Lipoprotein-associated phospholipase A2 activity and mass in relation to vascular disease and nonvascular mortality

Heart Protection Study Collaborative Group (Clinical Trial Service Unit, University of Oxford, Oxford, UK) Lipoprotein-associated phospholipase A(2) activity and mass in relation to vascular disease and nonvascular mortality. J Intern Med 2010; 268:348-358. Objectives. To assess whether associations of circulating lipoprotein-associated phospholipase A(2) (Lp-PLA(2)) with vascular disease are independent of other risk factors. Methods. Lp-PLA(2) activity and mass, lipids and other characteristics were measured at baseline in 19 037 individuals at high risk of vascular disease in a randomized trial of simvastatin with 5-year average follow-up. Results. Lp-PLA(2) activity and mass were correlated with each other (r = 0.56), lipids and other vascular risk factors. The moderate association of Lp-PLA(2) activity with occlusive coronary events (n = 2531) in analyses adjusted for nonlipid factors (hazard ratio per 1 SD [HR] 1.11, 95% CI 1.06-1.15) became nonsignificant after further adjustment for apolipoproteins (HR 1.02, 0.97-1.06). Such adjustment also attenuated HRs with Lp-PLA(2) mass from 1.08 (1.03-1.12) to 1.05 (1.01-1.09). By contrast, the HR with apolipoprotein-B100 of 1.15 (1.10-1.19) was only slightly attenuated to 1.14 (1.09-1.19) after further adjustment for apolipoprotein A(1) and Lp-PLA(2). Age- and sex-adjusted HRs for other cardiac events (n = 1007) with either Lp-PLA(2) activity or mass were about 1.20, but HRs reduced after adjustment for nonlipid factors (activity: 1.11, 1.04-1.18; mass: 1.08, 1.02-1.15). Adjusted HRs for ischaemic stroke (n = 900) were weak and nonsignificant and for nonvascular mortality (n = 1040) were 1.01 (0.94-1.09) with activity and 1.12 (1.05-1.19) with mass. Simvastatin reduced Lp-PLA(2) levels by about one-quarter, but simvastatin's vascular protection did not vary with baseline Lp-PLA(2) concentration. Conclusions. Associations of Lp-PLA(2) with occlusive coronary events depend considerably on lipid levels, whereas those with other cardiac events appear to reflect confounding from cardiovascular medication and prior vascular disease.