Efficacy and safety of dydrogesterone monotherapy versus estradiol–dydrogesterone combination therapy in perimenopausal women: a real-world cohort study

Objective: To compare the 12-week effectiveness and safety of dydrogesterone monotherapy versus estradiol-dydrogesterone (E2/DYD) combination therapy in women with perimenopausal symptoms in a real-world clinical setting. Methods: This retrospective cohort study included 150 women treated at a tertiary gynecological outpatient clinic between July 2022 and January 2025 (dydrogesterone: n=60; E2/DYD: n=90). The primary endpoint was change in the Kupperman Menopause Index (KMI) from baseline to week 12. Repeated measures were analyzed using linear mixed-effects models. Secondary outcomes included clinical response (≥50% KMI reduction), endometrial thickness change, and adverse events. Continuous outcomes were evaluated using adjusted linear regression models, and binary outcomes were analyzed using Firth penalized logistic regression. Results: At 12 weeks, KMI decreased by 8.0 ± 6.6 points in the dydrogesterone group and 12.5 ± 7.4 points in the E2/DYD group (adjusted mean difference -4.51, 95% CI -6.90 to -2.12; p<0.001). Clinical response occurred in 31.5% and 51.2% of participants, respectively (adjusted OR 2.08, 95% CI 1.02-4.45; p=0.042). Improvements in vasomotor and mood-related symptoms were more pronounced in the combination group. Endometrial thickness increased modestly in the E2/DYD group (+0.60 mm), with week-12 mean values remaining within physiologic ranges. Any adverse event was reported in 11.1% versus 31.7% of participants (adjusted OR 3.55, 95% CI 1.29-9.77; p=0.014), predominantly mild and self-limited events. Conclusion: In this real-world cohort, E2/DYD combination therapy was associated with greater short-term symptom reduction compared with dydrogesterone monotherapy. Although adverse events were more frequent with combination therapy, they were generally mild and clinically manageable, and no concerning endometrial safety signals were observed over 12 weeks. As a retrospective observational study, causal conclusions cannot be drawn, and these findings should be regarded as hypothesis-generating. They provide real-world contextual evidence to inform clinical discussions around hormone therapy selection in perimenopausal care, pending confirmation in prospective randomized trials.