Recent studies suggest that the distribution of

Transcription

1 1250 Waist to Hip Ratio in Middle-Aged Women Associations With Behavioral and Psychosocial Factors and With Changes in Cardiovascular Risk Factors Rena R. Wing, Karen A. Matthews, Lewis H. Kuller, Elaine N. Meilahn, and Pam Plantinga Waist to hip ratio (WHR) was measured in 487 middle-aged women participating in the Healthy Women Study. Upper body fat distribution was found to be associated with numerous behaviors that affect cardiovascular risk, including smoking, low exercise levels, weight gain during adulthood, and higher caloric intake. Moreover, WHR was also associated with higher levels of anger, anxiety, and depression and lower levels of perceived social support. Women with upper body fat obesity had higher systolic blood pressure, total cholesterol, low density lipoprotein cholesterol, triglycerides, and apolipoprotein B and lower levels of high density lipoprotein (HDL) and the HDL subfractions 2 and 3. These associations remained significant after adjusting for body mass index. Among 108 women who had repeat measurements of WHR, changes in WHR over a 3-year period were significantly correlated with changes in activity and with decreases in HDL,. Thus, WHR appears to be an integral component of the cardiovascular risk profile. WHR is related to those behaviors and psychosocial attributes that influence cardiovascular risk. (Arteriosclerosis and Thrombosis 1991;ll: ) Recent studies suggest that the distribution of body fat may be a stronger predictor of coronary heart disease (CHD) than the total amount of body fat. 1 " 4 For example, in the Paris Prospective study, men who had an upper body fat distribution were at increased risk for CHD, independent of body mass index (BMI). 1 Similarly, in a prospective study conducted in Sweden 2-3 the waist to hip ratio (WHR) was found to be positively associated with the incidence of myocardial infarction, stroke, and death from all causes in both men and women, after adjusting for BMI. The effect of upper body fat distribution on atherosclerotic disease may be mediated by the effect on cardiovascular risk factors. WHR has been associated with increased blood pressure, 5-6 increased triglycerides, and decreased high density lipoprotein (HDL) cholesterol. 7 " 10 Body fat distribution has also been related to fasting and stimulated levels of glucose and insulin and to increased rates of diabetes. 14 To date, there have been few studies that have systematically investigated the behavioral or psychosocial factors associated with WHR. Smoking has From the Departments of Psychiatry (R.R.W., K.A.M.) and Epidemiology (R.R.W., K.A.M., L.H.K., E.N.M., P.P.), University of Pittsburgh, Pittsburgh, Pa. Supported by a grant from the National Heart, Lung, and Blood Institute (HL-28266). Address for reprints: Rena R. Wing, PhD, Western Psychiatric Institute and Clinic, 3811 O'Hara Street, Pittsburgh, PA Received January 15, 1991; revision accepted May 29, been related to increased upper body fat distribution in both men and women, 15 " 18 and a recent study showed that men who started to smoke had an increase in WHR. 16 Activity level may also be related to WHR ; men and women who reported more intense leisure-time activity were found to have lower WHRs, 19 and exercise training has been shown to decrease WHR. 20 Alcohol intake has also been related to WHR, with both positive 17 and negative 18 associations noted. We are not aware of any previous studies that have examined the association between psychosocial variables and body fat distribution. Experimental studies with monkeys have shown that stress results in redistribution of fat to the abdominal area. 21 Bjorntorp 22 hypothesized that the same may be true for humans and that those individuals who experience a great deal of stress or who have difficulty coping with stress may experience higher cortisol levels and consequently may deposit more body fat in the abdominal area. The purpose of the present article is to examine the behavioral and psychosocial correlates of WHR in a population-based sample of middle-aged women who were participating in the Healthy Women Study. In addition, we assessed the relation between WHR and cardiovascular risk factors in these women. Moreover, 108 of these women also had measures of WHR completed 3 years previously. Consequently, we could determine the relation between changes in behavior and changes in WHR and the association

