A longitudinal population-based study of factors in adolescence predicting homelessness in young adulthood.

PURPOSE Almost everything known about risk factors for homelessness is based on cross-sectional studies of non-random samples. Furthermore, most studies have focused on a small number of risk factors and have not evaluated their relative importance. Our aim was to examine which factors, in a population-based sample of adolescents, independently predict homelessness in young adults. METHODS Participants (n = 10,433) in the US National Longitudinal Study of Adolescent Health (Add Health) were initially selected through systematic random sampling of US high schools. Interviews were conducted at home in 1994-1995 when the participants were 11-18 years of age and again in 2001 when participants were 18-28 years of age. We examined the relationships between a range of risk factors reported in adolescence (mood-related problems, substance involvement, delinquency, personality, quality of family relations, neighborhood quality, school adjustment, religious affiliation, perpetration of violence, and experiences of victimization) and experiences of homelessness reported in young adulthood, using regression analysis. RESULTS Each risk factor predicted homelessness. However, only family relationship quality (odds ratio [OR] = .79, 95% confidence interval [CI] = .69-.90), school adjustment problems (OR = 1.57, 95% CI = 1.35-1.82), and experiences of victimization (OR = 1.27, 95% CI = 1.11-1.45) were found to independently predict homelessness. CONCLUSIONS Among a range of well-established risk factors, a troubled family background, school adjustment problems and experiences of victimization were found to be the strongest predictors of homelessness in a general population of young people. Our findings suggest possibilities for the early identification of young persons at risk for homelessness through schools, agencies offering family-based support, and clinical services.