The relationship between body mass index and waist circumference: implications for estimates of the population prevalence of overweight

OBJECTIVE: Body mass index (BMI) based on self-reported height and weight i s a systematically biased, but acceptable measure of adiposity and is commo nly used in population surveys. Recent studies indicate that abdominal obes ity is more strongly associated with obesity-related health problems than i s adiposity measured by BMI. The purpose of this study was to determine the relationships of both measured and self-reported BMI with measured waist c ircumference in a randomly selected sample of Australian adults. DESIGN: Cross-sectional survey with self-reported and laboratory-based meas ures of adiposity. SUBJECTS: 1140 randomly-selected Australian adults aged 18-78 y resident in the city of Adelaide, South Australia. MEASUREMENTS: Data on self-reported and measured height and weight as well as measured waist circumference were drawn from the Pilot Survey of the Fit ness of Australians database. The proportion of men and women with acceptab le BMI (BMI less than or equal to 25 kg/m(2)) and with excess abdominal adi posity (greater than or equal to 94 cm for men and greater than or equal to 80 cm for women) was determined. Differences in the prevalence of overweig ht based on BMI alone or BMI and waist circumference were also determined. RESULTS: Compared with the prevalence based on self-reported BMI alone, the prevalence of overweight among men based on self-reported BMI and waist ci rcumference combined was 2.4%, 5.3%, 19.1% and 7.5% greater for men aged 18 - 39 y, 40 - 59 y, 60 - 78 y and for all men, respectively. Among women, c ompared with the prevalence based on self-reported BMI alone, the prevalenc e of overweight based on the combined measures was 9.9%, 24.0%, 33.3% and 2 0.6% greater for women aged 18 - 39 y, 40 - 59 y, 60 - 78 y and for all wom en, respectively. CONCLUSIONS: If waist circumference is used as the criterion, then the prev alence of overweight among Australian adults, and probably other Caucasian populations, may be significantly greater than indicated by surveys relying on self-reported height and weight. The development of valid self-reported measures of waist circumference for use in population surveys may allow mo re accurate epidemiological monitoring of overweight and obesity.