Abstract
This study examined whether having a parent with a mental illness or a parent who abuses substances predicts treatment outcomes for children receiving community-based services for exposure to violence. From 2001 to 2011, data were collected from 492 children from one-and-a-half to seven years old and their primary caregivers enrolled in Safe From the Start services. Results indicated significant improvements pre-intervention to post-intervention in child emotional and behavioural problems, as measured by the Child Behavior Checklist (CBCL). One-way bivariate analyses indicated that children of mothers or fathers with a mental illness and children of mothers who abused substances had higher CBCL scores at intake. Repeated-measures analyses of variance revealed a main effect such that maternal and paternal mental illness and maternal substance abuse were associated with poorer CBCL scores. The only parental risk factor to moderate the association between treatment and CBCL scores was paternal mental illness. Treatment was associated with greater improvement in CBCL scores for children of fathers with, relative to those without, mental illness, and the effect was due to higher CBCL scores at intake for children of fathers with mental illness rather than lower outcome scores. Results suggest that Safe From the Start services which provide early intervention can be effective in improving children's emotional and behavioural functioning. Additionally, the effectiveness of services appears to be robust to parental risk factors such as mental illness and substance abuse.
Acknowledgement
This project was supported by Grant #12-SFSE01-06 awarded by the Illinois Violence Prevention Authority to the University of Illinois at Chicago. Points of view or opinions in this document are those of the authors and do not necessarily represent the official position or policies of the Illinois Violence Prevention Authority.
Notes
1. Owing to the nature of the research question, respondents without both CBCL Time 1 and Time 2 data were not included in analyses. Because the sample was a subset of a larger sample collected for a state evaluation, the authors explored how the study sample may have differed from those excluded. By comparison, those children with both Time 1 and Time 2 CBCL data (n=630) were significantly more likely than those with Time 1 data only (n=1,876) and those with neither Time 1 nor Time 2 CBCL data (n=1,124) to be domestic violence witnesses and physical and sexual child abuse victims, have as the primary caregiver a father or non-parent rather than a mother, have a father without mental illness, and have a father who abuses substances (χ2, p < 0.05).
2. To contextualise these numbers, the 2012 Department of Health and Human Services defines poverty in the United States as an income of under $15,130 for two persons living in the household and under $19,090 for three persons in the household (US Department of Health and Human Services, Citation2012).