Genetically predicted frailty index and risk of stroke and Alzheimer's disease

BACKGROUND AND PURPOSE: Previous studies have reported the association between frailty and stroke or Alzheimer's disease (AD). However, the causality remains unclear. The aim of the present study was to evaluate whether genetically predicted frailty is associated with the risk of stroke or AD by a Mendelian randomization (MR) study. METHODS: Genetic variants associated with the frailty index (FI) were obtained from a large genome-wide association study (GWAS). Summary-level data for stroke and AD were adopted from the corresponding large GWAS of individuals of European ancestry. The inverse variance weighted method was used for estimating causal effects. Multivariable analysis was performed for further adjustment. RESULTS: The present MR study indicated a suggestive association between genetically predicted FI and a higher risk of any stroke (odds ratio 1.360, 95% confidence interval 1.006-1.838, p = 0.046). Regarding the subtypes of stroke, genetically predicted FI was associated with a higher risk of large artery atherosclerosis stroke (LAS) (odds ratio 2.487, 95% confidence interval 1.282-4.826, p = 0.007). No causal links were identified between genetically predicted FI and any ischaemic stroke, intracranial haemorrhage, cardioembolic stroke, small artery stroke, AD or AD-by-proxy. Multivariable MR analysis indicated that the association of genetically predicted FI with LAS was attenuated after adjustment for inflammatory bowel disease (p = 0.114). CONCLUSIONS: The MR study suggested that genetically predicted FI may be associated with an increased risk of any stroke. Subgroup analysis indicated a suggestive association between genetically predicted FI and the risk of LAS. The underlying mechanisms need further investigation.