Are There “Sleeper Effects” of Parenting Interventions on Reduced Disruptive Child Behavior: A Multilevel Meta-Analysis

Introduction: Parenting interventions aim to start a chain of change in which positive parent and child behavior reinforce each other in order to prevent or reduce disruptive child behavior. From a prevention perspective, it is vital to understand whether any successful change results in sustained effects. If parents and children continue to positively reinforce each other behaviors after interventions end, this would result in so called “sleeper effects” of parenting interventions, which reflects further improvement after intervention termination. If parents and children are not able to continue to positively reinforce each other, and even show a  fallback to previous behavior patterns, this may result in “fade-out” effects. Aim of this meta-analysis is to test the extent to which effects of parenting interventions on child behavior increase after intervention termination (i.e. sleeper effects), or whether they remain stable or even decrease (i.e., fade out effects)?

Methods: We aimed to identify sleeper-, sustained- or fade-out effects of parenting interventions on reduced disruptive behavior by evaluating the effect from posttest to follow-up. PsycINFO, Medline, and Eric were searched for randomized controlled trials reporting on follow-up effects of parenting interventions. In addition, reference lists of relevant systematic reviews and meta-analyses and experts in the field were consulted. Forty randomized controlled trials with a follow-up measurement of both intervention and control condition were included (n = 6,232), generating 93 standardized mean difference effect sizes over a follow-up period of one month to three years.

Results: A three-level-random-effects model estimated an overall small to moderate effect size of the immediate effect of the intervention (pretest – posttest):  d = .393 (95% CI = .278 to .508, p < .001). More important for the purpose of this study, the mean effect size of change in disruptive child behavior between posttest and follow-up was d = -.016 (95% CI = -.086 to .054, p= .642). This indicates that there was no further change in disruptive child behavior after immediate posttest, which means that there are no sleeper- or fade-out effects of parenting interventions. Instead, parenting interventions tend to lead to sustained effects until later follow-up.

Conclusions: Parenting interventions led to sustained effects, at least until three year follow-up. This is a hopeful conclusion for trials that are unable to incorporate follow-up measures in their design (e.g., due to a waitlist control design). It is also a hopeful conclusion for prevention science: families seem to be able to maintain initial improvements they make during intervention.