Early Childhood Sleep Interventions: Widening the Net

• In Event: 1-149 - Sleep in High-Risk Young Children: Disparities, Behavioral Correlates, and Intervention Results

Integrative Statement

Introduction
Although there are a growing number of early childhood sleep interventions with proven efficacy at least in the short term, relatively few have long-term follow-up, and many exclude children with medical, behavioral, or family factors that are commonly perceived to complicate treatment and/or reduce odds of success. As a result, it can be difficult to estimate the potential population health impact of broad implementation.

In the ongoing SHIP (Sleep Health in Preschoolers) study, 433 families of preschool children (2y6m – 5y11m) with behavioral sleep problems were randomized in a clinical trial of a parent coaching intervention that involved a home visit and follow-up phone calls for family-centered education, motivation, and problem solving.

Method
Families were recruited via community ads, and eligible families screened positive for behavioral sleep problems and/or inadequate sleep as reported on the Child Sleep Health Questionnaire (Owens, 2000). Children were excluded only if they had a serious medical condition requiring overnight treatment or monitoring (such as Type I Diabetes), a known diagnosis of autism or receptive language skills below the level of 1y6m, current stimulant medications, or a positive screen for sleep apnea. Of the 424 completing the home visit to begin intervention, 287 have completed the one-year follow-up assessment to date and are reported here. The Child Sleep Wake Scale (CSWS; LeBourgeois, 2016) is a validated parent-report of child sleep administered at baseline and follow-up; diary and actigraphy data are being cleaned and will be available for presentation. Linear regression analyses were conducted for overall CSWS and subscales, and were adjusted for the baseline measure of the scale or subscale. All beta coefficients and 95% confidence intervals were converted into standardized effect sizes (Cohen’s D).

Results
For the overall CSWS, the intervention group scored 0.72 SDs better (95%CI 0.49 – 0.93) at 12 month follow-up as compared to controls. Examining effect distribution across the population, 24% of controls demonstrated considerably worsening of sleep problems from baseline to follow-up (by > 1.0 SD), compared to 12% of the intervention; and marked improvement (by > 1.0 SD) in 52% in the intervention compared to 37% of controls, including 30% vs. 18% for gains > 2.0 SD over the year (p = 0.01). For CSWS subscales, strongest effects (in expected directions) were for going to bed (0.58 SDs, 95%CI 0.40 – 0.77), falling asleep (0.53 SDs, 95%CI 0.34 – 0.73), compared to maintaining sleep (0.17 SDs, 95%CI -0.01 – 0.35), returning to sleep (0.33 SDs, 95%CI 0.09 – 0.58), and returning to waking (0.38 SDs, 95%CI 0.21 – 0.55). Notably, when we examined results of the overall CSWS scale stratified by population groups, we observed no significant or meaningful differences in effect by child gender, age, comorbid conditions, co-sleeping status, parent education, household income, daytime napping, childcare use, or baseline marital tension.

Discussion
Results suggest that the benefits of early childhood sleep intervention can be sustained over a one-year period, and that a broad range of children and families can potentially benefit, including those who are sometimes screened out of research studies and treatment programs.

Authors

• Michelle Garrison, University of Washington
  Presenting Author

• Georgia Goldberg, Seattle Children’s Research Institute
  Non-Presenting Author

• Teresa Ward, University of Washington
  Non-Presenting Author

• Dimitri A. Christakis, University of Washington
  Non-Presenting Author