Despite consensus on the need for blood cholesterol reductions to prevent coronary heart disease (CHD), available evidence on optimal cholesterol levels or the added predictive value of additional lipids is sparse. After 10 years follow-up of 12 339 middle-aged participants free of CHD in the Atherosclerosis Risk in Communities Study (ARIC), 725 CHD events occurred. The lowest incidence was observed in those at the lowest LDL cholesterol (LDL-C) quintile, with medians of 88 mg/dL in women and 95 mg/dL in men, and risk accelerated at higher levels, with relative risks (RRs) for the highest quintile of 2.7 in women and 2.5 in men. LDL-C, HDL-C, lipoprotein(a) [Lp(a)], and in women but not men, triglycerides (TG) were all independent CHD predictors, providing an RR, together with blood pressure, smoking, and diabetes, of 13.5 in women and 4.9 in men. Lp(a) was less significant in blacks than whites. Prediction was not enhanced by HDL-C density subfractions or apolipoproteins (apo) A-I or B. Despite strong univariate associations, apoB did not contribute to risk prediction in subgroups with elevated TG, with lower LDL-C, or with high apoB relative to LDL-C. Optimal LDL-C values are <100 mg/dL in both women and men. LDL-C, HDL-C, TG, and Lp(a), without additional apolipoproteins or lipid subfractions, provide substantial CHD prediction, with much higher RR in women than men.
