Review. Regional Distribution of Body Fat, Plasma Lipoproteins, and Cardiovascular Disease

Review Regional Distribution of Body Fat, Plasma Lipoproteins, and Cardiovascular Disease Jean-Pierre Despres, Sital Moorjani, Paul J. Lupien, Angelo Tremblay, Andre Nadeau, and Claude Bouchard Several epldemlologlcal studies have reported that the regional distribution of body fat Is a significant and Independent risk factor for cardiovascular disease (CVD) and related mortality. Although these associations are well established, the causal mechanisms are not fully understood. Numerous studies have, however, shown that specific topographic features of adipose tissue are associated with metabolic complications that are considered as risk factors for CVD such as Insulin resistance, hyperinsullnemia, glucose Intolerance and type II diabetes mellltus, hypertension, and changes In the concentration of plasma llplds and lipoproteins. The present article summarizes the evidence on the metabolic correlates of body fat distribution. Potential mechanisms for the association between body fat distribution, metabolic complications, and CVD are reviewed, with an emphasis on plasma llpoproteln levels and plasma llpld transport From the evidence available, it seems likely that subjects with visceral obesity represent the subgroup of obese Individuals with the highest risk for CVD. Although body fat distribution Is now considered as a more significant risk factor for CVD and related death rate than obesity per se, further research Is clearly needed to Identify the determinants of body fat distribution and the causal mechanisms Involved In the metabolic alterations. It appears certain, however, that an altered plasma llpld transport Is a significant component of the relation between body fat distribution and CVD. (Arteriosclerosis 10:497-511, July/August 1990) There is much scientific evidence to warrant that the regional distribution of body fat is an important consideration in the relationship between obesity, metabolism, and health. Although recent evidence from several prospective studies has demonstrated significant associations between regional body fat distribution and cardiovascular disease (CVD) and related mortality, 1-8 the mechanisms responsible for this relationship have yet to be established. Numerous reports in the literature have confirmed the early clinical observations of Vague 8 and have shown that the topography of adipose tissue (AT) was associated with metabolic complications that are considered to be risk factors for cardiovascular disease From the Physical Activity Sciences Laboratory, Laval University, and the LJpid and Diabetes Research Units, Laval University Medical Center, Ste-Foy, Quebec, Canada. The research of the authors reported in this review received the financial support of the Quebec Heart Foundation, the Medical Research Council of Canada, the National Institutes of Health (U.SA), the Canadian Diabetes Association, the Fonds de la recherche en sarrte du Quebec, (FRSQ), the Natural Sciences and Engineering Research Council of Canada, the Canadian Fitness and Lifestyle Research Institute, and Merck Frosst Canada. Jean-Pierre Despr6s Is an FRSQ scholar. Address for correspondence: Jean-Pierre Despres, Ph.D., Physical Activity Sciences Laboratory, PEPS, Laval University, Ste-Foy, Quebec, Canada G1K 7P4. Received December 19, 1989; revision accepted March 7, 1990. (insulin resistance, hyperinsulinemla and diabetes, hypertension, and changes in the concentration of plasma lipids and lipoproteins). 789 Recently, three symposia have been entirely devoted to the clinical and metabolic aspects of body fat distribution, and the proceedings of these meetings have been published. 789 In September 1989, the National Institutes of Health (U.S.A.) organized a workshop in Bethesda, Maryland, addressing the issue of body fat distribution and its clinical implications. 10 These meetings undoubtedly reflect the increasing interest of the scientific community in the regional distribution of body fat as a significant biological phenomenon with important metabolic consequences and health implications. The aim of this review is to briefly summarize the current data on the metabolic correlates of body fat distribution and to review some potential mechanisms for the association between body fat distribution and metabolic complications, with an emphasis on lipoprotein metabolism. Relation of Regional Body Fat Distribution to Metabolism In the 1940s, Professor Jean Vague of the University of Marseille 6 presented clinical observations suggesting that the topography of AT was an important variable in the relation of obesity to diseases. He suggested that lowerbody obesity (gynoid) had rather minor effects on health, whereas upper-body obesity (android) was associated 497

498 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 with diabetes, gout, and atherosclerosis. More than 30 years later, Kissebah et al. 11 provided the first scientific evidence of a relationship between body fat distribution and glucose tolerance when they reported that upperbody obese women showed a higher plasma glucose response during a glucose tolerance test than women with lower-body obesity. Kissebah et al. showed that insulin resistance and high plasma triglyceride (TG) levels were also correlates of upper-body obesity, and they were the first to propose a metabolic link between regional fat accumulation, regional variation In adipocyte morphology and lipolytic activity, and the metabolic profile. Since this report, the results from five epidemiological studies 1 " 6 have shown that a high proportion of trunk or abdominal fat was associated with an increased probability of developing CVD and with an increased risk of mortality. These concordant results have also contributed to the marked interest of the scientific community in further characterizing the regional distribution of body fat and its association with metabolic events. In this section of the review, we will summarize the literature available on body fat distribution and metabolic complications. Since previous reviews have discussed in detail the associations between fat distribution and glucose-insulin metabolism and hypertension, 78912-17 these aspects will be only briefly discussed. In this review, greater emphasis will be given to the relation of body fat distribution to plasma lipoprotein levels and lipid metabolism. Body Fat Distribution, Insulin Resistance, Hyperlnsullnemla, and Diabetes Four decades ago, Vague 6 reported that women with a "male" pattern of fat distribution (he named it android obesity) were more metabolically affected than obese women with the typical female fat topography (gynoid obesity). Thirty-five years later, Kissebah et al. 11 were the first to show significant associations between the regional distribution of AT and disturbances in glucose-insulin metabolism when they reported a greater incidence of diabetes and glucose intolerance in "upper-body obese" women compared to "lower-body obese" women. In a large survey of 15 532 obese women, it was shown that obese subjects with excessive upper-body fat had a relative risk of diabetes that was 10-fold higher than normal women with peripheral fat accumulation. 18 Several investigators have reported that fat topography was associated with plasma insulin levels and with plasma glucose and insulin responses to a glucose tolerance test. 1920-21 Further analyses from the Gothenburg prospective study also demonstrated that the localization of body fat could predict the development of diabetes over a 13-year period. 22 These results cleariy demonstrated that body fat distribution is more important than obesity per se with regard to its effect on glucose and insulin metabolism. Thus, the link between fat distribution and diabetes could be considered as one of the mechanisms by which body fat topography is associated with CVD. The studies described above have, however, measured the regional distribution of body fat by using anthropometric measurements such as skinfolds and the waist-to-hip circumference ratio. Computed tomography (CT) is an alternative approach that allows the precise measurement of AT surface areas at any site of the body, 23 " 30 and particularly the delineation of the amounts of deep (or visceral) and subcutaneous fat It is presently not possible to obtain such information from anthropometric measurements. Using CT, Sparrow et al. 31 reported that the amount of intra-abdominal fat was significantly correlated with the concentration of plasma glucose at 2 hours after glucose ingestion. Accordingly, Fujioka et al. 32 reported that glucose tolerance was negatively correlated with the proportion of intraabdominal fat. In the latter study, the sample was, however, very heterogeneous for age, and it included both women and men as well as diabetic patients. To further investigate the associations between deep abdominal fat accumulation and glucose metabolism, we have studied a homogeneous sample of 52 obese nondiabetic premenopausal women. 33 Our results not only showed a significant association between the level of deep abdominal fat and glucose tolerance, but also that this relationship was independent from the level of obesity in these obese women. It has also been shown that regional adiposity indices that included CT-measured levels of deep abdominal fat had independent and additive contributions to alterations in glucose insulin homeostasis. 34 We have also reported that subcutaneous abdominal fat accumulation and abdominal fat cell hypertrophy were two conditions associated with a hypersecretion of insulin, whereas the level of deep abdominal fat was the best correlate of glucose tolerance. It therefore appears from these studies that various body fat distribution indicators display different levels of association with various indices of carbohydrate metabolism. Mechanisms responsible for the associations between regional body fat distribution, intra-abdominal fat accumulation, and glucose and insulin metabolism have been studied. Kinetic studies have shown that obesity was associated with an increased pancreatic insulin production, whereas a high proportion of abdominal fat was associated with a reduced metabolic clearance of insulin, due to a diminished hepatic insulin extraction. 3838 A high proportion of abdominal fat is also associated with a decreased insulinstimulated glucose disposal. 3637 The mechanism for the association between body fat distribution and hepatic insulin extraction has not been fully elucidated, but it has been suggested that the concentration of plasma free fatty acids (FFA) could play an important role. It has been reported that abdominal-obese women have a greater basal FFA turnover than peripheral-obese women. 38 Also, high FFA concentrations are associated with a reduction of glucose utilization through the glucose-fatty acid cyde, 3940 leading to glucose intolerance and to systemic hyperinsulinemia. Two recent studies 4142 have reported that a high concentration of FFA inhibits the binding and uptake of insulin in hepatocyte preparations. Thus, the high lipolytic activity of deep abdominal adipose cells, 43 combined with their resistance to the antilipolytic action of insulin, 44 could be conditions that would contribute to the exposure of the liver to high FFA concentrations through the portal circulation, leading to a reduction in hepatic insulin extraction. With increasing

