Real world pharmacovigilance assessment of drug related perinatal depression risks.

BACKGROUND

Perinatal depression (PND) is a prevalent and serious mental health concern affecting women during pregnancy and the postpartum period. Although several drugs have been implicated in PND etiology in previous studies, large-scale systematic evaluations of drug-associated PND using real-world pharmacovigilance data remain scarce.

METHODS

We conducted a retrospective pharmacovigilance study utilizing the FDA Adverse Event Reporting System data spanning from Q1 2004 to Q2 2025. Disproportionality analysis, logistic regression, LASSO regression, and time-to-onset analysis were employed to examine drug-PND associations from multiple analytical perspectives.

RESULTS

Disproportionality analyses identified 28 drugs with positive signals as potentially associated with PND. Multi-factor analysis identified age under 26 years, weight exceeding 74 kg, and use of sodium oxybate, levonorgestrel, ethinylestradiol/etonogestrel, or lurasidone as independent risk factors for drug-related PND. The model achieved a ROC-AUC of 0.907, indicating excellent discriminatory performance.

CONCLUSION

This study identified specific drugs and demographic factors associated with PND risk, offering valuable insights for pharmacovigilance and clinical medication management.