Exploring mediators of the Amaka Amasanyufu Multiple Family Group Intervention (2016-2022) on parenting stress reduction among caregivers of children with disruptive behavior disorders in Uganda

Abstract

Residing in poverty-impacted Ugandan communities and raising children with disruptive behavior disorders (DBDs) is associated with above heightened parenting stress. An evidence-based intervention designed to help parents tailor their parenting style to effectively manage children with DBDs was implemented in Uganda between 2017 and 2023. At the end of the intervention, caregivers displayed lower levels of parenting stress than controls, however the mechanisms by which stress reduction occurred is unknown. We investigated whether the intervention's effect on parenting stress was mediated by parenting practices, family cohesion, or mental health. Mediation was assessed using Mitchell and Maxwell's cross-lagged auto-regressive approach. Each mediation model regressed the outcome parenting stress score at each time point onto intervention group assignment (0 = Control; 1 = Multiple Family Group (MFG) intervention) and parenting stress score at the preceding time point. Each mediator was investigated one at a time and regressed onto intervention group assignment and the mediator score at the preceding time point. Parenting stress score was also regressed onto the preceding time point mediator score. The residuals for mediators and parenting stress were allowed to correlate at each time point. We included regression pathways from each covariate to the mediators and outcomes at subsequent waves and were included in the model as random variables. The findings show that the intervention had a significant direct effect on parenting stress at 16 weeks post intervention initiation and was partially mediated by parenting practices, caregiver mental health, and parenting stress at 8 weeks, but we found no evidence it was mediated by family cohesion, supporting two of our hypotheses. Equipping caregivers with parenting skills and strategies to improve mental health can go a long way in reducing overall parenting stress among caregivers with DBDs. Our findings could be leveraged to enhance stress reduction among caregivers of children with DBDs in the long-term.

Keywords: Caregivers; Disruptive behavior disorders; Low resource settings; Mediation effects; Parenting stress.

Fig. 1.. Simplified diagram of the unadjusted mediation models showing the pathways that were fitted in the unadjusted model.

In this model, covariates adjusted for are not shown for simplicity. Each mediator was investigated one at a time. For each model, the group variable exerts both direct and indirect effects on the outcome. There is one key indirect effect, which is the product of path a from the group to the mediator at wave 2 and path b from the mediator at wave 2 to the outcome at wave 3. Of note, pathways with the same letter or name are constrained to have the same estimates. Thus the autoregressive effects of the mediator on itself at later waves is assumed to be constant across time via pathway m as is the outcome via pathway y. The direct effects of the group on the mediator at each wave (pathway c) and the direct effects of the group on the outcome (pathway d) are assumed to be the same over time. We also allow the residuals of the outcome and mediator to correlate within each wave (cov). Residual variances for the mediator (mv1) and outcome (dv1) are assumed to be equivalent across time. These various equality constraints would be relaxed to improve global model fit if the fit of the initial model that assumes these various equalities is poor.