REGIONAL BODY FAT DISTRIBUTION AND LIPOPROTEINS Despres et al. 499 levels of total body fat, men generally accumulate more abdominal and visceral fat than women. 2628 We had previously suggested that this sex dimorphism in body fat topography could be responsible for the fact that obesity is more readily associated with metabolic complications and CVD risk in men than in women. 454647 Because steroid hormones are important determinants of the sex dimorphism in body fat distribution, their associations with body fat distribution and metabolic complications will be reviewed separately in this paper. Fat Distribution and Blood Pressure Numerous studies have reported associations between body fat localization and hypertension. 1920 ' 48-53 Indeed, Kalkhoff et al. 20 reported a significant association between the waist-to-hip circumference ratio and blood pressure. It has also been reported that central adiposity has more detrimental effects on blood pressure than peripheral fat accumulation. 48 Studies have also dealt with the respective contributions of body fat and fat distribution to variation in blood pressure, and an independent effect of fat localization on blood pressure has been reported. 195152 Similarly, when we controlled for the effects of biologic and lifestyle variables known to affect blood pressure, we observed a significant effect of trunk fat on diastolic blood pressure in men. 53 Although all these studies showed an effect of fat distribution on blood pressure, the mechanisms by which these variables are linked remain unclear. There is increasing evidence that variations in plasma insulin levels, which have been shown to be associated with body fat topography, could play a role in the regulation of blood pressure. 54-58 A model integrating obesity, body fat distribution, insulin levels, the activity of the sympathetic nervous system, and blood pressure has already been proposed. 59 Although additional research is warranted to elucidate the mechanisms by which body fat distribution is associated with hypertension, there is enough evidence to show that high blood pressure undoubtedly contributes to the increased risk of CVD in subjects with truncal-abdominal obesity. Body Fat Distribution and Plasma Lipoprotein-Lipids Disturbances in glucose homeostasis and increased susceptibility to diabetes and hypertension are not the only factors contributing to the association between fat distribution and CVD. Early studies reported that excess trunk fat was associated with high plasma TG levels. 6061 More recently, others have also shown that abdominal obesity was related to high plasma TG levels. 11. 1921 We have reported 45 that the regional distribution of body fat, especially an excessive deposition of abdominal fat, was associated with low concentrations of serum high density lipoprotein (HDL) cholesterol. We also showed that, since men generally have more abdominal fat than women, the association between obesity and plasma HDL cholesterol level was stronger in men than in women. 45 We further studied the associations between obesity, AT distribution, and serum HDL cholesterol levels in a sample of 429 men, and our results demonstrated that the effect of obesity on serum HDL cholesterol levels was no longer significant after controlling for abdominal fat deposition, 62 thereby indicating that the association that had been reported in the past between obesity and serum HDL cholesterol levels was mainly explained by the amount of abdominal fat. These findings on the association between abdominal fat and serum HDL cholesterol levels have since been confirmed by others. 1363-75 However, all these studies had measured the distribution of body fat by anthropometric measurements such as skinfolds and circumferences. Since it had been suggested that intra-abdominal fat was important in the etiology of the metabolic complications of obesity, 3132 further work was required to document the associations between deep abdominal fat deposition and lipoprotein levels and metabolism. Since CT accurately measures the amount of intraabdominal fat, we have used this technique to study the associations between body fat localization and plasma lipoprotein levels in a sample of 52 premenopausal obese women. Our results 76 clearly showed that a high proportion of intra-abdominal fat was associated with a reduction in the concentration of HDL cholesterol and with significant reductions in lipoprotein ratios (HDL cholesterol/low density lipoprotein [LDL] cholesterol, HDL apoprotein A-I/LDL apoprotein B, HDL 2 cholesterol/hdu cholesterol) that are frequently used in the estimation of the CVD risk (Table 1). Furthermore, these associations between intraabdominal fat accumulation and plasma lipoprotein ratios were independent from the effect of obesity. 76 Peiris et al. 77 simultaneously reported similar findings suggesting that visceral fat deposition may represent a greater risk factor for CVD than obesity per se. Thus, in a sample of premenopausal women with varying levels of body fatness, it was reported that the amount of intra-abdominal fat measured by CT explained a higher proportion of the variance in the HDL cholesterol/cholesterol ratio than body fat mass per se. 77 Since dyslipoproteinemic states have been associated with the development of CVD, the relation of fat distribution to plasma lipids and lipoproteins could be an additional factor linking abdominal fat to CVD. Therefore, the measurement of the level of intraabdominal fat becomes very important in the estimation of the CVD risk due to its independent association with alterations in glucose insulin homeostasis 31-34 and plasma lipoprotein levels. 7677 As men generally possess a higher relative accumulation of visceral fat than women, 2628 the gender difference in this depot could be a major factor in the sex difference noted in the magnitude of the association between the level of fatness and metabolic complications. As for diabetes, the mechanisms relating fat distribution to plasma lipoprotein physiology are poorly understood, but studies conducted in different laboratories, including ours, have allowed the identification of some factors potentially involved in the regional fat distribution-lipoprotein association. Mechanisms Involved in the Association between Body Fat Distribution and Plasma Lipoproteins Changes in plasma lipoprotein concentration and composition and in plasma lipid transport have been associated

500 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 Table 1. Correlation Coefficients between Body Fat Mass, Walst-to-HIp Ratio, Abdominal Fat Areas, and Metabolic Variables Variable Upoproteins VLDLTQ (log 10) LDL CHOL LDLapo B HDL CHOL HDU CHOL HDL, CHOL Upoprotein ratios HDL CHOL/LDL CHOL HDL A-I/LDL B HDLj CHOL/HDU CHOL LDL B/LDL CHOL Glucose tolerance test Glucose area Insulin area C-peptide area Fat mass 0.05 0.10 0.10-0.25-0.25-0.20-0.24-0.21-0.19-0.04 0.41t 0.47* 0.42t Walst-tc-hlp ratio 0.47* 0.10 0.27-0.47* -0.431- -0.44t -0.34* -0.33* -0.24 0.35t 0.44* 0.25 0.34t Abdominal fat* Subcutaneous 0.06 0.07 0.11-0.22-0.22-0.17-0.19 0.16-0.17 0.06 0.16 0.44* 0.29* Deep 0.24 0.23 0.26-0.35t -0.37t -0.27-0.40t -0.37t -0.32* -0.02 0.57 0.49* 0.52* Abdominal fat was measured by computed tomography in 52 obese premenopausal women. *p<0.05, tp<0.01, *p<0.001, p<0.0001. This table was adapted from references 33 and 76. VLDL=very low density Upoprotein, TG-trlglycerides, LDL=low density Upoprotein, CHOL-cholesterol, apo-apollpoprotein, HDL-hlgh density lipoproteln. with variations in the regional distribution of body fat This section will summarize the literature on this issue. When possible, underlying mechanisms will be discussed, and possible avenues for further research will be suggested. Role of Upoprotein Upases TG lipases (lipoproteln lipase [LPL] and hepatic-tg lipase [H-TGL]) play important roles in lipoproteln metabolism. In numerous clinical and experimental conditions, a high plasma postheparin LPL activity has been associated with high plasma HDL levels, 787980 whereas a high plasma postheparin H-TGL activity has been associated wtth low plasma HDL concentrations. 8182 We have studied the relationships between body fat distribution, plasma postheparin LPL and H-TGL activities, and plasma Upoprotein levels in a sample of 16 premenopausal obese women 83 (Table 2). Plasma postheparin LPL activity was correlated neither with total adiposity nor with the level of intra-abdominal fat. Intra-abdominal fat deposition was, however, positively correlated with H-TGL activity (r=0.66, p<0.005). Furthermore, covariance analysis showed that the association between intra-abdominal fat and H-TGL was independent from total adiposity. Plasma postheparin LPL (Table 2) and abdominal AT-LPL (results not shown) activities showed no significant correlation with plasma lipoprotein levels. We found, however, a significant association between femoral AT-LPL activity and the HDU cholesterol/hdu cholesterol ratio (r=0.51, p<0.05; result not shown in Table 2). H-TGL activity was negatively correlated with HDU cholesterol (r=-0.60, p<0.05), but not with HDL 3 cholesterol (r=-0.28, NS). Thus, these results suggested that the high H-TGL activity in obese Table 2. Correlations between Plasma Postheparin Upoprotein Upase and Hepatic Trlglycerlde Upase Activities, and Body Fatness, Fat Distribution, and Plasma High Density Upoprotein Cholesterol Levels Variable % Body fat Walst-tc-hlp ratio CT abdominal fat areas Subcutaneous Deep HDLCHOL HDU CHOL HDLjCHOL HDU CHOL/HDU CHOL Upoprotein lipase -0.33-0.20-0.06-0.29 0.37 0.29 0.40-0.02 Hepatic triglyceride lipase 0.55* 0.49 0.33 0.66t -0.47-0.60* -0.28-0.62* Sixteen obese premenopausal women were examined. This table was adapted from reference 83. *p<0.05, tp<0.01. CT=computed tomography. See the legend to Table 1 for other abbreviations. women with excess deep abdominal fat could be responsible for the reduction in plasma HDL 2 cholesterol levels observed in these subjects. 83 There is, however, no reason to believe that the accumulation of deep abdominal fat could be directly involved in the variation of H-TGL activity. As both viscera) fat accumulation and H-TGL are, however, sensitive to sex steroids, it had been suggested that an increased exposure to androgens could produce simultaneous increases in the level of abdominal fat and In H-TGL activity. 13 Our study confirms that, indeed, a

