Abstract 688: Differential Associations Between Novel HDL Markers and Incident Atherosclerotic Cardiovascular Disease by Gender and Vascular Territory: A Meta-analysis of Large Population-based Cohorts

Background: There has been conflicting evidence regarding the association between high-density lipoprotein cholesterol (HDL-C) and atherosclerotic disease, especially given the failure of several therapeutic strategies aimed at raising HDL-C in demonstrating cardiovascular benefits. HDL-P may better predict atherosclerotic cardiovascular disease (ASCVD) compared to HDL-C. The relationship between novel markers of cholesterol overload such as HDL-C/HDL-P and HDL-size/HDL-P and stroke in women is also less well established. Objective: To investigate whether novel HDL markers like HDL-P, HDL size/HDL-P and HDL-C/HDL-P predict stroke better than HDL-C in both men and women. Methods: We performed an individual level meta-analyses of two large patient cohorts - Dallas Heart Study (DHS) and Multi-Ethnic Study of Atherosclerosis (MESA) with HDL marker phenotyping available for 9503 subjects from both cohorts. Cox proportional hazards models were constructed for the HDL markers of interest with adjustment for traditional cardiovascular risk factors. Results: HDL-C was not significantly associated with ASCVD, MI or stroke whereas HDL-P was inversely associated with ASCVD events in both men (n=4378, 444 events) and women (n=5125, 270 events) (Figure 1). HDL-C/HDL-P was directly associated with ASCVD in women (HR 1.39, 95% CI 1-1.93). This was largely driven by the association between HDL-C/HDL-P and stroke (HR 1.99, 95% CI 1.42-2.79). HDL-size/HDL-P was directly associated with ASCVD (HR 1.24, 95%CI 1.1-1.39) in women with similar effect sizes for both MI and stroke. In contrast, among men, both HDL-C/HDL-P and HDL size/HDL-P were directly associated with ASCVD, largely driven by MI and less so by stroke. Conclusions: Reduced HDL-P and increased indices of cholesterol overload may predict atherosclerotic risk more accurately than HDL-C alone. These relationships may vary with gender and vascular territory