Bidirectional causal links between psychiatric disorders and orthopaedic diseases: a two-sample Mendelian randomization study

Abstract Background Psychiatric disorders and orthopaedic diseases frequently co-occur, but whether these associations are causal remains uncertain. We evaluated the bidirectional causal relationships between seven psychiatric disorders-major depressive disorder (MDD), bipolar disorder, schizophrenia, attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorder (ASD), obsessive–compulsive disorder (OCD), and anorexia nervosa (AN)-and four orthopaedic outcomes: osteoporosis, fractures, osteoarthritis, and osteomyelitis. Methods We performed a two-sample Mendelian randomization (MR) using summary-level genome-wide association study (GWAS) data; psychiatric traits were sourced from the Psychiatric Genomics Consortium and orthopaedic outcomes from the IEU Open GWAS resource. Causal effects were estimated primarily with inverse-variance weighted (IVW) models, complemented by MR-Egger and weighted median analyses. Standard sensitivity analyses were conducted to probe heterogeneity, pleiotropy, and robustness. Results Genetically proxied MDD was associated with higher risks of osteoarthritis (IVW OR = 1.010, 95% CI 1.006-1.014, P < 0.001) and osteomyelitis (OR = 1.457, 95% CI 1.258–1.687, P = 0.001). ASD showed a positive association with fractures (OR = 1.042, 95% CI 1.002-1.082, P = 0.039), whereas AN was inversely associated with fractures (OR = 0.960, 95% CI 0.928-0.993, P = 0.018). Reverse MR analyses did not identify significant causal effects of orthopaedic diseases on psychiatric disorders. Sensitivity analyses did not materially change these estimates. Conclusions These findings support a causal link between genetic liability to MDD and increased risks of osteoarthritis and osteomyelitis, and suggest potential associations between ASD/AN and fracture risk. Given multiple comparisons and possible residual biases, signals should be interpreted cautiously and validated in larger, multi-ancestry cohorts.