THE ASSOCIATION BETWEEN DEPRESSION AND DIABETES


Elizabeth Barrett-Connor, M.D., UCSD Dept. of Family and Preventive Medicine, La Jolla, CA, USA

Given the consistent inverse association of HDL cholesterol (HDL-C) with cardiovascular disease (CVD), there is immense interest in developing therapies that raise HDL-C. But HDL-C, the cholesterol carried by HDL particles, may not fully capture HDL-related cardioprotection. HDL-C is carried within HDL lipoprotein particles that are heterogeneous in size, density, charge, core lipid composition, specific apolipoproteins, and function. Recent failures of drugs targeting HDL-C have fueled interest in other metrics of HDL, including HDL particle number (HDL-P), size, and function. We hypothesized that HDL-C cardioprotection may at least partly be influenced by metabolic correlations with other lipoproteins, particularly atherogenic particles, but that HDL-P may be less affected. Using data from 5,598 individuals (ages 45-84 years) from the prospective Multi-Ethnic Study of Atherosclerosis (MESA), we compared the association of chemically-assayed HDL-C and NMR-spectroscopy-measured HDL-P with subclinical atherosclerosis and incident CVD before and after accounting for atherogenic particles. HDL-C was correlated with HDL-P (r = 0.73). Notably, HDL-C associations with atherosclerosis and incident CVD were substantially attenuated and became non-significant after adjusting for atherogenic lipoproteins. In contrast, HDL-P associations with atherosclerosis and incident events were relatively unaffected. In addition, a very high HDL-C also conferred increased risk (HR 2.59, 95% CI 1.11-6.02) after adjusting for HDL-P, while very high HDL-P remained inversely related to risk (HR 0.50, 95% CI 0.19-1.35). These results may have implications both for risk assessment and evaluating therapeutic interventions, particularly pharmacologic interventions which may differentially affect several lipid and lipoprotein parameters concurrently.