[Skip to Content]
[Skip to Content Landing]
Comment & Response
August 2013

Sleep Evaluation by Actigraphy for Patients With Alzheimer Disease—Reply

Author Affiliations
  • 1Department of Neurology, Washington University School of Medicine, St Louis, Missouri
  • 2The Charles F. and Joanne Knight Alzheimer’s Disease Research Center, Washington University School of Medicine, St Louis, Missouri
  • 3Hope Center for Neurological Disorders, Washington University School of Medicine, St Louis, Missouri
JAMA Neurol. 2013;70(8):1074-1075. doi:10.1001/jamaneurol.2013.3490

In Reply In our study of sleep in preclinical Alzheimer disease,1 we used actigraphy to objectively measure sleep in a middle-aged, healthy cohort. While polysomnography remains the gold standard for measuring sleep, it could not have provided the information required for our study, in which we sought data about sleep in the participants’ regular day-to-day lives. Attended polysomnography occurs in an unfamiliar environment and disrupts daily routines, especially when multiple nights’ recordings are performed. Because actigraphy is noninvasive and can be measured as individuals go about their usual activities, it was an appropriate method of sleep measurement for the research question. Actigraphy has been compared against polysomnography in many studies.2 Based on several level I studies validating actigraphy, the American Academy of Sleep Medicine’s evidence-based practice parameters for actigraphy recommend it at the guideline level as a “valid way to assist in determining sleep patterns” for normal healthy adults.3 Based on review of the existing literature, we elected a watch model that has high reliability for measuring sleep in healthy adults in a home environment.4 In general, a lower threshold for wakefulness provides improved accuracy; therefore, we chose a low activity count threshold of 20 for the best combination of sensitivity and specificity.5 Given the high correlation between sleep measured with polysomnography and actigraphy, even if the absolute measures are not identical, actigraphy would be able to discern relative differences between groups as would be measured by polysomnography.