International Journal of Obesity
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Contact Springer Nature
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
August 2001, Volume 25, Number 8, Pages 1183-1188
Table of contents    Previous  Article  Next   [PDF]
Waist circumference vs body mass index for prediction of disease risk in postmenopausal women
R E Van Pelt1,2, E M Evans1, K B Schechtman1,3, A A Ehsani1 and W M Kohrt1,2

1Department of Internal Medicine, Washington University School of Medicine, St Louis, Missouri, USA

2Department of Medicine, Division of Geriatric Medicine, University of Colorado Health Sciences Center, Denver, Colorado, USA

3Divisions of Geriatrics/Gerontology and Biostatistics, Washington University School of Medicine, St Louis, Missouri, USA

Correspondence to: W M Kohrt, Division of Geriatric Medicine, University of Colorado Health Sciences Center, 4200 E Ninth Ave, Campus Box B-179, Denver, CO 80262, USA. E-mail:


OBJECTIVE: To test the sensitivity of waist circumference (central adiposity) as an index of disease risk in postmenopausal women.

DESIGN: Retrospective analysis of postmenopausal women tested at Washington University School of Medicine.

SUSBJECTS: A total of 323 healthy postmenopausal (66±5 y; mean±s.d.) women not using any hormone replacement.

MEASUREMENTS: Body composition, hyperinsulinemia (insulin area), triglycerides and HDL-cholesterol.

RESULTS: Excess waist size had a stronger association with hyperinsulinemia and hypertriglyceridemia than body mass index (BMI; kg/m2) in otherwise healthy, postmenopausal women. After adjusting for BMI, a strong relation existed between waist circumference and insulin area, HDL-cholesterol and triglycerides (P<0.01). Conversely, after adjusting for waist circumference, no relation was apparent between BMI and the dependent variables of interest. The strength of the association between waist circumference and disease risk became most apparent when analyses were restricted to normal-weight women (BMI 24-28 kg/m2). When BMI was held constant, hyperinsulinemia and triglyceridemia increased dose-dependently with changes in waist size.

CONCLUSION: Waist circumference, an easily obtained index of central adiposity, is a more sensitive measure of relative disease risk than is BMI in middle-aged and older women, particularly in normal-weight individuals.

International Journal of Obesity (2001) 25, 1183-1188


waist circumference; disease risk; women


A high body mass index (BMI, kg/m2) is associated with increased mortality from all causes and from cardiovascular disease in middle-aged and older adults.1,2 Based on this association health professionals are currently using BMI as a means of categorizing a patient's relative risk of disease. However, it is becoming increasingly clear that body fat distribution is a more important determinant of disease risk than body mass. Individuals with a high proportion of abdominal fat have a greater risk of developing type 2 diabetes mellitus,3,4 coronary artery disease (CAD),5 and cardiovascular disease (CVD)-related mortality.6 Although the mechanism is unknown, it is postulated that central adiposity, specifically excess visceral adipose tissue, is associated with CAD and CVD-related mortality primarily through its impact on insulin resistance and dyslipidemia.7,8

As women progress through the menopause, changes in body fat distribution may contribute to an increase in disease risk. Abdominal adiposity increases markedly after the menopause,9 as does the incidence of cardiovascular and metabolic diseases.5,10 The menopause-related increase in adiposity is countered by an accelerated loss of lean mass,11 such that body weight does not change significantly.12 Thus, through these transitional years, BMI is likely to be an insufficient measure of relative risk of disease and mortality. Waist circumference appears to be the best simple anthropometric measure of abdominal fat and, importantly, visceral adipose tissue in younger women13 and may potentially be a more sensitive measure of disease risk than BMI in postmenopausal women. Furthermore, a large waist circumference has been linked to an impairment in general health and quality of life in young and middle-aged women.14 Therefore, we sought to determine in a large cohort of healthy postmenopausal women whether waist circumference: (1) is related to three important predictors of disease risk (hyperinsulinemia, triglycerides and high-density lipoprotein (HDL)-cholesterol); (2) is a more robust index of disease risk than BMI; and (3) maintains a linear (dose-dependent) relation with disease risk in normal-weight individuals.



