Validating Chronotype Questionnaires in Adolescents: Correlations with Actigraphy and the Dim Light Melatonin Onset

Abstract

Chronotype and diurnal preferences have consistently been shown to influence academic performance and many health-related issues and behaviors. A fundamental understanding of the impact of circadian timing is crucial for acting adequately on these findings. Large epidemiologic studies are required to obtain the necessary information. In order to conduct these studies, valid and reliable instruments are needed. Unfortunately, there are few known validation studies against objective measures in the adolescent age group even for widely used and well-known questionnaires such as the rMEQ-CA and the CSM. Our study was the first validation study of the MESSi in adolescents against an objective instrument. A total of 55 healthy 13- to 16-year-olds completed the MESSi, rMEQ-CA and CSM and provided information about their sleep-wake rhythm through a 7-day actigraphy monitoring and a sleep diary. Participants also completed the pediatric daytime sleepiness scale (PDSS). We examined correlations between sleep-wake and activity parameters and the questionnaires and analyzed the influence of chronotype classification on sleep-wake parameters, age and sex using uni- and multivariate analyses. We measured the evening rise in melatonin concentration in 24 adolescents and examined correlations with the questionnaires. The questionnaires had good internal consistency and convergent validity. Spearman correlations revealed earlier sleep onset and offset times and midpoints of sleep in more morning-oriented participants and later respective timings in participants with a stronger evening orientation. Due to technical problems, we were underpowered and could not examine correlations of the DLMO with the questionnaires. The cosinor activity parameter Acrophase showed stronger correlations with questionnaire scores and sleep onset and offset times than comparative non-parametric parameters. Chronotype classification differed significantly between questionnaires. Age and sex had no significant influence on questionnaire scores.