Examining the psychometric properties of the Patient Health Questionnaire-9 and Generalized Anxiety Disorder-7 among young urban South African women

Abstract

Background: Valid, reliable, and easy-to-administer scales are crucial for identifying mental health conditions, especially in LMICs where such scales tend not to be validated. This study aims to address this gap by investigating the psychometric properties and factorial structure of the PHQ-9 and GAD-7 in a sample of young women in Soweto, South Africa. Methods: The PHQ-9 and GAD-7 were administered to 6028 women aged 18–28 years old. Cronbach's alpha, Mokken scale analysis, and Confirmatory Factor Analysis were used to provide support for the internal consistency and construct validity of these scales. Results: Both scales demonstrated good internal consistency (α = 0.81 for PHQ-9 and α = 0.84 for GAD-7). Internal consistency reliability was further supported by positive inter-item correlations and item-by-scale correlations for all items on both measures. CFA of the PHQ-9 and GAD-7 showed a reasonable fit for the 1-factor model and 2-factor models (depression and anxiety with somatic and cognitive subtypes). Limitations: This study was limited to young African women in urban Soweto who were proficient in English, which may affect generalizability. Differences in language or cultural context may impact the accuracy and applicability of these scales to other African populations. Conclusion: The PHQ-9 and GAD-7 are valid and reliable for identifying psychological distress in the studied population. Despite showing good psychometric properties, further diagnostic assessment is needed to confirm clinical diagnoses. The scales are useful for identifying those at risk but not a substitute for comprehensive diagnostic evaluations.