¹Division of Preventive Medicine, Boston, Massachusetts, USA

²Department of Epidemiology, Harvard School of Public Health, Harvard, Massachusetts, USA

³Department of Ambulatory Care and Prevention, Harvard Medical School, Harvard, Massachusetts, USA

⁴Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Harvard, Massachusetts, USA

Correspondence to: K M Rexrode, Division of Preventive Medicine, 900 Commonwealth Avenue, East, Boston MA 02215, USA. Email: krexrode@partners.org

BACKGROUND: Waist circumference is a simpler measure of abdominal adiposity than waist/hip ratio (WHR), but few studies have directly compared the two measures as predictors of coronary heart disease (CHD) in men. In addition, whether the association of abdominal adiposity is independent of total adiposity as measured by body mass index (BMI) in men remains uncertain.

OBJECTIVE: To compare waist circumference and WHR as predictors of CHD in men, and to determine whether the association is independent of BMI.

DESIGN: Prospective cohort study.

METHODS: We compared WHR, waist circumference and BMI with risk of CHD (myocardial infarction or coronary revascularization) among men in the Physicians' Health Study, a randomized trial of aspirin and beta-carotene among 22 071 apparently healthy US male physicians, aged 40-84 y at baseline in 1982. Men reported height at baseline, and weight, waist and hip measurements on the 9 y follow-up questionnaire.

RESULTS: Among the 16 164 men who reported anthropometric measurements and were free from prior CHD, stroke or cancer, a total of 552 subsequent CHD events occurred during an average follow-up of 3.9 y. After adjusting for age, randomized study agent, smoking, physical activity, parental history of myocardial infarction, alcohol intake, multivitamin and aspirin use, men in the highest WHR quintile (0.99) had a relative risk (RR) for CHD of 1.50 (95% CI 1.14-1.98) compared with those in the lowest quintile (<0.90). Men in the highest waist circumference quintile (103.6 cm) had a RR of 1.60 (CI, 1.21-2.11) for CHD compared with men in the lowest quintile (<88.4 cm). Further adjustment for BMI substantially attenuated these associations: men in the highest WHR and waist circumference quintiles had relative risks for CHD of 1.23 (CI, 0.92-1.66) and 1.06 (CI, 0.74-1.53), respectively. Men in the highest BMI quintile (27.6 kg/m²) had a multivariate RR of CHD of 1.73 (CI, 1.29-2.32), after adjustment for WHR. No significant effect modification by age of the relationship between either measure of abdominal adiposity and risk of CHD was observed.

CONCLUSIONS: These data support a modest relationship between abdominal adiposity, as measured by either WHR or waist circumference, and risk of CHD both in middle-aged and older men. However, abdominal adiposity did not remain an independent predictor of CHD after adjustment for BMI.

International Journal of Obesity (2001) 25, 1047-1056

obesity; coronary heart disease; fat distribution; anthropometry; epidemiology