Geographic and Socioeconomic Predictors of Postpartum Depression Identified Through Machine Learning Analysis of Pakistani Health Data

Authors

  • Sabira Feroz Senior Lecturer, Bahria University, Islamabad, Pakistan Author
  • Sidra Irfan Senior Registrar, Preventive Medicine, Amiri Hospital, Kuwait City, Kuwait Author
  • Fariha Muhammad Iqbal Doctor, Muhammad College of Medicine, Peshawar, Pakistan Author
  • Noorulain Tariq Siddiqui BSc (Hons), Prosthetic and Orthotic Sciences, PIPOS, Khyber Medical University, Peshawar, Pakistan; Quality Assurance Manager, Vital Medical Associates, PC, USA Author
  • Zubia Shahid Lecturer, Quaid-i-Azam University, Islamabad, Pakistan Author
  • Saeed Wali MBBS, Bolan Medical College, Quetta, Pakistan Author
  • Iraj Siddiqui MBBS, ZMC, Karachi, Pakistan; House Officer, Ziauddin University Hospital, Karachi, Pakistan Author

DOI:

https://doi.org/10.61919/thq1qy88

Keywords:

Edinburgh Postnatal Depression Scale; geographic variation; household income; machine learning; maternal education; overcrowding; Pakistan; postpartum depression

Abstract

Background: Postpartum depressive symptoms are frequently under-recognised in resource-constrained settings, and their geographic and socioeconomic distribution within Pakistani hospital populations remains insufficiently characterised. Objective: To identify hospital-city and socioeconomic predictors of postpartum depressive symptom severity using conventional regression and random forest analysis. Methods: This multicentre retrospective record-based analytical study included 336 postpartum women aged 18–45 years receiving care at tertiary hospitals in Karachi, Lahore, and Rawalpindi between January 2021 and December 2023. The primary outcome was the continuous Edinburgh Postnatal Depression Scale score. Geographic and socioeconomic predictors were examined using analysis of variance, multivariable linear regression, and random forest regression. Results: The mean EPDS score was 11.2 ± 5.8, and 187 women (55.7%) screened positive at an EPDS threshold of ≥10. Mean scores were highest in Rawalpindi (13.4 ± 5.9), followed by Lahore (11.1 ± 5.6) and Karachi (9.1 ± 5.2). Compared with the respective reference categories, Rawalpindi hospital-city site was associated with a 3.9-point higher score, household income below PKR 30,000 with a 5.8-point increase, no formal education with a 4.9-point increase, and overcrowding with a 3.2-point increase. The random forest model achieved a test-set R² of 0.51. Conclusion: Socioeconomic disadvantage and hospital-city context were associated with greater postpartum depressive symptom severity. Targeted screening may be warranted among women from low-income, poorly educated, and overcrowded households

References

1. Abdullah MA, Shaikh BT, Sattar NY, Sarwar B, Ahmed AS, Fatima SS. Are social determinants associated with depression among married women of reproductive age? A mixed methods study from urban slums of Islamabad, Pakistan. PLOS Glob Public Health. 2024;4(7):e0003463.

2. Muhammad FS, Shahabudin SM, Talib MBA. Measuring spatial inequalities in maternal and child mortalities in Pakistan: evidence from geographically weighted regression. BMC Public Health. 2024;24(1):2229.

3. Rahman A. Facilitators, barriers and requirements for addressing perinatal depression in urban slums through mobile phone-based interventions in Bangladesh: a qualitative study [dissertation]. Liverpool: University of Liverpool; 2025.

4. Emikpe AO, Obeng H, Bonsu ZM, Tsiko EA, Owusu LB, Dwumfour CK, et al. Predictors of postpartum depression among postnatal women in Kumasi, Ghana. 2025;25(1):1179.

5. Bhatti MMA, Ashraf MS, Bhatti H, Jamil M, Mukhdooma K, Ibrahim Z. Prevalence and social factors associated with postpartum depression in urban areas of Pakistan: a cross-sectional study. J Rawalpindi Med Coll. 2024;28(4).

6. Zafar A, Akram BA, Wasim M, Pires IM, Coelho PJ. PERI_DEP: a dataset of mothers’ mental health in Pakistan. Data Brief. 2025;60:111621.

7. Sajjad A, Shah S, Abbas G, Aslam A, Randhawa F, Khurram H, et al. Treatment gap and barriers to access mental healthcare among women with postpartum depression symptoms in Pakistan. 2024;12:e17711.

