ApoA-1 versus HDL-C as markers of cardiovascular risk.

BACKGROUND: Conflicting results have been reported as to the relative importance of apolipoprotein A-1 (apoA-1) vs high-density lipoprotein cholesterol (HDL-C) as markers of atherosclerotic cardiovascular disease (ASCVD) risk. OBJECTIVE: To compare the apoA-1 and HDL-C as predictors of cardiovascular risk. METHODS: Residual discordance analysis with Cox proportional hazard models comparing apoA-1 and HDL-C as markers of ASCVD risk was applied to a sample of 291,995 UK Biobank participants, followed for a median of 11 years. Interaction tests for the 2 markers and estimation of the effects of partitioning HDL-C into apoA-1, log-triglyceride, and the remaining residual were also performed. RESULTS: ApoA-1 and HDL-C had similar associations with ASCVD risk (hazard ratios [HRs] of 0.85, P value <.001 for both). The residual of HDL-C added significantly to the risk associated with apoA-1, as did the residual of apoA-1 to HDL-C. There was a statistically significant interaction between apoA-1 and HDL-C (HR = 1.05, 95% CI: (1.04-1.06); P < .001). Decomposing HDL-C into the 3 components, apoA-1 accounted for the largest portion of the effect, with an HR of 0.85 (95% CI: 0.83-0.86), with smaller effects for ln triglycerides: 1.04 (95% CI: 1.02-1.06) and the residual of HDL-C: 0.98 (95% CI: 0.96-0.995). CONCLUSION: HDL-C and apoA-1 have associations of equivalent strength with ASCVD risk, with significant interaction modifying the effect of one by the other. Upon decomposition, apoA-1 retained more of the effect of HDL-C as compared with log-triglycerides. While only observational, the results are consistent with the relation of HDL to risk not being determined by the concurrent level of triglyceride.

Duke Scholars

Author Michael J Pencina Biostatistics & Bioinformatics, Division of Biostatistics