Reliability of accelerometer scores for measuring sedentary and physical activity behaviors in persons with multiple sclerosis

This brief research note examined the reliability of scores from an accelerometer as measures of sedentary and physical activity behaviors in persons with multiple sclerosis (MS). The analysis was performed on a combined data set from 2 previous longitudinal investigations of physical activity in MS. We focused on the number of days required to reliably estimate sedentary behavior, based on time spent in sedentary behavior per day and number of sedentary breaks, number of long sedentary bouts, and average length of sedentary bouts per day. We further examined the number of days required to reliably estimate physical activity behavior, based on time spent in light and moderate-to-vigorous physical activity and average length of activity bouts per day. Between 4-6 days of monitoring and 3-7 days of monitoring were necessary for good reliability of scores from all sedentary outcomes and physical activity outcomes, respectively. These results should guide research and practice examining sedentary and physical activity behaviors using accelerometry in persons with MS.