Genetic evidence on the association between pulmonary function and cognitive impairment: A 2-sample Mendelian randomization study

Some observational studies have suggested that lower pulmonary function increases the risk of cognitive decline or dementia; however, the evidence remains inconclusive. We performed 2-sample Mendelian randomization (MR) analyses to investigate the potential associations between forced vital capacity (FVC) and a range of dementia- and cognition-related outcomes. FVC was selected as the primary indicator of pulmonary function because it is less effort- and cognition-dependent and better reflects overall lung capacity. Outcomes included 6 dementia types: all-cause dementia, Alzheimer disease (AD), dementia with lewy bodies, Parkinson disease dementia, frontotemporal dementia, and vascular dementia, and 6 cognitive domains, including intelligence, fluid intelligence (reasoning and problem-solving ability independent of acquired knowledge), cognitive performance, numeric memory, executive function, and prospective memory. All genetic associations were reported per 1-standard-deviation increase in genetically predicted FVC – expressed as log-odds ratios (log-ORs) for dementia outcomes and standard-deviation changes for cognitive outcomes. The inverse-variance weighted method was used as the primary analysis, complemented by MR-Egger, weighted median, weighted mode, simple mode and MR-PRESSO for sensitivity analyses. False discovery rate (FDR) correction, colocalization, and reverse MR analyses were also performed. This study provides genetic evidence supporting an association between reduced pulmonary function and cognitive impairment. Further studies are needed to clarify the underlying mechanisms. Higher genetically predicted FVC was associated with a lower risk of AD (log-OR per 1-SD increase = −0.24; P = .002; FDR-adjusted P = .011). An inverse association was also observed with all-cause dementia (log-OR per 1-SD increase = −0.37; P = .031), but it did not remain significant after FDR correction (FDR-adjusted P = .094). No significant associations were observed for other dementia subtypes or cognitive outcomes. The results were robust in sensitivity analyses, with no significant findings in reverse MR. Colocalization analysis did not support shared causal variants between FVC and AD (PP.H 4 .abf <0.75).