Self-esteem and social relationships: A meta-analytic review of longitudinal studies with continuous time modeling

This preregistered systematic review investigated reciprocal effects between people's self-esteem and the quality of their social relationships by meta-analyzing the available longitudinal data. Analyses were based on 82 samples, including 86,850 participants. Mean age of samples ranged from 8 to 63 years. We used continuous time modeling to account for between-study differences in the time intervals between assessments, which ranged from 1 week to 6 years. Moreover, we controlled for stable between-person variance in the constructs via latent trait variables. The results showed positive cross-lagged effects between self-esteem and relationship quality. The effects peaked at an interval of approximately 3 months, with medium to large effect sizes of β = .13 for the effect of self-esteem on relationship quality and β = .09 for the effect of relationship quality on self-esteem. The effects had largely faded after 1 year. We tested for moderator effects of type of relationship partner, age, proportion of female participants, cultural context, and type of relationship measure. Cross-lagged effects were largest for general social relationships, followed by relationships with parents, but were smaller and close to zero for peer and romantic relationships. Age, proportion of female participants, and cultural context showed significant effects, but the magnitude of these moderator effects was small. The effects also differed by type of relationship measure, with self-reports showing larger effects than other measures. In sum, the findings support theories that emphasize the importance of social relationships for self-esteem and theories suggesting that people's self-esteem shapes the quality of their relationships. (PsycInfo Database Record (c) 2026 APA, all rights reserved).