Interval setting selection affects ambulatory activity outputs in children with cerebral palsy

Accelerometer based devices have been widely used to assess the ambulatory activity of children with and without functional disabilities. Many researchers who utilize the StepWatch Activity Monitor (SAM) collect at a 60second (60sec) interval setting. The purpose of this study was to assess the effect of SAM interval settings on ambulatory activity outputs in children with cerebral palsy (CP) and typically developing youth.

Participants wore a SAM which recorded the number of strides every 10seconds (10s) for one week. Raw 10s data was downsampled to combine strides into 60sec intervals. Strides were ensembled into walking bouts with the Intensity/Duration calculated as a percentage of Total Ambulatory Time (TAT).

Twenty-eight children with CP (14 boys; avg. 12 yrs. 4 mths.; GMFCS Level I n=4, Level II n=19, Level III n=5) completed testing and 28 age matched typically developing youth (14 boys; avg. 12 yrs. 6 mths.) were included. Using the 10sec interval, ∼80% of walking bouts in both groups were less than or equal to 60s. Data recorded at 60sec intervals had higher daily TAT but fewer walking bouts. In children with CP, daily steps were higher using the 60sec interval. At the Easy intensity, the 60sec interval reported an increased volume of Long duration walking, and it rarely identified any Moderate+ intensity activity.

60sec interval data overestimated low intensity and long duration ambulatory activity. It is imperative that investigators choose a finer interval setting (10sec) to maximize the detection of gait transitions and rest periods which are critical in describing community ambulation of patients with cerebral palsy.