Lack of validity of free-living physical activity energy expenditure estimated from ActiGraph equations in severely obese women: A doubly labelled water study

The current study aimed to investigate the validity of three ActiGraph predictive equations that are available to estimate free-living physical activity energy expenditure (PAEE) in women with severe obesity. The study included 20 women with class III obesity (age: 22–38 years). During 14 days of free-living conditions, total energy expenditure was measured using the doubly labelled water method; in addition, participants wore a triaxial accelerometer (model GT3X+) on the hip. The resting metabolic rate was measured by indirect calorimetry. At group level, the Freedson VM3 Combination was found to be more precise (bias = −61 kcal/day) than the Williams Work-Energy (bias = −283 kcal/day) and the Freedson Combination equations (bias = −186 kcal/day) for estimating PAEE. However, the three predictive equations had a wider limit of agreement (Williams Work-Energy [258, −824 kcal/day], Freedson Combination equations [324, −697 kcal/day] and Freedson VM3 Combination [424, −546 kcal/day]), indicating great uncertainty of the estimate. In conclusion, a wide variation was observed in the performance of different ActiGraph equations in estimating free-living PAEE among women with class III obesity. Therefore, our data do not support the use of these equations, and more studies are needed to improve predictive performance in free-living conditions.