Consistent use of bedtime parenting strategies mediates the effects of sleep education on child sleep: secondary findings from an early-life randomized controlled trial

Abstract

Objective: Using data from a randomized controlled trial, this study investigated whether parent's consistent use of recommended bedtime strategies with infants was a mediating mechanism for improved child sleep.

Methods: Expectant mothers were allocated to 4 groups: usual care (Control); additional support regarding Food, (physical) Activity, and Breastfeeding (FAB); advice on infant sleep through 1 group educational session during the antenatal period and 1 home visit when the child was 3 weeks of age (Sleep), or both FAB and Sleep interventions (Combination). An index relating to parent's consistent use of strategies to encourage infant sleep self-settling was developed from data collected when infants were 4 and 6 months of age. Child sleep self-control was measured at 3.5 years of age through a behavior rating scale. Child overnight sleep duration was measured using accelerometers at 1, 2, 3.5, and 5 years of age. Analyses examined whether any association between intervention group and child sleep self-control or sleep duration was mediated by consistent use of bedtime strategies at 4 and 6 months.

Results: Compared to Controls, Sleep group parents had significantly higher odds of using more intervention strategies consistently (1.63; 95% confidence interval [CI] 1.14-2.33), as did Combination group parents (1.45; 95% CI 1.01-2.07). Consistent strategy use was significantly associated with a decrease in child bedtime behavioral difficulties (0.97; 95% CI 0.95-0.98) and increased sleep duration (0.152; SE = 0.017). Sleep group assignment reduced child sleep self-control difficulties and improved sleep duration indirectly via parent's consistent use of bedtime strategies.

Discussion: Consistent use of appropriate bedtime strategies in infancy is an important factor that influences child sleep self-control in later development.

Trial registration: ClinicalTrials.gov NCT00892983.

Copyright © 2019 National Sleep Foundation. Published by Elsevier Inc. All rights reserved.

Figures

Fig. 1.

Path diagrams illustrating mediation effects of Bedtime Parenting Strategies. OR = odds ratio, and is the exponential of an ordinal logistic regression estimate (presented in Table 2). RGM = ratio of geometric means, and is the exponential of linear regression estimate of a log-transformed dependent variable (presented in Table 3). β = the mediation effect computed using the product-of-coefficients method from nonexponentiated estimates from ordinal logistic and linear regression models. * Indicates statistically significant difference from reference group at P < .05. ** Indicates statistically significant difference from reference group at P < .01. *** Indicates statistically significant difference from reference group at P < .001.

Fig. 2.

Mediation model predicting trait and state overnight sleep duration. Circles represent state and trait latent variables for overnight sleep duration (latent variable indicators not depicted). Dashed lines represent latent variable loadings. Boxes represent observed variables. Solid lines and corresponding numbers represent estimated regression coefficients. Only significant effects are depicted. a = regression coefficients predicting the mediator. b = regression coefficient for mediator predicting outcomes. c = regression coefficients for direct effects on outcome. c’ = regression coefficients for mediation effects on outcome. * Indicates statistically significant difference from reference group at P < .05. ** Indicates statistically significant difference from reference group at P < .01. *** Indicates statistically significant difference from reference group at P < .001.