Intersecting Risks: Behavioral Health Outcomes Beyond Single Predictors Among Children In the Child Welfare System

Abstract

Background:
Children involved in the child welfare system face elevated risk for behavioral health difficulties with lasting consequences, including poor social functioning, academic difficulties, substance use, and poor mental health in adulthood. While demographic factors such as age, gender, race/ethnicity, and poverty have been studied, their interactions with service- and system-related factors have been limited. This study addresses that gap by jointly modeling these factors in a nationally representative sample to identify vulnerable subgroups and inform targeted intervention.

Methods

Data were drawn from the National Survey of Child and Adolescent Well-Being II (NSCAW II), a nationally representative longitudinal study of children involved in the child welfare system. The analytic sample included N=5,872 children ages 0–17, with predictors measured at baseline and behavioral health outcomes assessed at Wave 3. To account for the complex survey design, weighted logistic regression was used to estimate the likelihood of scoring in the clinical range on the Child Behavior Checklist (CBCL) Total Problems scale. Predictors included types of initial child maltreatment report, the proportion of days spent in out-of-home care, and the receipt of CPS services while controlling for child race/ethnicity, gender, age and poverty status.

Results:

The results showed that children whose initial CPS involvement was due to supervisory neglect had higher odds of behavioral problems (β = 0.204, p = .034), and spending more days in out-of-home placement was associated with increased odds of clinical-level problems (β = 0.06, p < .001).

A significant interaction emerged between race/ethnicity and poverty status. Black children living in poverty had substantially lower odds of clinical-level problems (β = –0.78, p = .002), whereas children in the “Other” ((American Indian/Alaska Native, Native Hawaiian/Pacific Islander, and Asian) racial/ethnic group living in poverty had markedly higher odds (β = 1.56, p < .001). Additionally, female children had lower odds of clinical-level problems than males (β = –0.29, p = .003), and older age at baseline was associated with increased odds of scoring in the clinical range on the CBCL (β = 0.16, p < .001).

Among the different types of services provided by CPS (economic, family counseling, and mental health), only receipt of parental substance abuse treatment was significantly associated with reduced odds of children scoring in the clinical range on the CBCL (β = –0.715, p < .001).

Implications

Findings call for screening practices that account for intersecting vulnerabilities rather than demographic factors in isolation. Children who are older, male, or in out-of-home placement should be prioritized for behavioral health services. Culturally responsive services are needed for underrepresented minority subgroups living in poverty whose risk profiles may be obscured in aggregate analyses. The protective effect of parental substance abuse treatment suggests that caregiver-level interventions are a high-leverage investment for improving child behavioral health outcomes. Together, these findings support a more intersectional and family-centered approach to service planning in child welfare.