Using accelerometer to estimate energy expenditures with four equations in four training sessions

This study compares estimated energy expenditure (EE) from four equations using accelerometer countsin Zumba, interval 4x4 spinning, interval 4x4 running and pyramid running. The study also characterizesdifferences in EE and accelerometer counts during activity and recovery periods for these activities.Twenty six men and women (21.8±2.4 years) completed four training sessions. Vector magnitude countsalong three axes were measured using ActiGraph GT3X accelerometers. EE was estimated using fourequations. Results show that EE varied by 34.2%, 19.7%, 18.0% and 20.0% depending on which equationwas used in Zumba, 4x4 running, 4x4 spinning and pyramid running, respectively (p<0.001). Comparedwith 4x4 running, Zumba had 22.0% lower EE and 4x4 spinning had 47.8% lower EE in kcal/min(p<0.0001). There was no significant difference in EE between 4x4 running and pyramid running. Themean VM/min (vector magnitude counts per min) for Zumba was 22.1% and 20.4% lower than for 4x4running and pyramid running, respectively (p<0.0001). An 85.3% higher VM/min was found in 4x4running compared to 4x4 spinning (p<0.0001). The various equations caused substantial differences in theestimation of EE, particularly in Zumba, which is explained. Interval running provided the highest EE andcounts/min. When 4x4 spinning was carried out both in sitting and standing positions, the underestimationin EE from accelerometer was about 50% compared with 4x4 running