Abstract
BACKGROUND: Loneliness is increasingly recognized as a critical yet understudied determinant of adolescent mental health, particularly in low- and middle-income countries. While its prevalence and impact have been well-documented in Western contexts, little is known about its role within the psychosocial networks of youth in many African contexts, where social structures and gender norms may diverge sharply. METHODS: We conducted a cross-sectional network analysis using data from 1,445 Kenyan secondary school students. Participants completed validated self-report measures of depression, anxiety, loneliness, social support, optimism, happiness, gratitude, and key demographic variables. Mixed graphical models were employed to examine the global structure and centrality of variables within the adolescent psychosocial network. Gender-stratified analyses and network comparison tests were used to identify sex-specific differences in network architecture and key pathways. RESULTS: Loneliness emerged as the most central psychological variable, directly bridging depressive symptoms, diminished wellbeing, and social support. Gender-stratified networks revealed notable divergences: the positive association between depression and loneliness was significant for girls (edge weight = 0.15) but was not significant appeared in the male network (a statistically significant difference, p = 0.040), while peer support more strongly buffered loneliness for boys. Furthermore, family support was more central for girls, whereas support from friends was more central for boys. CONCLUSIONS: These findings highlight loneliness as a pivotal and gender-contingent node within adolescent psychosocial networks in Kenya. Network-based approaches reveal unique pathways of distress and resilience, underscoring the need for contextually and gender-sensitive interventions.