Background:
Social, living conditions and school level predictors significantly affect mental health literacy of learners preventing early identification and intervention of mental health disorders. This study applies hierarchical logistic regression model to determine the effects of social, living conditions and school level predictors on mental health literacy (MHL) among high school learners.

Methodology:
A secondary data analysis were conducted utilizing 529 secondary school learners selected using a stratified random sampling approach, from five different school, four in Townships and one in urban area. Hierarchical logistic regression with four levels was conducted to determine significant predictors of MHL and a p-value less than 0.05 was considered statistically significant.

Results:
The hierarchical multiple logistic regression analysis was conducted in four sequential blocks to examine significant predictors of MHL among learners. In Models 1 and 2, which included individual-level variables and family living conditions, no statistically significant predictors of MHL were identified. However, the inclusion of school-level variables in Model 3 significantly improved model fit, with both school location and grade level emerging as significant predictors. Learners attending urban schools were over three times more likely to exhibit higher MHL compared to those in township schools (OR = 3.03; 95% CI: 2.02, 4.55; p < .001). Similarly, learners in Grade 12 had significantly higher odds of MHL than those in lower grades (OR = 6.24; 95% CI: 2.24, 17.39; p < .001). In Model 4, the addition of parental employment status further enhanced the model fit, revealing that learners with employed parents had significantly greater odds of higher MHL (OR = 1.65; 95% CI: 1.30, 2.09; p < .001).

Conclusion:
The hierarchical multiple logistic regression analysis revealed that individual-level variables were not significant predictors of MHL in the initial models. However, the inclusion of school-level variables in later models significantly improved the overall model fit. Learners in urban location and those in Grade 12 demonstrated higher odds of MHL. Furthermore, parental employment also emerged as a significant predictor, indicating that school level and socioeconomic context plays a critical role in shaping learners’ MHL. This study introduces a multilevel approach to understanding MHL, revealing that school level and socioeconomic predictors outweigh individual-level predictors. By highlighting the structural determinants of MHL, this study urges practitioners to move beyond individual-based models and advocate for systemic, school-based interventions that address socioeconomic and educational inequities and promote mental well-being among learners.

Keywords: Mental health literacy, Socio-economic inequalities, Access to services, educational inequalities.