Background: Homeless youth in the United States have high rates of substance use. Existing research has identified social network composition and street-associated stressors as contributing factors. Incarceration is a highly prevalent stressor for homeless youth. Its effect on youth’s social network composition and substance use, however, has been neglected. Aims: This study investigated the direct and indirect associations between incarceration history and substance use (through social networks) among homeless youth in Los Angeles, California. Methods: A sample of 1047 homeless youths were recruited between 2011 and 2013. Computerized self-administrated surveys and social network interviews were conducted to collect youth’s sociodemographic characteristics, incarceration history, social network composition, and substance use. Bootstrapping was used to identify the direct and indirect associations between youth’s incarceration history and substance use. Results: Incarceration history was positively associated with youth’s cannabis, methamphetamine, and injection drug use. The percentage of cannabis-using peers partially mediated the associations between incarceration history and youth’s cannabis, cocaine, and heroin use. The percentage of methamphetamine-using peers partially mediated the associations between incarceration history and youth's methamphetamine, cocaine, and injection drug use. The percentage of heroin-using peers partially mediated the association between incarceration history and youth’s heroin use. Moreover, the percentage of peers who inject drugs partially mediated the associations between incarceration history and youth’s methamphetamine, heroin, and injection drug use. Discussion: Incarceration history should be taken to a more central place in future research and practice with homeless youth in the United States.

Despite growing research on women’s offending trajectories, knowledge on the imprisonment experiences of women who are mothers is limited. This study used a nationally representative dataset of state and federal prisoners to identify and characterize subgroups of mothers based on incarceration histories. Group-based trajectory modeling identified four groups with distinct incarceration trajectories: stable escalating group, moderate declining group, adolescence-peak group, and young-adulthood-peak group. Bivariate analyses then suggested that different incarceration trajectories were associated with adverse life experiences (i.e., foster care placement, sexual abuse experience) and confinement experiences (i.e., number of episodes of prison confinement). Discussed are the future research directions and implications for criminal justice policies and interventions targeting imprisoned mothers.