You are here: Home / Publications / Calibration of the RT3 Accelerometer for Ambulation and Nonambulation in Children

Calibration of the RT3 Accelerometer for Ambulation and Nonambulation in Children

Chu, Eva Y. W.; McManus, Alison M.; & Yu, Clare C. W. (2007). Calibration of the RT3 Accelerometer for Ambulation and Nonambulation in Children. Medicine & Science in Sports & Exercise, 39(11), 2085.

Chu, Eva Y. W.; McManus, Alison M.; & Yu, Clare C. W. (2007). Calibration of the RT3 Accelerometer for Ambulation and Nonambulation in Children. Medicine & Science in Sports & Exercise, 39(11), 2085.

Octet Stream icon 1127.ris — Octet Stream, 2 kB (2242 bytes)

Purpose: The purpose of this study was to provide RT3 accelerometer thresholds to distinguish nonambulation from ambulation, as well as distinguish between low, moderate, and vigorous intensity ambulation in Chinese children. Methods: Thirty-five 8- to 12-yr-olds completed a laboratory-based experiment to calibrate the RT3 device across a range of intensities. Oxygen uptake during sedentary tasks and ambulation was aligned, with corresponding RT3 output, to one of four intensity categories (sedentary, < 2 METs; low, >= 2 but < 3 METs; moderate, >= 3 but < 6; and vigorous, >= 6). Threshold values were derived using receiver operator curves to distinguish the sedentary from low intensity, low from moderate intensity, and moderate from vigorous intensity. These were confirmed in an independent sample of 46 similarly aged children. Results: RT3 movement counts increased in a linear manner with scaled oxygen uptake from stationary to vigorous movement (r2 = 0.83). The ROC-derived thresholds showed a good ability to discriminate between nonambulatory and ambulatory tasks, as well as distinguish between low, moderate, and vigorous ambulation (sensitivity values of 87-100% and specificity values of 97-100%). ROC analyses in the independent sample showed sensitivity and specificity values ranging from 72 to 98%, indicating that the thresholds provided an accurate distinction between the four intensity categories. A Cohen's kappa of [kappa] = 0.92 confirmed that the lower threshold had near-perfect agreement with the independent sample, whereas the moderate-intensity and vigorous-intensity thresholds showed good agreement with the independent sample ([kappa] = 0.63 and 0.65, respectively). Conclusion: The RT3 accelerometer provides an acceptable marker of both sedentary and ambulatory behavior in Chinese children.




JOUR



Chu, Eva Y. W.
McManus, Alison M.
Yu, Clare C. W.



2007


Medicine & Science in Sports & Exercise

39

11

2085






0195-9131

10.1249/mss.0b013e318148436c



1127