REGIONAL BODY FAT DISTRIBUTION AND UPOPROTEINS Despr6s et al. 501 Table 3. Full and Partial Correlations between Body Fat Mass, Proportion of Abdominal Fat, Plasma VLOL TG and HDL CHOL Levels, and Upoproteln-Upld Composition Full correlations Partial correlations Variable VLDLTG HDL CHOL LDL CHOL/LDL TQ HDL CHOL/HDL TG Fat mass 0.56* -0.47* -0.42* -0.27* Walst-to-hlp ratio 0.49* -0.45* -0.29* -0.46* Fat mass (adjusted for WHR) 0.33f -0.27* -0.31* -0.01 A group of 76 premenopausal women was examined. This table was adapted from reference 85. *p<0.05, tp<0.01, *p<0.001. WHR=waist-to-hlp ratio. See the legend to Table 1 for other abbreviations. Waist-to-hip ratio (adjusted for fat mass) 0.34t -0.30* -0.09-0.41* high intra-abdominaj fat deposition is associated with a high H-TGL activity and with low HDL S cholesterol levels. In our sample of obese women, we reported that plasma postheparin LPL activity was not related to body fat distribution. To further investigate the potential associations between body fatness, fat distribution, and plasma postheparin LPL activity, we have studied a sample of 29 premenopausal women with body fatness levels ranging from lean to massively obese. 84 In this sample, LPL activity was negatively correlated (r=-0.45, p<0.05) with the level of total body fat. In addition, a low LPL activity was associated with high plasma TG levels and with low HDL cholesterol levels. These results suggest that obesity per se is associated with a reduction in plasma postheparin LPL activity, whereas obesity combined with a high level of visceral fat is associated with a high H-TGL activity leading to reduced HDL cholesterol levels. These results may explain the reported independent and additive effects of obesity and of fat distribution on the concentration of plasma TG and HDL cholesterol. Role of Plasma Thglyceride Levels Our observation that plasma TG levels were not associated with intra-abdominal fat deposition per se in a sample composed of obese subjects alone led to further investigations to measure the independent effects of obesity and of the proportion of abdominal fat on plasma lipoprotein concentration and llpid composition in a sample of women with wide variations in their levels of total body fat. 86 Indeed, obese women had significantly higher plasma levels of very low density lipoprotein (VLDL)-TG, LDL cholesterol, LDL-TG, LDL apolipoprotein (apo) B, and reduced HDL cholesterol levels compared to nonobese controls (p<0.01). However, plasma concentrations of HDL apo A-l and HDL-TG were not different between obese and nonobese women. 86 Partial correlation analyses revealed that both fat mass and abdominal fat accumulation significantly contributed to VLDL-TG and HDL cholesterol variances (Table 3). After control for body fat mass, the waist-to-hip circumference ratio (WHR) remained significantly correlated with the LDL apo B/LDL cholesterol ratio (r=0.3q, p<0.01). M Body fat mass was, however, associated with TG enrichment of LDL (p<0.01). After control for WHR, body fat mass showed no significant association with the relative TG content of HDL, whereas the WHR was associated with a relative TG enrichment of HDL, as reflected by the HDL cholesterol/tg ratio. Partial correlation analyses presented in Table 4 Indicate that adjustment for fat mass or for the WHR failed to eliminate the associations between plasma VLDL-TG levels and lipoprotein lipld composition. These results emphasize the importance of plasma VLDL-TG level as a correlate of plasma LDL and HDL lipld composition in abdominal obesity. It therefore appears that obesity and abdominal fat accumulation (irrespective of the subcutaneous or visceral accumulation of fat) are associated with hypertriglyceridemia and that high plasma TG levels are associated with a TG enrichment of LDL and HDL. 88 It is very likely that the high plasma TG levels observed in abdominal obesity represent an indirect risk for CVD, as they reflect disturbances in plasma lipid transport such as a reduced catabolism of TG-rich Iipoprotein8 and a greater lipid exchange among llpoproteins leading to TG enrichment of LDL and HDL 86 This condition is associated with higher concentrations of dense, small LDL and with low HDL cholesterol levels in the plasma. It has been suggested that an altered composition of LDL, namely an increased ratio of LDL apo B to LDL cholesterol, was associated with an increased CVD risk. 86 Indeed, many patients with coronary heart disease have high concentrations of smaller and denser LDL particles that contained less cholesterol than normal. 87 Furthermore, as hypertriglyceridemic LDL contains less cholesterol, cellular cholesterol synthesis is Inhibited to a lesser extent by dense LDL. 88 In addition, because dense LDL particles have a reduced cholesterol content, less cholesterol is cleared from the plasma through the LDL receptor pathway, and a higher proportion of cholesterol is transported by VLDL in subjects with high TG levels, in comparison with individuals with normal plasma lipid transport. 88 Such cholesterol transported by VLDL may be cleared from the plasma by potentially atherogenic routes such as macrophages and smooth muscle cells. 89 The reduction in plasma HDL cholesterol observed in subjects with high plasma TG levels may first be the consequence of an increased exchange of lipids among lipoproteins through the action of lipid transfer proteins. 9091 ' 82 The negative association between plasma TG and HDL cholesterol levels may also be attributed to a reduced catabolism of TG-rich lipoproteins, caused by a

502 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 Table 4. Correlations between Plasma VLDL Trlglyceride Concentration and Plasma Upoproteln Levels and Upid Composition without and with Adjustment for Fat Mass or for Waist-to- Hlp Ratio Effect of Obesity Per Se on Plasma Upoprotelns Most observations that have been reported so far on body fat distribution and metabolism have been derived from samples of obese individuals or samples including obese individuals. Landin et al. 95 have recently studied the associations between the proportion of abdominal fat estimated by the WHR and metabolism in lean and obese postmenopausal women. Whereas they reported significant associations between the WHR and alterations in metabolism in obese women, no relationship of fat distribution to metabolic complications was observed in lean women. These results suggested that obesity is a necessary condition for the documented associations between fat distribution and metabolic alterations. To further dissociate the effects of obesity from those related to fat distribution, we have studied the potential associations between fat distribution, regional variation in fat cell size, and plasma lipoprotein levels in a sample of 22 lean women with a mean body mass index (BMI) of 21 kg/m 2 " (Table 5). In this sample, very few significant associations were observed between the WHR, intra-abdominal fat accumulation, and plasma lipoprotein levels. However, we observed in these nonobese women that a high trunkabdominal fat accumulation measured by CT was associated with increased LDL cholesterol (r=0.56,p<0.01) and LDL apo B levels (r=0.70, p<0.01). Furthermore, abdom- Variable LDL CHOL LDLTG LOLapo B LDL apo B/LDL CHOL LDL CHOL/LDL TG HDL CHOL HDLTG HDL apo A-l HDL apo A-I/HDL CHOL HDL CHOL/HDL TG Full correlations VLDLTG 0.34t 0.664 0.45* 0.32f -0.45* -0.59* 0.35t -0.15 0.58* -0.59* Partial correlations VLDLTG (adjusted for fat mass) 0.11 0.54* 0.32t 0.49* -0.30* -0.48* 0.46* -0.07 0.51* -0.56* A sample of 76 premenopausal women was studied. From reference 85, reproduced with the permission of Clinical and Investigative Medicine. *p<0.05, tp<0.01; *p<0.001. decreased LPL activity. In numerous experimental conditions, it has been reported that the plasma HDU cholesterol concentration was positively correlated with plasma LPL activity and with the TG clearance rate. 81 Thus, HDU levels could be markers of disturbances in lipoproteln metabolism without having a direct role in the atherosclerotic process. In our study, plasma VLDL-TG levels were negatively correlated with HDU cholesterol (r=-0.52, p<0.001) and HDL3 cholesterol (r=-0.51, p<0.001) levels, and to a lesser extent with the HDL 2 cholesterol/hdla cholesterol ratio (r=-0.33, p<0.01). Our results suggest that the high VLDL-TG levels observed in abdominal obesity are associated with a reduced cholesterol content of both HDL subtractions (due to a TG enrichment) and also with a reduced conversion of HDL3 to HDL2 during the intravascular lipotytjc process. These metabolic alterations indicate disturbances in plasma lipid transport that are associated with an increased risk of cardiovascular complications. VLDLTG (adjusted for WHR) 0.16 0.59* 0.29* 0.33t -0.36+ -0.47* 0.33+ -0.03 0.51* -0.47* Role of Carbohydrate Metabolism Because insulin resistance and glucose intolerance are associated with abdominal obesity, we have investigated the potential role of alterations in carbohydrate metabolism in mediating the effects of fat distribution on lipoprotein levels. Numerous studies have indicated that plasma lipid transport and lipoprotein levels are altered in diabetes, and the evidence for associations between glucose tolerance, insulin resistance, plasma insulin levels, and plasma lipoproteins is abundant (for reviews, see references 14, 17, 93). It is thus relevant to verify to what extent the alterations in plasma lipid transport observed in abdominal obesity are related to changes in glucoseinsulin homeostasis. In a previous study in men, we found that, during an oral glucose tolerance test (OGTT), the proportion of trunk fat was positively correlated with the plasma insulin response, as well as with plasma LDL cholesterol and LDL apo B levels, suggesting that these variables are all interrelated. 04 In women, we reported that the hyperinsulinemia associated with a high proportion of abdominal fat was a significant correlate of plasma TG and VLDL levels, but that other mechanisms explained the negative association between abdominal obesity, visceral obesity, and plasma HDL cholesterol levels (Feriand et a), unpublished observations). In the San Antonio Study, it was found that insulinemia and glycemia could mediate the association between abdominal fat accumulation and TG levels, but that the relation of fat distribution to HDL and HDL 2 cholesterol levels was independent from variations in plasma glucose and insulin concentrations. 72

REGIONAL BODY FAT DISTRIBUTION AND LIPOPROTEINS Despres et al. 503 Table 5. Correlations between Fat Mass, Waist-to-HIp Ratio, Deep Abdominal Fat, and Plasma Upoprotelns in Lean Women Variable VLDLTG LDL CHOL LDL apo B HDL CHOL HDU CHOL HDL3 CHOL at mass -0.29 0.46* 0.51* -0.24-0.26-0.10 A sample of 22 lean menopause! women was studied. This table was adapted from reference 96. *p<0.05, tp<0.01. See the legend to Table 1 for an explanation of the abbreviations. Walst-to-hip ratio -0.20 0.21 0.36 0.03 0.05 0.16 CT-measured deep abdominal fat -0.11 0.38 0.52t -0.18-0.27 0.03 inal fat cell weight was positively correlated with the LDL apo B/LDL cholesterol ratio (r=0.58, p<0.01) and negatively correlated with plasma HDL apoa-l (r=-0.51, p<0.05) and HDU cholesterol levels (r=-0.51, p<0.05). On the other hand, femoral fat accumulation and femoral fat cell weight showed no association with plasma lipoprotein levels. 96 These results show that in nonobese premenopausal women, subcutaneous trunk-abdominal fat accumulation and abdominal fat cell hypertrophy are related to changes in plasma lipoprotein levels, whereas the WHR and the level of deep abdominal fat show little association with the metabolic profile of lean women. We have also compared obese women with similar levels of total body fat who markedly differ in their level of intra-abdominal fat to nonobese women (Despres et al., unpublished observations). This approach with three distinct groups (25 nonobese women, 10 obese women with low levels of Intra-abdominal fat, and 10 obese women with high levels of intra-abdominal fat) has allowed us to dissociate the independent effects of obesity and of intra-abdominal fat accumulation on risk factors for CVD (Table 6). Although obese women with low levels of intra-abdominal fat had much higher adiposity than nonobese controls, increased plasma TG and insulin levels were the only two metabolic alterations observed in these obese women. In contrast, obese women with high levels of intra-abdominal fat displayed numerous metabolic complications compared to nonobese women, including increased plasma TG, cholesterol, LDL cholesterol, LDL apo B, glucose and insulin areas under the curve measured during an OGTT, reduced HDL cholesterol levels, and reduced ratios of HDL cholesterol/ldl cholesterol and HDL apo A-I/LDL apo B. Therefore, despite comparable levels of obesity In the two obese groups, obese women with low levels of intra-abdominal fat had fewer metabolic alterations than obese women with visceral obesity (Despres et al., unpublished observations). Results from these studies emphasize two notions. First, since no significant correlation was observed between deep abdominal fat deposition and plasma HDL cholesterol levels in nonobese women, these results indicate that, at least in premenopausal women, obesity has to be present for there to be significant associations between intra-abdominal fat accumulation and plasma lipoprotein concentrations. However, among obese subjects, the level of intra-abdominal fat is a more important correlate of a potentially atherogenic profile than obesity per se. It should be emphasized, however, that these notions have not been verified in men, as the association between fatness and visceral fat deposition is greater in men than in women. It is possible that obesity is not a prerequisite for the metabolic complications of visceral fat accumulation in men. Effects of Sex Steroids and Glucocortlcolds Several studies performed by Kissebah and colleagues 13219798 have suggested that the association between abdominal obesity and metabolic complications might not represent a cause-and-effect relationship but could rather reflect the influence of sex steroids on both body fat distribution and metabolism. The association between androgenic activity, abdominal fat accumulation, and metabolism was investigated in a sample of 80 premenopausaj women with adiposity levels ranging from lean to massively obese. 97 Although a high androgenic activity was associated with both an increased deposition of abdominal fat and a decreased insulin sensitivity, 97 it has been found that, although of lower magnitude, the correlations between body fat distribution and alterations in glucose metabolism remained significant after controlling for steroid hormone levels. 21 After adjustment for the effect of abdominal fat accumulation as estimated by the WHR, no significant correlation remained between androgenic activity and metabolic indices, suggesting that the relationship was largely mediated by the regional accumulation of body fat. 21 Recent results have indicated that the systemic hyperinsulinemia associated with abdominal obesity was partjy attributed to a reduced hepatic extraction of insulin. 35 Sex hormone binding globulin (SHBG) level was positively correlated with the hepatic insulin extraction, whereas the percentage of free testosterone was negatively correlated with the hepatic insulin extraction. 98 After correction for the effects of percent free testosterone and of SHBG, correlations between the proportion of abdominal fat (as estimated by the WHR) and the hepatic extraction of insulin remained significant suggesting that body fat distribution could, per se, affect insulin metabolism

