Adolescence in social context: Longitudinal associations of 15 social factors with health and well-being.

Concern is growing that social disconnection poses serious risks to health and well-being, especially for adolescents. Yet without clear evidence on which relational factors (e.g., parent-child, sibling, romantic, and friend relationships; connectedness to families, schools, and neighborhoods) matter most, designing effective, developmentally attuned interventions remains difficult. To address this question, we used data from a large, nationally representative sample of U.S. adolescents (Add Health) and an outcome-wide approach to prospectively test whether one-year changes (between Wave I: 1994-1995 and Wave II: 1996) in 15 social factors, spanning structural, functional, and quality domains, were associated with 34 indicators of health and well-being 11.37 years later (in Wave IV: 2008, N = 11,040) or 20.64 years later (in Wave V: 2016-2018, N = 9003), adjusting for an extensive set of covariates. The outcome-wide framework fits a series of regression models for the relationship between one predictor and a set of outcomes (one outcome analyzed at a time), adjusting for the same set of covariates in each regression model. Six factors (school connectedness, loneliness, emotional support, neighborhood cohesion, religious service attendance, and family cohesion) showed associations with at least one-quarter of outcomes. Associations were most consistent in psychological well-being, mental health, and civic/prosocial domains. Strengthening school connectedness and emotional support, and reducing loneliness, appear especially promising for fostering long-term health.