Objective: To use standardized cut-offs of body mass index (BMI), waist circumference, waist-to-hip ratio, and fasting insulin levels to predict the development of metabolic disorders and metabolic syndrome.
Research methods and procedures: We performed an 8-year follow-up study of 628 non-Hispanic whites and 1340 Mexican Americans, ages 25 to 64 years, from the second cohort of the San Antonio Heart Study. We defined metabolic disorders as dyslipidemia (triglycerides > or =2.26 mM or high-density lipoprotein <0.91 mM in men and <1.17 mM in women), hypertension (blood pressure > or =140 or >=90 mm Hg, or receiving antihypertensive medications), and type 2 diabetes (fasting glucose > or =7.0 mM, 2-hour test glucose > or =11.1 mM, or receiving anti-diabetic medications). People with at least two metabolic disorders were defined as having metabolic syndrome.
Results: High waist-to-hip ratio and fasting insulin levels were significant predictors of developing metabolic syndrome. High anthropometric indices remained significant predictors of metabolic syndrome after adjusting for fasting insulin. Waist circumference, BMI, and insulin had similar areas under the receiver operating characteristic curves (0.74 to 0.76). Further multivariate analyses combining these indices showed minimal increase in prediction. Of subjects who had a combination of high BMI (> or =30 kg/m(2)) and high waist circumference (above "Action Level 2"), 32% developed metabolic syndrome, compared with 10% of subjects with both low BMI and low waist circumference.
Discussion: These findings support the National Institutes of Health recommendations for reducing the risk of metabolic syndrome. Adjustment for baseline fasting insulin levels had only a small effect on the ability of anthropometric indices to predict the metabolic syndrome.