PERI_DEP: A dataset of mother's mental health in Pakistan

Area of Science:

• Mental Health Research
• Clinical Psychology
• Public Health

Background:

• Perinatal depression (PND) significantly affects women's mental health during pregnancy and postpartum.
• Existing datasets often suffer from class imbalance, leading to biased research outcomes.
• There is a need for comprehensive datasets in Pakistan to study PND accurately.

Purpose of the Study:

• To introduce the novel PERI_DEP dataset for PND research in Pakistan.
• To address class imbalance issues in perinatal mental health data.
• To provide insights into the socio-demographic factors influencing PND in urban and rural Pakistani women.

Main Methods:

• Development of the PERI_DEP dataset including 14,008 samples from Lahore and Gujranwala.
• Utilized Patient Health Questionnaire (PHQ-9), Edinburgh Postnatal Depression Scale (EPDS), and socio-demographic surveys.
• Employed Generative Adversarial Network (GAN) oversampling to mitigate class imbalance.

Main Results:

• The PERI_DEP dataset offers a balanced view of PND prevalence and related factors.
• Identified a significant socio-demographic divide between women in urban and rural areas of Pakistan.
• Demonstrated the effectiveness of GAN oversampling in handling imbalanced data for PND research.

Conclusions:

• The PERI_DEP dataset is a valuable resource for advancing PND research in Pakistan.
• Findings underscore the importance of considering socio-demographic contexts in PND interventions.
• This dataset can enhance clinical screening, intervention customization, and risk factor identification for PND.