Dual kindness, less distress: A longitudinal and dual-paradigm study on self-compassion, prosocial behavior, and depressive symptoms in Chinese adolescents.

Abstract

Self-compassion (kindness to oneself) and prosocial behavior (kindness to others) are important protective resources against adolescents' depressive symptoms; however, their bidirectional dynamic relationships with depressive symptoms and heterogeneous trajectories of depressive symptoms remain unclear. Clarifying these relationships can provide empirical evidence for precision interventions targeting adolescents' depressive symptoms. This study surveyed 1214 Chinese adolescents (M_age = 15.46 ± 0.71, 60.7% male) three times (with approximately 6-month intervals) and integrated variable-centered and person-centered approaches to (1) examine the bidirectional longitudinal relationships between dual kindness and depressive symptoms using the random intercept cross-lagged panel model (RI-CLPM) and (2) identify heterogeneous developmental trajectories of depressive symptoms using the latent class growth model (LCGM) and subsequently test whether baseline dual kindness predicts trajectory membership and whether depressive symptom trajectories predict T3 kindness qualities. RI-CLPM results showed that at the within-person level, increases in positive self-compassion and prosocial behavior effectively buffered depressive symptoms, with positive self-compassion and depressive symptoms exhibiting bidirectional dynamic relationships, whereas fluctuations in negative self-compassion were unrelated to depressive fluctuations. At the between-person level, global and positive self-compassion as well as prosocial behavior were stably negatively correlated with depressive symptoms, while negative self-compassion was positively correlated with depressive symptoms. LCGM identified three developmental trajectories: a low-symptom declining group, a threshold-symptom increasing group, and a high-symptom remitting group. Baseline self-compassion significantly predicted trajectory group membership across all pairwise comparisons, whereas prosocial behavior predicted membership only in the low-symptom declining group relative to the other two. Different trajectory groups also showed significant differences in subsequent kindness qualities. Together, the two analytical paradigms revealed the complex interplay between dual kindness and depressive symptoms at both the variable level and the individual heterogeneity level. These findings suggest that prosocial behavior may be better suited for universal prevention, whereas self-compassion (enhancing its positive components and reducing its negative components) should serve as a core intervention target for risk groups.