Actigraphy data in pediatric research : the role of sleep diaries

Abstract

Background. When assessing children's sleep using actigraphy, researchers usually rely on a sleep diary completed by a parent as an aid in scoring actigraphic data. However, parental nonadherence in completing the sleep diary may significantly reduce the amount of available data. The current study examined the agreement between actigraphic data scored with and without a sleep diary to evaluate the impact of not using a sleep diary when studying children's sleep with actigraphy. Methods. Sixty children (aged 6–10 years; 36 girls) wore an actigraph for three to seven consecutive nights, and mothers were asked to complete a diary of their child's sleep during the same period. Actigraphy data were scored under two conditions (with and without diary) rated independently for each child by two different research assistants, who each scored 50% of the files in each condition. Results. Group-level analyses and intraclass correlations revealed very strong convergence between the two scoring conditions: on all sleep variables (sleep duration, wake duration, and sleep efficiency), average mean differences were very small and intraclass correlations very high. Bland and Altman’s (1999) approach allowed for a child-by-child examination of agreement between the two conditions and revealed that, although they cannot be considered interchangeable, the two conditions produce quite minimal differences in the estimation of sleep variables. Conclusions. The findings suggest that it is possible to use some actigraphy data for which no corresponding diary data are available, although this approach should be used sparingly.