2 between changes in WHR and changes in cardiovascular risk factors. Methods Subjects Subjects in this study were participants in the Healthy Women Study, a prospective investigation of changes in biologic and behavioral characteristics during the climacteric. Women were recruited into the study during from a random sample of licensed drivers in Allegheny County, Pennsylvania. Eligibility criteria included age years; menstruating within the past 3 months; no surgical menopause; diastolic blood pressure less than 100 mm Hg; and not taking insulin, thyroid medication, lipidlowering drugs, estrogens, antihypertensive drugs, or psychotropic drugs. Eighty-nine percent of women who were contacted agreed to a telephone interview, and 60% of those who were eligible volunteered to participate. A total of 541 women entered the study. Comparisons of participants with those who declined participation have been published previously 23 ; participants were better educated and were employed in higher-paying jobs than either those who were ineligible or those who declined participation. All women were invited to be restudied in 1987 after an average of 3 years in the study. The 487 women who completed the full reexamination form the sample for the present study. Of these 487 women, 282 were still premenopausal at follow-up and 62 were considered to be naturally postmenopausal, defined as not having menstruated for at least 12 months, not having undergone surgical menopause, and not having received hormone therapy in the past year. The remaining women were in the perimenopausal period (n=92), were using hormones (n=32), or had undergone surgical menopause (n=19). WHR was added late to the baseline protocol and was measured on the final 120 women. One hundred eight of these women completed the follow-up assessment as well, and they form the subsample for the analyses of changes in WHR over time. The women who had WHR assessed at baseline were older (/?<0.0001), had lower HDL cholesterol (/?<0.05), and had higher apolipoprotein (apo) A-II (p<0.0l) and apo B (/><0.0001) levels than did those for whom WHR was not assessed. Procedures Subjects were evaluated in the morning after a 12-hour fast for the following measures. Lipids and apolipoproteins. Serum cholesterol, 24 total HDL cholesterol, 25 the subfractions 2 and 3 of HDL (HDL 2 and HDL3), 26 and triglycerides 27 were assessed. Low density lipoprotein (LDL) cholesterol levels were estimated with the Friedewald equation (Friedewald et al 28 ). Apos A-I and B 29 were assessed Wing et al WHR and CV Risk Factors 1251 by electroimmunoassay, and apo A-II was assessed by enzyme-linked immunoabsorbent assay. 30 Glucose and insulin. Glucose and insulin were measured in the fasting state and 2 hours after a 75-g oral glucose load. Glucose was analyzed by enzymatic assay (Yellow Springs Glucose Analyzer, Yellow Springs, Ohio), and insulin was analyzed by radioimmunoassay. 31 Blood pressure. Blood pressure was measured twice by the random-zero muddler method 32 by observers certified according to the Multiple Risk Factor Intervention Trial protocol. 33 Height, weight, and body mass index. Height and weight were obtained with a balance beam scale, and BMI, a measure of body fat, was obtained by dividing body weight in kilograms by height in meters squared. Subjects were asked to report what they had weighed at age 20. Waist to hip ratio. The waist was measured at the smallest circumference and the hips at the largest circumference. A nonstretchable tape measure was used. Measures of health habits. Cigarette smoking was defined by the total number of cigarettes smoked per day. The Paffenbarger activity questionnaire (Paffenbarger et al 34 ) was used to obtain a measure of kilocalories per week spent in leisure-time activity. A 24-hour recall was used to obtain a measure of calories consumed and percent of calories from fat; the data were analyzed by use of a computerized nutrient data base Standardized tests of personality and behavior. Subjects also completed a battery of tests of personality and behavior, including the Framingham Type A, Anger-Discuss, and Tension (Haynes et al 37 ) scales; the Beck Depression Inventory (Beck et al 38 ); and the Spielberger Trait Anger Anxiety Questionnaire (Spielberger et al 39 ). Results Table 1 shows the basic characteristics of the sample. WHR averaged 0.77± (mean±sd), with a range from 0.45 to 0.99, and was significantly correlated with body weight (r=0.35, /?