INTRODUCTION: Post-stroke depression (PSD) affects approximately 30% of stroke survivors and worsens functional outcomes, yet its biological basis remains poorly understood. METHODS: We integrated cross-disorder genomics, developmental spatial transcriptomics, and causal neuroimaging in 16 PSD patients and 14 matched controls. Polygenic overlap was quantified using MiXeR and conditional FDR (condFDR). Shared risk variants were mapped using genetically informed spatial mapping of cells for complex traits (gsMap) onto the embryonic mouse brain (E16.5). Resting-state fMRI with group independent component analysis (ICA), functional network connectivity (FNC), and spectral dynamic causal modeling (spDCM) characterised circuit-level alterations. RESULTS: Stroke and depression showed robust polygenic overlap (Dice = 0.09). condFDR prioritized five pleiotropic loci implicating HDAC9-mediated neurovascular inflammation and PITX2-driven cardio-cerebral signaling. gsMap revealed enrichment of shared genetic risk in developing cortical regions at E16.5. PSD exhibited selective default mode network (DMN)-sensorimotor network (SMN) decoupling and reduced directed DMN→SMN influence (BPA = -0.12 Hz; Pp > 0.99). Auditory network (AN) outflow to DMN, SMN, and ventral attention network (VAN) was broadly attenuated, and AN→DMN effective connectivity scaled with depressive severity (BPA = -0.19 Hz; Pp > 0.99). CONCLUSION: PSD reflects a unified developmental-acquired pathophysiology where latent developmental genetic vulnerability, revealed at E16.5, is unmasked by stroke-triggered circuit decompensation. HDAC9/PITX2 pathways and AN-DMN circuitry are mechanistically grounded targets for potential biomarker development and circuit-informed interventions.
Z et al. (2026) studied this question.