504 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 Table 6. Metabolic Characteristics of Nonobese Women, and of Obese Women with Low or High Levels of Deep Abdominal Fat Variable % Body fat Deep abdominal fat area (cm 2 ) TQ (mmol/l) CHOL (mmol/l) LDL CHOL (mmol/l) LDL apo B (mg/dl) HDL CHOL (mmol/l) HDL apo A-l (mg/dl) HDL CHOL/LDL CHOL HDL apo A-I/LDL apo B Fasting Insulin (pmol/l) Glucose area ([mmol/l/180 mlnlxio" 3 ) Insulin area ([pmol/1/180 mlnjxio" 3 ) Nonobese (n-25) 28.0±5.6 50.3±16.8 0.79±0.35 4.59±0.88 3.00±0.87 68.4±23.2 1.36±0.24 116.8±13.5 0.50±0.23 1.90 ±0.74 39.0±12.4 1.07±0.19 46.6±19.4 Obese: level of deep abdominal fat Low (n=10) High (n-10) 47.0±6.4* 49.8±3.2* 107.0±33.4* 186.7±36.8*,t 1.47 ±0.79* 2.57±2.41*,t 5.18+0.93 5.65±1.23* 3.56±0.92 3.81±1.10* 77.8±21.8 90.9±24.7* 1.25±0.18 0.96±0.12*,t 126.5±24.1 105.6±15.0t 0.37±0.09 0.27±0.07* 1.71 ±0.47 1.22±0.29* 91.5±59.9* 150.3±57.3*,t 1.14±0.22 1.40±0.19*,t 82.1 ±48.3* 121.0±39.5*,t The values represent means ± SD. The number of subjects Is shown In parentheses. Triglyceride data were log 10 transformed for statistical analysis. Significantly different from nonobese women, tslgnificantty different from obese women with low levels of deep abdominal fat, p<0.05. TG=triglycerides, CHOL=cholesterol, LDL«low density lipoprotein, HDL=high density lipoproteln, apo-apoproteln. Glucose and Insulin areas represent Integrated plasma concentrations for 3 hours after oral glucose (75 g) Ingestion. From Depris et al., unpublished observations. through a mechanism partty independent from variations in sex hormone levels. 98 Numerous studies have also shown that sex steroids have significant effects on lipoprotein metabolism. 82 In a sample of 73 healthy men, Stefanick et al. 98 have shown that plasma estradiol levels were correlated negatively with total cholesterol as well as with LDL cholesterol, whereas testosterone levels were positively correlated with apo B levels. SHBG correlated positively with HDL 2 mass determined by analytical centrifugation and with plasma apo A-l levels. However, after adjustment for the amount of abdominal fat, as estimated by the WHR, no significant correlations remained between SHBG and TG, HDLj mass, and apo A-l. These results suggest that some of the effects of sex steroids on plasma lipoprotein levels are mediated by variations in regional body fat distribution. Soler et al. 75 also studied the associations between abdominal adiposity (estimated by the WHR), fasting insulin and sex hormone levels, and plasma lipoprotein levels in a sample of 75 postmenopausal women. They reported that, after adjustment for insulin, SHBG, and estrone level, abdominal obesity was still significantly associated with plasma TG, HDL cholesterol, and apo A-l and B levels. 75 These results also support the notion that mechanisms other than sex hormone levels are involved in the fat distributjon-lipoprotein association. There is, however, experimental evidence indicating that sex hormones alter lipoprotein metabolism through additional mechanisms. As indicated in a recent review, 82 sex steroids produce changes in plasma HDL 2 cholesterol concentration by up- or down-regulating the H-TGL activity, as steroids with androgenic activity increase H-TGL activity and reduce plasma HDU cholesterol levels. Administration of an androgenic steroid, stanozolol, was associated with an increase in H-TGL activity that preceded the decrease in plasma HDL 2 cholesterol levels, suggesting a cause-and-effect relationship between the enzyme and plasma HDU cholesterol levels. 100 In summary, it appears that sex steroid levels, through regulation of H-TGL activity, play an important role in the regulation of plasma HDL cholesterol levels in abdominal obesity, particularly in situations of visceral fat accumulation. We believe that variations in sex steroid levels may contribute to the association that we have recently reported between intra-abdominal fat deposition, H-TGL activity, and plasma HDU cholesterol levels. 83 Mechanisms responsible for these associations dearly require further investigation. In addition to sex steroid hormones, glucocorticoids have also been proposed as a factor involved in the body fat distribution-metabolism association. It is well known that Cushing's disease is associated with a preferential accumulation of abdominal fat and with body fat distribution-related metabolic complications such as insulin resistance, hypertension, and hyperiipidemia. 101 These results suggest a significant role for cortisol in the control of regional fat accumulation and carbohydrate and lipid metabolism. A higher cortisol production rate was reported in subjects with abdominal obesity compared to individuals with peripheral fat accumulation. 102103 A significant positive correlation between the cortisol production rate and the plasma insulin response during an OGTT has also been reported. 102 The effects of glucocorticoids on lipid metabolism have been recently reviewed. 104 Glucocorticoids stimulate hepatic VLDL secretion 10 *- 107 and decrease the binding and degradation of LDL by rat hepatocytes. 108109 The high cortisol levels frequently observed in abdominal obesity could

REGIONAL BODY FAT DISTRIBUTION AND LIPOPROTEINS Despres et al. 505 lead to an increased VLDL production and to a decreased clearance of LDL through the apo B,E receptor. Clearly, more research is warranted on the interaction between body fat distribution, glucocorticoids, and metabolism in humans. It is, however, very likely that glucocorticoids play a significant role in the metabolic complications associated with abdominal obesity. Role of Regional Variation In Adipose Tissue Metabolism There are well-established regional differences in AT metabolism. 110 It has been suggested that the high lipolytic activity of abdominal, 43111112113 especially irrtraabdominal fat cells, 43114 could stimulate hepatic TG synthesis by exposing the liver to high FFA concentrations. 12-17116 In addition, it has been reported that the retroperitoneal fat depot has metabolic characteristics similar to those of subcutaneous abdominal fat, whereas the adipose depots drained by the portal vein show differences in metabolic activity in comparison with the nonportal adipose depots. 116 Additional in vitro studies are, therefore, needed to further address the issue of the role of variation in the metabolic activities of the various deep abdominal adipose depots in the etiology of the metabolic complications of visceral obesity. Direct experimental evidence for this highly plausible model is, however, lacking. In vivo studies that we have conducted have shown that fasting plasma FFA levels measured in the morning were significant correlates of glucose tolerance 118 and of VLDL-TG content. 117 An obvious limitation of these studies is that systemic, rather than portal, plasma FFA levels were measured. We have also observed that the lipolytic response of subcutaneous abdominal fat cells to 0-adrenergic stimulation was positively correlated with plasma insulin and TG levels, 118 suggesting that abdominal adipose cell lipolysis may be etiologicalty involved in the metabolic complications of abdominal obesity. No data are, however, available on the potential associations between omental fat cell lipolysis and in vivo carbohydrate and lipid metabolism. In addition, human fat cells specifically bind HDL, 119 and regional variation in HDL binding to human adipose cells has been reported. 120121 Furthermore, the HDL binding to isolated human abdominal adipose cells is proportionate to fat cell size, 121 and it is increased in obesity. 122 It has also been reported that the binding of HDL to the fat cell plasma membrane leads to a disproportionate uptake of HDL cholesterol ester by the hypertrophled adipose cell. 123 Such a high uptake of cholesterol ester by large abdominal fat cells may contribute to the reduction in plasma HDL cholesterol levels observed in abdominal obesity. No information is, however, available on the cellular HDL metabolism of femoral adipose cells, and further studies are needed to address the role of the regional variation in adipose cell HDL metabolism in the reduction in plasma HDL cholesterol levels observed in abdominal obesity. Weight loss has, however, reportedly decreased HDL binding to abdominal adipose cell plasma membranes and increased plasma HDL cholesterol levels. 124 Whether these reciprocal changes represent cause-and-effect relationships or simultaneous variations will require further experimentation. Finally, the response of AT-LPL to insulin is reportedly blunted in obesity. 120 Preliminary results suggest that there is no regional variation in the AT-LPL response to insulin. 126 We have recently reported that, whereas abdominal AT-LPL activity was not associated with plasma HDL levels, the LPL activity measured in the femoral adipose tissue was positively correlated with the plasma HDLj cholesterol/hdlg cholesterol ratio, 8384 suggesting that a high femoral fat accumulation and LPL activity could contribute to a generation of more HDL particles in individuals with peripheral obesity. Further studies are needed to document the potential associations between fat distribution, regional variation in AT-LPL activity, and plasma lipoprotein levels. Classification of Obesities and Cardiovascular Risk The genetic, morphological, and clinical evidences support the notion that obesity is a heterogeneous phenotype. There is some confusion in the fat distribution literature regarding measurements and indices used to assess regional fat distribution. Subcutaneous skinfolds, skinfold ratios, circumferences, or circumference ratios (including WHR) have been used and, more recently, CT has been used to distinguish between measurements of subcutaneous and deep fat accumulation at any site of the body. With regard to the commonly used WHR, we do not know exactty what it represents, and this measurement cannot be clearly interpreted, because two variables are involved. The WHR is generally correlated with age, with the level of total body fat, and with the level of visceral fat. It is, therefore, not surprising that the WHR has emerged as a significant correlate of the metabolic complications of obesity. Important limitations of the WHR are that a high ratio may be obtained in an individual with a small hip circumference or a low ratio in an individual with a large hip circumference, for the same abdominal girth. These two individuals may not necessarily differ with regard to their absolute amount of abdominal fat and in their risk profile, but it may be falsely concluded that they do on the basis of different WHR values. In addition, the WHR is not necessarily responsive to weight loss or weight gain, whereas the waist circumference is obviously so. The significant correlations between total body fat, visceral fat, and the WHR is also a phenomenon that should be carefully considered. The relationship is too low to allow the use of WHR as a surrogate for the other two fatness variables. However, the interrelationships should always be controlled when the associations between total body fat, body fat distribution assessed by the WHR, and metabolism are studied. We believe that the practice of using the WHR as an independent indicator of regional fat distribution should be abandoned in favor of more direct and absolute measurements of truncal-abdominal fat, such as those provided by truncal and abdominal skinfolds or by the waist circumference. CT or nuclear magnetic resonance

