Differential impact of life-long lower LDL-C levels on ischaemic stroke subtypes and coronary heart disease risk among 1 million individuals

Abstract

Abstract Background The relative impact of low density lipoprotein cholesterol (LDL-C) lowering on risk of ischaemic stroke [IS] and coronary heart disease [CHD] has been debated. The relevance of aetiologically distinct IS subtypes may contribute to this complex relationship, but the causal relationship and relative importance of LDL-C for risk of large artery, cardioembolic and small vessel stroke and CHD remains unclear. Purpose To disentangle the relative impact and quantify the causal effects of life-long differences in LDL-C on risk of cardioembolic, small vessel, and large artery stroke, and CHD. Methods We undertook a multi-ethnic Mendelian randomization (MR) study of 1,008,662 individuals including 66,218 IS cases (with TOAST subtyping available in 3 times more cases than previously available), and 103,536 CHD cases. IS results were based on a meta-analysis of summary and individual participant data from the MEGASTROKE Consortium and UK Biobank respectively, and CHD results on data from the CARDIoGRAMplusC4D Consortium and UK Biobank. We created a weighted genetic risk score based on 175 independent variants identified in previous genome-wide studies of LDL-C that included individuals of multiple ethnicities. This genetic instrument was used to estimate the causal effects of life-long differences in LDL-C for risk of IS, IS subtypes and CHD. In addition, the associations between functionally informed genomic regions representing LDL-lowering drug targets including HMGCR (for statins), NPC1L1 (for ezetimibe), and PCSK9 (for PCSK9-inhibitors) and the disease outcomes were examined. Results Life-long lower genetically determined LDL-C was associated with a 25% lower risk of CHD per standard deviation of the genetic score (odds ratio [OR] 0.75, 95% confidence interval [CI] 0.73–0.77; p=2×10–91), but had a significantly weaker effect on IS (OR 0.92, 95% CI 0.88–0.95, p=8x10–6, p for heterogeneity=7.3x10–18). Underlying the relationship with IS was a strong association with large artery stroke (OR 0.75, 95% CI 0.68–0.83; p=2×10–8), comparable to that for CHD, but no significant effects on cardioembolic (OR 1.02, 95% CI 0.95–1.10; p=0.60) or small vessel strokes (OR 0.95, 95% CI 0.88–1.03; p=0.23). Results from multivariable MR analyses that accounted for potential pleiotropy with HDL-C and triglycerides showed no material difference. Additional analyses to examine sensitivity of these findings to the genetic risk score, ethnicity, and wider sources of pleiotropy were performed, and effects of genetic proxies for individual LDL-lowering drug targets assessed. Conclusion This study, the largest of its type to date, demonstrates that life-long lower genetically determined LDL-C levels have a materially weaker effect on IS than on CHD, and differences in the relative importance of life-long lower LDL-C for aetiologically distinct IS subtypes contribute to the differences observed. Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): British Heart Foundation