Being able to track energy (caloric) expenditure is of great importance for those pursuing weight loss and weight maintenance goals. But the technology to track energy expenditure (EE) with a high degree of accuracy is still lacking. Research results are mixed on the ability of activity trackers to accurately compute TEE in controlled laboratory settings, during semi-structured activities, and in free-living environments. Although the correlation between activity tracking accelerometers is reported to be moderate to strong, significant underestimations of the reference values are common. Accelerometers, categorized as research-grade (ActiGraph GT3X+, BodyMedia Core, Body Media SenseWear), underestimate energy expenditure in comparison to the gold standard, indirect calorimetry (Bai et al. 2016; Ferguson et al. 2015; Imboden et al. 2017; Kim and Welk 2015).
Consumer-targeted devices worn during a variety of activities produce large differences and variable estimates of EE and tend to underestimate reference values of EE (Bai et al 2016; Ferguson et al. 2015; Imboden et al. 2017; Kim and Welk 2015; Price et al. 2017; Sasaki et al. 2015). Typically, the proprietary algorithms developed by the manufacturers account for differences between devices. Although some accelerometers perform better during moderate- to fast-paced activities (ActiGraph GT3X+, BodyMedia SenseWear, Core Armband), others perform better during slow-paced activities (activPAL). Lyden and associates (2017) reported that the activPAL accurately categorizes sedentary behaviors as well as light-intensity and MVPA exercise compared with direct observation. Triaxial and multisensory devices tend to provide more accurate estimates of TEE than uniaxial devices (Van Remoortel et al. 2012).