506 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 clearly remain the best techniques available to measure the amount of visceral fat. From the evidence available, obese individuals can be classified into two main groups: 1) gluteal-femoral-obese individuals, and 2) abdominal-obese individuals. However, it has been recently suggested that abdominalobese individuals may be further divided into two subgroups displaying different metabolic profiles that may be suggestive of variable risks of CVD. 10127 Gluteal-femoral obesity is characterized by mild insulin resistance, hyperinsulinemia, and increased plasma TG levels, due to a reduced plasma postheparin LPL activity and/or to an increased VLDL production. The first subgroup of abdominal-obese individuals would be composed of obese subjects with high levels of subcutaneous trunkabdominal fat (trunk-abdominal obesity). Evidence for a role of upper-trunk fat per se in the metabolic complications associated with body fat distribution is primarily derived from studies that have shown a significant relationship between the level of trunk fat and metabolic variables after control for the level of total body fat and the WHR. 6574 Metabolic complications observed in this group have been associated with a high proportion of upper-trunk or abdominal fat (as estimated by skinfolds or by circumferences) but not preferentially associated with an accumulation of visceral fat. These complications would include further insulin resistance, leading to a decrease in plasma postheparin LPL activity and to a further increase in plasma TG levels. HDL is enriched with TG, leading to a reduction in the cholesterol content of HDL Increased levels of small (dense) LDL are also observed. The proportion of abdominal fat estimated by the WHR ratio has been positively correlated with the concentration of denser small LDL mass determined by analytic ultracentrifugation. 74 Using subfractionation by equilibrium density ultracentrifugation, others have also reported that a high WHR ratio was associated with heavy and dense LDL as compared to subjects with low WHR. 12a Whether these compositional changes are preferentially associated with deep or subcutaneous abdominal fat is presently not clear. On the other hand, subcutaneous trunk fat accumulation was shown to be negatively correlated with the large LDL mass, and this association was independent from the level of obesity and from the WHR. 74 These results suggest that subcutaneous abdominal-trunk fat accumulation may be associated with LDL compositional changes. Although the data suggest an independent association between trunk fat and LDL metabolism, further research is warranted. The other subgroup of abdominal-obese individuals would include obese subjects with high levels of deep abdominal fat (visceral obesity). This condition is associated with further insulin resistance probably of hepatic origin, leading to glucose intolerance, and with the proper genetic susceptibility, to noninsulin-dependent diabetes mellitus (NIDDM). H-TGL activity is also increased, leading to further reduction in plasma HDL cholesterol levels. Obese subjects with gluteal-femoral or peripheral fat accumulation do not show substantial changes in their CVD risk profile and in their death rate. These individuals are, however, prone to varicose veins and exhibit the orthopedic problems commonly associated with excess body weight. It is proposed that the risk of CVD in gluteal-femoral-obese individuals is only moderately different from that in nonobese subjects, whereas CVD risk increases in trunk-abdominal-obese individuals and would be the highest in the subgroup of abdominalobese subjects with high levels of deep abdominal fat. From the literature available, there is little doubt that subjects with visceral obesity represent the subgroup of obese individuals with the highest risk for CVD. Obese subjects with both high levels of trunk-abdominal fat and high levels of deep abdominal fat will also be found. These obese subjects should presumably be at greater risk for CVD than obese subjects with visceral obesity only. In our opinion, this working classification of obesities is supported by a reasonable body of data The metabolic correlates of each obesity subtype reviewed in this article, especially for plasma lipoprotein metabolism and plasma lipid transport, clearly warrant additional clinical and experimental research. This field has, however, the potential to provide better answers to an old and difficult problem, namely the role of obesity in the etiology of dyslipoproteinemias. Further progress in this area will require the use of proper body fatness and fat distribution variables for the assessment of the various obesity components and for the proper classification of the metabolic correlates of obesity. Adipose Tissue Distribution and Dysllpoprotelnemia. Relevance of Genetic Variation to Metabolic Heterogeneity of Abdominal Obesity Although significant associations have consistently been found between the regional distribution of AT and changes in plasma lipid transport, the various metabolic profiles measured in abdominal-obese subjects suggest the presence of a metabolic heterogeneity among these individuals. A good example of this situation is provided by NIDDM; this issue has been extensively reviewed by Wssebah and co-workers. 15 ' 19 ' 17 Indeed, although these investigators have shown that a high proportion of abdominal fat is associated with disturbances in glucose homeostasis, insulin resistance, and hyperinsulinemia, not every abdominal-obese person showing these metabolic complications will become diabetic. These observations suggest that the genetic susceptibility to NIDDM is an important component of the fat distribution-niddm association. 151617 Unsuppressed hepatic glucose production and decreased pancreatic insulin secretory capacity have been proposed as two important sites in the genetic control of the susceptibility to NIDDM in subjects with abdominal obesity. Regarding plasma lipid transport and plasma lipoprotein levels, variation in "candidate genes" relevant to lipoprotein metabolism (e.g., genes coding for LPL H-TGL, lipoprotein receptors, and apoproteins 129 ) could potentially alter the relation of AT distribution to plasma lipoproteins. Since apoprotein E polymorphism is known

REGIONAL BODY FAT DISTRIBUTION AND UPOPROTEINS Despr6s et al. 507 Conclusions From this review, it is likely that the strong independent effect of body fat distribution on CVD can be explained by the hormonal and metabolic disturbances observed in abdominal obesity. In addition to the independent effects of insulin resistance and diabetes, of the altered plasma lipoprotein-lipid profile, and of hypertension on CVD, it should be emphasized that several of these metabolic complications will often be simultaneously observed in subjects with abdominal obesity. Therefore, the interac- Table 7. Correlations between Body Fat Mass, Deep Abdominal Fat Deposition, and Plasma VLDL Cholesterol and LDL Cholesterol In Women with the Apo E2 or the Apo E4 Isoprotein Apo E2 (n=22) Apo E4 (n=17) Variable VLDL cholesterol LDL cholesterol Fat mass 0.57t 0.09 Deep abdominal fat 0.64t 0.20 Fat mass 0.10 0.16 Deep abdominal fat 0.11 0.51* *p<0.05, tp<0.01. This table was adapted from reference 131. VLDL=very low density llpoprotein, LDL=low density lipoprotein, apo=apolipoprotein. to affect plasma lipoprotein metabolism (for a review, see reference 130), we have used this approach to study the associations between obesity, AT distribution, and plasma lipoprotein levels in three groups of women defined on the basis of their apo E phenotypes. 131 Our results indicated that the associations typically observed between total body fat, AT distribution, and plasma lipoprotelns were altered in the subgroups of women carrying either the apo E2 or the apo E4 isoprotein (Table 7). In women with the apo E2 isoprotein, which has been shown to have a lower affinity for the apo E and B,E receptors than the apo E4 isoprotein, body fatness variables were significantly correlated with plasma VLDL cholesterol levels, whereas in contrast, no relationship was found between total body fat, the level of deep abdominal fat, and LDL cholesterol levels (Table 7). In the apo E4 group, obesity was not significantly correlated with the level of plasma VLDL cholesterol, whereas a high level of deep abdominal fat was positively correlated with plasma LDL cholesterol levels. The mechanisms by which lipoprotein metabolism is related to the apo E polymorphism have been previously discussed in this journal, and the reader is referred to the comprehensive review of Davignon et al. 130 Figure 1 illustrates the effects of the increased hepatic VLDL synthesis observed in abdominal obese subjects on plasma VLDL and LDL metabolism in relation to apo E2 and apo E4 phenotypes. Thus, in abdominal-obese women, subjects carrying the apo E2 isoprotein showed no increase in plasma cholesterol or LDL cholesterol levels, but higher plasma VLDL cholesterol concentrations. Abdominal-obese women in the E4 group had higher levels of plasma cholesterol and LDL cholesterol, but showed no increase in plasma VLDL levels. This study provides the first indication that a variation in genes relevant to lipoprotein metabolism may substantially alter the association between obesity, regional AT distribution, and plasma lipoprotein levels. We therefore propose that obesity and abdominal-visceral fat accumulation are conditions that may exacerbate the disturbances in lipoprotein metabolism and plasma lipid transport. We believe that the analogy with the genetic susceptibility to NIDDM in abdominal obesity is relevant to the issue of the susceptibility to dyslipoproteinemia and CVD in abdominal obesity. This topic is challenging AfOOMMAlOOM WMCT i t f FFA CUJCOM MUUf QESOAMCE ns Attmoaoa J Figure 1. Metabolism of very low density lipoprotein (VLDL) and low density lipoprotein (LDL) in abdominal-obese individuals carrying either the apoprotein (apo) E2 or the apo E4 isoproteins. The higher hepatic VLDL synthesis observed in abdominal obesity affects lipoprotein metabolism differently in the apo E2 phenotype (A) than In the apo E4 phenotype (B). In the E2 phenotype, a higher concentration of VLDL remnants Is observed, due to the lower affinity of apo E2 isoprotein for the remnant and apo B,E receptors. The LDL levels are "normal" due to the reduced cholesterol delivery to the liver, thus, resulting in the up-regulation of the hepatic apo B,E receptors. In the apo E4 phenotype, normal plasma levels of VLDL remnants are observed, due to the higher affinity of apo E4 Isoprotein for the remnant and apo B,E receptors. However, plasma LDL levels are increased due to the greater cholesterol delivery to the liver, thus resulting In the down-regulation of hepatic apo B,E receptors. For a comprehensive model of the effects of apo E polymorphism on lipoprotein metabolism, the reader is referred to Davignon et al. 130 and promising, and additional studies are clearly warranted for a further understanding of the interactions between fat distribution, genes relevant to lipoprotein metabolism, and plasma lipid transport.

