Causal interplay between major depressive disorder and cognitive impairment: A bidirectional Mendelian randomization study with brain imaging mediation.

Background

Major depressive disorder (MDD) features widespread cognitive impairments, some pre-existing before onset. However, the causal associations between MDD and multidimensional cognitive impairments, and the structural and functional neural underpinnings mediating these relationships, remain to be elucidated.

Method

We integrated summary-level data from large-scale genome-wide association studies of MDD, cognitive impairments, and neuroimaging phenotypes. Bidirectional mendelian randomization (MR) explored causal associations between MDD and cognitive impairments, with two-step MR assessing neuroimaging phenotypes as mediators.

Results

MR analyses revealed bidirectional causal relationships between MDD and cognitive impairments. Specifically, MDD exerted causal effects on Trail Making Test Part B performance, assessed by the duration to complete the alphanumeric path (β = 0.25, 95% CI 0.05 to 0.45, P = 0.016), and Pairs Matching task performance, assessed by the number of incorrect matches in round (β = 0.11, 95% CI 0.06 to 0.16, P = 2.17 × 10-5). These associations were partially mediated by brain structural and functional measures, including aparc-DKTatlas_lh_area_medialorbitofrontal (16.00%), IDP_T1_FAST_ROIs_R_mid_front_gyrus (18.18%), and rfMRI connectivity (ICA100 edge 645) (27.27%). Conversely, the duration to complete the alphanumeric path (OR = 1.77, 95% CI 1.30 to 2.40, P = 2.48 × 10-4) and the number of incorrect matches in round (OR = 2.10, 95% CI 1.43 to 3.07, P = 1.48 × 10-4) also causally influenced MDD. However, no significant mediation through macro-scale brain imaging phenotypes was observed.

Conclusions

This study provides evidence for potential bidirectional causal relationships between MDD and impairments in executive function and visual memory. The effect of MDD on these cognitive domains may be partially mediated by specific neuroimaging phenotypes.