Abstract
BACKGROUND: Observational studies suggest that statins may affect amyotrophic lateral sclerosis (ALS). However, they are limited by confounding and reverse causality biases. Therefore, we aimed to investigate the potential causal associations between statins and ALS using a mendelian randomization (MR) approach. METHODS: Two-sample MR and drug-target MR were performed. Exposure sources included GWAS summary statistics of statin use, low-density-lipoprotein cholesterol (LDL-C), HMGCR-mediated LDL-C and LDL-C response to statins. RESULTS: Genetic predisposition to statin medication was associated with increased ALS risk (OR = 1.085, 95% CI = 1.025-1.148, p = 0.005). After removing SNPs significantly associated with statin use from the instrumental variables (IVs), LDL-C-related higher ALS risk was absent (before removing: OR = 1.075, 95% CI = 1.013-1.141, p = 0.017; after removing: OR = 1.036, 95% CI = 0.949-1.131, p = 0.432). HMGCR-mediated LDL-C (OR = 1.033, 95% CI = 0.823-1.296, p = 0.779) and blood LDL-C response to statins (OR = 0.998, 95% CI = 0.991-1.005, p = 0.538) had no association with ALS. CONCLUSIONS: Here, we show that statins may be a risky exposure that increases ALS risk independent of the lowering effect of LDL-C in peripheral circulation. This provides insights into ALS development and prevention.