Genetic Variants, Metabolic Dysfunction-Associated Fatty Liver Disease, and Major Health Outcomes in Older Adults.

: Multiple genetic variants have been associated with disease prevalence and outcomes in middle-aged people with metabolic dysfunction-associated fatty liver disease (MAFLD). However, genetic studies in older adults have been lacking. We aimed to understand their clinical relevance in healthy older persons. : A secondary analysis of the ASPREE (ASPirin in Reducing Events in the Elderly) randomized trial involving community-dwelling older adults ≥ 70 years without prior cardiovascular disease events or life-limiting illness at enrolment. The Fatty Liver Index (FLI) was used to identify MAFLD at baseline. We assessed the associations between six previously reported MAFLD-associated genetic variants with prevalent MAFLD at baseline, and the associations of these variants with cardiovascular disease events and all-cause mortality. : A total of 8756 participants with genetic data were stratified according to the FLI, with 3310 having MAFLD at baseline. The follow-up was for a median of 8.4 (IQR 7.3-9.5) years. Variants in two genes ( and ) were associated with prevalent MAFLD ( < 0.05); , , , and were not. , , , , and were associated with major adverse cardiovascular events (MACEs) or mortality in the overall cohort or in participants with MAFLD during the follow-up (all > 0.05). Within the MAFLD group, homozygosity for the rs641738 C > T variant in the gene was associated with a reduced risk of MACEs (HR 0.68 [95% CI 0.48-0.97]), but not all-cause mortality (HR 1.14 [95% CI 0.89-1.47]). This protective association remained significant after adjusting for multiple key covariates (aHR 0.64 [95% CI 0.44-0.92]). The results were similar when using the metabolic dysfunction-associated steatotic liver disease definition rather than MAFLD. : The rs641738 C > T variant in may confer protection against MACEs in older adults with MAFLD, independent of other clinical risk factors. Further validation using external cohorts is needed.