Predictors of insulin resistance in the obese with metabolic syndrome

Abstract

Background: In the obese, the metabolic syndrome (MetS) is assumed to reflect insulin resistance.

Objective: To determine the predictors of insulin resistance in obese subjects with MetS.

Design: We used the 90th percentile of the homeostasis model assessment (HOMA) to define insulin resistance in 4958 nondiabetic adults evaluated in the National Health and Nutrition Examination Surveys, 1999-2004, and compared the 373 obese subjects who were insulin-resistant (HOMA 9.52+/-5.73) to a control group of 373 obese who had the highest sensitivity to insulin (HOMA 1.79+/-0.44).

Measurements: MetS was present in 312 (83.6%) obese with insulin resistance and in 156 (41.8%) obese from the insulin-sensitive control group. Demographic, metabolic, and lifestyle variables were analyzed with logistic regression.

Results: In a logistic model of insulin resistance given the presence of MetS, the significant predictors were triglycerides (P=0.0021), body mass index (P=0.0096), HDL-cholesterol (P=0.0098), age (P=0.0242) and smoking (P=0.0366).

Limitations: Cross-sectional design prevents elucidation of causality for the association between insulin resistance and MetS.

Conclusions: Insulin resistance is not an obligatory correlate of MetS in the obese. Its likelihood can be predicted by cigarette smoking and by the severity of obesity and dyslipidemia.