Long-term dementia risk in metabolic dysfunction-associated steatotic liver disease: a population-based study.

Metabolic dysfunction-associated steatotic liver disease (MASLD) has been implicated in cognitive decline through vascular and metabolic pathways, yet evidence linking it to dementia remains inconsistent. We aimed to determine this association by addressing implications of selection, confounding and reverse causation. We conducted a nationwide, registry-based matched cohort study including all individuals diagnosed with MASLD in Denmark between 2000 and 2020, each with 5 age- and sex-matched references without liver disease. Dementia diagnoses were ascertained through hospital-, prescription-, and death registries. We estimated hazard ratios (HR) using cause-specific Cox regression, adjusting sequentially for comorbidities and socioeconomic status. Sensitivity analyses included age restrictions, delayed entry to address reverse causations, and a negative control outcome (Chronic Obstructive Pulmonary Disease (COPD)) to assess residual confounding from misclassification of alcohol consumption. We included 8,398 individuals with MASLD of which 174 developed dementia during follow-up and 41,990 references of which 641 developed dementia during follow-up. In unadjusted analyses, a MASLD diagnosis was associated with an increased risk of dementia (HR 1.40; 95% CI 1.17 - 1.67). However, the association attenuated after adjustment for comorbidities and socioeconomic factors (HR 1.12; 95% CI 0.94-1.34). Delayed entry analyses suggested possible reverse causation as HRs tended to increase, suggesting underestimation. Using COPD as a negative control outcome showed no increased risk in MASLD patients. Although MASLD was not independently associated with increased dementia risk, our findings suggest possible reverse causation, theoretically reflecting that cognitive decline may cause weight loss, leading to underestimating the association.