Measurement properties of the SUNRISE Study Parent Questionnaire for assessing sleep and sleep-related family practices

Abstract

Background: Few questionnaires with established measurement properties can globally measure sleep in preschoolers and sleep-related family practices. Objective: To examine (1) concurrent validity of the SUNRISE parent questionnaire against an accelerometer for measuring sleep in preschoolers and (2) test-retest reliability of the questionnaire for sleep and related family practices. Methods: Sleep was measured using the questionnaire and Actigraph GTX3+ accelerometer using a decision-tree algorithm and the Sadeh algorithm in 1737 preschoolers (4.4 ± 0.6 years) from 30 countries. Concurrent validity was examined using correlation analysis (duration, timing, and quality), paired t test or the Wilcoxon signed-rank test, Bland-Altman plot (duration, timing), and analysis of sensitivity, specificity, and accuracy (variability). Test-retest variability was examined for sleep and family practice variables in a subsample of 163 participants (4.3 ± 0.6 years) from eight countries. Results: Questionnaire measures of sleep timing and duration were correlated with the accelerometer measures (r = 0.43-0.75; p < .001). Although statistically significant mean differences were observed between questionnaire and accelerometer measures of sleep timing and duration variables, the difference in nighttime sleep duration had a small effect size (−14 min/d; Cohen's d = −0.2). The questionnaire was less able to provide adequate measurement for sleep quality and variability. High levels of reliability were observed for sleep (ICC = 0.63-0.83; Kappa = 0.53-0.62) and family practice (ICC = 0.81-0.94; Kappa = 0.73-0.86) variables. Conclusion: The SUNRISE questionnaire appears reliable in assessing preschooler sleep characteristics and related family practices, particularly in disadvantaged settings. It could be used in global surveillance of nighttime sleep duration and in studies examining associations of sleep timing and duration with health indicators in preschoolers.