Abstract P181: Atherogenic Lipoproteins and Incident Atherosclerotic Cardiovascular Disease in the Framingham Offspring Study

Abstract

Introduction: Elevated plasma low-density lipoprotein cholesterol (LDL-C), small-dense LDL-C (sdLDL-C), LDL-triglyceride (LDL-TG), triglycerides (TG), remnant-lipoprotein cholesterol (RLP-C), triglyceride-rich lipoprotein-C (TRL-C), very low-density lipoprotein cholesterol (VLDL-C), and lipoprotein(a) [Lp(a)] levels have been associated with increased atherosclerotic cardiovascular disease (ASCVD) risk. However, these parameters have not been included in risk factors for ASCVD in the pooled cohort equation (PCE). Hypothesis: We assessed the hypothesis that these atherogenic lipoprotein parameters add significant information for ASCVD risk prediction in the Framingham Offspring Study. Methods: We evaluated 3,147 subjects without ASCVD at baseline (mean age 58 years) from participants of Framingham Offspring Study cycle 6, 677 (21.5%) of whom developed inclusive ASCVD over 16 years. Biomarkers of risk were assessed in frozen plasma samples. Total cholesterol, TG, HDL-C, direct LDL-C, sdLDL-C, LDL-TG, Lp(a), RLP-C, and TRL-C were measured by standardized automated analysis. Calculated LDL-C, large buoyant low-density lipoprotein cholesterol (lbLDL-C), VLDL-C, and non-HDL-C values were calculated. Data were analyzed using Cox proportional regression analysis and net reclassification improvement (NRI) analysis to identify parameters significantly associated with the incidence of ASCVD after controlling for standard ASCVD risk factor and applying the PCE model. Results: All specialized lipoprotein parameters were significant ASCVD risk factors on univariate analysis, but only direct LDL-C, sdLDL-C, and Lp(a) were significant on multivariate analysis with standard risk factors in the model. Together these parameters significantly improved the model c statistic (0.716 vs 0.732, P < 0.05) and net risk reclassification (mean NRI 0.104, P < 0.01) for ASCVD risk. Using the ASCVD risk pooled cohort equation, sdLDL-C, TG, LDL-TG, LDL-C, RLP-C, and TRL-C individually added significant information, but no other parameter added significant information with sdLDL-C (hazard ratio 1.30 for 75th vs 25th percentile, P < 0.0001) in the model. Conclusions: In multivariate analysis, sdLDL-C, direct LDL-C, and Lp(a) contributed significantly to ASCVD risk, but only sdLDL-C added significant risk information to the PCE model, indicating that sdLDL-C may be the most atherogenic lipoprotein particle.