Comparison of Step Count and Activity Assessment with the activPAL (PAL Technologies), Dynapaort (McRobert’s) and Step Activity Monitor (Orthocare Innovations)

January 2011

Rosenbaum, D; Winter, C; Muller, C

2nd ICAMPAM

Introduction: In our lab, we usually assess activity patterns with the Dynaport Activity Monitor (‘DAM’ McRoberts) and step numbers with the Step Activity Monitor (‘SAM’, Orthocare Innovations) [1-3]. The activPAL (‘PAL’, PAL Technologies) so far has only been used sporadically so that we were interested to find out whether its combined information content would make it a good compromise between the other two devices. Therefore, we applied all three devices in a group of normal subjects during the working hours to record their number of steps and activity patterns. The aim of the project was to assess the differences between the results obtained with three devices in order to determine their applicability and potential source of error. Methods: Currently, 7 healthy subjects (4 male, 3 female; 41±13 yrs., 176±7 cm, 68+12 kg, BMI 22±3) were included and agreed to participate (aiming to include >20). Subjects were asked to wear the SAM, DAM and PAL according to manufacturers’ instructions for one day during the working hours. The SAM was worn around the right ankle where it recorded the number of gait cycles in one-minute intervals. The DAM was worn around the waist and a thigh sensor was worn on the left leg; it was used to determine the time spent in lying, sitting, standing posture and locomotor activities. The PAL was worn on the front of the right thigh and reported the number of steps and time spent lying or sitting, standing and stepping. Results: Subjects wore the devices for an average of 7±1 hours. The number of daily gait cycles and percent activities reported by the three devices is given in Tab. 1. The PAL counted only 1507±931 gait cycles and underestimated the SAM numbers by 1-39%. The distribution of activities recorded with the DAM and the PAL compared well and stayed below 5% difference. Parameter SAM DAM PAL Rel. Difference No. of gait cycles 2570±1120 n. a. 1957±932 16% Percent lying/sitting n. a. 71.7±16.3 70.4±16.6% 2% Percent standing n. a. 19.0±11.7 19.9±12.1% 5% Percent stepping n. a. 9.3±5.1 9.7±5.0% 4% DISCUSSION AND CONCLUSION The PAL recorded a lower number of gait cycles than the SAM which was previously reported to be highly reliable and accurate (e.g. McDonald, Karabulut). This effect might be due to a threshold in the PAL software that does not count step activities shorter than 20 continuous steps. The percentage of activities was highly comparable between DAM and PAL with the PAL reaching comparable values. Therefore, the PAL appears to be a good solution for step and activity pattern assessments with only one device if the additional level of information provided by either the SAM or the DAM is not deemed necessary.

http://www.umass.edu/sphhs/documents/ConferencebookICAMPAM2011.pdf

Rosenbaum, D; Winter, C; Muller, C Comparison of Step Count and Activity Assessment with the activPAL (PAL Technologies), Dynapaort (McRobert’s) and Step Activity Monitor (Orthocare Innovations) 2nd ICAMPAM 2011. http://www.umass.edu/sphhs/documents/ConferencebookICAMPAM2011.pdf.