Heterogeneity in the relationship between ethnicity, BMI, and fasting insulin.

OBJECTIVE To determine whether the association of BMI and fasting insulin is modified by ethnicity. RESEARCH DESIGN AND METHODS Non-Hispanic black (black), non-Hispanic white (white), and Mexican-American men and women aged 20-80 years from the Third National Health and Nutrition Examination Survey (1988-1994) were included in this study. Linear regression models with an interaction term were used to test whether ethnicity modified the association between BMI and fasting insulin. RESULTS Fasting insulin was 19, 26, 20, and 19% higher in black women than white women with BMI levels of <22, 22-24, 25-27, and 28-30 kg/m(2), respectively. These differences between black and white women converged at BMI levels >30 kg/m(2). Mexican-American women had fasting insulin levels that were 17, 22, 20, and 16% higher than those of white women at BMI levels of 25-27, 28-30, 31-33, and >34 kg/m(2), respectively, but were not different in individuals with BMI levels <25 kg/m(2). Adjusting for established risk factors did not attenuate these associations in women. Differences in fasting insulin among men were not as apparent. CONCLUSIONS These findings suggest that the effect of obesity on insulin sensitivity is different for Americans in ethnic minorities. In black subjects, fasting insulin is higher at lean weight when compared with white and Mexican-American subjects. In Mexican-American subjects, fasting insulin is higher in overweight individuals when compared with white and black subjects. These findings are more pronounced in women than in men. This result reinforces the importance of designing prevention programs that are tailored to meet the needs of specific populations. Investigation of possible explanations for these differences seems warranted.