteria. As large as a 14-cm difference in waist girth existed between 25th and 75th percentiles in women with regard to attaining the visceral adiposity cutoff. A BMI criterion (30 kg/m2) might be needed in Turkish women as an alternative for the WC measurement in identifying MetS with the aim of enhancing the detection of an associated insulin resistance, a critical parameter somewhat dissociated from visceral fat in women. Second, this finding supports the view that overall fatness also has adverse metabolic activity. Recognizing an adiposity-related gendermodulated cardiometabolic risk may facilitate the understanding of why there may be a sex difference in the pathway from obesity to diabetes and cardiovascular disease [29]. Prediction of MetS in women by fat mass rather than VAT is an unexpected finding, given that WC was a good predictor of MetS in a large female sample [28]. However, the better risk predictability of diabetes by BMI than waist girth in women was shown to be mediated largely by insulin levels with which present findings concur. This suggests that insulin resistance is of greater relevance than excess pure visceral fat in the pathogenesis of MetS and diabetes in women, whereas visceral adiposity is the driver of cardiometabolic risk in men.

388

A. Onat et al. / Nutrition 26 (2010) 382–389

Thus the partial lack of overlapping between visceral adiposity and insulin resistance emphasized by the ATP III in the pathogenesis of MetS [12] is illustrated strikingly across genders of the same ethnicity. Limitations and strengths Although MetS was fully prospectively analyzed, the fact that only just over half the participants having the composite endpoint was elicited prospectively is a limitation, but crosssectional associations may also provide valid inferences and need not be discarded. The studied sample size may be considered limited but is larger than most published ones using abdominal imaging techniques and was adequate to yield significant findings, yet not large enough to separately analyze the different endpoints. Conclusions reached may not be fully applicable to a population in which the prevalence of MetS is far lower than in the present one. The strengths of this study may be cited as the inclusion of both sexes, its being representative of a middle-aged general population, the concomitant availability of data on body fat distribution, anthropometric indices, and related biochemical determinations, and the largely prospective nature of the study. We conclude that gender, presumably by being a determinant in the insulin sensitivity of adipose tissue, modulates cardiometabolic risk conferred by visceral adiposity and overall fatness: visceral adiposity in men versus body fat mass (forming the major pool of insulin resistance) in women appears of greater relevance in cardiometabolic risk among Turks. A cutoff of 130 cm2 of VAT is useful in predicting the development of diabetes and CHD in both sexes; a fat mass of 27 kg (or BMI 30 kg/m2) as an alternative criterion may improve the accuracy of women at risk. Clinical implications of these gender differences may be related to directing more specific preventive measures; targeting reduction of subcutaneous fat might be adequate in women in some ethnic groups, whereas reduction of visceral fat in men is essential. Acknowledgments The authors are indebted to the computed tomographic findings obtained by Dr. M. M. Barlan and to the contributions of Prof. G. Sx. Avci. They deeply appreciate the works of Dr. Z. Ku¨c¸u¨kdurmaz, Dr. S. Bulur, and Dr. S. Ordu in the survey teams. References [1] Despre´s J-P, Brewer HB. Metabolic syndrome: the dysmetabolic state of dysfunctional adipose tissue and insulin resistance. Eur Heart J 2008; 10(suppl B):B1–3. [2] Vega GL, Adams-Huet B, Peshock R, Willett DW, Shah B, Grundy SM. Influence of body fat content and distribution on variation in metabolic risk. J Clin Endocrinol Metab 2006;91:4459–66. [3] Despre´s J-P, Poirier P, Bergeron J, Tremblay A, Lemieux I, Alme´ras N. From individual risk factors and the metabolic syndrome to global cardiometabolic risk. Eur Heart J 2008;10(suppl B):B24–33.