Body composition and CVD risk factors were measured in 323 healthy postmenopausal (66±5 y) women recruited from the greater St Louis metropolitan area. All women were at least 2 y past menopause (18±7 y), not using any type of hormone replacement, non-smokers and free of overt heart disease, as assessed by resting and exercise 12-lead ECG, and diabetes mellitus, as assessed by an oral glucose tolerance test (OGTT).15 All of the participants provided written informed consent to participate in the study, which was approved by the Washington University Institutional Review Board.

Body composition

Hydrodensitometric and anthropometric measures of body composition were performed as previously described.16 Briefly, body density was measured using underwater weighing and body fat percentage was estimated from body density using the equation of Brozek et al.17 Waist circumference was measured in triplicate at the mid-point between the distal border of the ribs and the top of the iliac crest. Weight and height were measured using a physician's balance scale and a stadiometer respectively. All measurements were made by trained research technicians.

Blood lipids and lipoproteins

Measurement of serum lipid and lipoprotein concentrations was performed in the Core Laboratory for Clinical Studies at Washington University. Cholesterol and glycerol-blanked triglycerides were measured by automated enzymatic commercial kits (Miles/Technicon, Tarrytown, NY). HDL-cholesterol was measured in plasma after precipitation of apolipoprotein B-containing lipoproteins by dextran sulfate (50 000 MW) and magnesium.18 Low-density lipoprotein (LDL)-cholesterol was calculated using the Friedewald equation.19 These methods are continuously standardized by the Lipid Standardization Program of the Centers for Disease Control and Prevention.

Glucose tolerance test

Glucose tolerance was determined using a 75 g OGTT, performed in the morning after an overnight fast. Diet was monitored for 3 days prior to the OGTT to ensure an intake of >150 g of carbohydrate per day. Blood samples (3.0 mL) were obtained for the determination of glucose (glucose oxidase method; Beckman glucose analyzer) and insulin20 immediately before and 30, 60, 90, 120 and 180 min after the glucose challenge. The total areas under the glucose and insulin curves were calculated using the trapezoidal rule.


The primary outcome variables (insulin area, triglyceride and HDL-cholesterol) for analysis in this study were chosen a priori. Data were analyzed using two approaches to determine whether waist circumference or BMI was a more important predictor of the outcome measures of interest. The first used partial correlation coefficients between: (1) BMI and the outcome measures after adjusting for waist size, and (2) waist size and the outcome measures after adjusting for BMI. The relative value of these coefficients and of the associated significance levels provided an assessment of whether one predictor was more important than the other. The second approach began by dividing waist size into tertiles, thereby producing a middle tertile with a narrow range that, for all practical purposes, produced a sample of women with similar and normal waist size. Within this tertile of women with normal waist size, the significance of BMI as a predictor of the outcome measures was assessed using regression analysis. Similarly, using the middle tertile of essentially similar and normal BMI values, regression analysis was used to evaluate the significance of waist size. The contrast between (1) the significance of BMI when waist size was held constant and (2) the significance of waist size when BMI was held constant provided a second assessment of the relative importance of the two predictors. Because the variables were skewed, all analyses involving insulin area and triglycerides were performed following a logarithmic transformation. All data were analyzed using SAS and results are presented as mean±s.e. unless stated otherwise.


Body composition, lipid and glucose tolerance tests were performed in 323 postmenopausal women. Subject characteristics are provided in Table 1. Both BMI and waist circumference were strongly and significantly (P<0.01) correlated with percentage body fat (r=0.91 and 0.85, respectively).

To assess the relative importance of waist circumference and BMI as predictors of three prospectively chosen variables (OGTT insulin area, triglycerides and HDL-cholesterol), two related types of analyses were performed. First, Table 2 summarizes the results of partial correlation analyses between waist (after adjusting for BMI) or BMI (after adjusting for waist) and the three dependent variables of interest. After adjusting for BMI, a strong relation existed between waist circumference and insulin area, HDL-cholesterol and triglycerides (P£0.0l). Conversely, after adjusting for waist circumference, no significant relation was apparent between BMI and the dependent variables of interest.

