Youth on the Street and Youth Involved with Child Welfare: Maltreatment, Mental Health and Substance Use

Citation: Goldstein, A.L., Amiri, T., Vilhena, N., Wekerle C., Thornton, T., & Tonmyr, L.

Toronto, ON: University of Toronto.

Authors:

• Abby L. Goldstein, University of Toronto
• Touraj Amiri, University of Toronto
• Natalie Vilhena, University of Toronto
• Christine Wekerle, McMaster University
• Tiffany Thornton, Health Canada
• Lil Tonmyr, Public Health Agency of Canada

Funded by:

• Health Canada, Office of Research and Surveillance
• Public Health Agency of Canada, Health Surveillance and Epidemiology Division

Table of Contents

• Acknowledgements
• Executive Summary
• 1 Introduction
  • Purpose and Scope of the Study
  • About This Report
• 2 Research Design and Methodology
  • Study Participants
  • Measures
  • Summary of Key Variables
  • Procedures
• 3 Context
  • Homelessness in Canada
  • Maltreatment and Violence Victimization
  • Substance Use Issues
  • Mental Health Issues
  • Concurrent Mental Health Symptoms and Substance Use
  • Homelessness and Child Welfare Involvement: Shared Characteristics
  • Pathways to Substance Use and Mental Health Issues
• 4 Results
  • Child Maltreatment Among Youth Who Are Homeless
  • Housing Status for Youth Who Are Homeless and Youth Involved with CAS
  • Perceptions of Housing Safety and Youth Experiencing Partner Violence
  • Mental Health: Internalizing and Externalizing Symptoms
  • Substance Use and Other Issues Across Sub-Populations
  • Looking Across the Issues
• 5 Discussion
• 6 Practice and Policy Recommendations
  • Recommendations
  • Conclusion
• References
• Appendix A: Description of Analyses
• Appendix B: Tables

List of Figures and Tables

Figures

• Figure 1: Percentage of youth who are homeless that has experienced maltreatment, by history of CAS involvement
• Figure 2: Percentage of youth who are homeless that has experienced physical abuse, sexual abuse and/or maltreatment, by gender
• Figure 3: Percentage of youth involved in partner violence, as victim and/or perpetrator within the past 12 months, by gender
• Figure 4: Percentage of youth that met the externalizing symptoms criteria, by youth homelessness/CAS status
• Figure 5: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who reported past month alcohol use, by youth homelessness/CAS status
• Figure 6: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who reported binge drinking, by youth homelessness/CAS status
• Figure 7: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who reported smoking cigarettes, by youth homelessness/CAS status
• Figure 8: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who reported marijuana use, by youth homelessness/CAS status
• Figure 9: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who met the cut-off for problem alcohol use, by youth homelessness/CAS status
• Figure 10: Percentage of youth that met the cut-offs for internalizing or externalizing symptoms, and who met the cut-off for problem drug use, by youth homelessness/CAS status

Appendix B: Tables

• Table B-1: Logistic regression coefficients, standard errors and adjusted odds ratios for youth homelessness/CAS status predicting odds of maltreatment
• Table B-2: Logistic regression coefficients, standard errors and adjusted odds ratios for gender predicting odds of physical abuse
• Table B-3: Logistic regression coefficients, standard errors and adjusted odds ratios for gender predicting odds of sexual abuse
• Table B-4: Logistic regression coefficients, standard errors and adjusted odds ratios for gender differences in overall experience of maltreatment
• Table B-5: Logistic regression coefficients, standard errors and adjusted odds ratios for differences in shelter status within the past seven days, by youth homelessness/CAS status
• Table B-6: Logistic regression coefficients, standard errors and adjusted odds ratios for differences in nights spent with parents within the past seven days, by youth homelessness/CAS status
• Table B-7: Summary of one-way ANOVA examining differences in perceptions of safety within the past seven days, by youth homelessness/CAS status
• Table B-8: Logistic regression coefficients, standard errors and adjusted odds ratios for gender differences in spending nights without shelter within the past seven days
• Table B-9: Logistic regression coefficients, standard errors and adjusted odds ratios for gender differences in spending nights with parents within the past seven days
• Table B-10: Summary of independent samples t test for gender differences in perceptions of safety within the past seven days
• Table B-11: Multiple logistic regression coefficients, standard errors and adjusted odds ratios youth homelessness/CAS status and gender predicting the likelihood of partner violence victimization within the past 12 months
• Table B-12: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for youth homelessness/CAS status and gender predicting the likelihood of partner violence perpetration within the past 12 months
• Table B-13: Adjusted odds ratios with 95% confidence intervals for the likelihood of meeting the cut-off for internalizing and externalizing symptoms by youth homelessness/CAS status
• Table B-14: Adjusted odds ratios with 95% confidence intervals for the likelihood of past year substance use by gender, by youth homelessness/CAS status and by internalizing/externalizing symptoms
• Table B-15: Adjusted odds ratios with 95% confidence intervals for the likelihood of past month substance use by gender, by youth homelessness/CAS status and by internalizing/externalizing symptoms
• Table B-16: Adjusted odds ratios with 95% confidence intervals for the likelihood of problem alcohol and drug use by gender, by youth homelessness/CAS status and by internalizing/externalizing symptoms
• Table B-17: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with past month alcohol use
• Table B-18: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with past month binge drinking
• Table B-19: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with cigarette smoking
• Table B-20: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with past month marijuana use
• Table B-21: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with past month illicit drug use
• Table B-22: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with past month polysubstance use, for youth who are homeless only
• Table B-23: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with problem alcohol use
• Table B-24: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of concurrent mental health symptoms with problem drug use
• Table B-25: Multiple logistic regression coefficients, standard errors and adjusted odds ratios for the indicator variables predicting the likelihood of any substance use with concurrent mental health symptoms, for youth who are homeless only