ApoB/apoA-I ratio: An independent predictor of insulin resistance in US non-diabetic subjects

Background: Recently, the apoB/apoAI ratio has been associated with the metabolic syndrome; however, is unclear if its association with insulin resistance is mediated through traditional risk factors or if it adds an independent risk by itself. The aim of this study was to assess the independent association between apoB/apoAI ratio and insulin resistance in the US non-diabetic population. Methods: We examined the association between high apoB/apoAI ratio and insulin resistance among 2955 adults (mean age 47 years; 1457 women) without diabetes (fasting glucose ≤7 mmol/L and not taking diabetes medication), who participated in the Third National Health and Nutrition Examination Survey. Insulin resistance was estimated using the computer homeostatic model assessment (HOMA2) and defined as the upper quartile. The updated ATP-III definition of the metabolic syndrome was used. First, logistic regression was applied to estimate the cross-sectional association between apoB/apoAI (highest quartile vs. lowest quartile) and insulin resistance adjusting for metabolic syndrome components excluding glucose. Finally, multiple linear regression was used to assess the relationships between apoB/apoAI and insulin sensitivity. Results: Overall, median of apoB/apoAI ratio was significantly higher in subject with insulin resistance than without (0.85, IQR 0.69-0.99 vs. 0.69, IQR 0.56-0.85; P < 0.0001). High apoB/apoAI ratio was independently associated with insulin resistance after adjustment for age and race, and remained significant after further adjustment for metabolic syndrome components, traditional and inflammatory risk factors (in men: OR, 4.12-95% CI, 1.97-8.81; in women: OR, 3.69-95% CI, 1.94-7.27). When apoB/apoAI was considered as a quantitative trait rather than dichotomized, use of the ratio improved the prediction of HOMA2 independently of metabolic syndrome components, traditional and inflammatory risk factors (in men: additional R ² = 0.09, P < 0.001; in women: additional R² = 0.05, P < 0.001). Conclusion: In the US population, apoB/apoAI ratio is significantly associated with insulin resistance in non-diabetic subjects, independently of the traditional risk factors, metabolic syndrome components, and inflammatory risk factors. Important clinical risk information provided by apoB/apoAI ratio should be recognized and implemented in future clinical guidelines.