For the second part of the analyses, waist circumference and BMI data were grouped by tertiles (Table 3). As anticipated, a strong dose-dependent relation existed between waist circumference or BMI and the outcome measures (P<0.001 in all cases) when these variables were considered separately. However the increased predictive value of waist circumference, as compared to BMI, is apparent in the fig Women who fell into the middle waist and BMI tertiles were further divided into sub-tertiles to test the hypothesis that waist circumference is a more sensitive predictor of disease risk than BMI within a narrow range of body size. Figure 1A contains data for the 110 subjects in the middle tertile for BMI (range of 24.5-27.9 kg/m2) and indicates a consistent and marked increase in insulin area with increasing waist size, (7100±687 vs 8950±512 vs 11 709±1031 µU/mL min, P<0.001). Figure 1B, on the other hand, contains the data for the 109 subjects in the middle waist tertile (range of 78.8-88.0 cm). In contrast with Figure 1A, insulin areas were not different across groups and actually declined slightly, from 9112±714 µU/mL min in the lowest BMI group to 8950±400 and 8036±676 µU/mL min in the middle and highest BMI, respectively. Similar relations were observed between waist or BMI group and serum triglycerides. Among the subjects in the middle tertile for BMI, serum triglycerides ranged from 128±21 mg/dL in the lowest to 142±13 and 149±16 mg/dL in the middle and highest waist group, respectively (Figure 1C). In contrast, there was no association between BMI and triglycerides for women in the middle waist tertile (Figure 1D). Serum HDL-cholesterol levels were not influenced by either waist size or BMI within the middle tertile groups (Figure 1E and 1F).


We investigated the relation between waist size and markers of insulin resistance and dyslipidemia in a large cohort of postmenopausal women. The results of this study indicate that: (1) waist size is more strongly associated with disease risk than is BMI in healthy, postmenopausal women; (2) this association is linear, such that serum insulin area and triglycerides are positively related, and HDL-cholesterol concentration is negatively related with waist size; and (3) the strong positive association between waist size and both insulin area and triglycerides continues to hold when the analyses are restricted to normal weight women within a narrow range of BMI (24-28 kg/m2). That is, increased waist size indicates increased risk, when BMI is held constant and normal. However, in marked contrast, when analyses are restricted to women within a narrow range of waist size, BMI is not an important predictor of risk.

Body mass tends to increase with age in adults and is associated with increased risk of disease and mortality.1,2 However, weight gain is not dramatic among women at the time of menopause. This was previously demonstrated in a longitudinal study by Wing et al in which the body weight of 485 middle-aged women was followed over a 3 y period.12 After 3 y, weight gain was not significantly different in women who remained premenopausal compared with those who had undergone natural menopause (+2.07 vs+1.35 kg). On the other hand, changes in body composition become more readily apparent as women progress through the menopause. A 6 y study by Poehlman et al found that premenopausal women who became postmenopausal during the period of the study had a 3.0±1.1 kg decrease in fat-free mass and a 2.5±2.0 kg increase in fat mass.11 Age-matched women who remained premenopausal through the period of the study had non-significant changes in fat-free mass and fat mass (-0.5±0.5 and +1.0±1.5 kg, respectively). Similarly, a study by Bjorkelund and colleagues found that women going through the menopause had greater increases in waist circumferences and waist-hip ratios than women who remained pre-menopausal. Moreover, both groups had identical changes in BMI within this transitional period.21

Clearly increases in fat mass, rather than body mass per se, cause deleterious effects on the metabolic profile of aging women. Fat mass accumulation further adds to the risk of disease as abdominal obesity is associated with greater incidence of type 2 diabetes mellitus, disturbances in lipid metabolism7,8 and increased mortality6 than is total degree of adiposity. The body fat distribution of women changes significantly at the time of menopause, with a shift from preferential storage in gluteal/femoral regions to abdominal depots. Studies suggest that it is declining sex hormones, rather than aging, that trigger the menopause-related increase in abdominal adiposity22,23 and the accompanying unfavorable changes in the metabolic profile.24,25 However, prospective studies are needed to determine whether the increase in abdominal adiposity with the menopause involves a preferential deposition of fat in visceral regions. It is likely that greater central fat accumulation is due to increases in visceral adipose tissue and that this contributes to the increase in risk for cardiovascular and metabolic abnormalities that occurs after the menopause.

