Comparison of Motionlogger Watch and Actiwatch actigraphs to polysomnography for sleep/wake estimation in healthy young adults

Sleep/wake identification and sleep parameter estimates from Motionlogger Watch and Actiwatch-64 actigraphs were compared to polysomnography (PSG). Following one night of baseline sleep, 29 volunteers remained awake for 36 h, followed by 11 h of recovery sleep in the laboratory. Two sets of analyses were performed: (1) epoch-by-epoch agreement and discriminability index (d') calculations, and (2) sleep parameter concordance with repeated measures ANOVAs. Sensitivity (sleep identification), specificity (wake detection), and overall agreement with PSG, as well as d', were higher for the Motionlogger than for Actiwatch. Relative to PSG, the Actiwatch-estimated total sleep time and sleep efficiency were underestimated and the number of awakenings was overestimated for baseline and recovery; sleep latency was underestimated on the baseline night. On the other hand, the Motionlogger-estimated total sleep time and sleep efficiency estimates were underestimated, and the sleep latency was overestimated on recovery, versus PSG. Despite these misestimations, it was concluded that the Motionlogger provided nominally better agreement with PSG, and that actigraphy generally constitutes a reasonably reliable tool for producing objective measurements of sleep/wake, but that users should remain mindful of its limitations.