Comparison between actigraphy and polysomnography in a real world environment.
Walsh-L; Barger-LK; Flynn-Evans-EE; Lockley-SW
Sleep 2009 Jun; 32(Abstract Suppl):A367-A368
Introduction: The detrimental extent of sleep restriction in society is becoming increasingly prevalent. Actigraphy has been widely deployed to estimate sleep metrics in real-world situations where polysomnography (PSG) is not suitable. Its use for sleep assessment provides quantitative statistics, although PSG remains the gold standard for sleep measurement. The logistics of research in field settings may require the use of longer epochs to maximize the length of data collection in remote locations (e.g., during spaceflight). The aim of this analysis was to examine the accuracy of the sleep/wake algorithm provided by the Minimitter 'Actiware® - Sleep' scoring program with data collected using different epoch lengths. Methods: Four PGY-1 medical residents working extended duration shifts wore two Actiwatch-Ls for at least 4 days (Mini-Mitter, Bend, OR); one was set to collect data in 1-minute epochs and the other in 2-minute epochs. Ambulatory PSG was also measured simultaneously (Vitaport, TEMEC Instruments, The Netherlands). Results: PSG sleep duration was consistently underestimated by 4.16%+/-6.07% SD in actigraphy data collected using 1-minute epochs whereas the actigraphy data collected using 2-minute epochs overestimated sleep duration by 6.05%+/-8.09% SD. Actigraphy-estimated sleep durations were consistently longer using data collected with 2-minute epochs compared with those data collected with 1-minute epochs (10.21%+/-8.33% SD). Conclusion: In this small pilot study, sleep duration measured by actigraphy was underestimated by four percent using one-minute epochs and overestimated by six percent as compared to PSG scored every 30 seconds. Results from this limited study population suggest that actigraphy is a reliable estimator of sleep duration in a field environment and two minute epochs can be used when logistics demand long continuous collection episodes.
Sleep-deprivation; Sleep-disorders; Circadian-rhythms; Psychomotor-function; Humans; Men; Women; Age-groups; Medical-personnel; Statistical-analysis; Analytical-processes
Brigham and Women's Hospital