Comparison of pedometer and accelerometer measures of free-living physical activity

Purpose: The purpose of this investigation was 1) to evaluate agreement between dual-mode CSA accelerometer outputs and Yamax pedometer outputs assessed concurrently under free-living conditions; 2) to determine the relationship between pedometer-steps per day and CSA-time spent in inactivity and in light-, moderate-, and vigorous-intensity activities; and 3) to identify a value of pedometer-steps per day that corresponds with a minimum of 30 CSA-min·d^-1 of moderate ambulatory activity. Methods: Data were analyzed from 52 participants (27 men, 25 women; mean age = 38.2 ± 12.0 yr; mean BMI = 26.4 ± 4.5 kg·m^-2) who were enrolled in the International Physical Activity Questionnaire study and wore both motion sensors during waking hours for 7 consecutive days. Results: Participants averaged 415.0 ± 159.5 CSA-counts·min^-1·d^-1, 357,601 ± 138,425 CSA-counts·d^-1, 11,483 ± 3,856 CSA-steps·d^-1, and 9,638 ± 4,030 pedometer-steps·d^-1. There was a strong relationship between all CSA outputs and pedometer outputs (r = 0.74-0.86). The mean difference in steps detected between instruments was 1845 ± 2116 steps·d^-1 (CSA > pedometer; t = 6.29, P < 0.0001). There were distinct differences (effect sizes > 0.80) in mean CSA-time (min·d^-1) in moderate and vigorous activity with increasing pedometer-determined activity quartiles; no differences were noted for inactivity or light activity. Approximately 33 CSA-min·d^-1 of moderate activity corresponded with 8000 pedometer-steps·d^-1. Conclusion: Differences in mean steps per day detected may be due to differences in set instrument sensitivity thresholds and/or attachment. Additional studies with different populations are needed to confirm a recommended number of steps per day associated with the duration and intensity of public health recommendations for ambulatory activity.