Causal Associations Between Smoking, Brain Structural Alterations and Psychiatric Disorders: Evidence From a Mediation Analysis

TL;DR

This study shows that smoking causes psychiatric disorders like depression and schizophrenia, with brain structural changes and a specific gene playing key roles.

Contribution

The study identifies a causal link between smoking and psychiatric disorders, mediated by brain microstructure and SH2B2 gene activity.

Findings

Smoking has a bidirectional causal relationship with schizophrenia and major depression.

Microstructural disorganization in the left uncinate fasciculus mediates 19.6% of smoking's effect on depression risk.

SH2B2 gene is implicated in linking smoking to corticolimbic dysfunction through epigenetic changes.

Abstract

Both epidemiological and Mendelian randomization (MR) studies have confirmed the association between smoking and psychiatric disorders, yet the underlying mechanism remains poorly understood. To address this gap, this study aimed to evaluate causal relationships between smoking, brain structural alterations, and psychiatric disorders and to identify genetic and neuroimaging mediators. We analysed summary data from the genome‐wide association study (GWAS) and multimodal neuroimaging data, using linkage disequilibrium score regression (LDSC) to quantify genetic correlations and two‐sample bidirectional MR to assess causality between smoking and three psychiatric disorders: schizophrenia (SCZ), major depressive disorder (MDD) and bipolar disorder (BD). Additionally, we conducted mediation analysis to identify brain structural mediators and colocalization and pathway analyses to elucidate shared genetic architecture. LDSC analysis revealed significant genetic correlations between smoking and MDD (r

g
 = 0.41), SCZ (r

g
 = 0.16) and BD (r

g
 = 0.15). Consistently, bidirectional MR confirmed a bidirectional causal relationship between smoking and SCZ/MDD and a unidirectional causal effect of smoking on BD. Mediation analysis further revealed that microstructural disorganization in the left uncinate fasciculus mediated 19.6% (95% CI: 3.36%‐35.8%) of the effect of smoking on MDD risk. Moreover, colocalization analysis implicated SH2B2 as a pleiotropic locus linking UF orientation dispersion (OD) to prefrontal–amygdala gene expression, suggesting that smoking may exacerbate corticolimbic dysfunction by inducing SH2B2 hypermethylation. Taken together, our findings establish smoking as a causal risk factor for SCZ, MDD and BD, with corticolimbic white matter degeneration serving as a key mediator, and they highlight the SH2B2‐TrkB axis as a mechanistic conduit for genetic and environmental risk, pointing to therapeutic targets to disrupt smoking related psychiatric disorders.

This study establishes smoking as a causal risk factor for schizophrenia, major depression and bipolar disorder using genetic evidence. We identified microstructural disorganization in the left uncinate fasciculus, a key corticolimbic white matter tract, as a significant mediator of this risk, explaining 19.6% of the effect on depression. Genetic analyses further implicated the SH2B2 gene, suggesting a mechanism whereby smoking may induce epigenetic changes that disrupt prefrontal‐amygdala circuitry. These findings highlight white matter degeneration and the SH2B2‐TrkB axis as pivotal contributors to smoking‐related psychiatric pathology, revealing new potential therapeutic targets.

Figures