The relationship between lipoprotein(a) and risk of cardiovascular disease: a Mendelian randomization analysis

Abstract

Background: Lipoprotein(a) [Lp(a)] is one of the residual risk factors for cardiovascular disease (CVD) in the setting of optimal low-density lipoprotein cholesterol (LDL-C). The association between Lp(a) and CVD is still in the exploratory phase, with few studies indicating a causal connection between Lp(a) and various CVD.

Methods: Lp(a) (n = 377,590) was a genome-wide association study (GWAS) based on European populations from Neale Lab. Large GWAS datasets for CVD, including aortic aneurysm(AA) (n = 209,366), atrial fibrillation(AF) (n = 1,030,836), coronary heart disease(CHD) (n = 361,194), secondary hypertension(HBP) (n = 164,147), heart failure(HF) (n = 208,178), ischemic stroke (IS) (n = 218,792), large artery atherosclerosis stroke(ISL) (n = 150, 765), small vessel stroke(ISS) (n = 198,048), lacunar stroke(LIS) (n = 225,419), and pulmonary embolism(PE) (n = 218,413) were also based on European populations. We performed separate univariate two-sample Mendelian randomization (MR) analysis for Lp(a) and CVD as described above. We evaluated this connection mainly using the random-effects inverse variance weighted technique(IVW1) with a 95% confidence interval (CI) for the odds ratio (OR). This was supplemented by MR-Egger, weighted median, maximum likelihood, penalized weighted median, and fixed-effects inverse variance weighted methods. MR-PRESSO offers another means of statistical detection.

Results: Our two-sample MR, which was predominately based on IVW1, revealed a causal relationship between Lp(a) and AA (OR = 1.005, 95%CI: 1.001-1.010, P = 0.009), CHD (OR = 1.003, 95%CI 1.001-1.004, P = 0.010), and ISL (OR = 1.003, 9 5%CI 1.002-1.004, P = 9.50E-11), in addition, there is no causal association with AF, HBP, HF, IS, ISS, LIS, or PE. Similar conclusions were reached by the MR-PRESSO method.

Conclusion: This MR study suggested a causal relationship between Lp(a) and CHD, AA, and ISL, but not associated with AF, HF, IS, LIS, ISS, HBP, or PE. Our work further verifies the association between Lp(a) and various CVD, resulting in improved Lp(a) management and a reduction in the prevalence of CVD.

Keywords: Cardiovascular disease; Lipoprotein(a); Mendelian randomization.