Psychometric properties of the patient health questionnaire-8 and general anxiety disorder-7 in adolescents and young adults from three Latin American cities: Internal structure, invariance, internal consistency and divergent validity.

BACKGROUND

Depression and anxiety are the most common mental disorders in the world. Screening tools allow early diagnosis and intervention, preventing disease progression and reducing years lost to disability. Internationally, the most widely used screening instruments for depression and anxiety symptoms are the Patient Health Questionnaire (PHQ-8) and the Generalised Anxiety Disorder (GAD-7). Both have been validated in various countries; however, these studies mostly focused on the adult population and clinical settings.

OBJECTIVE

This study assessed the psychometric properties of the PHQ-8 and GAD-7 in adolescents and young adults in deprived urban areas of three of the largest Latin American cities. We evaluated the internal structure, measurement invariance, internal consistency, and divergent validity through correlations with the Manchester Short Assessment of Quality of Life (MANSA).

METHODS

The study included 1056 adolescents (ages 15-16) and 1306 young adults (ages 20-24) from Bogotá (Colombia), Buenos Aires (Argentina), and Lima (Peru). Confirmatory factor analysis was performed using the Weighted Least Squares Mean and Variance (WLSMV) adjusted estimator with polychoric matrices to assess the internal structure of the model. Measurement invariance was evaluated through multi-group factor analysis. Divergent validity was examined using Spearman's rho by correlating MANSA with the PHQ-8 and GAD-7. Finally, internal consistency was assessed with Cronbach's alpha and McDonald's omega.

RESULTS

Our study found that the PHQ-8 and the GAD-7 exhibited good goodness-of-fit indices for a one-dimensional model. Both scales achieved measurement invariance across different sociodemographic variables, including gender, country, education level, and age group. On the other hand, the PHQ-8 (rs = -0.52) and GAD-7 (rs = -0.46) showed a negative correlation with the MANSA, suggesting divergent validity between the higher scores on depressive and anxious symptoms are associated with lower quality of life. Finally, the PHQ-8 and GAD-7 showed adequate internal consistency in all cases evaluated (ω and α > 0.8).

CONCLUSIONS

Both the PHQ-8 and the GAD-7 showed good goodness-of-fit indices for a unidimensional theoretical model, adequate psychometric properties, and evidence of invariance for gender, age group, education level, and country. These findings highlight the reliability and versatility of these instruments in identifying urban young people with emotional distress in Latin America.