<0.001) and BMI (r=0.40,/?<0.001). WHR was higher in those with less education, but after adjusting for differences in BMI, there was no effect of education on WHR. WHR was not significantly related to age (r=0.04). Determinants of Waist to Hip Ratio WHR differed significantly as a function of menopausal status. Women who had remained premenopausal (n=282) had a WHR of compared with a WHR of in women who had experienced natural postmenopause and who were not receiving hormone therapy (n=62) (/?<0.01). Thus, postmenopausal women had more upper body fat distribution. The difference remained significant (p<0.05) after adjusting for BMI. WHR was also affected by smoking, as has been reported previously The correlation between

3 1252 Arteriosclerosis and Thrombosis Vol 11, No 5 September/October 1991 TABLE 1. Characteristics of Sample (n=487) Variable Mean Waist to hip ratio BMI (kg/m2) Weight (lb) Age (yr) SBP (mm Hg) DBP (mm Hg) Cholesterol (mg/dl) HDL cholesterol (mg/dl) HDL 2 (mg/dl) HDL 3 (mg/dl) Triglycerides (mg/dl) LDL cholesterol (mg/dl) Fasting glucose (mg/dl) 2-Hour glucose (mg/dl) Fasting insulin (/i.units/ml) 2-Hour insulin (/i,units/ml) Apo A-I (mg/dl) Apo A-II (mg/dl) Apo B (mg/dl) Wam/Hip Kalio SD Former Never Smoking Status 9.7 Waltt/Hlp Ratio BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; apo, apolipoprotein. number of cigarettes smoked per day and WHR was r=0.13,/?<0.01. The correlation increased to r=0.14, /><0.001, after adjusting for BMI. Current smokers had an average WHR of 0.780± versus a WHR of 0.765± in former smokers and of 0.762± in those who had never smoked (/?<0.05). The effect become more significant (/?<0.01) after adjusting for BMI (Figure 1). Another correlate of WHR was weight change since age 20 (r=-0.38, p<0.0001), indicating that women who gained the most weight since age 20 had the highest WHR. The correlation was of borderline significance (r=-, p=) after adjusting for current BMI. WHR was also correlated with current caloric intake, r=0.15, /?< Caloric intake was unrelated to BMI (r=0.04), and hence the correlation between caloric intake and WHR remained significant after adjusting for BMI (r=0.15, p<0.001). Analysis of variance comparing women in the four quartiles of caloric intake showed that women who consumed more calories had higher WHRs (Figure 1), even after adjusting for BMI (p<0.01). The relation between activity and WHR is also shown in Figure 1. Women were divided into quartiles according to their self-reported activity levels (0-500, , 1,000-1,999, and >2,000 kcal/ week). As shown, women who reported the lowest levels of activity had the highest WHR. After adjusting for differences in BMI, the differences between quartiles remained marginally significant (p=0.08). The difference between the lowest exercise quartile ) 2213 Calorie Intake Walsl/Hlp Ratio ?Vv 100G i>9 y, icgc Calories Expended in Exercise FIGURE 1. Bar graphs showing waist/hip ratio as a function of smoking status (upper panel), calorie intake after adjusting for body mass index (middle panel), and activity level (lower panel). and the other three quartiles was significant (0.784 vs , p<0.05, after adjusting for BMI). Stepwise multiple regression examining the effect of these behavioral attributes on WHR indicated that change in weight since age 20, number of cigarettes smoked per day, and calorie intake all contributed independently to the prediction of WHR. However, together these variables explained only 17% of the variance in WHR. Waist to Hip Ratio and Psychosocial Variables Table 2 shows the correlations between WHR and psychosocial variables. As shown in the table, upper body fat distribution was associated with more tension and anxiety, more frequent angry feelings and expression of those feelings, higher levels of depressive symptomatology, and lower levels of perceived social support. Most of these associations remained significant when nonparametric correlations were used, after adjusting for BMI, or when the analyses

