BackgroundAdolescent psychological crisis—encompassing depressive disorder, anxiety disorder, and suicidal ideation—is a major public health challenge. One in five adolescents experiences a clinically significant mental health condition globally. Loneliness is a well-documented, modifiable risk factor for these outcomes. However, the exact cognitive-emotional pathways linking loneliness to acute psychological crisis remain unclear. This is particularly true in China, where rapid social change and intense academic pressure increasingly strain adolescent mental health. We do not yet fully understand how protective resources might buffer these specific pathways.Objectives and methodsThis pilot study tests a moderated mediation model based on the Evolutionary Theory of Loneliness and the stress-vulnerability framework. We investigate ruminative thinking as a cognitive-emotional mediator between loneliness and psychological crisis. We also test whether perceived social support and psychological resilience moderate distinct stages of this pathway. Based on a pilot sample of N = 312 Chinese adolescents (ages 12–18), we use structural equation modelling (SEM), confirmatory factor analysis (CFA), and Hayes’ PROCESS Macro (Model 14) with 5, 000 bootstrap replications. Two-wave longitudinal data (n = 187, 8-week interval) were available for the depression outcome only; the anxiety and suicidal-ideation pathways were examined cross-sectionally.ResultsLoneliness is significantly associated with depressive symptoms (β = 0.43, p < 0.001), anxiety (β = 0.38, p < 0.001), and suicidal ideation (β = 0.27, p < 0.001). Ruminative thinking accounts for 40–46% of the total indirect association across all three outcomes. Social support buffers the transmission from loneliness to rumination (b = −0.16, p = 0.002, ΔR2 = 0.023), while resilience buffers the transmission from rumination to crisis (b = −0.19, p < 0.001, ΔR2 = 0.031).ConclusionsThese preliminary, exploratory findings are consistent with the possibility that ruminative thinking partially accounts for the cross-sectional association between loneliness and adolescent psychological crisis in China, and with social support and resilience attenuating this association at potentially distinct stages. Because the pilot is underpowered for the moderated mediation model and the available longitudinal evidence covers depression only, the moderation and stage-specificity results should be read as hypothesis-generating rather than confirmatory. A planned full-scale study (N ≥ 1, 200) will test these mechanisms further. We also outline initial practical implications for school-based interventions.
Radically different