Comparison of Activity Monitors to Count Steps and Estimate Energy Expenditure During Treadmill Ambulation

Abstract

PURPOSE: To compare and validate the steps counts, activity energy expenditure (EE), and total EE outputs from Actigraph GT1M (AG) and Kenz Lifecorder EX (KL) activity monitors during treadmill ambulation. In addition, the KL was placed bilaterally on the right (KL-R) and left (KL-L) sides of the anterior aspect of the waistline to evaluate potential side differences in output. METHODS: Ten male and 10 female participants performed 10 min walking bouts on a treadmill at speeds of 54, 80, and 107 m/min and running bouts at speeds of 134, 161, and 188 m/min. During each bout, step counts were hand tallied by two observers and indirect calorimetry (IC) was used to validate the activity monitors' estimates of activity and total EE. Resting EE was measured via IC. Actigraph total EE was calculated using the Freedson et al. (1998) equation. Paired sample t-tests and ANOVA were used to analyze the data. RESULTS: There were no significant differences between KL-R and KL-L monitors in step counts, total EE and activity EE at any speed. At the slowest walking speed, the AG counted 58±10% of the observed steps, whereas the KL-R counted 92±5% and the KL-L counted 93±8% of the observed steps. At all other walking and running speeds, all activity monitors undercounted compared to observed steps, by 3% or less. Compared to the AG, the KL-R and KL-L counted significantly greater steps at 54 m/min (p<0.001), but counted significantly less steps at all remaining speeds (p</=0.047). At walking speeds the KL-R and KL-L significantly overestimated total EE (p<0.013) and underestimated total EE while running at the fastest speed (p</=0.013). The AG underestimated total EE at running speeds of 161 and 188 m/min. Compared to the AG, the KL-R and KL-L overestimated total EE at the slowest walking speed and at all running speeds (p</=0.021). The KL-R and KL-L underestimated activity EE at running speeds of 161 and 188 m/min. CONCLUSION: Side placement had no effect on KL output. The KL provided the most accurate step counts at the slowest walking speed; both KL and AG monitors yielded acceptable step counts at the remaining walking and running speeds. The AG yielded the best estimate of total EE at walking speeds. The KL yielded accurate activity EE estimates at walking and moderate running speeds.