Siblings, friends, course-mates, club-mates: how adolescent health behavior homophily varies by race, class, gender, and health status.

Many unhealthy behaviors develop during adolescence, and these behaviors can have fundamental consequences for health and mortality in adulthood. Social network structure and the degree of homophily in a network affect how health behaviors and innovations are spread. However, the degree of health behavior homophily across different social ties and within subpopulations is unknown. This paper addresses this gap in the literature by using a novel regression model to document the degree of homophily across various relationship types and subpopulations for behaviors of interest that are related to health outcomes. These patterns in health behavior homophily have implications for which behaviors and ties should be the subjects of future research and for predicting how homophily may shape health programs focused on specific subpopulations (gender, race, class, health status) or a specific social context (families, peer groups, classrooms, or school activities).