Actigraphy in Children with Autism Spectrum Disorders: Strategies for Success

Background:  The literature has been highly informative regarding actigraphy and its validity in pediatric research. However, minimal literature exists on how to perform actigraphy, especially in special populations including children with autism spectrum disorders.

Objectives:  We determined whether providing comprehensive actigraphy training to parents increased the nights of actigraphy data that could be scored in a clinical trial. The training program covers the proper use of actigraphs in the home and recording on the sleep diary.

Methods:  We compared two studies in children with autism spectrum disorders, both containing a behavioral sleep intervention in which actigraphy was the major outcome variable. The single site trial (Study 1) consisted of 20 children ages 4-10 years (80% male) whose parents received sleep education in a group setting at Vanderbilt University. The multisite trial (Study 2) consisted of 80 children ages 3-10 years (80% male) whose parents received either group or individual education at Vanderbilt University, University of Colorado Denver, or University of Toronto.  Study 1 provided a basic level of actigraphy training while Study 2 provided more comprehensive actigraphy training using a variety of educational tools. Educational tools developed and implemented in Study 2 included presentation of actigraphy data to the parents, use of a quiz testing the parents' knowledge of actigraphy procedures, and a practice session in which the parents recorded data on sleep diaries and demonstrated proper use of an actigraphy watch, including the event marker. Feedback was given to the parents on their use of the watch.

One week of actigraphy was performed at baseline and post-intervention in both trials, and scored by a single analyst. The proportion of scorable nights of actigraphy was compared between Study 1 and Study 2 using Mann-Whitney U test.

Results:  Scorable nights of actigraphy were higher in Study 2 than Study 1 at baseline (P = 0.04) and post-intervention (p = 0.002). The mean proportion of scorable nights (standard deviation) in the baseline and post-intervention phases  was 0.91 (0.20) and 0.90 (0.17) for Study 2 compared to 0.73 (0.35) and 0.58 (0.39) for Study 1. One-way analysis of variance showed no difference in scorable nights between the three research sites in Study 2 [F= 0.93; p = 0.4].

Conclusions: Comprehensive training resulted in an increase in scorable nights of actigraphy.  Our results support the use of educational tools to enhance actigraphy data collection in research studies involving children with autism spectrum disorders.