Mendelian randomization and transcriptome analysis reveal depression-driven regulatory patterns of the immune microenvironment in myocardial infarction and heart failure.

BACKGROUND

Major depressive disorder (MDD) and cardiovascular diseases (CVD) are mutually amplifying global health burdens, yet the causal directions and immune-determined molecular substructures that link MDD to myocardial infarction (MI) and heart failure (HF) remain poorly resolved.

METHODS

Bidirectional two-sample Mendelian randomization (MR) was applied to large-scale GWAS (1.35 million MDD; 361 K MI; 977 K HF) followed by replication in 11,004 NHANES 2005-2020 participants using restricted cubic splines and multivariable logistic regression. Multi-cohort transcriptomics (peripheral blood microarray n = 447; in-house RNA-seq n = 14; left-ventricular tissue from dilated cardiomyopathy (DCM) patients (n = 332) were integrated to identify MDD-driven expression signatures. LASSO regression, CIBERSORT, ssGSEA, consensus clustering and GSVA were employed to derive diagnostic gene panels and immune endotypes.

RESULTS

MR analyses provided genetic evidence consistent with a directional effect of MDD on MI (IVW β = 0.01, P = 4.6 × 10⁻6) and HF (IVW β = 0.19, P = 1.3 × 10⁻6) without reverse causation. Depression (PHQ-9 ≥ 10) has a dose-dependent nonlinear association with MI and HF (P<0.0001), with adjusted odds ratios (OR) of 1.80 (95% CI: 1.07-3.05) and 2.41 (95% CI: 1.45-4.00), respectively. A total of 202 MDD-related genes were identified through integrated transcriptomic analysis. After cross validation with the MI/HF dataset, six robust biomarkers (TMEM43, C1orf174, L3MBTL4, OR52N4, SLC25A20, MISP3) were screened. Risk-score models discriminated MI (AUC = 0.90-1.00) and HF (AUC = 0.95) in peripheral blood, but HF discrimination in cardiac tissue was modest (AUC = 0.60). Consensus clustering on 184 MDD-correlated genes stratified each CVD into two reproducible subtypes: a "homeostatic/pro-fibrotic" cluster enriched for ribosomal and cell-cycle pathways and an "inflammatory-metabolic" cluster characterized by NF-κB, TNF-α, IL-6, complement and coagulation activation.

CONCLUSIONS

Genetic, epidemiological, and multi-omic evidence supports a directional association between MDD and increased risk of MI and HF. We deliver reproducible blood-based gene panels and immune endotypes that dissect biologically distinct MDD-CVD substructures, offering actionable targets for precision immunomodulatory therapy in cardio-depressive comorbidity.