Mental health measures among adolescents in 12 low‐ and middle‐income countries: Measurement invariance and cross‐sectional analyses of Disrupting Harm survey data

Abstract

Abstract Background Nationally representative mental health data in adolescents from low‐ and middle‐income countries (LMICs) are scarce. This study aimed to examine mental health and wellbeing indicators amongst adolescents in 12 LMICs across Eastern and Southern Africa and Southeast Asia. Methods We conducted a secondary analysis of data involving 12,169 internet‐using adolescents, aged 12–17 years, from Ethiopia, Kenya, Mozambique, Namibia, Tanzania, Uganda, Cambodia, Indonesia, Malaysia, Philippines, Thailand, and Vietnam. We explored cross‐country measurement invariance of the multi‐item mental health scales, computed country‐level estimates for life satisfaction, psychological wellbeing, anxiety, depression, loneliness, self‐harm and suicidal ideation/attempts, and examined socio‐demographic variations (across age, gender and food insecurity). Results Measurement invariance was not established, limiting cross‐country comparisons. No consistent patterns of mental health estimates emerged across countries. The greatest variation was observed for loneliness (ranging from 18.3% in Vietnam to 71.3% in the Philippines) and for suicidal ideation/attempts (ranging from 7% in Vietnam to 52.7% in Uganda). Gender and age disparities were present, but their magnitude and direction varied by country. The experience of food insecurity was the most consistent correlate of mental health outcomes, with significant associations with life satisfaction and psychological wellbeing in seven countries each, with anxiety in six, and with depression and loneliness in four. Conclusion Overall, our findings provide key insights into the validity and use of a range of mental health measures among adolescents across 12 LMICs, as well as valuable within‐country observations that may help inform and tailor interventions, policies and future research on adolescent mental health to local contexts. Key points What's known? Nationally representative mental health data on adolescents from low‐ and middle‐income countries are scarce. Understanding variations by gender, age, and food insecurity is essential to identify particularly vulnerable groups. What's new? Measurement invariance was not supported for the multi‐item scales, limiting between‐country comparisons. No consistent patterns of mental health estimates emerged across countries; however, food insecurity was the most consistent correlate of mental health outcomes. What's relevant? Our findings offer valuable within‐country insights into measurement of adolescent mental health across 12 LMICs, with respect to the validity and use of a range of measures that can help tailor research on adolescent mental health to local contexts.