4 Wingetal WHR and CV Risk Factors 1253 TABLE 2. Relation Between Waist to Hip Ratio and Psychosocial Measures (n=487) WHR BMI WHR, controlling for BMI Bortner Type A Social anxiety Low self-esteem * * 0.08* Spielberger trait anger 0.14$ 0.12t O.llt Spielberger trait anxiety 0.14$ t Anger expression Anger in 0.10* Anger out 0.12t 0.11* 0.08* Framingham Anger-Discuss Social support -0.10* * Framingham Tension 0.14* 0.09* O.llt Pessimism 0.09* Stress Depression 0.17* 0.16* 0.12t WHR, waist to hip ratio; BMI, body mass index. V<0.05, t/><0.01, :*p< BMI, controlling for WHR o.iot were restricted to premenopausal women only. The psychosocial variables appeared to be more closely associated with WHR than with BMI. Waist to Hip Ratio and Coronary Heart Disease Risk Factors Pearson correlation coefficients were computed to examine the relation between WHR and CHD risk factors. These analyses were restricted to the women who remained premenopausal at the final clinic visit (n=282), as menopausal status appears to affect both WHR and CHD risk factors. 40 As shown in Table 3, WHR was positively correlated with blood pressure, triglycerides, cholesterol, apo B, glucose, and insulin (the latter two measures both in the fasting state and 2 hours after the glucose load). WHR was negatively correlated with HDL cholesterol and its subfractions. Partial correlations showed that both WHR and BMI were independently related to systolic blood pressure, total HDL cholesterol and its subfractions, triglycerides, 2-hour glucose, and fasting and 2-hour insulin. Figures 2 and 3 show the effects of both BMI and WHR on cholesterol HDL and insulin levels, after dividing the women into tertiles for each of these measures. It is notable that WHR was independently associated with both total cholesterol and LDL cholesterol, but BMI was not. Change in Waist to Hip Ratio From Baseline to Final Follow-up One hundred eight women had measures of WHR completed at entry into the study and again at the final followup. For these 108 subjects there was a nonsignificant increase in WHR over the 3 years (0.007±0.065). However, there were significant increases in both the waist circumference (79.1 ±12.1 to TABLE 3. Correlation of Waist to Hip Ratio With Coronary Heart Disease Risk Factors in Premenopausal Women WHR, BMI, controlling controlling Variable WHR BMI for BMI for WHR SBP 0.19* t 0.24 DBP 0.13t Total cholesterol * 0.00 Total HDL HDL * HDL * -0.18* -0.13t -0.14* Triglyce rides* LDL cholesterol* 0.19* * 0.04 Fasting glucose 0.15* Hour glucose * 0.21* Fasting insulin* * Hour insulin* * 0.36 Apo A-I * * Apo A-II Apo B * 0.16* 0.12t *These variables were logarithmically transformed for analysis. WHR waist to hin ratio- RMT hndv mass index:?rp svstnlif! blood pressure; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; Apo, apolipoprotein. tp<0.05, *p<0.01, p< ±12.8 cm,^<0.01) and in the hip circumference (101.0±11.0 to cm, /?<0.05) considered separately. Of these 108 women, 58 had remained premenopausal and 31 had experienced natural menopause. There were nonsignificant changes in WHR over time for both groups, and there was no evidence that menopausal status affected the change. On average, the women gained 5.1 ±9.8 lb during the 3-year interval. Change in weight was unrelated to change in WHR (r=). However, change in weight was significantly related to change in waist and hip circumferences considered separately (r=0.42 and 0.53, respectively;/><0.001 for both). Table 4 presents the correlation between changes in WHR and changes in CHD risk factors. Pearson correlation coefficients, adjusting for changes in BMI, showed that those women who had the greatest increases in WHR had the greatest decreases in HDL 2 cholesterol (r=-0.24,/?<0.01) and the greatest increases in triglycerides (r=0.16,/?=0.053). The correlation between change in WHR and change in HDL 2 cholesterol remained significant when only women who were still premenopausal were considered (r=-0.31, p<0.01, unadjusted; and r=-0.32, p<0.0l, adjusted for BMI). Change in activity level was significantly correlated with change in WHR (r= -0.18, p< 0.05). The correlation remained significant after adjusting for change in BMI (r=-0.18,/><0.05). Change in intake was not related to change in WHR. Onlyfivewomen changed from smoking to nonsmoking status, precluding analysis of the effect of this factor on WHR.

