Modifiable lifestyle factors influencing psychiatric disorders mediated by plasma proteins: A systemic Mendelian randomization study

Psychiatric disorders are emerging as a serious public health hazard, influencing an increasing number of individuals worldwide. However, the effect of modifiable lifestyle factors on psychiatric disorders remains unclear. Genome-wide association studies (GWAS) summary statistics were obtained mainly from Psychiatric Genomics Consortium and UK Biobank, with sample sizes varying between 10,000 and 1,200,000. The two-sample Mendelian randomization (MR) method was applied to investigate the causal associations between 45 lifestyle factors and 13 psychiatric disorders, and screen potential mediator proteins from 2992 candidate plasma proteins. We implemented a four-step framework with step-by-step screening incorporating two-step, univariable, and multivariable MR. We found causal effects of strenuous sports or other exercise on Tourette's syndrome (OR [95%CI]: 0.0047 [5.24E-04–0.042]); lifelong smoking index on attention-deficit hyperactivity disorder (10.53 [6.96–15.93]), anxiety disorders (3.44 [1.95–6.05]), bipolar disorder (BD) (2.25 [1.64–3.09]), BD II (2.89 [1.81–4.62]), and major depressive disorder (MDD) (2.47 [1.90–3.20]); and educational years on anorexia nervosa (AN) (1.47 [1.22–1.76]), and MDD (0.74 [0.66–0.83]). Five proteins were found to have causal associations with psychiatric disorders, namely ADH1B, GHDC, STOM, CD226, and TP63. STOM, a membrane protein deficient in the erythrocytes of hereditary stomatocytosis patients, may mediate the effect of educational attainment on AN. The mechanisms underlying the effects of lifestyle factors on psychiatric disorders require further investigation. These findings could help assess the risk of psychiatric disorders based on lifestyle factors and also support lifestyle interventions as a prevention strategy for mental illness.