Cognitive function and knee osteoarthritis: A bidirectional Mendelian randomization study with BMI mediation

Abstract

This study investigated the causal relationship between cognitive function and knee osteoarthritis (KOA) and whether body mass index (BMI) mediates this effect. A bidirectional Mendelian randomization (MR) analysis was conducted using genome-wide association study summary statistics. The primary analysis used the inverse-variance weighted method, with MR-Egger, weighted median, robust adjusted profile score, and MR-PRESSO applied for sensitivity analyses. Forward MR analyses showed a significant inverse causal effect of cognitive function on KOA risk, with intelligence (OR = 0.843, 95% CI: 0.755–0.941, P = .002) and cognitive performance (OR = 0.850, 95% CI: 0.771–0.937, P = .001) associated with reduced risk. Reverse MR analyses found no evidence that KOA causally affects cognitive function (Intelligence: β = −0.020, 95% CI:−0.048 to 0.008, P = .167; Cognitive performance: β = −0.006, 95% CI:−0.032–0.020, P = .661). Mediation analysis showed that BMI partially mediated the effect of intelligence on KOA (indirect β = −0.047, 95% CI:−0.110 to−0.006, P = .023), accounting for 27.29% of the total effect. In conclusion, these findings support a unidirectional causal effect of cognitive function on KOA, partially mediated by BMI, highlighting the relevance of cognitive function and weight management in high-risk individuals.