Association of fat-to-muscle ratio with MoCA-defined cognitive impairment in older adults

The fat-to-muscle ratio (FMR) may reflect an imbalance between adiposity and muscle mass; however, its association with Montreal Cognitive Assessment (MoCA)-defined cognitive impairment remains unclear. To examine the associations of FMR with MoCA-defined cognitive impairment and cognitive domains among community-dwelling older adults in Shenzhen, China. This cross-sectional study included adults aged ≥ 65 years who were registered at community health service centers in Nanshan District, Shenzhen, China. The primary analysis examined the association between total FMR and MoCA-defined cognitive impairment using multivariable logistic regression. Regional, sex-stratified, cognitive-domain, and lipid indirect-effect analyses were exploratory. The Benjamini–Hochberg false discovery rate (FDR) procedure was applied separately within three testing families: four fully adjusted logistic-regression tests for total, trunk, leg, and arm FMR; 28 cognitive-domain tests (four FMR measures × seven cognitive domains); and 12 lipid indirect-effect tests (four lipid markers × three models). FDR correction was performed separately within each family for the overall sample and the male and female subgroups. The prevalence of MoCA-defined cognitive impairment was 47.93%. In the fully adjusted primary analysis, total FMR was not significantly associated with MoCA-defined cognitive impairment in the overall sample (OR per 1-SD increase = 1.203, 95% CI: 0.957–1.511). Regional FMR measures were also not significantly associated with the outcome in the overall sample. In exploratory sex-stratified analyses, each 1-SD increase in total FMR (OR = 1.350, 95% CI: 1.022–1.782) and trunk FMR (OR = 1.335, 95% CI: 1.032–1.726) was associated with higher odds of MoCA-defined cognitive impairment among women, but neither remained statistically significant after FDR correction (both FDR-adjusted P = 0.068). In exploratory cognitive domain analyses, inverse associations with visuospatial/executive function remained significant after FDR correction, whereas nominal positive associations with attention did not. Exploratory analyses suggested TG-related indirect associations between FMR and MoCA scores in the overall sample and among women, but these findings did not persist after full adjustment. Total and regional FMR were not significantly associated with MoCA-defined cognitive impairment in the overall sample. The cognitive domain and lipid indirect effect findings were exploratory and require confirmation in prospective studies.