University of Turku and Turku University Hospital
Our office will remain closed through Friday, September 18th as we continue to assess the damages caused by Hurricane Sally. ActiGraph team members are working remotely, however shipping delays should be expected at this time. We expect to resume regular business hours on Monday, September 21st. If you need immediate assistance, please contact us by email at email@example.com and we will respond as quickly as possible. Thank you for your continued support.
Utility of ActiGraph wear sensor in assessing wear-time in 24-hour wrist measurement
- Published on Jun 21, 2017
There are no standard procedures for identifying wake wear, sleep wear and non-wear time from the 24-h accelerometer recordings. Participant diaries increase participant burden and frequently have missing values. The Choi algorithm identifies 90-min consecutive zero counts as non-wear time, allowing for 2 minutes of non-zero counts, thus possibly ignoring short non-wear bouts and over-estimating nonwear time during sleep. The ActiGraph wActiSleep-BT has a wear sensor which may offer solution for distinguishing between wear and non-wear time. The objective of this study was to compare wake and sleep wear time identified by the sensor to those identified by the Choi algorithm and by participant diaries. In the Finnish Retirement and Aging Study, 477 aging workers (mean age 62, SD 1.2, 18% male) were asked to wear wActiSleep-BT on their non-dominant wrist continuously for 7 days and 6 nights. Sleep and wake time were defined by participant diaries and divided into wear and non-wear time by three methods: participant diaries indicating start and end time of the measurement, Choi algorithm and sensor. The sensor identified significantly less wear time than the other two methods. Mean (SD) minutes of wake and sleep wear time were 804 (263) and 399 (137) by sensor, 903 (179) and 429 (99) by Choi algorithm and 922 (158) and 448 (89) by participant diaries. However, for the majority of participants the correlation of wake wear time identified by Choi algorithm and sensor was very high (Figure 1). The observed discrepancy between Choi algorithm and sensor was mostly due to underestimation of wear time by the sensor possibly owing to technical errors. In the conference we will present methods to identify those participants for whom sensor works appropriately. In addition, we will present a tool to calculate wear time for those for whom wear sensor does not work.
- Anna Pulakka 1
- Eric Shiroma 2
- Tamara Harris 2
- Sari Stenholm 1
National Institute on Aging
ICAMPAM 2017 Abstract Booklet