508 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 tion of these risk factors could markedly increase the probability of a cardiovascular accident in these individuals. Results from CT indicate that intra-abdominal fat accumulation is a critical variable in the study of the relation of body fat distribution to metabolic complications. As there are marked gender differences in regional body fat distribution, sex steroids are likely candidates for an important role in the determination of individual differences in body fat distribution and related metabolism. Issues that will require further investigation include 1) the role of regional variation in AT metabolism, 2) the mechanisms involved in the lipoprotein compositional changes associated with high plasma TG levels in abdominal obesity, 3) lipoprotein turnover studies in relation to body fat distribution, 4) the role of sex steroids and glucocorticoids in the regulation of fat distribution and lipoprotein metabolism, and in lipoprotein metabolism per se. Additional research on the genetic and nongenetjc determinants of body fat distribution and on the proper measurement of fat topography is also warranted. Finally, the study of genetic variation in apolipoprotein, enzymes and receptor genes involved in lipoprotein metabolism could contribute to a better understanding of the metabolic heterogeneity of abdominal obesity. Studies should be undertaken to clarify the potential relationship between DNA variation for these genes and the lipoprotein phenotypes with proper controls over obesity and regional fat distribution. It is likely that DNA polymorphism will contribute to individual differences in the sensitivity of lipoprotein metabolism and other processes involved in the development of atherosclerosis in the presence of an excessive accumulation of abdominal fat. This area of research represents an important challenge to those of us involved in the body fat distributjon-cvd issue. Acknowledgments Thanks are expressed to the subjects and to the dedicated staff of the Physical Activity Sciences Laboratory for their excellent collaboration. The authors express their gratitude to Judith Maheux, Jaclnthe Hovlngton, Jean-Fangois Landry, Marie- Christine Poullot, Marine Marcotte, Henri Bessette, and Claude Leblanc for their help at various stages of the studies. We also thank Nicole Roy, Yolande Montreuil, Marie Martin, Rachel Duchesne, Suzanne Brulotte, and Jean-Guy Bertrand for their assistance in data collection. The excellent assistance of the LJpfd Research Center and the Diabetes Research Unit staff Is also acknowledged. References 1. Larsson B, Svardaudd K, Wolln L, Wllhelmsen L, B]6rntorp P, TIbblln 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 1984;288:1401-1404 2. Lapldus L, Bengtsson C, Larsson B, Panned K, Rybo E, SJdstrdm 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 1984;289:1261-1263 3. Stokes J III, Garrison RJ, Kannel WB. The Independent contributions of various Indices of obesity to the 22-year incidence of coronary heart disease. The Framingham Heart Study. In: Vague J, BjOrntorp P, Guy-Grand B, Rebufte-Scrive M, Vague P, eds. Metabolic complications of human obesities. Amsterdam: Elsevier Science, 1985:49-57 4. DuclmatJere P, Richard J, CamWen 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 5. Donahue RP, Abbott RD, Bloom E, Reed DM, Yarto K. Central obesity and coronary heart disease In men. Lancet 1987; 1:822-824 6. Vague J. La differentiation sexuelle, facteur determinant des formes de robeslte. Presse Med 1947;30:339-340 7. Vague J, B]6rntorp P, Guy-Grand B, Rebuffe-Scrtve M, Vague P, eds. Metabolic complications of human obesities. Amsterdam: Elsevier Science, 1985:287 8. BJomtorp P, Smith U, Lonnroth P, eds. Health implications of regional obesities. Stockholm: Almqvtst & Wlksell International, 1988:237 9. Bouchard C, Johnston FE, eds. Fat distribution during growth and later health outcomes. New York: Alan R LJss, 1988:363 10. Bouchard C, Bray GA, Hubbard VS. Basic and clinical aspects of regional fat distribution. Am J Clln Nutr (In press) 11. Klsssbah AH, Vydellngum N, Murray R, et al. Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab 1982;54:254-260 12. B]omtorp P. Obesity and the risk of cardiovascular disease. Ann Clln Res 1985;17:3-9 13. Klssebah AH, Evans DJ, Pelrls A, Wilson CR. Endocrine characteristics in regional obesities: role of sex steroids. In: Vague J, BJorntorp P, Guy-Grand B, Rebuffe-Scrive M, Vague P, eds. Metabolic complications of human obesities. Amsterdam: Elsevier Science, 1985:115-130 14. Stern MP, HafTner SM. Body fat distribution and hyperinsullnemla as risk factors for diabetes and cardiovascular disease. Arteriosclerosis 1986:6:123-130 15. Klssebah AH, Pelrls AN, Evans DJ. Mechanisms associating body fat distribution to glucose intolerance and diabetes mellitus: window with a view. Acta Med Scand 1988;S723:79-89 16. Klssebah AH, Freodman DS, Pelrls AN. Health risks of obesity. Med Clln North Am 1989;73:111-138 17. Klssebah AH, Pelrls AN. Biology of regional body fat distribution: Relationship to non-insulin-dependent diabetes mellitus. Diabetes Metab Rev 1989:5:83-109 18. Hartz AJ, Rupley DC, Kalkhoff RD, Rlmm AA. Relationship of obesity to diabetes: Influence of obesity and body fat distribution. Prev Med 1983; 12:351-357 19. Krotklewakl M, B 6mtorp P, SJ&strom L, Smith U. Impact of obesity on metabolism In men and women. Importance of regional adipose tissue distribution. J Clln Invest 1983; 72:1150-1162 20. Kalkhoff RK, Hartz AH, Rupley D, Klssebah AH, Kelber S. Relationship of body fat distribution to blood pressure, carbohydrate tolerance, and plasma llplds In healthy obese women. J Lab Clin Med 1983;102:621-627 21. Evans DJ, Hoffman RG, Kalkhoff RK, Klssebah AH. Relationship of body fat topography to insulin sensitivity and metabolic profiles In premenopausal women. Metabolism 1984;33:68-75 22. Ohlson LO, Larsson B, Svardaudd K, et al. The influence of body fat distribution on the Incidence of diabetes mellitus 13.5 years of follow-up of the participants in the study of men bom in 1913. Diabetes 1985;34:1055-1058 23. Borkan GA, Gerzof SG, Robblns AH, Hurts DE, Sllbert CK, Sllbert JE. Assessment of abdominal fat content by computed tomography. Am J Clin Nutr 1982;36:172-177 24. Grauer WO, Moss AA, Cann CE, Goldberg HI. Quantification of body fat distribution in the abdomen using computed tomography. Am J Clln Nutr 1984;39:631-637 25. SJostrdm L, Kvlst H, Cederblad A, Tylen U. Determination of total adipose tissue and body fat in women by computed tomography, 40 K, and tritium. Am J Physiol 1986;250:E736-E745

