Research Study Abstract

Contemporaneous Assessment Of Physical Activity, Sedentary Behavior And Sleep Using An Actigraph GT3X+ Accelerometer

  • Presented on June 18, 2013

Introduction Physical activity (PA), sedentary behavior (SB) and sleep form the behavioral tripartite of energy expenditure and appear inter- and independently related to cardiometabolic disease and cancer. The ability to collect valid data on all three behaviors contemporaneously using one type of accelerometer is important because it may improve our understanding of behavioral inter-relationships while lowering research costs and participant burden.

Purpose (i) Examine the feasibility of using an ActiGraph GT3X+ accelerometer to measure PA, SB and sleep during 7 days and 7 nights of continuous monitoring; (ii) measure the day-to-day variability in PA, SB, and sleep habits, and (iii) examine associations between PA, SB and sleep.

Methods Twenty three women (mean age = 54.7 ±15.5 yrs; BMI = 31.8 ±3.4) wore an ActiGraph GT3X+ accelerometer on their hip for 7 consecutive days during waking hours, and the same device on their non-dominant wrist 24 hr/day for 7 days. PA and SB data were scored using validated cut-points (SB = <100 counts per min [CPM]; moderate intensity PA = 1952-5724 CPM; vigorous PA = >5724 CPM. Non wear time was defined as 90 consecutive minutes of zero count data. Sleep data were scored using the Cole-Kripke algorithm in ActiLife v6.5.1. Self-reports of bedtime and out-of-bed time were used to establish the sleep analysis window for each actigraphy file.

Results Compliance to wearing the hip-mounted device for the full monitoring period was 83% (mean wear time = 5.8±1.3 days, 13.4±1.8 hr/day). Compliance to wearing the wrist-device was 102% (7.13 ±0.87 nights). Across all participants, 133 days of matched PA, SB and sleep data were collected. The within-person daily stability (intraclass correlation, R) of SB and moderate-to-vigorous PA was 0.78 and 0.86, respectively. The nightly stability of sleep parameters were R = 0.93 (Total Sleep Time), 0.89 (Sleep Efficiency), 0.13 (Sleep Latency), and 0.84 (Wakefulness After Sleep Onset). Daily PA and SB were not significantly associated with subsequent nightly sleep parameters (range, r = -0.2 to 0.1), although associations were stronger (range , r = -0.45 to 0.33) when weekly averages were used.

Conclusion It is feasible to collect PA, SB and sleep data contemporaneously using an ActiGraph GT3X+ accelerometer. A 5-7 day monitoring period was sufficient to yield stable estimates of all measured variables, except sleep latency. Future research should attempt to examine associations between PA, SB and sleep in a larger sample and validate the use of the GT3X+ for sleep assessment using polysomnography. This study was supported by funding from the NIH/National Cancer Institute (TREC 1U54 CA155435-01)


  • Simon Marshall
  • Jacqueline Kerr
  • Suneeta Godbole
  • Jacqueline Chen
  • Katherine Ellis
  • David Wing


  • University of California, San Diego, La Jolla, CA

Presented at



, , , , ,