Comparison of cardiovascular disease risk associated with 3 lipid measures in Japanese adults

Background To assess dyslipidemia, measurement of low-density lipoprotein cholesterol via either Friedewald equation (LDL-F) or direct assay (LDL-D), and non–high-density lipoprotein cholesterol (non–HDL-C) are recommended with some guidelines showing preference to direct over calculated measurements. However, direct comparisons of their respective associations with cardiovascular disease (CVD) risk are currently unavailable. Objective In this study, we evaluated the clinical effectiveness of LDL-F and non–HDL-C vs LDL-D and their associations with CVD. Methods This retrospective cohort study comprised apparently healthy Japanese individuals who underwent an annual health check-up between 2005 and 2007 and completed a 5-year follow-up visit. The incidence of CVD, including coronary and cerebrovascular diseases, during a 5-year follow-up period was evaluated using multivariate logistic regression. Results At baseline, 26,739 participants (mean age, 47 years; 49.0% men) were enrolled, and 292 (1.09%) incidents of CVD were identified at follow-up. Baseline LDL-F, LDL-D, and non–HDL-C were all significantly associated with CVD, although the effect appeared higher for LDL-F, particularly for coronary heart disease. Increased risks of CVD were observed for high LDL-F (≥130 mg/dL), despite being categorized into the lower LDL category based on LDL-D (odds ratio [OR], 1.85; 95% confidence interval [CI], 1.19–2.87) and non-HDL-C (OR, 1.75; 95% CI, 1.22–2.52). Without high LDL-F, no CVD associations were found for high LDL-D ( P  = .62) or non-HDL-C ( P  = .93). Conclusion Despite growing availability of direct assays and increasing evidence of non-HDL-C utility, the Friedewald equation may offer better clinical utility for CVD prevention, especially in the screening of apparently healthy individuals.