The relationship between sleep quality and daytime sleepiness with chronotype latent constructs: An exploratory and confirmatory factor analysis in Chinese college students

Abstract

Introduction: The Composite Scale of Morningness (CSM) has been suggested to consist of two to three latent constructs. Yet, the role of each latent construct in relating to other sleep behaviors was unclear. This study examined the factor structure of CSM by exploratory and confirmatory factor analyses from structural equation modeling. We aim to elucidate the relationship between each CSM latent construct with daytime sleepiness and sleep quality with a Chinese sample. Methods: Participants (n=661, 18-25, Chinese undergraduates, 32.1% male) completed an online survey including the validated Chinese version of Epworth Sleepiness Scale, Pittsburgh Sleep Quality Index(PSQI), and CSM, translated (with back translation) by our research team. Results: The CSM fits into a 3-factor model (CFI=.982, RMSEA=.047, SRMR=.039) with much better model fit indexes than a 2- or 1-factor model. The derived latent constructs captured individuals’ voluntary sleep-wake schedule, self-reported morningness/eveningness chronotype and alertness after waking-up. The 10th and 90th percentile scores are 23 and 40 indicating eveningness and morningness chronotypes, respectively. Correlational analyses revealed that a late voluntary sleepwake schedule, self-reported eveningness-chronotype and low alertness after waking-up were significantly correlated (p<.05) with a higher level of daytime sleepiness, worse subjective sleep quality, shorter sleep, lower sleep efficiency and greater daytime dysfunction. However, a longer sleep latency was not related to self-reported chronotype but the other two CSM latent constructs. Sleep disturbances and use of sleep medication were only correlated with low alertness after waking-up. Conclusion: A 3-factor model of CSM is validated with good psychometric properties. While the cutoff score for eveningness-chronotype is comparable with the original paper on Caucasians, a lower cut-off for morningness-chronotype is found among Chinese undergraduates. Regarding the factor structure, these 3 factors differ in their relationships with sleep quality dimensions. A 3-factor scoring of CSM is suggested for better delineation of the different aspects of circadian rhythm for future studies.