Twelve-month trajectories of depressive symptoms and associated socio-structural factors among adolescents and young adults in Soweto and Durban, South Africa

J Affect Disord. 2025 Oct 9:120409. doi: 10.1016/j.jad.2025.120409. Online ahead of print.

ABSTRACT

INTRODUCTION: Globally, approximately 14 % of adolescents and young adults live with health disorders such as depression. We investigated longitudinal (12-months) trajectories of depressive symptoms and associated socio-structural correlates, among South African youth aged 16-24.

METHODS: 425 youth were enrolled in AYAZAZI, a prospective cohort study (2014-2019) assessing intersecting socio-behavioural, structural, and biomedical HIV risk-factors among youth. Median age was 19 [IQR 18-21]; 60 % were women. The Centre for Epidemiological Studies Depression (CES-D-10) scale measured depressive symptoms; a CES-D-10 score of ≥10 indicating probable depression. A group-based trajectory modeling (GBMT) approach was used to identify trajectory classes of depressive symptoms over 12-months. We used prediction modeling, utilizing multinomial regression, to estimate baseline candidate predictors (n = 9) of depression trajectory class.

RESULTS: At baseline, 43 % (n = 181) had probable depression (49 % women, 35 % men), decreasing to 35 % (p < 0.01) at 12-months (39 % women, 30 % men). GBTM modeling indicated four distinct probable depression trajectories comprising ‘high-stable’ (n = 39), ‘moderate-stable’ (n = 182), ‘low-declining’ (n = 165), and ‘lowest-declining’ (n = 35). In multinomial regression, being a woman was associated with higher, persistent symptoms and lower likelihood of symptom decline; being a student and higher resiliency were associated with a lower likelihood of higher or moderate symptoms trajectories; food insecurity predicted high persistent depressive symptoms across follow-up.

DISCUSSION: The prevalence of probable depression decreased over the 12-month among both genders. Given evident longitudinal associations between depression trajectories and markers of sociostructural inequity, including gender, food insecurity, our results emphasize the need for holistic approaches to address fundamental health determinants for South African youth.

PMID:41076165 | DOI:10.1016/j.jad.2025.120409