Abstract
BACKGROUND: Previous studies have suggested that antidiabetic drug use may be associated with amyotrophic lateral sclerosis. However, these studies are limited by many confounding and reverse causality biases. We aimed to determine whether antidiabetic drug use has causal effects on ALS. METHODS: Drug-target Mendelian randomization analysis was conducted to evaluate the association between genetic variation in the targets of antidiabetic drugs and ALS risk. The antidiabetic drugs included sulfonylureas, GLP-1 analogues, thiazolidinediones, insulin/insulin analogues, metformin, and SGLT2 inhibitors. Summary statistics for ALS were retrieved from previous genome-wide association studies comprising 27,205 ALS patients and 55,058 controls. The instrumental variables for these drugs are from previous published articles. RESULTS: Genetic variation in SGLT2 inhibition targets was associated with lower risk of ALS (odds ratio [OR] = 0.32, 95% CI = 0.14-0.74; p = 0.008). We did not find that genetic variation in metformin targets was associated with ALS (OR = 1.61, 95% CI = 0.94-2.73; p = 0.081). Nevertheless, mitochondrial complex I, a target of metformin, was associated with a higher risk of ALS (OR = 1.83, 95% CI = 1.01-3.32; p = 0.047). The analysis showed that genetic variation in sulfonylureas, GLP-1 analogues, thiazolidinediones, insulin or insulin analogues targets was not associated with ALS (all p > 0.05). CONCLUSIONS: The complex interaction between hypoglycemic, antioxidation, and anti-inflammatory effects may account for the different results across antidiabetic drug types. These findings provide key evidence to guide the use of antidiabetic drugs and will help to identify novel therapeutic targets in ALS.