Abstract
HDL-C is an established risk marker for coronary heart disease. We investigated sociodemographic, lifestyle, anthropometric/physiologic, and other predictors of HDL-C over 30 years of follow-up in the Coronary Artery Risk Development in Young Adults (CARDIA) study, a multicenter, longitudinal cohort with a baseline exam in 1985-86 and follow-up exams at least every five years through 2016. During exams, participants completed various questionnaires, anthropometric measurements, and blood collection. We performed multiple linear regression of HDL-C at each CARDIA exam in a cross-sectional analysis, and linear mixed-effects regression to assess the longitudinal change of HDL-C across exams, treating time as a linear predictor, separately in women and men. The sample size ranged from 5114 participants at baseline to 3358 at the Year 30 Exam. Just over half of the participants were women and just under half were Black. Several factors were positively associated with HDL-C longitudinally in both women and men: age, self-identified Black race, alcohol intake, physical activity, and energy intake, along with sex-specific associations with current oral contraceptive use, current hormone therapy, and post-menopausal status (women only), and homeostasis model assessment-estimated insulin resistance (HOMA-IR) (men only). Other factors were inversely associated with HDL-C concentration longitudinally in both women and men: cigarette smoking, intakes of carbohydrates and total fat, body mass index, and low-density lipoprotein cholesterol and triglyceride concentrations, along with sex-specific associations with HOMA-IR and parity >0 (women only). We corroborated associations of several factors with HDL-C concentration and provided evidence of associations with several factors not extensively investigated in previous studies.