Depression and anxiety among young people are a major public health concern in South Africa. The emergency of the COVID-19 pandemic increased young people's exposure to multiple social determinants of mental health problems. There is a need to identify young people at risk of depressive and anxiety symptoms and link them to appropriate care. There is scarce research on the predictors of young people's mental health at the individual, household, and community level in sub-Saharan Africa (SSA). As a first step, there is need to understand the predictors of depressive and anxiety symptoms to inform community-based screening tools among young people. This knowledge can help us better design positive mental health interventions such as screening tools and community-based programmes to reduce the burden of mental health conditions in this population.
The aim of this study is to analyze the predictors of depressive and anxiety symptoms among young people aged 14-24 years using existing data from three surveillance sites in South Africa.
The study will use a cross-sectional South African Population Research Infrastructure Network (SAPRIN) datasets, as well as a study population of young people aged 14-24. Multilevel predictor models will be used to determine the individual-, household-, community- level factors associated with depressive and anxiety symptoms. The selection of factors will be informed by previous literature, Wellcome Trust active ingredients framework (<https://wellcome.org/sites/default/files/active-ingredients-insights> RFP.pdf), and feedback by a local Community Advisory Board (CAB) established by the Youth Health Economics Focal Area of the SAMRC. The Wellcome Trust active ingredients framework is a list of interventions that help prevent, manage, or treat depressive and anxiety symptoms among young people that were identified by research teams during the first round of Wellcome Trust funding for this grant call. Some of the factors to be assessed include social connection, age, sex, health status, employment status, head of household, household financial status, household food insecurity, and community violence against women, men, and children. We will implement the Generalized Estimating Equations (GEE) as our primary prediction model. The best model at predicting depressive and anxiety symptoms will be assessed using likelihood ratio test. All predictor variables with a p-value < 0.05 in the parsimonious model will be considered the top potential predictors of depressive and anxiety symptoms.
The outputs of this study will help inform the development of the digital tool that will help to screen and identify young people at risk of developing depressive and anxiety symptoms.