It is important to emphasize that an increase in abdominal fat may not be an inevitable consequence of aging or menopause. Lifestyle factors, such as regular physical activity, may prevent or at least attenuate gains in abdominal adiposity. This is supported by the observation that women who are physically active demonstrate less of an age-related increase in trunk fat mass compared with their sedentary peers.26 Furthermore, menopause-related increases in abdominal fat are potentially reversible. This is supported by evidence that postmenopausal women can improve their pattern of fat distribution with regular exercise.27 There are also data to suggest that certain hormone replacement regimens may attenuate accumulation of abdominal fat after menopause.23,25,28

Advances in technology have allowed investigators to distinguish between centrally located visceral and subcutaneous adipose tissue using computed tomography (CT) or magnetic resonance imaging (MRI). However, measurement of visceral adipose tissue by CT or MRI is impractical for health screening. Both BMI and waist circumference are good 'field measures' of adiposity that are easily measured in clinical settings. Results of the current study suggest that waist circumference is a more sensitive indicator of disease risk than BMI in postmenopausal women. It might be argued that waist circumference is simply a better marker of total adiposity than BMI and that this accounts for its sensitivity as a predictor of disease risk. However, in our postmenopausal women both waist circumference and BMI were equally strong predictors of fat mass, suggesting that the difference between the two measures lies in the ability of the waist circumference to capture abdominal adiposity. This possibility is supported by the fact that waist circumference has been validated against CT as the best simple estimate of visceral adipose tissue.13

In the current study, risk of disease increased as waist size increased from lowest (<78.5 cm) to middle (78.5-88.0 cm) to highest (>88.0 cm) in postmenopausal women. These waist tertiles are consistent with levels previously reported in a study by Lean et al14 in which larger waist circumferences were associated with impairments in health and quality of life indexes in young and middle-aged men and women. In that study, women who had waist circumferences in excess of 88.0 cm had greater incidence of type 2 diabetes as well as increased difficulties with lower back pain and physical functioning than women with lower waist circumferences. Although it may be clear that women with waist girths greater than 88.0 cm are at a much higher risk of disease and functional impairment, the same could be said for women whose BMI is greater than 28 kg/m2. The present study emphasizes the impact of waist size on disease risk independent of BMI. Women in this study who did not differ with respect to BMI, but differed in waist size demonstrated significant differences in metabolic profile. On the other hand, the disease risk profiles of women with similar waist size, but different BMI were not distinguishable. It is important that clinicians understand the negative impact that increases in waist size can have on disease risk independent of changes in body weight or BMI.

In summary, because waist size is (1) associated with metabolic markers of disease risk in postmenopausal women and (2) easy to measure, we recommend that healthcare professionals monitor changes in waist circumference in postmenopausal women in conjunction with body weight and standard metabolic screening. This practice should not be limited to obese women because even small increases in abdominal adiposity in normal-weight women may contribute to unfavorable changes in the metabolic profile.


This research was supported by the following awards from the National Institutes of Health: Claude Pepper Older Americans Independence Center, AG13629; Research Career Development, AG00663 (Kohrt); General Clinical Research Center, RR00036; and Diabetes Research and Training Center, DK20579.


1 Manson J, Willet W, Mier J et al. Body weight and mortality among women. New Engl J Med 1995; 333: 677-685, MEDLINE

2 Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The effect of age on the association between body-mass index and mortality. New Engl J Med 1998; 338: 1-7, MEDLINE

3 Ohlson L-O, Larsson B, Sva¨rdsudd 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 born in 1913. Diabetes 1985; 34: 1055-1058, MEDLINE

4 Seidell J, Han T, Feskens E, Lean M. Narrow hips and broad waist circumferences independently contribute to increased risk of non-insulin-dependent diabetes mellitus. J Intern Med 1997; 242: 401-406, MEDLINE

5 Rexrode K, Carey V, Hennekens C et al. Abdominal adiposity and coronary heart disease in women. JAMA 1998; 280: 1843-1848, MEDLINE

6 Folsom A, Kaye S, Sellers T et al. Body fat distribution and 5-year risk of death in older women. JAMA 1993; 269: 483-487, MEDLINE

7 Bjo¨rntorp P. Body fat distribution, insulin resistance, and metabolic diseases. Nutrition 1997; 13: 795-803, Article MEDLINE

8 Despres J-P. The insulin resistance-dyslipidemic syndrome of visceral obesity: effect on patient's risk. Obes Res 1998; 6: S8-S17,

9 Ley CJ, Lees B, Stevenson JC. Sex- and menopause-associated changes in body-fat distribution. Am J Clin Nutr 1992; 55; 950-954,

10 Kannel WB, Wilson PW. Risk factors that attenuate the female coronary disease advantage. Arch Intern Med 1995; 155: 57-61, MEDLINE

