Association of body mass index and waist-hip circumference ratio with cardiovascular disease risk factors in Micronesian Nauruans
Hodge, A.M.; Dowse, G.K.; Zimmet, P.Z.
International Journal of Obesity and Related Metabolic Disorders Journal of the International Association for the Study of Obesity 17(7): 399-407
1993
ISSN/ISBN: 0307-0565 PMID: 8395475 Document Number: 413558
The relationship of body mass index (BMI) and waist-hip circumference ratio (WHR) with cardiovascular disease (CVD) risk factors was examined in adult Nauruans, a population with a particularly high prevalence of noninsulin dependent diabetes mellitus and obesity, with mean BMI 34.2 kg/m-2 in males and 34.9 kg/m-2 in females aged gtoreq 20 years. In univariate analysis for men, fasting and 2 h insulin (after a 75 g glucose load), total cholesterol, high density lipoprotein (HDL)/total cholesterol ratio and uric acid levels varied significantly across tertiles of BMI but not WHR, while fasting and 2 h glucose and fasting triglyceride levels were significantly associated with both BMI and WHR. Neither systolic nor diastolic blood pressure were related to BMI tertiles, but diastolic blood pressure was associated with WHR. High density lipoprotein cholesterol was not associated with tertiles of BMI or WHR in men. In women, BMI and WHR had similar univariate associations. When the effects of BMI and WHR were studied simultaneously by analysis of covariance, adjusting for age. BMI explained more of the variance in risk factor levels in men than did WHR for fasting and 2 h insulin, total cholesterol, HDL cholesterol, HDL/total cholesterol ratio and uric acid, while in women, BMI was more important for 2 h insulin, 2 h glucose and uric acid, and equal with WHR for fasting insulin. Waist-hip circumference ratio was more important for fasting plasma glucose, fasting triglycerides and both systolic and diastolic blood pressure in men and women. There was little evidence for statistically important interaction between BMI and WHR in relation to risk factor levels. Regression of BMI and WHR on risk factors using a quadratic model indicated that the curvilinear model fitted the data better than the linear model for plasma lipids in men but not women. Overall neither BMI nor WHR explained much of the variance in risk factor levels. The results suggest that in Nauruan men BMI is at least as closely associated with an adverse cardiovascular disease risk factor profile as WHR, while in Nauruan women WHR appears more important. The weaker association of BMI and WHR with risk factor levels compared with results from other populations may indicate ethnic variation in fat distribution and/or its metabolic consequences, or could be related to the extreme obesity found in Nauruans.