5 1254 Arteriosclerosis and Thrombosis Vol 11, No 5 September/October 1991 HDL-T so mm mm II mni HDL II ill ii illi1 53J I ll ill! IS n-47 low Merf BUI (23J-25.8) High BM ) iii ill...11 I 21J2 n-17 Body Mass Index m I17.35 mi II m n-m J B12.79 Low BMI ((23.1) Mod BMI (23J-2S.8) High BMI 025.8) Body Mass Index W/H WOUI/HIp,ov W/H «.732) ^Med W/H ( ) CD High W/H 0.783) Discussion This study suggests that WHR, independent of BMI, is related to cardiovascular risk factors. WHR was found to be significantly and independently related to systolic blood pressure, total cholesterol, LDL cholesterol, HDL cholesterol and its subfractions, glucose, insulin, triglycerides, and apo B. Moreover, this study is the first to suggest that changes in WHR, independent of changes in BMI, may be related to changes in HDL 2 cholesterol; however, the change in WHR was not related to change in other cardiovascular risk factors. Second, this study shows that various behaviors, including smoking, activity, and caloric intake, are related to WHR and that the change in activity is related to a change in the WHR. Upper body fat obesity was also associated with higher tension and anxiety scores and greater reports of anger, lower social support, and more depressive symptomatology. Previous studies with samples of premenopausal and postmenopausal 710 women have shown that WHR is positively associated with triglycerides, blood pressure, apo B, glucose, and insulin levels. Moreover, these studies have shown a negative FIGURE 2. Bar gra/?/w showing total high density lipoprotein cholesterol (HDL-T) and HDL 2 (HDL-2) as a function of body mass index (BMI) and waist to hip ratio (WIH). association between WHR and HDL cholesterol and the HDL 2 subfraction. Although these correlations are not observed in all studies, the majority of the evidence does seem to suggest that WHR, independent of BMI, affects these CHD risk factors. The present study confirms these findings in one of the largest premenopausal cohorts of women studied to date. The correlation between WHR and BMI was 0.40, p< 0.001, and was extremely similar to the relation observed by others Moreover, those who had the greatest weight gain since age 20 had greater upper body fat distribution. Previous studies have shown that abdominal obesity is associated with fat cell hypertrophy, 43 characteristic of those with adultonset obesity. 44 However, we found no evidence that the weight changes exhibited over the 3 years of the study were related to changes in WHR, despite the fact that the women had gained 5 lb on average and 24 of the 108 women had gained 10 lb or more. In contrast, we found that there were significant increases in both the waist and hip circumferences considered separately and that the changes in these circumference measures were related to weight

6 Wingetal 20 WHR and CV Risk Factors 1255 Fasting Insulin IS ) m 7.99 i a-17 low BUI «231) J Med BUI ( ) n-47 High BUI ()25.8) Body Mass Index 2 Hoar Insulin FIGURE 3. Bar graphs showing fasting insulin (upper panel) and 2-hour insulin (lower panel) as a function of body mass index (BMI) and waist to hip ratio (WIH) Low BUI ((23J) Ued BUI (23J-25J) High BUI Q2S.8) Body Mass Index 8LOWW/HC<732) ^ M e f l W / H ( ) CDHlgh W/H <>.769> W/H Wam/Hlp change (usually weight gain). Thus, women during the perimenopausal period appear to be gaining weight but to be depositing it in both their waist and hip areas. Several other studies have examined the effect of weight change on WHR, with conflicting results Wadden et al,46 for example, studied 68 female participants in a weight control program after a weight loss of 12.3 kg. These women showed decreases in WHR, from to Others have found no change in WHR with weight loss.45 Results regarding the effect of menopause on WHR were unclear. When the total sample of women who had undergone natural menopause was compared with the sample who had remained premenopausal, significant differences were observed, indicating a higher WHR in postmenopausal women. However, when we used a within-subject analysis and examined the 31 women who were measured initially when premenopausal and then again later when postmenopausal, we found no significant change in WHR over time. The small sample