Does Rural and Urban Residence Matter?: Mental Health Need Among Child Welfare Involved Children

Background and purpose: Very little is known about the 20% of child welfare involved children who live in rural areas. The few existing studies suggest differences between these populations; hence, different types of services may be needed in rural areas.  A single-state study found that household factors may be more important in rural areas than in urban areas, but no national studies have analyzed these differences. This study uses nationally representative data to answer:  What are the differences in child and household characteristics as predictors of mental health need for child welfare involved children in urban areas compared to rural areas?

Methods: Data were from the Second Cohort of the National Study of Child and Adolescent Well-Being and included children 18 months to 18 years who remained in-home following an investigation of maltreatment (n=2,135).  The sample was 20.4% African American, 43.4% White, 28.4% Hispanic; 50.5% were male.  The outcome variables of interest, from the Child Behavior Checklist (CBCL) a common proxy for mental health need, were the total score (mean=53.1), the internalizing scale (mean=52.1) and the externalizing scale (mean=53.5).  The inclusion of significant interactions of urbanicity and other independent variables was driven by prior studies and data exploration.  Specifically, for each outcome, a base model including the selected variables was run. Second, a model that included the base model and the interaction of urbanicity with other independent variables (i.e. employment, number of children, primary maltreatment type, parent education, and being at least 200% above the poverty level), was run (one interaction term included in each model).  Significant interactions were retained in the final models.

Results: For the CBCL total score, male gender, child’s age, employment, number of children, parent’s age, physical abuse as the primary type of maltreatment and the interaction between urbanicity and employment were significant (F(10,65)=16.70(p<.001),R²=.08).  For internalizing problems, urbanicity, being white, physical abuse, poverty and the interaction between sexual abuse and urbanicity were significant (F(13,62)=6.12(p<.001),R²=.06).  For externalizing problems, male gender, child’s age, employment, physical abuse, poverty, and the interactions between employment and urbanicity (rural higher) and sexual abuse and urbanicity were significant (F(13,62)=9.89(p<.001),R²=.10).

Conclusions and implications:  There were differences in predictors of mental health need between rural and urban areas.  For the total and externalizing scores, children whose parents were employed less than full-time had higher scores; however, parental employment had a larger effect on children’s mental health need in rural areas.  Sexual abuse as the primary type of abuse was related to a higher CBCL score for internalizing and externalizing problems in urban but not rural areas.  These differences were consistent with the previous finding that household factors, such as employment, are more of a problem in rural areas.  These findings suggest that different types of services may be needed in rural and urban areas to address the underlying family issues related to children’s mental health needs.  To better serve both populations, there is also a need for additional research to further illuminate the differences between urban and rural child welfare involved populations.