Should triglycerides and the triglycerides to high-density lipoprotein cholesterol ratio be used as surrogates for insulin resistance?

The aims of the present study were to examine whether triglycerides (TG) and the triglyceride to high-density lipoprotein cholesterol ratio (TG/HDL-C) could predict insulin resistance in healthy African Americans and whites. This cross-sectional study included 99 African American and 50 white men and women between 18 and 45 years of age with body mass indexes between 18.5 and 38.0 kg/m(2). Anthropometric measures were obtained; and overnight fasting blood was collected for TG, HDL-C, glucose, and insulin. Insulin resistance was defined by fasting insulin concentration of at least 13.13 microU/mL and homeostasis model assessment of insulin resistance (HOMA-IR) of at least 2.5. Receiver operating characteristic curves were used to analyze the data. African Americans and whites had comparable demographic and anthropometric measures. Fasting insulin was higher in African Americans (12.4 +/- 7.8 microU/mL) than whites (10.2 +/- 7.5 microU/mL), but HOMA-IR did not differ significantly (African Americans, 2.9 +/- 2.0; whites, 2.4 +/- 1.9). Triglycerides and TG/HDL-C were significantly lower in African Americans (TG, 68.2 +/- 43.3 mg/dL; TG/HDL-C, 1.8 +/- 2.1) compared with whites (TG, 105.4 +/- 55.2 mg/dL; TG/HDL-C, 2.8 +/- 1.8). Area under the receiver operating characteristic curves revealed that both TG and TG/HDL-C were acceptable markers of insulin resistance, as defined by fasting insulin concentration, in whites, 0.770 and 0.765, respectively, but poor predictors in African Americans, 0.633 and 0.651, respectively. Similarly, TG and TG/HDL-C were acceptable in predicting insulin resistance, as measured by HOMA-IR, in whites, 0.763 and 0.770, respectively, but poor in predicting HOMA-IR in African Americans, with areas of 0.625 and 0.639, respectively. In conclusion, the relationship between TG and TG/HDL-C with insulin resistance differs by ethnicity; and using TG and TG/HDL-C to predict insulin resistance in African Americans would not be appropriate.