Data-driven derivation of a multidimensional sleep health framework in adolescents

Multidimensional sleep health captures the 24-hour experience of sleep as regularity, satisfaction, alertness, timing, efficiency, and duration. However, most evidence comes from adults. Adolescents have unique sleep needs that may not be fully represented. This study aimed to empirically identify and validate dimensions of adolescent sleep health and evaluate methods relevant to their assessment.

We analyzed subjective and objective sleep measures from 347 adolescents in the Penn State Child Cohort (16.3±2.2 years; 45.8% female; 22.2% race and ethnic minority). Exploratory factor analysis was conducted in a random 40% subsample to identify latent dimensions of self-reported, actigraphy, and polysomnography measures, followed by confirmatory factor analysis in the remaining 60% to test model validity.

Exploratory factor analysis identified 6 dimensions: Breathing, Efficiency, Duration, Regularity/variability, Satisfaction, and Timing, collectively referred to as the BEDReST framework, explaining 63% of the variance. Confirmatory factor analysis of individual factors within the 6-factor structure demonstrated good to excellent fit (Comparative Fit Index = 0.95-1.00; Tucker-Lewis Index = 0.86-1.00; Root Mean Square Error of Approximation = 0.06-0.15; Standardized Root Mean Square Residual < 0.08). All dimensions, except Satisfaction, included objective measures.

We empirically derived and validated 6 dimensions of adolescent sleep health, the BEDReST framework, using a multimethod approach. This approach highlighted the importance of including objective sleep data to comprehensively capture adolescent sleep health. The BEDReST framework extends existing models by incorporating sleep-disordered breathing and provides a foundation for refining the conceptualization and measurement of sleep health during adolescence.