C-26 Free Communication/Poster - Objective Measures: Accelerometry and Pedometry: JUNE 3, 2010 7:30 AM - 12:30 PM: ROOM: Hall C
Better assessment of non-active and activity-related energy expenditure in older adults would likely provide important information about the effects of physical activity on their health. The SenseWear Pro armband (BodyMedia Inc., Pittsburgh, PA) is a novel multisensor device used to measure energy expenditure during free-living activities. It is considerably less expensive than doubly labeled water, more accurate than simple accelerometers, easy to use in free-living environments and comfortable to wear. However, validation is needed in older populations.
PURPOSE: To examine the validity of the SenseWear Pro armband to assess energy expenditure compared to the gold standard of measured oxygen consumption in free-living older adults at different walking speeds.
METHODS: Ten older adults (mean age ± SD: 70.4 ± 6.4 yrs) participating in an exercise (>150 min/wk of treadmill and over ground walking) and weight loss pilot study were recruited. Participants wore the SenseWear Pro armband and a validated portable oxygen consumption device (COSMED K4b2) concurrently during 2 walking bouts done over ground on a flat surface: a usual pace (mean ± SD gait speed: 1.1 ± 0.1 m/s) 40m walk and a fast pace (1.4 ± 0.2 m/s) 400m walk, both at a self-selected speed. Gold standard MET values were obtained from breath-by-breath sampling using the COSMED. The SenseWear Pro armband recorded data in 1 minute epochs and METs were derived from a proprietary algorithm. Comparisons between the mean METs obtained from both devices were made for each walking bout.
RESULTS: METs from SenseWear were highly correlated with METs from COSMED, but only for the fast paced walk (r = 0.88, p<0.001). SenseWear slightly over estimated METs for the fast walk by 6% (4.31 vs 4.05), while SenseWear overestimated METs for the usual paced walk by nearly 17% (2.91 vs 2.49).
CONCLUSIONS: The SenseWear Pro armband overestimated energy expenditure during both walking bouts. The over-estimation was 3-fold greater during the usual paced compared to the fast paced walk. These data suggest that a better estimate of 1 MET (resting metabolic rate) is needed for older adult populations to obtain accurate estimates for exercise energy expenditure.
Source of Funding: Pittsburgh Claude D. Pepper Center NIH P30 AG 024826