F19o Free Communication/Poster Evaluation of Physical Activity Monitors
For prevention and treatment of obesity, it is important to assess not only volitional activity (active exercise) but also involuntary activity (non-structured activity). Recently, we tested the performance of an activity monitor based on two-axial accelerometry sensor (Lifecorder, Suzuken Co.) during running and walking by comparison with the value of energy expenditure measured by the Douglas bag method. The purpose of this study was to further investigate the accuracy of the activity monitor assessment of energy expenditure over longer time periods in a respiratory chamber.
In thirteen healthy Japanese (36+/−13 y of age, 21.1+/−2.0 kg/m2) we measured 24-h energy expenditure (TEE- chamber) in a respiratory chamber. The subjects were prescribed two walking exercises on the treadmill and the sleeping period was controlled (for 8-h). The accelerometer was put on the belt. Linear regressions were calculated between TEE-chamber and estimated 24-h energy expenditure by accelerometer (TEE-acc), and between activity levels (0 to 9) assessed by acceleration and measured energy expenditure (EE-chamber) after these values were averaged each fifteen minutes.
TEE-acc was lower than TEE-chamber (−151+/−139kcal/d; p < 0.01, i.e.92+/−6% of the chamber value). However, there was a highly significant correlation between the two (r = 0.877; p < 0.01). Furthermore, there was a highly significant relationship between acceleration levels and EE-chamber including walking periods (r = 0.838; p < 0.01) and excluding obligatory walking (r = 0.565; p < 0.01).
The accelerometer understimated TEE. However, it could well assess EE not only during exercise but also during non-structured activity. These data suggest that the accelerometer method is useful for assessment of daily physical activity in real life, provided individual calibration factors are developed.