Association between changes in estradiol, serotonin, and BDNF and antidepressant response in perimenopausal depression: a retrospective cohort study comparing integrated TCM-acupuncture with SSRIs

Background Perimenopausal depression (PMD) involves complex neuroendocrine interactions among estradiol (E 2 ), serotonin (5-HT), and brain-derived neurotrophic factor (BDNF). Integrated traditional Chinese medicine and acupuncture (TCM-Acu) is widely used clinically, but its biological mechanisms relative to selective serotonin reuptake inhibitors (SSRIs) remain poorly characterized. Methods We conducted a single-center retrospective cohort study in 240 perimenopausal women (TCM-Acu, n = 120; SSRIs, n = 120) treated for a minimum of eight weeks. Treatment effects were estimated using multivariable-adjusted regression (ANCOVA) accounting for baseline imbalances in age, disease duration, HAMD-17 score, and other covariates. Biomarker associations were examined via treatment × ΔBiomarker interaction terms. Results In multivariable-adjusted analyses, TCM-Acu was associated with significantly greater HAMD-17 improvement relative to SSRIs (adjusted β = 3.79 points, 95% CI 2.79–4.79; P < 0.001; unadjusted mean change: 7.07 vs. 3.24 points). However, absolute treatment response rates (≥50% HAMD-17 reduction) were low in both groups (14.2% vs. 5.0%), indicating the observed difference reflects a relative rather than absolute clinical advantage. TCM-Acu was also associated with greater increases in 5-HT (adjusted β = 9.24 ng/mL, 95% CI 6.92–11.56; P < 0.001) and BDNF (adjusted β = 1.60 ng/mL, 95% CI 0.95–2.25; P < 0.001). Unadjusted estradiol levels did not differ significantly between groups (P = 0.592). Across the full cohort, changes in 5-HT and BDNF were positively associated with HAMD-17 improvement (Spearman ρ = 0.31 and 0.37, respectively), with the BDNF–HAMD-17 association being more pronounced in the TCM-Acu group on interaction testing. Conclusion In multivariable-adjusted analyses, TCM-Acu was associated with greater HAMD-17 improvement and more pronounced increases in 5-HT and BDNF compared to SSRIs, while both treatments showed minimal differential effects on estradiol. However, absolute response rates were low in both groups, and the between-group difference reflects a relative rather than absolute clinical advantage. These observational findings are consistent with—but do not establish—the hypothesis that TCM-Acu’s antidepressant associations may be mediated primarily through serotonergic and neurotrophic pathways; this requires prospective confirmation. Key limitations include the retrospective single-center design, baseline imbalances, absence of a sham acupuncture control, and differential treatment durations, which collectively preclude causal inference.