Activity counts from accelerometers do not add value to energy expenditure predictions in sedentary overweight individuals during weight loss interventions

Background: Knowing the total energy expenditure (TEE) of overweight adults is important for prescribing weight loss interventions. However, objective measurements of TEE may not always be readily available and can be expensive. This study aimed to investigate the validity of RT3 accelerometers in predicting the TEE of sedentary overweight adults, and to identify any sensitivity to anthropometric changes. Methods: The analysis used data from a 12-week weight loss study. At baseline and 12-week, TEE was predicted using RT3 accelerometers during whole room calorimeter stays. Bias between 2 methods was compared at and between the baseline and 12-week measurement points. Multiple regression analyses of TEE data were conducted. Results: Predicted and measured values for TEE were not different at baseline (P = .677) but were significantly different after weight loss (P = .007). However, the mean bias between methods was small (/d) and was not significantly different between 2 time-points. RT3 activity counts explained an additional 2% of the variation in TEE at 12-week but not at baseline. Conclusion: RT3 accelerometers are not sensitive to body composition changes and do not explain variation in TEE of overweight and obese individuals in a sedentary environment. Keywords: energy expenditure, calorimetry, accelerometry