11 Poehlman ET, Toth MJ, Gardner AW. Changes in energy balance and body composition at menopause: a controlled longitudinal study. Ann Intern Med l995; 123: 673-675,

12 Wing R, Matthews K, Kuller L, Meilahn E, Plantinga P. Weight gain at the time of menopause. Arch Intern Med 1991; 151: 97-102, MEDLINE

13 Poiliot MC, Despres J-P, Lemieux S et al. Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 1994; 73: 460-468, MEDLINE

14 Lean ME, Han TS, Seidell JC. Impairment of health and quality of life in people with large waist circumference. Lancet 1998; 351: 853-856, Article MEDLINE

15 American Diabetes Association. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 1998; 21: S5-S19,

16 Kohrt WM, Malley MT, Dalsky GP, Holloszy JO. Body composition of healthy sedentary and trained, young and older men and women. Med Sci Sports Exerc 1992; 24: 832-837, MEDLINE

17 Brozek J, Grande F, Anderson JT, Keys A. Densitometric analysis of body composition: revision of some quantitative assumptions. Ann NY Acad Sci 1963; 110: 113-140,

18 Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantification of high-density lipoprotein cholesterol. Clin Chem 1982; 28: 1379-1388, MEDLINE

19 Friedewald W, Levy R, Fredrickson D. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without the use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499-502, MEDLINE

20 Morgan DR, Lazarow A. Immunoassay of insulin: two antibody system. Diabetes 1963; 12: 115-126,

21 Bjorkelund C, Lissner L, Andresson S, Lapidus L, Bengtsson C. Reproductive history in relation to relative weight and fat distribution. Int J Obes Relat Metab Discord 1996; 20: 213-219,

22 Espeland MA, Stefanick ML, Kritz-Silverstein D et al. Effect of postmenopausal hormone therapy on body weight and waist and hip girths. J Clin Endocrinol Metab l997; 82: 1549-1556,

23 Gambacciani M, Ciaponi M, Cappagli B et al. Body weight, body fat distribution, and hormonal replacement therapy in early postmenopausal women. J Clin Endocrinol Metab 1997; 82: 414-417, MEDLINE

24 The Writing Group for the PEPI Trial. Effects of hormone replacement therapy on endometrial histology in postmenopausal women. The Postmenopausal Estrogen/Progestin Inverventions (PEPI) Trial. JAMA 1996; 275: 370-375, MEDLINE

25 Samaras K, Hayward C, Sullivan D, Kelly R, Campbell L. Effects of postmenopausal hormone replacement therapy on central abdominal fat, glycemic control, lipid metabolism, and vascular factors in type 2 diabetes: a prospective study. Diabetes Care 1999; 22: 1401-1407, MEDLINE

26 Van Pelt R, Davy K, Stevenson E et al. Smaller differences in total and regional adiposity with age in regularly exercising compared with sedentary women. Am J Physiol (Endocrin Metab) 1998; 275; E626-E634,

27 Kohrt WM, Obert KA, Holloszy JO. Exercise training improves fat distribution patterns in 60- to 70-yr-old men and women. J Gerontol 1992; 47: M99-M105, MEDLINE

28 Kohrt WM, Ehsani AA, Birge SJ. HRT preserves increases in bone mineral density and reductions in body fat after a supervised exercise program. J Appl Physiol 1998; 84: 1506-1512, MEDLINE


Figure 1 Insulin area (A, B), triglycerides (C, D) and HDL-cholesterol (E, F) for the middle BMI tertile stratified by waist (left panel) and the middle waist tertile stratified by BMI (right panel) are presented (mean±s.e.). This figure demonstrates that, when BMI is constant (24.4-27.9 kg/m2) among individuals, differences in waist circumference predict insulin area and triglycerides. However, when waist is constant (78.5-88.0 cm) among individuals, differences in BMI do not predict insulin area and triglycerides.


Table 1 Subject characteristics

Table 2 Relative importance of waist circumference and BMI as predictors of serum OGTT insulin area, triglycerides and HDL-cholesterol. Entries are partial correlation coefficients

Table 3 Association between tertiles of waist circumference or BMI and insulin area, triglycerides, and HDL-cholesterol

Received 29 July 2000; revised 5 December 2000; accepted 23 January 2001
August 2001, Volume 25, Number 8, Pages 1183-1188
Table of contents    Previous  Article  Next    [PDF]