Measuring Antenatal Depressive Symptoms Across the World: A Validation and Cross-Country Invariance Analysis of the Patient Health Questionnaire-9 (PHQ-9) in Eight Diverse Low-Resource Settings

Public Significance Statement Antenatal depression is a significant global problem; however, there has been limited research into symptom measures in a diversity of low-resource contexts. In this study, common measure of antenatal depressive symptoms (PHQ-9) gave scores with good reliability across eight different countries and nine languages; however, scores cannot always be compared across settings. We provide recommendations for making symptom measures more relevant within and comparable across diverse country contexts. Measures that produce valid and reliable antenatal depressive symptom scores in low-resource country contexts are important for efforts to illuminate risk factors, outcomes, and effective interventions in these contexts. Establishing the psychometric comparability of scores across countries also facilitates analyses of similarities and differences across contexts. To date, however, few studies have evaluated the psychometric properties and comparability of the most widely used antenatal depressive symptom measures across diverse cultural, political, and social contexts. To address this gap, we used data from the Evidence for Better Lives Study-Foundational Research (EBLS-FR) project to examine the internal consistency reliability, nomological network validity, and cross-country measurement invariance of the nine-item version of the Patient Health Questionnaire (PHQ-9) in antenatal samples across eight low-resource contexts. We found that the PHQ-9 scores had good internal consistency across all eight countries. Correlations between PHQ-9 scores and constructs conceptually associated with depression were generally consistent, with a few exceptions. In measurement invariance analyses, only partial metric invariance held and only across four of the countries. Our results suggest that the PHQ-9 yields internally consistent scores when administered in culturally diverse antenatal populations; however, the meaning of the scores may vary. Thus, interpretation of PHQ-9 scores should consider local meanings of symptoms of depression to ensure that context-specific conceptualizations and manifestations of antenatal depressive symptoms are adequately reflected.