Causal relationship and mediating role between depression and cognitive performance - 09/05/25

Doi : 10.1016/j.tjpad.2025.100196

Xinyu Hao ^a,^{^{1
Xinyu Hao, Fuyang Cao, Ziyao Xu and Shaohua You contributed equally to this work.}}, Fuyang Cao ^b,^{^{1
Xinyu Hao, Fuyang Cao, Ziyao Xu and Shaohua You contributed equally to this work.}}, Ziyao Xu ^c,^{^{1
Xinyu Hao, Fuyang Cao, Ziyao Xu and Shaohua You contributed equally to this work.}}, Shaohua You ^d,^{^{1
Xinyu Hao, Fuyang Cao, Ziyao Xu and Shaohua You contributed equally to this work.}}, Tianyue Mi ^e, Lei Wang ^f, Yongxin Guo ^a, Zhuoning Zhang ^a, Jiangbei Cao ^a, Jingsheng Lou ^a, Yanhong Liu ^a, Xianyang Chen ^e, Zhikang Zhou ^a, Weidong Mi ^a,^⁎ , Li Tong ^a,^⁎
^a Department of Anesthesiology, The First Medical Center of Chinese PLA General Hospital, Beijing, PR China
^b Department of Anesthesiology, The Sixth Medical Center of Chinese PLA General Hospital, Beijing, PR China
^c Department of General surgery, The First Medical Center of Chinese PLA General Hospital, Beijing, PR China
^d Department of Pain, The First Medical Center of Chinese PLA General Hospital, Beijing, PR China
^e Department of Health Promotion, Education, and Behavior, University of South Carolina, Columbia, USA
^f Biomedical Big Data Center, Zhongguancun Big Data Industry Alliance, Beijing, PR China

^⁎Corresponding authors.

connectez-vous ou créez un compte

Bienvenue sur EM-consulte, la référence des professionnels de santé.
Article gratuit.

Connectez-vous pour en bénéficier!

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Friday 09 May 2025

Abstract

Background

Recent studies have increasingly emphasized the robust correlation between depression and cognitive function. However, it remains unclear whether this relationship is causal or merely coincidental. To address this uncertainty, we conducted two-sample bidirectional Mendelian randomization (MR) analyses to investigate the connection between depression and cognitive performance.

Methods

We sourced genome-wide association study (GWAS) data for depression (N_SNPs=21,306,230) from the FinnGen (R10) and for cognitive performance (N_SNPs=10,049,954) from the IEU GWAS database. Causal effects employed methodologies such as Inverse variance weighted (IVW), weighted median, MR Egger, simple mode and weighted mode. Two-step analysis determined the contribution of the mediator variable to the outcomes. To determine stability and reliability, sensitivity analyses were performed that included an assessment of heterogeneity, horizontal pleiotropy, and the leave-one-out techniques.

Results

This MR analysis identified 8 independent significant SNPs associated with depression and 81 SNPs linked to cognitive performance. Our findings revealed that depression increases the risk of developing deteriorating cognitive performance (IVW β, -0.11; 95 % confidence interval (CI), -0.18 – -0.05; P_IVW value= 5.97E-04). Conversely, cognitive performance decline could also predispose individuals to depression [odds ratio (OR)_IVW, 0.85; 95 % CI, 0.76 – 0.95; P_IVW value=0.004]. Multivariate MR analysis confirmed the robustness of this bidirectional association. A two-step MR mediation analysis indicated that the pathway from depression to cognitive performance is mediated by pain, with a mediation effect size of -0.022 and a mediation ratio of 28.95 %. The pathway from cognitive performance to depression is mediated by frailty, with a mediation effect value of -0.028, representing 22.40 % of the mediation proportion.

Conclusion

A two-way causal relationship between depression and cognitive performance, with pain and frailty being mediating factors, respectively. Future research should prioritize mechanistic studies, targeted interventions, and personalized approaches to disentangle and mitigate the bidirectional effects of depression and cognitive performance.

Le texte complet de cet article est disponible en PDF.

Keywords : Mendelian randomization, Depression, Cognitive performance, Bidirectional, Two-sample, Mediation effect