Validity of a novel activity monitor as an objective measure of walking in older adults

Grant, M; Dall, PM; Ryan, CG; Granat, MH

4th World Congress of the International Society of Physical Medicine and Rehabilitation

METHODS: Twenty older adults (10 male; 10 female; aged 71 ± 5 years) were recruited from participants of community-based exercise classes and their relatives. The small, lightweight (20g) activity monitor (activPAL) is based on a uni-axial accelerometer, and is worn on the anterior aspect of the thigh. Outcome measures include duration of sedentary, upright and walking periods, step number, and cadence. Wearing an activity monitor, participants walked on a treadmill for five minutes at five different speeds (0.67, 0.9, 1.12, 1.34 and 1.56 m/s) and walked outdoors over a 500m course at three self-selected speeds (slow, preferred and fast). All trials were videoed and step number and cadence were calculated from the recordings. Agreement between activPAL and observation (the criterion measure) was assessed using the Bland and Altman method. RESULTS: The mean difference was less than 1% between observation and the activPAL monitor for step number and cadence for all walking speeds on both surfaces (treadmill and outdoors). The limits of agreement were narrow (no more than ± 2%) for all conditions. CONCLUSIONS: The level of agreement between the activPAL actvitiy monitor and observation in this group of elderly people, was excellent for step count and for cadence over a range of speeds regardless of walking surface. In community-dwelling older adults the activPAL monitor is a valid and reliable measure of walking and its accuracy is unaffected by walking speed.

NA

Grant, M; Dall, PM; Ryan, CG; Granat, MH Validity of a novel activity monitor as an objective measure of walking in older adults 4th World Congress of the International Society of Physical Medicine and Rehabilitation 2007. NA.