Total Energy Expenditure: Combined Heart Rate and Motion Sensor vs Uni-axial Accelerometry Models

Objective To compare total daily energy expenditure (TDEE) from accelerometry models and combined heart rate with motion sensor.

Methods Twenty male subjects were involved in this study (25.5±5.3 years; 73.2±9.9 kg). TDEE was measured using a combined heart rate monitor with a motion sensor (Actiheart), and the corresponding equations for the Actigraph uni-axial accelerometer. Paired-Samples T-Tests, regression models, and agreement between methods were used.

Results At a group level analysis, there were significant differences between TDEE from actiheart and all accelerometry models (actiheart and all accelerometry models did not differ from the line of identity, and the Pearson coefficient of correlation among these methods ranged from 0.72 to 0.82. Individual results demonstrated large limits of agreement, particularly in the Freedson and Crouter regressions presenting values below actiheart of 1752 kcal/d and 1331 kcal/d, respectively. Hendelman (p=0.075) and Swartz (p=0.084) models were the only to present no association between the difference and the mean of the methods (TDEE from accelerometry models-TDEE from actiheart).

Discussion These results indicate that at an individual level all the four models present a poor accuracy to predict energy expenditure comparing to a motion sensor combined with HR. It is noteworthy that the TDEE provided by these models were found to be highly correlated with the TDEE determined by the actiheart. These results highlight the importance of additional recording HR in estimating individual TDEE in free-living conditions, where subjects may perform activities that uni-axial accelerometry alone might not detect.