The structural and functional characteristics of hippocampal subfields have been extensively studied in dementia, with findings indicating stronger associations with cognitive performance than those observed in the whole hippocampus (HP). However, the impact of post-stroke dementia (PSD) on the structural and functional connectivity between hippocampal subfields and cortical regions remains unclear. The objective of this study is to examine alterations in the functional and structural connectivity between hippocampal subfields and cortical regions in PSD. We collected resting-state functional magnetic resonance imaging (fMRI) and diffusion tensor imaging (DTI) data from 24 PSD patients, 36 post-stroke non-demented (PSND) patients, and 21 normal controls (NC). These data were used to estimate fractional amplitude of low-frequency fluctuations (fALFF), fractional anisotropy (FA), and diffusivity maps in the hippocampal subfields. Additionally, we constructed functional and structural connectivity matrices between hippocampal subfields and cortical regions for each participant, highlighting group-specific connectivity alterations. Statistical analyses were conducted using a linear mixed model to compare group differences and assess the relationship between MRI measures and clinical evaluations. Our results revealed distinct PSD-related changes in functional connectivity, particularly with the temporal-occipital cortex, within hippocampal subfields compared to the whole HP. Notably, different subfields contributed differently to connectivity changes within the entire HP. Furthermore, we identified positive correlations between diffusivity in the bilateral hippocampal tails and illness duration in PSND patients, which were not observed in PSD. These findings highlight the significant impact of PSD on hippocampal subfields, with subfield analysis offering new insights into the underlying mechanisms of PSD.
16 September 2026 in Journal of Psychiatric Research
Emerging evidence suggests that habenula (Hb) dysfunction plays an important role in major depressive disorder (MDD), with prior functional and structural studies implicating Hb-related networks in emotional and reward processing. However, it remains unclear whether structural covariance of Hb differs between first-episode drug-naive (FEDN) and recurrent MDD (RMDD). T1-weighted structural MRI data were obtained from the REST-meta-MDD consortium, including 839 MDD patients (346 FEDN and 261 RMDD) and 788 healthy controls. Single-subject structural covariance networks were constructed via Kullback-Leibler divergence, with bilateral Hb as seed regions. Group comparisons were conducted using linear mixed-effects models. For right Hb, FEDN group showed reduced structural covariance with the superior/inferior frontal gyrus, cerebellum, and insula, whereas RMDD group exhibited decreased structural covariance in the basal ganglia and temporoparietal junction (TPJ). Both MDD subgroups demonstrated decreased structural covariance between right Hb ventral prefrontal cortex, FEDN specifically exhibited reductions with the thalamus and ventrolateral prefrontal cortex (vlPFC), while the RMDD group specifically showed reductions with the occipital gyrus. In contrast, for left Hb, only RMDD group showed decreased structural covariance with the basal ganglia and TPJ, while no significant alterations were observed in FEDN group. Furthermore, structural covariance between the bilateral Hb and the posterior cingulate cortex was positively correlated with depressive, but not anxious, symptom severity. These findings reveal hemispheric- and subtype-specific patterns of Hb-centered single-subject structural covariance (intra-individual GMV distribution similarity) alterations in MDD, which were associated with depressive symptomatology. These findings provide hypothesis-generating evidence for Hb-centered structural covariance alterations in MDD and may inform future multimodal studies of subtype-specific pathophysiology.