of menopausal women may have obscured an effect of menopause; alternatively, differences between the group of premenopausal and postmenopausal women, other than their menopausal status, may have affected the between-group comparison. Furthermore, length of time since cessation of menses may also affect WHR. Previous data on the effect of menopause on WHR have also been unclear. Haffner et al47 found a weak relation between menopausal status and WHR, with menopausal status explaining 0.4% of the variance in WHR. Lanska et al48 found that menopausal status explained % of the variance. Recent studies suggest that when female monkeys are exposed to standardized stressors, they respond with hyperinsulinemia, hyperlipidemia, hypertension, and increased abdominal body fat distribution.21 Likewise, Bjorntorp22 reported that men and women with abdominal body fat distribution have a number of symptoms suggestive of difficulty coping with stress, including psychosomatic disease. From these data, Bjorntorp hypothesized that increased stress, resulting in increased levels of catecholamines and adrenal cortical hormones, may lead to symptoms of hypercortisolism, including abdominal body fat. We examined this hypothesis indirectly by looking at the relation between WHR and several standard psychosocial questionnaires measuring variables related to stress. In support of the

7 1256 Arteriosclerosis and Thrombosis Vol 11, No 5 September/October 1991 TABLE 4. Correlation Between Change in Waist to Hip Ratio and Change in Coronary Heart Disease Risk Factors CHD risk factor Total cholesterol Total HDL HDL 2 HDL 3 L.DL cholesterol Triglycerides SBP DBP Apo A-I Apo A-II Apo B Fasting insulin 2-Hour insulin Fasting glucose 2-Hour glucose Correlation with change inwhr * Correlation with change in WHR adjusting for change in BMI * t CHD, coronary heart disease; WHR, waist to hip ratio; BMI, body mass index; HDL, high density lipoprotein; LDL, low density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure; Apo; apolipoprotein. *p=0.007, tp= hypothesis, we found that WHR, independent of BMI, was associated with tension, anger, depressive symptomatology, and low social support. However, we did not find that WHR was correlated with a measure of current perceived stress. Negative emotions and lack of social support have been related to the risk for CHD morbidity and mortality. 49 " 51 The finding that WHR is related to standardized test scores measuring similar behaviors raises the possibility that the pathogenic influences of negative emotions and lack of social support may be due in part to their effects on body fat distribution. Finally, we find that activity, caloric intake, and smoking habits are all related to body fat distribution and that those individuals who increased their activity level showed decreases in WHR. These findings are of interest because they suggest that by changing behavior, it may be possible to improve WHR. However, the effect of a change in WHR on a change in cardiovascular risk factors is less clear. Change in WHR was associated with decreased HDL 2 cholesterol and marginally with increased triglycerides. It is possible that with greater changes in weight or activity, there would be larger changes in WHR, making an association between change in WHR and change in risk factors more apparent. Alternatively, it is possible that WHR is to some extent modifiable, but that it is the genetic 52 or nonmodifiable component of WHR that is related to cardiovascular disease. Further studies are needed to investigate the relation between changes in WHR and changes in cardiovascular risk factors. Thus, in conclusion, WHR seems to be an integral part of the cardiovascular risk profile and is affected by behavioral and psychosocial variables previously found to be associated with CHD. References l. Ducimetiere P, Richard J, Cambien F: The pattern of subcutaneous fat distribution in middle-aged men and the risk of coronary heart disease: The Paris Prospective Study. IntJObes 1986; 10: Larsson B, Svardsudd K, Welin L, Wilhelmsen L, Bjorntorp P, Tibblin G: Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: 13-Year follow-up of participants in the study of men born in Br Med J 1984;288: Lapidus