Free-living activity counts-derived breaks in sedentary time: Are they real transitions from sitting to standing?

Abstract

Background: Previous research has demonstrated a link between free-living accelerometer-measured breaks in sedentary time and health related variables. Breaks in sedentary time are typically inferred from time-stamped accelerometer data indicating a transition from lack of movement (recording of <100 activity counts/min) to relatively more movement (≥100 activity counts/min). However, it remains unknown whether these breaks actually represent sit-to-stand postural transitions in free-living. Thus, the purpose of this study was to compare free-living accelerometer-derived and posture-derived estimates of breaks in sedentary time using the ActiGraph GT3X+ (AG) and the activPAL™ (AP), respectively.

Methods: A total of 15 participants concurrently wore an AG at their waist and an AP on their right thigh for 7 consecutive days (24h/day - removing them only when in contact with water). Data from both devices were matched on minute-by-minute timestamps while also applying a 3-min allowance window to account for clock drift. Dependent t-test was used to evaluate differences in mean breaks between AG and AP.

Results: The AG detected 74±4.1 breaks/day (mean±SEM) while the AP detected 39±3.1 breaks/day (P<0.001). On average, the AG detected 67% of the AP breaks while 65% of the AG breaks did not correspond with AP breaks. Of the non-corresponding AG breaks, 52% occurred when participants were sitting, 42% when standing, and 6% when transitioning from standing to sitting.

Conclusion: The AG detected a significantly higher number of breaks in sedentary time, the majority of which do not correspond to sit-to-stand transitions as measured by the AP.

Keywords: Accelerometry; Agreement; Posture; Sitting; Validity.