Comparative effectiveness of interventions for mental health during pregnancy: a systematic review and network meta-analysis

Background: Perinatal mood and anxiety disorders (PMADs) affect approximately 10% of pregnant women globally, with prevalence rising to 15.6% in low- and middle-income countries (LMICs). The global burden has nearly doubled over the past decade, yet only 20% of affected women receive timely, appropriate treatment. Despite the existence of 19 systematic reviews, evidence remains fragmented, with 79% rated as low or critically low quality. No network meta-analysis (NMA) has simultaneously compared all major intervention types within a single analytical framework. Objective: To conduct a comprehensive systematic review and network meta-analysis comparing the comparative efficacy of psychological, physical (exercise-based), psychosocial, and pharmacological interventions for preventing and treating perinatal mental health disorders during pregnancy. Methods: We searched MEDLINE, Embase, PsycINFO, CINAHL, Cochrane CENTRAL, and trial registries from inception to May, 2024. We included randomized controlled trials (RCTs) evaluating any intervention for perinatal mental health in pregnant women, assessing outcomes of depression, anxiety, stress, and quality of life. Two independent reviewers screened records, extracted data, and assessed risk of bias using Cochrane RoB 2.0. We performed random-effects NMA within a frequentist framework using netmeta in R. The primary outcome was standardized mean difference (SMD) in depressive symptoms. Confidence in evidence was evaluated using the CINeMA framework. Results: We included 87 RCTs involving 24,312 pregnant women. Network meta-analysis demonstrated that physical/exercise-based interventions were the most effective for reducing depressive symptoms (SMD = -0.45, 95% CI: -0.62 to -0.28; P-score = 0.89), followed by psychological interventions (SMD = -0.30, 95% CI: -0.44 to -0.16; P-score = 0.72), psychosocial interventions (SMD = -0.24, 95% CI: -0.40 to -0.08; P-score = 0.58), and pharmacological interventions (SMD = -0.11, 95% CI: -0.27 to 0.05; P-score = 0.31). Similar rankings were observed for anxiety and stress outcomes. Subgroup analyses revealed larger effect sizes in high-income countries and for interventions lasting ≥8 weeks. Certainty of evidence was moderate for physical and psychological interventions, low for psychosocial, and very low for pharmacological interventions. Conclusions: Physical/exercise-based interventions demonstrate the strongest comparative effect and should be considered first-line scalable options for mild-to-moderate perinatal mental health symptoms. These findings inform clinical guidelines, identify optimal interventions for resource-limited settings, and directly address a WHO-identified priority gap in perinatal evidence synthesis.