[4] Alberti KG, Zimmet P, Shaw J. The metabolic syndrome—a new world-wide definition. Lancet 2005;366:1059–62. [5] Jensen MD. Is visceral fat involved in the pathogenesis of the metabolic syndrome? Human model. Obesity 2006;14(suppl 1). 20S–4. [6] Ferrannini E, Sironi AM, Iozzo P, Gastaldelli A. Intra-abdominal adiposity, abdominal obesity, and cardiometabolic risk. Eur Heart J 2008;10(suppl B):B4–10. ¨ , Erer B, Toprak S, Sansoy V. Metabolic [7] Onat A, Ceyhan K, Basar O syndrome: major impact on coronary risk in a population with low cholesterol levels—a prospective and cross-sectional evaluation. Atherosclerosis 2002;165:285–92. [8] Onat A. Risk factors and cardiovascular disease in Turkey. Atherosclerosis 2001;156:1–10. [9] Onat A, Avcı GS x, Barlan MM, Uyarel H, Uzunlar B, Sansoy V. Measures of abdominal obesity assessed for visceral adiposity and relation to coronary risk. Int J Obes Relat Metab Disord 2004;28:1018–25. [10] Ferland M, De´spres JP, Tremblay A, Pineault S, Nadeau A, Moorjani S, et al. Assessment of adipose tissue distribution by computed axial tomography in obese women: association with body density and anthropometric measurements. Br J Nutr 1989;61:139–48. [11] Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, et al. Expert committee on the diagnosis and classification of diabetes mellitus. Follow-up report on the diagnosis of diabetes mellitus: the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 2003;26:3160–7. [12] 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. [13] Grundy SM, Brewer HB, Cleeman JI, Smith SC, Lenfant C. Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation 2004;109:433–8. [14] Onat A, Uyarel H, Hergenc¸ G, Karabulut A, Albayrak S, Can G. Determinants and definition of abdominal obesity as related to risk of diabetes, metabolic syndrome and coronary disease in Turkish men: a prospective cohort study. Atherosclerosis 2007;191:182–90. [15] Mather KJ, Hunt AE, Steinberg HO, Paradisi G, Hook G, Katz A, et al. Repeatability characteristics of simple indices of insulin resistance: implications for research applications. J Clin Endocrinol Metab 2001; 86:5457–64. [16] Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular survey methods. 2nd ed. Geneva: World Health Organization; 1982; p. 124–7. [17] Demerath FW, Sun SS, Rogers N, Lee M, Reed D, Choh AC, et al. Anatomical patterning of visceral adipose tissue: race, sex, and age variation. Obesity 2007;15:2984–93. [18] Despre´s J-P, Couillard C, Gagnon J, Bergeron J, Leon AS, Rao DC, et al. Race, visceral adipose tissue, plasma lipids, and lipoprotein lipase activity in men and women: the Health, Risk Factors, Exercise Training, and Genetics (HERITAGE) family study. Arterioscler Thromb Vasc Biol 2000;20:1932–8. [19] Hallgreen CE, Hall KD. Allometric relationship between changes of visceral fat and total fat mass. Int J Obes Relat Metab Disord 2008; 32:845–52. [20] Drolet R, Richard C, Sniderman AD, Maillout J, Fortier M, Huot C, et al. Hypertrophy and hyperplasia of abdominal adipose tissues in women. Int J Obes 2008;32:283–91. [21] Cartier A, Lemieux I, Alme´ras N, Tremblay A, Bergeron J, Despre´s JP. Visceral obesity and plasma glucose-insulin homeostasis: contributions of interleukin-6 and tumor necrosis factor-alpha in men. J Clin Endocrinol Metab 2008;93:1931–8. [22] de Luis DA, Gonzales Sagradeo M, Conde R, Aller R, Izaola O, Romero E. Circulating adipocytokines in obese nondiabetic patients in relationship with cardiovascular risk factors, anthropometry and resting-energy expenditure. Ann Nutr Metab 2007;51:416–20. [23] Goodpaster BH, Krishnaswami S, Haris TB, Katsiaras A, Kritchevsky SB, Simonsick EM, et al. Obesity, regional body fat distribution and the

A. Onat et al. / Nutrition 26 (2010) 382–389 metabolic syndrome in older men and women. Arch Intern Med 2005; 165:777–83. [24] Snijder MB, Dekker JM, Visser M, Bouter LM, Stehouwer CDA, Yudkin JS, et al. Trunk fat and leg fat have independent and opposite associations with fasting and postload glucose levels: the Hoorn Study. Diabetes Care 2004;37:372–7. [25] Nicklas BJ, Penninx BW, Cesari M, Kritchevsky SB, Newman AB, Kanaya AM, et al. Health, Aging and Body Composition Study. Association of visceral adipose tissue with incident myocardial infarction in older men and women: the Health, Aging and Body Composition Study. Am J Epidemiol 2004;160:741–9. _ Can G. Discordance [26] Onat A, Hergenc¸ G, Tu¨rkmen S, Yazıcı M, Sarı I, between Insulin resistance and metabolic syndrome: features and

389

associated cardiovascular risk in adults with normal glucose regulation. Metabolism 2006;55:445–52. [27] Onat A, Can G, Ayhan E, Kaya Z, Hergenc¸ G. Impaired protection against diabetes and coronary heart disease by high-density lipoproteins in Turks. Metabolism 2009. doi:10.1016/j.metabol.2009.05.004. _ Hergenc¸ G, Yazıcı M, Can G, Uyarel H, Sansoy V. [28] Onat A, Sarı I, Predictors of abdominal obesity and high susceptibility of cardiometabolic risk to its increments among Turkish women: a prospective population-based study. Metabolism 2007;56:348–56. _ Dog˘an Y, Tu¨rkmen S, Sansoy V. Sex [29] Onat A, Hergenc¸ G, Kelesx I, difference in development of diabetes and cardiovascular disease on the way from obesity and metabolic syndrome: prospective study of a cohort with normal glucose metabolism. Metabolism 2005;54:800–8.