Caregiver Response Trajectories and Their Predictors in a Digital Parenting Program for Child Maltreatment Prevention

Purpose Although parenting programs have been shown to be effective in reducing child maltreatment, previous evaluations have predominantly adopted group-level analysis, thereby obscuring potential heterogeneity in caregiver responses. This study examined the response trajectories to an online, self-directed parenting program—the Chinese version of Parenting for Lifelong Health for Young Children—and identified predictors that distinguished responders from non-responders. Method Data were obtained from a two-arm, cluster-randomized controlled trial conducted in a kindergarten in Hebei Province, China. A total of 89 caregivers in the intervention group were included in the trajectory analyses. Latent class growth analysis was performed to identify distinct trajectories of change in child maltreatment, followed by multinomial logistic regression analyses to examine predictors of trajectory membership. Results Four trajectories were identified: non-response class (18%), low-decreasing class (25%), high-decreasing class (53%), and worsening class (4%). Compared with the non-response class, higher program session attendance and lower baseline parenting stress were associated with more favorable trajectories. Although no significant predictors were identified for the worsening trajectory group, further analyses suggested that chronic family-level risks might hinder intervention effectiveness. Conclusion The considerable heterogeneity in response underscores the importance of sustained engagement, parenting stress alleviation, and consideration of broader family contexts to optimize the impact of parenting interventions.