Resting-state functional connectivity of hippocampal subregions in predicting escitalopram treatment outcome in major depressive disorder

Abstract

To examine whether pretreatment resting-state functional connectivity (rsFC) of hippocampal subregions is associated with 12-week escitalopram treatment outcomes in major depressive disorder (MDD). A total of 87 patients with MDD received 12 weeks of escitalopram monotherapy and underwent resting-state functional MRI before treatment. Based on the 17-item Hamilton Depression Rating Scale (HAMD-17) score at week 12, patients were classified into the remission depression group (RD, n = 42) or the nonremission depression group (NRD, n = 45). Seed-based whole-brain rsFC analyses were performed using eight bilateral hippocampal subregions as seeds. Group differences in rsFC were examined, and exploratory logistic regression and receiver operating characteristic analyses were conducted to assess the discrimination of remission. Partial correlations between differential rsFC and the HAMD-17 reduction rate were calculated controlling for age, sex, and education level. Compared with NRD, RD showed higher pretreatment rsFC between the left hippocampal tail and the right inferior parietal lobule and lower rsFC between the right hippocampal body and the left middle occipital gyrus and between the right hippocampal tail and the right cuneus. An exploratory combined model integrating the three differential rsFC measures showed discrimination between the RD and NRD groups, with an area under the curve of 0.841 (95% confidence interval: 0.755–0.915). The HAMD-17 reduction rate was negatively associated with right hippocampal body–left middle occipital gyrus rsFC and right hippocampal tail–right cuneus rsFC, but positively associated with left hippocampal tail–right inferior parietal lobule rsFC. Patterns of pretreatment rsFC involving the hippocampal body and tail and parietal and visual cortical regions were associated with 12-week escitalopram treatment outcomes in MDD. These connectivity measures may represent candidate imaging features for further investigation, although validation in larger independent samples is required.