REGIONAL BODY FAT DISTRIBUTION AND LIPOPROTEINS Despr6s et al. 509 26. Enzl Q, Gasparo M, Blondettf PR, Ror D, S«mlsa M, Zurio F. Subcutaneous and visceral fat distribution according to sex, age, and overweight, evaluated by computed tomography. Am J Clin Nutr 1986;44:739-746 27. Soldoll JC, Oosteriee A, TTilJssen MAO, at al. Assessment of Intra-abdominal and subcutaneous abdominal fat: relation between anthropometry and computed tomography. Am J Clin Nutr 1987;45:7-13 28. Seldell JC, Oosteriee A, Deurenberg P, et al. Abdominal fat depots measured with computed tomography: Effects of degree of obesity, sex, and age. Eur J Clin Nutr 1988;42:805-815 29. Baumgartner RN, Heymsfleld SB, Roche AF, et al. Abdominal composition quantified by computed tomography. Am J Clin Nutr 1988;48:936-945 30. Fertand M, Despres JP, Tremblay A, et al. Assessment of adipose tissue distribution by computed axial tomography in obese women: Association with body density and anthropometrlc measurements. Br J Nutr 1989; 61:139-148 31. Sparrow D, Borkan GA, Gerzof SG, Wlsnlewskl C, Sllbert CK. Relationship of body fat distribution to glucose tolerance. Results of computed tomography in male participants of the nomiative aging study. Diabetes 1986; 35:411-415 32. Fujloka S, Matsuzawa Y, Tokunaga K, Tarul S. Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism In human obesity. Metabolism 1987;36:54-59 33. Despres JP, Nadeau A, Tremblay A, et al. Role of deep abdominal fat in the association between regional adipose tissue distribution and glucose tolerance in obese women. Diabetes 1989:38:304-309 34. Pelrls AN, Sothmann MS, Hennes Ml, et al. 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 35. Pelrls A, Mueller RA, Smith GA, Struve MF, Klssebah AH. Splanchnic insulin metabolism in obesity: Influence of body fat distribution. J Clin Invest 1986;78:1648-1657 36. Pelrls AN, Struve MF, Klssebah AH. Relationship of body fat distribution to the metabolic clearance of insulin In premenopausal women. Int J Obes 1987:11:581-589 37. Krotkiewskl M. B]6rntorp P. Muscle tissue in obesity with different distribution of adipose tissue. Effects of physical training. Int J Obes 1986:10:331-341 38. Jensen MD, Haymond MW, Rlzza RA, Cryer PE, Miles JM. Influence of body fat distribution on free fatty acid metabolism In obesity. J Clin Invest 1989;83:1168-1173 39. Randle PJ, Garland PB, Hales CN, Newsholme EA. The glucose-fatty acid cycle; Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1963; 1:785-789 40. Ferrannlnl E, Barrett EJ, Bevllacqua S, DeFronzo RA. Effect of fatty acids on glucose production and utilization In man. J Clin Invest 1983;72:1737-1747 41. Svedberg J, BJorrrtorp P, Smith U, Lonnroth P. Insulin binding, degradation and action in rat hepatocytes is reduced in a dose-dependent way by free fatty adds [abstr]. Int J Obes 1989;13:213A 42. Hennes MM, Klssebah AH. Receptor and postreceptor effects of free fatty acids (FFA) on hepatocyte Insulin dynamics [abstr]. Int J Obes 1989;13:556A 43. Ostman J, Amer P, Engfeldt P, Kager L Regional differences in the control of lipotysis in human adipose tissue. Metabolism 1979;28:1198-1205 44. Bollnder J, Kager L, Ostman J, Amer P. Differences at the receptor and postreceptor levels between human omental and subcutaneous adipose tissue In the action of insulin on llpolysis. Diabetes 1983;32:117-123 45. Despres JP, Allard C, Tremblay A, Talbot J, Bouchard C. Evidence for a regional component of body fatness In the association with serum lipfds In men and women. Metabolism 1985;34:967-973 46. Despres JP, Tremblay A, Bouchard C. Regional adipose tissue distribution and plasma llpoprotelns. In: Bouchard C, Johnston FE, eds. Fat distribution during growth and later health outcomes. New York: Alan R Uss, 1988:221-241 47. Despres JP, Tremblay A, Nadeau A, Bouchard C. Physical training and changes in regional adipose tissue distribution. Acta Med Scand 1988;S723:205-212 48. Blair D, Hablcrrt JP, Sims EAH, Sylwester D, Abraham S. Evidence for an Increased risk for hypertension with centrally located body fat and the effect of race and sex on this risk. Am J Epidemiol 1984; 119:526-540 49. Berglund G, Hetlstrom A, Hartford M, LJungman S, Wlkstrand J. Adipose tissue distribution in relation to blood pressure in the population. In: Vague J, BJdmtorp P, Guy-Grand B, Rebuffe-Scrtve M, Vague P, eds. Metabolic complications of human obesities. Amsterdam: Elsevier Science, 1985:61-66 50. Ralson J, Guy-Grand B. Body fat distribution in obese hypertensives. In: Vague J, Bjorntorp P, Guy-Grand B, Rebuffe-Scrtve M, Vague P, eds. Metabolic complications of human obesities. Amsterdam: Elsevier Science, 1985:67-75 51. Welnsler RL, Norris DJ, Birch R, et al. Serum Insulin and blood pressure in an obese population. Int J Obes 1986; 10:11-17 52. Welnsler RL, Norris DJ, Birch R, et al. The relative contribution of body fat and fat pattern to blood pressure level. Hypertension 1985:7:578-585 53. Despres JP, Tremblay A, Therlault G, Perusse L, Leblanc C, Bouchard C. Relationships between body fatness, adipose tissue distribution, and blood pressure In men and women. J Clin Epidemiol 1988:41:889-897 54. Sims EAH, Berchtold P. Obesity and hypertension; mechanisms and implications for management. JAMA 1982; 247:49-52 55. Krotkiewskl M, Mandroukas K, SJostrom L, Sullivan L, Wetterqvtst H, B 6rntorp P. Effects of long-term physical training on body fat, metabolism, and Wood pressure in obes. Metabolism 1979;28:650-658 56. DeFronzo RA. Insulin and renal sodium handling: clinical implications. Int J Obes 1981;5(suppl 1):93-104 57. Sims EAH. Mechanisms of hypertension in the syndromes of obesity. Int J Obes 1981;5(suppl 1):9-18 58. Christileb AR, Krolewskl AS, Warram JH, Soeldner JS. Is insulin the link between hypertension and obesity. Hypertension 1985;7(suppl ll):ll-54-ll-57 59. Landsberg l_ Diet, obesity and hypertension: an hypothesis involving insulin, the sympathetic nervous system, and adaptive thermogenesis. Q J Med 1986;61:1081-1090 60. Albrink MJ, Melgs JW. Interrelationship between skinfold thickness, serum liplds and blood sugar in normal men. Am J Clin Nutr 1964;15:255-261 61. Allard C, Goulet C. Serum lipids: An epidemiological study of an active Montreal population. Can Med Assoc J 1968:98:627-637 62. Despres JP, Tremblay A, Perusse L, Leblanc C, Bouchard C. Abdominal adipose tissue and serum HDLcholesterol: Association independent from obesity and serum triglyceride concentration. Int J Obes 1988;12:1-13 63. Haffner SM, Stern MP, Hazuda HP, Rosenthal M, Knapp AJ. The role of behavioral variables and fat patterning in explaining ethnic differences in serum liplds and lipoprotelns. Am J Epidemiol 1986:123:830-839 64. Foster C, Weinsler RL, Birch R, et al. Obesity and serum lipids: An evaluation of the relative contribution of body fat and fat distribution to lipid levels. Int J Obes 1987; 11:151-161 65. Haffner SM, Stern MP, Hazuda HP, Pugh J, Patterson JK. Do upper-body and centralized adiposity measure different aspects of regional body-fat distribution? Relationship to non-insulin-dependent diabetes mellitus, lipids, and lipoproteins. Diabetes 1987;36:43-51

510 ARTERIOSCLEROSIS VOL 10, No 4, JULY/AUGUST 1990 emia (increased protein but normal cholesterol levels In human plasma low density (B) lipoproteins). Proc Natl Acad Sci USA 1980;77:604-608 87. Kwtterovlch PO Jr, Snlderman AD. Atherosclerosis and apoproteins B and A-l. Prev Med 1983;12:815-834 88. Klelnman Y, Oschry Y, Elsenberg S. Abnormal regulation of LDL receptor activity and abnormal cellular metabolism of hypertriglyceridaemlc tow density lipoprotein: normalization with bezafibrate treatment. Eur J Clin Invest 1987; 17:538-543 89. Elsenberg S. Lipoprotein abnormalities In hypertrigiyceridemla: Significance in atherosclerosis. Am Heart J 1987; 113:555-561 90. Chajek T, Raiding CJ. Isolation and characterization of a human serum cholesteryl ester transfer protein. Proc Natl Acad Sci USA 1978:75:3445-3449 91. Hopkins GJ, Barter PJ. Transfer of esterified cholesterol and triglyceride between high density and very low density lipoproteins: In vitro studies of rabbits and humans. Metabolism 1980;29:546-550 92. Deckelbaum R, Elsenberg S, Oschry Y, Ollvecrona T. Reverse modification of human plasma low density lipoprotein toward triglyceride rich precursors: a mechanism for losing excess cholesterol ester. J Biol Chem 1982; 257:6509-6517 93. Howard BV. Lipoprotein metabolism in diabetes mellitus. J Upid Res 1987;28:613-628 94. Despres JP, Moorjanl S, Tremblay A, et al. Heredity and changes in plasma lipids and lipoproteins after short-term exercise training In men. Arteriosclerosis 1988:8:402-409 95. Landin K, KrotMewskl M, Smith U. Importance of obesity for the metabolic abnormalities associated wtth an abdominal fat distribution. Metabolism 1989:38:572-576 96. Poullot MC, Despres JP, Moorjanl S, et al. Computed tomography-measured trunk fat and plasma lipoprotein levels In nonobese women. Metabolism 1989:38: 1244-1250 97. Evans DJ, Hoffmann RG, Kalkhoff RK, Klssebah AH. Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolic aberrations in premenopausal women. J Clin Endocrinol Metab 1983; 57:304-310 98. Pelrls AN, Mueller RA, Struve MF, Smith GA, Klssebah AH. Relationship of androgenic activity to splanchnic insulin metabolism and peripheral glucose utilization In premenopausal women. J Clin Endocrinol Metab 1987; 64:162-169 99. Stefanick ML, Williams PT, Krauss RM, Terry RB, Vranlzan KM, Wood PD. Relationships of plasma estradtol, testosterone, and sex hormone-binding globulin wtth llpoproteins, apolipoproteins, and high density lipoprotein subtractions in men. J Clin Endocrinol Metab 1987; 64:723-729 100. Applebaum-Bowden D, Haffner SM, Hazzard WR. The dysllpoproteinemia of anabolic steroid therapy: Increase In hepatic triglyceride lipase precedes the decrease In high density lipoprotein cholesterol. Metabolism 1987; 36:949-952 101. Rebuffe-Scrive M, KrotMewskl M, Eitverson J, BJ6mtorp P. Muscle and adipose tissue morphology and metabolism in Cushing's syndrome. J Clin Endocrinol Metab 1988:67:1122-1128 102. Vague J, Vague P, Boyer J, Clolx MC. Anthropometry of obesity, diabetes, adrenal and beta-cell functions. In: Rodriguez RR, Vallance Owen J, eds. Proceedings of the VII congress of the International diabetes federation. International Congress Series Excerpta Med lea, Amsterdam 1970;231:517-525 103. Vague J, Vague P. Obesity and diabetes. In: Alberti KGMM, Krall LP, eds. The diabetes annual/4. Amsterdam: Elsevier Science, 1988:311-338 104. Brlndley DN, Holland Y. Possible connections between stress, diabetes, obesity, hypertension and altered lipopro- 66. Ralson J, Basdevant A, Guy-Grand B. Relation entre la repartition du tissu adlpeux et les lipldes circulants chez la femme obese. Presse Med 1987;16:431-433 67. Baumgartner RN, Roche AF, Chumlea WC, Slervogel RM, Glueck CJ. Fatness and fat patterns: associations with plasma lipids and blood pressures in adults, 18 to 57 years of age. Am J Epidemiol 1987; 126:614-628 68. Relchley KB, Mueller WH, Hanls CL, et al. Centralized obesity and cardiovascular disease risk in Mexican Americans. Am J Epidemiol 1987; 125:373-386 69. Anderson AJ, Soboclnakl KA, Freedman DS, Barboriak JJ, Rlmm AA, Gruchow HW. Body fat distribution, plasma lipids and llpoproteins. Arteriosclerosis 1988;8:88-94 70. Soler JT, Folsom AR, Kushl LH, Prlneas RJ, Seal US. Association of body fat distribution wtth plasma lipids, lipoprotelns, apollpoproteins Al and B in postmenopausaj women. J Clin Epidemiol 1988;41:1075-1081 71. Seldell JC, Menslnk RP, Katan MB. Measures of fat distribution as determinants of serum lipids In healthy volunteers consuming a uniform standardized diet. Eur J Clin Invest 1988; 18:243-249 72. Haffner SM, Fong 0, Hazuda HP, Pugh JA, Patterson JK. Hyperinsulinemia, upper-body adiposity, and cardiovascular risk factors in non-diabetics. Metabolism 1988; 37:338-345 73. van Qaal L, Vansant G, van Campenhout C, Lepoutre L, de Le«uw I. Apolipoproteln concentration in obese subjects wfth upper and lower body fat mass distribution. Int J Obes 1989;13:255-264 74. Terry RB, Wood PD, Haakell WL, Stefanlck ML, Krauss RM. Regional adiposity pattern in relation to lipids, lipoprotein cholesterol, and lipoprotein subtraction mass in men. J Clin Endocrinol Metab 1989;68:191-199 75. Soler JT, Folsom AR, Kaye SA, Prlneas RJ. Associations of abdominal obesity, fasting insulin, sex hormone binding globulin, and estrone with lipids and lipoproteins In postmenopausal women. Atherosclerosis 1989;79:21-27 76. Despres JP, Moor anl S, Ferland M, et al. Adipose tissue distribution and plasma lipoprotein levels in obese women: Importance of intra-abdomlnal fat. Arteriosclerosis 1989; 9:203-210 77. Pelrls AN, Sothmann MS, Hoffmann RG, et al. Adiposity, fat distribution, and cardiovascular risk. Ann Intern Med 1989; 110:867-872 78. Kekkl M. LJpoproteln-llpase action determining plasma high density lipoprotein cholesterol levels In adult normolipaemics. Atherosclerosis 1980;37:143-150 79. Elsenberg S. Lipoproteins and lipoprotein metabolism. A dynamic evaluation of the plasma fat transport system. Win Wochenschr 1983;61:119-132 80. Elsenberg S. High density lipoprotein metabolism. J Upld Res 1984;25:1017-1058 81. Nlkklla EA, Tasklnen M-R, Sane T. Plasma high-density lipoprotein concentration and subtraction distribution in relation to triglyceride metabolism. Am Heart J 1987; 113:543-548 82. Tlkkanen MJ, Nlkklla EA. Regulation of hepatic lipase and serum lipoproteins by sex steroids. Am Heart J 1987; 113:562-567 83. Despres JP, Ferland M, Moorjanl S, et al. Role of hepatic-trlglyceride lipase activity in the association between Intra-abdominal fat and plasma HDL cholesterol in obese women. Arteriosclerosis 1989:9:485-492 84. Poullot MC, Despres JP, Moorjani S, et al. Regional variation in adipose tissue lipoprotein lipase activity: association wtth plasma lipoprotein levels In women [abstr]. FASEB J 1990;4:918A 85. Despres JP, Moorjanl S, Tremblay A, et al. Relation of high plasma triglyceride levels associated with obesity and regional adipose tissue distribution to plasma lipoproteinlipid composition in premenopausal women. Clin Invest Med 1989;12:374-380 86. Snlderman A, Shapiro S, Marpole D, et al. Association of coronary atherosclerosis with hyperapobetalipoprotein-

