Evaluation of a Portable Armband Device to Assess Resting Energy Expenditure in Patients With Anorexia Nervosa

Abstract

Background: In women with anorexia nervosa (AN), resting energy expenditure (REE) is decreased due to reduced energy intake and severe underweight. The assessment of REE allows estimating individual metabolic downregulation and better understanding body weight regulatory mechanisms in severely underweight patients with AN. However, REE predictive equations are known to have considerable shortcomings in patients with AN. Our aim was to evaluate a portable armband device (SenseWear armband [SWA]; BodyMedia, Inc, Pittsburgh, PA) for the assessment of REE against the measurement with indirect calorimetry (IC) as the reference method. Methods: We assessed REE simultaneously by IC and SWA in 50 women with AN at the start of inpatient therapy and calculated REE using 2 predictive equations. Results: Reliable data for IC measurement were obtained for 34 patients (age: 27.0 ± 8.0 years; body mass index: 14.4 ± 2.0 kg/m²). REE assessed with SWA was overestimated by 23% ± 27% compared with REE measured by IC (1166 ± 174 vs 979 ± 198 kcal/d, P < .001). REE estimation with SWA gave an accurate prediction within 10% deviation of REE measured with IC in 35% of the patients. In contrast, REE calculated with 2 predictive equations underestimated REE measured with IC by −26% ± 17% and −5% ± 20%, respectively. Conclusions: A mean difference of 187 kcal/d between both techniques for the assessment of REE may be of methodological relevance. Therefore, SWA and IC are not interchangeable methods for the assessment of REE in underweight females with AN.