Abstract
Purpose : Our study evaluated the intermonitor reliability of the SENS accelerometer for raw acceleration, counts, steps, physical activity intensity, and physical activity types/postures in a laboratory setting. Methods: A total of 61 apparently healthy adults (54% female, 21.6 ± 4.2 years) completed ∼10 activities of daily living in a single laboratory visit while wearing two SENS devices side-by-side at one or more of the following body locations: right anterior thigh, right hip, nondominant wrist, and dominant wrist. Outcomes (raw acceleration, counts, steps, and activity intensity for all locations, plus activity types/postures for thigh location) were compared between two devices placed at the same body location using correlations, kappa, percent agreement, and linear mixed models. Results: Strong agreement was found between side-by-side devices worn at the thigh and dominant wrist and, to a lesser extent, the nondominant wrist and hip for mean acceleration ( r = .80–.97), counts ( r = .94–1.00), total steps ( r = .88–.94 for thigh, r = .58–.92 for wrists, and r = .17–.39 for hip), and activity intensity (κ = .92–.95 for thigh, κ = .74–.75 for dominant wrist, κ = .47–.57 for nondominant wrist, and κ = .48–.58 for hip) and activity type/posture for the thigh ( r = .75–.94). Time spent in sedentary, light, and moderate- to vigorous-intensity physical activity was not significantly different between devices for any tested placement. Conclusion: The SENS device, especially when thigh-worn, has strong intermonitor reliability for several physical activity outcomes and appears suitably reliable for use in apparently healthy adults.

This publication has 23 references indexed in Scilit: