Behavioral treatments for sleep problems in youth: A meta-analytic review

More than 50% of children in the general population have sleep problems at some point during their childhood (Bates, Viken, Alexander, Beyers, & Stockton, 2002; Donaldson & Owens, 2006; Gregory, Rijsdijk, & Eley, 2006; Owens, 2004). Sleep problems put children at risk for a variety of cognitive, academic, affective, and psychological difficulties (Owens, 2004; Sadeh et al., 2000; Stores, 1999). Evidence-based interventions are clearly needed to address these sleep problems; however, the literature in general, and with respect to treatment-outcome, is limited as compared to the breadth published for adults (Gregory et al., 2006; Sadeh et al., 2000; Stein et al., 2001). Further, research disseminating the treatment outcome literature is limited to reviews as a meta-analysis has not yet been conducted for behavioral treatments of pediatric sleep disorders. The present research attempts to build upon these reviews (Kuhn & Elliot, 2003; Mindell, 1999) by conducting a comprehensive meta-analysis with an increased number of research studies included compared to the reviews. This present meta-analysis identified 33 usable studies (27 published journal articles and six dissertations) with 11 additional treatment groups, resulting in 44 usable treatment groups included in this meta-analysis. While most interventions utilized unique combinations of the behavioral techniques, positive routines (66%) was the most frequently included component, followed by graduated extinction and written information (both 36%). A large overall within sleep treatment group unweighted effect size was found (M = 1.11; SD = 0.60), which indicates that after receiving treatment for sleep problems, children and parents demonstrated positive changes on the outcome measures. Further, a significantly larger overall within sleep treatment group unweighted effect size was also found at follow-up (M= 2.02; SD = 0.97), which indicates that significantly more treatment gains were made at follow-up compared to post-treatment. Effect sizes were also examined by intervention type, number of behavioral techniques used, construct used to measure change, duration of treatment, and setting of treatment. A small to medium positive effect size was found for the overall comparison of sleep treatment to control groups (M = 0.42; SD = 0.24). Limitations and implications for school psychologists are discussed.