Abstract 5145: Predictors of Coronary Heart Disease Events among Asymptomatic Individuals with Low LDL: The Multi-Ethnic Study of Atherosclerosis
Even among individuals with low LDL, some will still experience coronary heart disease events (CHD-E) and may benefit from more aggressive therapies. Our aim was to identify risk factors for CHD-E among asymptomatic individuals with LDL≤100 mg/dl. The MESA study is a prospective cohort of 6811 participants (pp) free of clinical cardiovascular disease who were followed for a median of 4.1 years. Of 5701 pp with data on LDL; 2034 (30%) had LDL ≤100 mg/dl. 524 pp on lipid lowering therapy were excluded. For each risk factor of interest, unadjusted hazard ratios were calculated to assess the association with CHD-E (model 1). Traditional risk factors and biomarkers which had a significant univariate association were simultaneously included in a multivariate model to assess the independent association with CHD-E (model 2). To determine if subclinical atherosclerosis markers provided additional information, CAC and CIMT were each separately added to model 2. Final study population consisted of 1510 pp (age 61±11, 47% male) with LDL≤100 mg/dl who were not on lipid lowering therapy at baseline; 34 CHD-E ((2.3%) were observed in this group. In unadjusted analysis, age, male gender, hypertension, diabetes, low HDL, high TG and subclinical atherosclerosis markers (CAC>0; CIMT ≥1mm) predicted CHD-E. In model 2, age, male gender, hypertension, and low HDL remained independent predictors of CHD-E. The relationship of these risk factors remained robust even after adjusted for CAC or IMT (models 3&4). After accounting for these traditional risk factors & biomarkers, the predictive value of CAC was reduced but remained significant; whereas CIMT was not significant. Among individuals with low LDL, older age, male gender, low HDL and presence of CAC are associated with adverse CHD-E and potentially may serve as a base for considering more aggressive therapies. Given the small event rate, additional predictors may be identified over a longer follow-up.