Comparison of activity monitors accuracy to estimate energy expenditure of daily living activities

Accurate assessment of the total amount of physical activity (PA) or the energy expenditure (EE) in free living condition is a big challenge. A priority has been placed on the study of valid and reliable measures for all levels of PA. New technologies are available for objectively measuring PA and inferring EE: the Actiheart -AH-(Cambridge Neurotechnology, UK) and the SenseWear Pro2 ArmbandTM -SWA-(BodyMedia, USA) are an example of these new devices. PURPOSE The purpose of the present study was to examine the validity of these two new devices during light (<3 METS) to moderate (3-6 METS) and vigorous (>6 METS) intensity daily living activities in laboratory and feld settings. METHODS: 8 physically active women volunteered to participate in the study and performed in two separate days different activities: resting, occupation, housework, conditioning and recreation. Oxygen consumption was measured continuously throughout the routine by indirect calorimetry -IC- (K4b2, Cosmed, Italy) and participants wore the Actiheart on the chest and the Armband on the right arm for simultaneous estimation of EE (METS). RESULTS The combined AH model (Branched model: activity counts+HR) had the strongest relationship with PAEE (r=0785, p<0001) compared with those from the single-measure models (r=.748 and .719, p<0001 for the activity model and the HR model respectively), so that we only used this equation model to compare data with SWA and IC. The two analyzed methods were highly correlated with IC (AH r=.785 and SWA r=.795, p<0001) even if PAEE expressed in METS resulted significantly different for both devices (p<0.001 for AH and p<0.001 for SWA). The positively correlated errors with measured PAEE in the plots of both AH and SWA models indicate a systematic error in these equations (mean difference between methods= AH 1.52±2.38, SWA -0.11±2.48; 95% limits of agreement= AH -3.24 to 6.28, SWA -5.07 to 4.85). On the contrary the explained variances from the AH were lower if light intensity exercises are considered. CONCLUSION The AH provides reliable estimates of EE for light intensity activities; on the contrary it underestimates moderate and vigorous activities. The SWA has a lower systematic error in moderate and vigorous exercises. Both devices can provide a valid measure of the time spent in various intensity categories.