8. Chachar AS, Mian AI. A review of the intersection of social determinants and child and adolescent mental-health services: a case for social psychiatry in Pakistan. World Soc Psychiatry. 2022;4(2):69–77.

9. Makinde OA, Uthman OA, Mgbachi IC, Ichegbo NK, Sule FA, Olamijuwon EO, et al. Vulnerability in maternal, newborn, and child health in low- and middle-income countries: findings from a scoping review. PLoS One. 2022;17(11):e0276747.

10. Mah B, Brown A, Eades S, Pringle K, Rae K, et al. Psychological distress, stressful life events and social disadvantage in pregnant Indigenous Australian women residing in rural and remote New South Wales: a longitudinal cohort study. 2022;9(6):2197–207.

11. Hirani R, Noruzi K, Khuram H, Hussaini AS, Aifuwa EI, Ely KE, et al. Artificial intelligence and healthcare: a journey through history, present innovations, and future possibilities. 2024;14(5):557.

12. Iqbal F, Satti MI, Irshad A, Shah MA. Predictive analytics in smart healthcare for child mortality prediction using a machine-learning approach. Open Life Sci. 2023;18(1):20220609.

13. Raza SH, Khan L. Beyond numbers: mapping poverty disparities in Pakistan through a spatial lens. 2025;18(1):28.

14. Wuermli A, Hiott MC, Ugarte E, Rahman MS, Elahi M, Rahim A, et al. Cohort profile: a prenatal birth cohort study of intergenerational risk and resilience after conflict and forced displacement. BMJ Open. 2025;15(11):e099052.

15. Sah LK. Social determinants of mental health of pregnant women in Nepal: a sequential exploratory mixed-methods research study [dissertation]. Canterbury: Canterbury Christ Church University; 2024.

16. Holland N. An exploration of the factors that influence seeking treatment from the perspective of Muslim women who have experienced postpartum depression [dissertation]. Cape Town: University of the Western Cape; 2024.

17. Shukla A, Purohit N. Mental health matters: investigating the association of maternal psychological factors with pregnancy outcomes in Central Uttar Pradesh, India. 2026:1–11.

18. Hassan R, Jafree S. Exploring the challenges and coping strategies faced by women street vendors in Central Punjab, Pakistan. Lahore: Department of Sociology, Forman Christian College; 2024.

19. Rahman MM, Sorkar MNU, Nakamura R, Islam MM, Alam MA, Azmat SK, et al. Sociodemographic disparities in health-related quality of life among hypertensive patients in Bangladesh: a comprehensive survey analysis. 2025;34(5):1457–71.

20. Park K. Addressing social isolation and suicide risk in South Korea’s permanent rental housing: an empirical investigation of intervention strategies. 2025;11(1).

21. Karim HT. Maternal perceptions of postpartum depression and its influence on children in early years [dissertation]. Dhaka: BRAC University; 2025.

22. Khan T. Household, neighbourhood, and community contexts and the mental health of immigrants and immigrant mothers: an ecological perspective. 2026.

23. Hamidi AR, Ford JD, Novo P, Paavola J, Malik IH. Rethinking flood resilience: systemic risk, governance failures, and the social production of vulnerability in the Himalayan–Indus region. J Disaster Stud Manag. 2026;2(1):8.

24. Arshed M, Kiran M, Ashraf W, Virk MKS, Qamer S, Umer MF, et al. Non-adherence to oral antidiabetic medications among patients with type 2 diabetes: evidence from a nationwide multicentre study in a lower-middle-income country. 2025;25(1):4072.

25. Muqaddam I, Kayani SN, Gul K, Khan M, Murad S, Kanwal A, et al. Effectiveness of community-based health education in reducing home birth rates.

26. Sahni PS, Dewan S, Sachdeva M, Ranathunga N, Fadhullah W, Abdullah JM, et al. Climate cognition crisis: an urgent call to address the effects of air pollution and extreme heat on learning environments for children in South and Southeast Asia. 2025;32(5).

Downloads

Published

2026-06-30

How to Cite

Geographic and Socioeconomic Predictors of Postpartum Depression Identified Through Machine Learning Analysis of Pakistani Health Data. (2026). Link Medical Journal, 4(1), 1-11. https://doi.org/10.61919/thq1qy88

Most read articles by the same author(s)

Similar Articles

1-10 of 158

You may also start an advanced similarity search for this article.