L, Bengtsson C, Larsson B, Pennert K, Rybo E, Sjostrom L: Distribution of adipose tissue and risk of cardiovascular disease and death: 12-Year follow-up of participants in the population study of women in Gothenburg, Sweden. Br MedJ 1984;289: Donahue RP, Abbott RD, Bloom E, Reed DM, Yano K: Central obesity and coronary heart disease in men. Lancet 1987;l: Kalkhoff RK, Hartz AH, Rupley D, Kissebah AH, Kelber S: Relationship of body fat distribution to blood pressure, carbohydrate tolerance, and plasma lipids in healthy women. J Lab Clin Med 1983;102: Baumgartner RN, Roche AF, Chumlea C, Siervogel 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: Soler JT, Folsom AR, Kushi LH, Prineas RJ, Seal US: Association of body fat distribution with plasma lipids, lipoproteins, apolipoproteins AI and B in postmenopausal women. J Clin Epidemiol 1988;41: Despres JP, Allard C, Tremblay A, Talbot J, Bouchard C: Evidence for a regional component of body fatness in the association with serum lipids in men and women. Metabolism 1985;34: Haffner SM, Stern MP, Hazuda HP, Rosenthal M, Knapp AJ: The role of behavioral variables and fat patterning in explaining ethnic differences in serum lipids and lipoproteins. Am J Epidemiol 1986;123: Ostlund RE Jr, Staten M, Kohrt WM, Schultz J, Malley M: The ratio of waist-to-hip circumference, plasma insulin level, and glucose intolerance as independent predictors of the HDL 2 cholesterol level in older adults. N Engl J Med 1990; 332: Peiris AN, Sothmann MS, Hennes MI, Lee MB, Wilson CR, Gustafson AB, Kissebah AH: Relative contribution of obesity and body fat distribution to alterations in glucose insulin homeostasis: Predictive values of selected indices in premenopausal women. Am J Clin Nutr 1989;49: Sparrow D, Borkan GA, Gerzof SG, Wisniewski C, Silbert CK: Relationship of fat distribution to glucose tolerance: Results of computed tomography in male participants of the normative aging study. Diabetes 1986;35: Evans DJ, Hoffmann RG, Kalkhoff RK, Kissebah AH: Relationship of body fat topography to insulin sensitivity and metabolic profiles in premenopausal women. Metabolism 1984; :68-75 Ohlson LO, Larsson B, Svardsudd K, Welin L, Eriksson H, Wilhelmsen L, Bjorntorp P, Tibblin G: 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 born in Diabetes 1985;34: Barrett-Connor E, Khaw KT: Cigarette smoking and increased central adiposity. Ann Intern Med 1989;111: Shimokata H, Muller DC, Andres R: Studies in the distribution of body fat: III. Effects of cigarette smoking. JAMA 1989;261: Laws A, Terry RB, Barrett-Connor E: Behavioral covariates of waist-to-hip ratio in Rancho Bernardo. Am J Public Health 1990;80:

8 18. Kaye SA, Folsom AR, Prineas RJ, Potter JD, Gapstur SM: The association of body fat distribution with lifestyle and reproductive factors in a population study of postmenopausal women. IntJ Obes 1990;14: Krotkiewski M: Can body fat patterning be changed? Ada Med Scand 1987;723: Tremblay A, Despres JP, Leblanc C, Craig CL, Ferris B, Stephens T, Bouchard C: Effect of intensity of physical activity on body fatness and fat distribution. Am J Clin Nutr 1990;51: Kaplan JR, Adams MR, Clarkson TB, Koritnik DR: Psychosocial influences on female protection among cynomolgus macaques. Atherosclerosis 1984;53: Bjorntorp P: Abdominal obesity and the development of noninsulin-dependent diabetes mellitus. Diabetes Metab Rev 1988;4: Matthews DA, Kelsey SF, Meilahn EN, Kuller LH, Wing RR: Educational attainment and behavioral and biological risk factors for coronary heart disease in middle aged women. Am J Epidemiol 1989;129: Allain CC, Poon LS, Chan CS, Richmond W, Fu PC: Enzymatic determination of total serum cholesterol. Clin Chem 1974;20: Warnick GR, Albers JJ: A comprehensive evaluation of the heparin-manganese precipitation procedure for estimating high density lipoprotein cholesterol. J Lipid Res 1978;19: Gidz LI, Miller GJ, Burstein M, Eder HA: Analyses of plasma high density lipoprotein subclasses by a precipitation procedure: Correlations with preparative and analytical centrifugation, in Lippel K (ed): Report of the High Density Lipoprotein Methodology Workshop, San Francisco, California, March 12, 14, and 14, US Government Printing Office, Bethesda, Md, Department of