Caregiver Problematic Drug Use and Child Behavior Problems in a National Probability Sample of Child Welfare Families

Background/Purpose:Caregiver problematic drug use (CPDU) has been associated with an increased risk for child internalizing behaviors (IB) and externalizing behaviors (EB) among children involved with child protective services (CPS).  However, research is needed to explain the pathways from CPDU to child behavior problems in order to identify intervention points when children remain in the home (87% of CPS population).  Among families investigated by CPS whose child remain in the home, do parental monitoring, harsh discipline, emotional maltreatment, and exposure to violence serve as mediators and caregiver depression, domestic violence, and criminal involvement serve as moderators in the relationship from CPDU at baseline to child IB and EB at 18-month follow-up?

Methods: Using data from the National Survey of Child and Adolescent Well-Being II, this analysis utilized a sub-sample of families in which children remained in the home following a CPS investigation.  A random half sample (n=1015) was used for each of two path models to examine direct and indirect pathways from CPDU to, separately, child IB and EB.  Results were confirmed on the second half of the sample.  CPDU was measured continuously with the 20-item Drug Abuse Screening Test.  Four parallel mediators were examined continuously: exposure to violence (Violence Exposure Scale), parental monitoring (Supervision-Child Scale), physical assault (Parent-Child Conflict Tactics Scale; CTSPC), and psychological aggression (CTSPC).  Three moderators were examined: domestic violence (Conflict Tactics Scale-Physical Assault), criminal involvement (self-report of criminal conviction), and caregiver depression (Composite International Diagnostic Interview-Short Form).  The internalizing and externalizing subscales of the Child Behavior Checklist were used continuously as separate dependent variables in two different models.  Control variables were child age and gender, poverty, and family CPS history.  Analyses were conducted in Mplus 7.0 and standardized indirect effects are reported.  A model building approach was utilized and model fit was compared using fit indices.

Results: In the EB model, significant single mediators were physical assault (μ=.037, σ=.015, p<.05) and psychological aggression (μ=.053, σ=.023, p<.05). Comparing fit indices of single and multiple mediator models, the strongest model for EB (RMSEA=.022, p=.90, CFI=.98) was a single mediator model through psychological aggression.  Physical assault (μ=.018, σ=.009, p<.05) and psychological aggression (μ=.037, σ=.017, p<.05) each individually mediated the relationship from CPDU to IB.  Two separate models were strong for IB.  First, a single mediator model through psychological aggression (RMSEA=.022, p=.90, CFI=.96) fit the data well.  When testing depression as a moderator, the path controlling for depression separate from the interaction was significant.  Supported by theory and consistent with model building, a single mediator model through caregiver depression was tested and found to be a significant mediating pathway (μ=.045, p<.05) which fit the data well (RMSEA=.000, p=.98, CFI=1.00).  Confirmation supported depression as a single mediator in the pathway from CPDU to IB.

Conclusions/Implications:Analyses indicate increases in CPDU were associated with an increased likelihood of caregiver depression.  An increased likelihood of depression was associated with increased IB scores.  Further research is needed to disentangle how genetics and parenting behaviors are separately impacting this pathway.