REGIONAL BODY FAT DISTRIBUTION AND LIPOPROTEINS Despr6s et al. 511 tein metabolism that may result in atherosclerosis. Clin Sci 1989;77:453-461 105. Brlndley DN, Brown NF, Sattar AM, Fisher SC, Fears R. Role of Insulin and counter-regulatory hormones In the control of hepatic glycerolipid synthesis and low density llpoprotein catabolism in diabetes. Biochem Soc Trans 1988;17:43-46 106. Manglapana EH, Brlndley DN. Effects of dexamethasone and Insulin on the synthesis of triacylglycerols and phosphatidylcholine and the secretion of very-low-density lipoprotelns and lysophosphatidylcholine by monolayer cultures of rat hepatocytes. Biochem J 1986:233:151-160 107. BartJett SM, Gibbons QF. Short- and longer-term regulation of very-low-density lipoprotein secretion by insulin, dexamethasone and llpogenic substrates In cultured hepatocytes. A biphasic effect of insulin. Biochem J 1988; 249:37-43 108. Sattar AM, Fischer SC, Brlndley DN. Interactions of triiodothyronine, insulin and dexmethasone on the binding of human LDL to rat hepatocytes In monolayer culture. Atherosclerosis 1988:71:77-80 109. Saltor AM, Fischer SC, Brlndley DN. Binding of lowdensity lipoprotein to monolayer cultures of rat hepatocytes Is Increased by insulin and decreased by dexamethasone. FEBS Lett 1987;220:159-162 110. Lelbel RL, Edens NK, Fried SK. Physiologic basis for the control of body fat distribution In humans. Annu Rev Nutr 1989:9:417-443 111. Smith U, Hammersten J, BJdmtorp P, Krai JG. Regional differences and effect of weight reduction on human fat cell metabolism. Eur J Clin Invest 1979:9:327-332 112. Lafontan M, Dang-Tran L, Berian M. Alpha-adrenerglc antjlipolytic effect of adrenaline in human fat cells of the thigh: comparison with adrenaline responsiveness of different fat deposits. Eur J Clin Invest 1978:9:261-266 113. Rebuffe-Scrtve M, Enk L, Crona N, et al. Fat cell metabolism indifferent regions in women. Effect of menstrual cycle, pregnancy, and lactation. J Clin Invest 1985; 75:1973-1976 114. Rebutie-Scrive M, Andersson, Olbe L, BJomtorp P. Metabolism of adipose tissue in intraabdominal depots of nonobese men and women. Metabolism 1989;38:453-458 115. B 6rntorp P. Morphological classifications of obesity: what they tell us, what they don't. Int J Obes 1984;8:525-533 116. Poullot MC, Despres JP, Nadeau A, et al. Associations between regional body fat distribution, fasting plasma free fatty add levels and glucose tolerance in premenopausal women. Int J Obes (in press) 117. Despres JP, Moorjanl S, Poullot MC, et al. Correlates of plasma very-low-density lipoprotein concentration and composition In premenopausal women. Metabolism (In press) 118. Mauriege P, Despres JP, Marcotte M, et al. Role of adipose tissue metabolism in the association between the regional distribution of body fat and metabolic variables in premenopausal women [abstr]. Int J Obes 1989;13:565A 119. Fong BS, Rodrigues PO, Salter AM, et al. Characterization of high density lipoprotein binding to human adipocyte plasma membranes. J Clin Invest 1985:75:1804-1812 120. Salter AM, Fong BS, Jimenez J, Rotsteln L, Angel A. Regional variation in high-density lipoprotein binding to human adipocyte plasma membranes of massively obese subjects. Eur J Clin Invest 1987:17:16-22 121. Despres JP, Fong BS, Jullen P, Jimenez J, Angel A. Regional variation In HDL metabolism in human fat cells: effect of fat cell size. Am J Physiol 1987;252:E654-E659 122. Jimenez JG, Fong B, Jullen P, Despres JP, Rotsteln L, Angel A. Effect of massive obesity on low and high density llpoprotein binding to human adipocyte plasma membranes. Int J Obes 1989;13:699-709 123. Despres JP, Fong BS, Jimenez J, Jullen P, Angel A. Selective uptake of HDL cholesterol ester by human fat cells. Am J Physiol 1988;254:E667-E675 124. Jimenez JG, Fong BS, Jullen P, Despres JP, Angel A. Weight loss in massive obesity: Reciprocal changes In plasma HDL cholesterol and HDL binding to human adipocyte plasma membranes. Metabolism 1988;37:580-586 125. Eckel RH. Adipose tissue llpoprotein lipase. In: Borensztajn J, ed. Lipoprotein llpase. Chicago: Evener, 1987: 79-132 126. Eckel RH. Lipoprotein lipase. A multifunctional enzyme relevant to common metabolic diseases. N Engl J Med 1989:320:1060-1068 127. Bouchard C. Variation in human body fat: The contribution of the genotype. In: Bray G, Ricquier D, Spiegelman B, eds. Obesity: towards a molecular approach. New York: Wlley-Uss, 1989:17-28 128. Harris Peoples U Carpenter JW, Israel RG, Barakat HA. Alterations in low-density lipoproteins in subjects with abdominal obesity. Metabolism 1989:38:1029-1036 129. Lusls AJ. Genetic factors affecting blood lipoproteins: the candidate gene approach. J Upid Res 1988:29:397-429 130. Davtgnon J, Gregg RE, Sing CF. Apollpoprotein E polymorphism and atherosclerosis. Arteriosclerosis 1988; 8:1-21 131. Poullot MC, Despres JP, Moorfanl S, Luplen PJ, Tremblay A, Bouchard C. Apollpoprotein E polymorphism alters the association between body fatness and plasma lipoproteins in women. J Upid Res (in press) Index Terms: obesity body fat distribution deep abdominal fat lipoprotein metabolism lipoprotein lipase hepatic lipase high density lipoproteins sex steroids glucocorticoids adipose tissue metabolism

Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. J P Després, S Moorjani, P J Lupien, A Tremblay, A Nadeau and C Bouchard Arterioscler Thromb Vasc Biol. 1990;10:497-511 doi: 10.1161/01.ATV.10.4.497 Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright 1990 American Heart Association, Inc. All rights reserved. Print ISSN: 1079-5642. Online ISSN: 1524-4636 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://atvb.ahajournals.org/content/10/4/497 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answerdocument. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online at: http://atvb.ahajournals.org//subscriptions/