Effects of body mass index and tilt angle on output of two wearable activity monitors.

Abstract

Accelerometer-based activity monitors have been used to provide objective measures of physical activity and energy expenditure (EE) in free-living individuals. However, output from these devices has not been compared among normal, overweight, and obese individuals. The purpose of this study was to examine the effects of body mass index (BMI) and device tilt angle on activity counts recorded by wearable monitors in a controlled laboratory setting. A secondary aim was to examine the effects of these variables on estimated EE. Seventy-one healthy adults wore an Actical and an ActiGraph GT1M on the right and left hip, respectively, while walking at 40, 67, and 94 m·min. EE was measured by indirect calorimetry and compared with estimated values using published equations. Three-way repeated-measures ANOVA were used to examine differences in outcome variables (activity counts and EE) between speeds, BMI, and tilt angle for each device. No significant differences in activity counts were observed among BMI categories for either the Actical or ActiGraph (P>0.05). For the Actical, however, among those with an absolute tilt angle <10°, the obese group recorded higher activity counts than the normal weight group (P=0.01). Using the Heil two-regression model, the Actical overestimated EE by up to 35% at the intermediate speed and up to 12% at the fastest speed (P<0.001). The Freedson METs regression equation yielded closer estimates of EE than the Freedson kilocalorie regression equation. Our findings indicate that the Actical has limitations when comparing individuals with varying BMI and tilt angles in a controlled laboratory environment. The ActiGraph seems to be a more suitable device for making these comparisons.