Health, Education, and Welfare, Public Health Service, National Institutes of Health (NIH publication No ), 1979, pp Bucolo G, David H: Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem 1973;19: Friedewald WT, Levy RI, Fredrickson DS: Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18: Medoza SG, Zerpa A, Carrasco H, Colmeneres O, Rangel A, Gartside KS, Kashyap ML: Estradiol, testosterone, apolipoproteins, lipoprotein cholesterol, and lipolytic enzymes in men with premature myocardial infarction and angiographically assessed coronary occlusion. Artery 1983;12:l Stein EA, DiPersio L, Pesce AJ: Enzyme-linked immunoabsorbant assay of apolipoprotein All in plasma, with use of a monoclonal antibody. Clin Chem 1986;32: Herbert V, Lau K-S, Gottlieb CW, Bleicher SJ: Coated charcoal immunoassay of insulin. / Clin Endocrinol Metab 1965;25: Garrow JS: Zero-muddler for unprejudiced sphygmomanometry. Lancet 1963;2: Dischinger P, DuChene AG: Quality control aspects of blood pressure measurements in the Multiple Risk Factor Intervention Trial. Controlled Clin Trials 1986;7(suppl):137S-157S 34. Paffenbarger RS Jr, Wing AL, Hyde RT: Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol 1978;108: Wing et al WHR and CV Risk Factors Adams CF: Nutritive Value of American Foods in Common Units. Washington, DC, US Government Printing Office, Agricultural Research Service, Department of Agriculture, Agriculture handbook No. 456, Nutrition Coordinating Center: Reference Food Table, version 10. Minneapolis, Minn, University of Minnesota, Haynes SG, Levine S, Scotch N, Feinleib M, Kannel WB: The relationship of psychosocial factors to coronary heart disease in the Framingham study. Am J Epidemiol 1978;107: Beck AT, Beamesderfer A: Assessment of depression: The depression inventory, in Pinchot P (ed): Psychological Measurements in Psychopharmacology, Volume 7. Basel, Karger, 1974, pp Spielberger CD, Johnson EH, Russell SF, Crane RJ, Jacobs GA, Wordon RJ: The experience and expression of anger, in Chesney MA, Goldston SE, Rosenman RH (eds): Anger and Hostility in Behavioral Medicine. New York, Hemisphere/ McGraw-Hill Book Co, Matthews KA, Meilahn E, Kuller LH, Kelsey SF, Caggiula AW, Wing RR: Menopause and risk factors for coronary heart disease. N Engl J Med 1989;321: Haffner SM, Stern MP, Hazuda HP, Pugh J, Patterson JK: Do upper-body and centralized adiposity measure different aspects of regional body-fat distribution? Diabetes 1987;36: Seidell JC, Bakx JC, De Boer E, Deurenberg P, Hautvast JGAJ: Fat distribution of overweight persons in relation to morbidity and subjective health. Int J Obes 1985;9: Kissebah AH, Vydelingum N, Murray R, Evans DJ, Hartz AJ, Kalkhoff RK, Adams PW: Relation of body fat distribution to metabolic complications of obesity. / Clin Endocrinol Metab 1982;54: Sjostrom L: Fat cells and body weight, in Stunkard AJ (ed): Obesity. Philadelphia, WB Saunders Co, 1980, pp Garrow JS: Is body fat distribution changed by dieting? Ada Med Scand 1987;723: Wadden TA, Stunkard AJ, Johnston FE, Wang J, Pierson RN, Van Itallie TB, Costello E, Pena M: Body fat deposition in adult obese women: II. Changes in fat distribution accompanying weight reduction. Am J Clin Nutr 1988;47: Haffner SM, Stern MP, Hazuda HP, Pugh J, Patterson JK, Malina R: Upper body and centralized adiposity in Mexican Americans and Non-Hispanic whites: Relationship to body mass index and other behavioral and demographic variables. IntJ Obes 1986;10: Lanska DJ, Lanska MJ, Hartz AJ, Rimm AA: Factors influencing anatomic location of body fat in 52,953 women. Int J Obes 1985;9: Booth-Kewley S, Friedman HS: Psychological predictors of heart disease: A quantitative review. Psychol Bull 1987;101: Haynes SG, Feinleib M, Kannel WB: The relationship of psychosocial factors to coronary heart disease in the Framingham study: III. Eight-year incidence of coronary heart disease. Am J Epidemiol 1980;lll: Matthews KA: Coronary heart disease and type A behaviors: Update on and alternative to the Boot-Kewley and Friedman (1987) quantitative review. Psychol Bull 1988;104: Bouchard C: Genetic factors in obesity. Med Clin North Am 1989;73:67-81 KEY WORDS menopausal status cardiovascular risk factors waist to hip ratio psychosocial factors behavioral factors