Search

The lipid profile is consistently associated with coronary artery disease (CAD) and ischemic stroke (IS). However, the lipoprotein subfractions have not been deeply explored in stroke subtypes, especially in IS outcome. We performed two-sample Mendelian randomization (MR) analysis using 92 lipid traits measured by nuclear magnetic resonance in 115,000 subjects from the UK Biobank. Data for genetic associations with IS, its subtypes, and long-term outcome (LTO) were obtained from three cohorts of European ancestry: GIGASTROKE (73,652 cases, 1,234,808 controls), GODS (n = 1791) and GISCOME (n = 6165). Results obtained using CARDIoGRAMPlusC4D were used to identify differences with CAD. Genetically determined low concentration of medium high-density lipoprotein (HDL) particles (odds ratio (OR) = 0.92, 95% CI 0.88–0.96; p = 3.6 × 10−4) and its cholesterol content (OR = 0.92, 95% CI 0.88–0.96; p = 1.9 × 10−4) showed causal associations with an increased risk of stroke. Genetic predisposition to high apolipoprotein (apo)B to apoA-I ratio was causally associated with an increased risk of IS (OR = 1.12, 95% CI 1.06–1.18, p = 1.1 × 10−4), and a highly suggestive association was found between non-esterified cholesterol in low-density lipoprotein (LDL) and increased risk of atherothrombotic stroke (LAS) (OR = 1.35, 95% CI 1.10–1.66; p = 4.0 × 10−3). Low cholesterol in small and medium LDL was suggestively associated with poor LTO. Our results support that low medium HDL concentration was causally associated with an increased stroke risk, while high levels of non-esterified cholesterol in LDL were suggestively associated with an increased risk of LAS and with a better LTO. [Display omitted] • Causal association of low HDL-C with stroke risk is due to the medium subfraction. • Free cholesterol content is implicated in the association of high LDL with stroke risk. • The association of triglyceride-rich lipoproteins with stroke remains inconclusive. • ApoB-lipoproteins play a more important role in CAD risk than in stroke risk. • The association of low LDL-C with a poor long-term outcome needs further studies. [ABSTRACT FROM AUTHOR]