Validation of Actigraph accelerometer estimates of total energy expenditure in young children

Purpose. To assess the validity of two equations based on the Actigraph CSA/MTI accelerometer for prediction of total energy expenditure (TEE). Research methods and procedures. The criterion was TEE measured using the doubly labeled water method in 85 children, mean age 4.6 years (SD 1.1), over 7 days in the pre-schoolers and 10 days in the school-age participants. Children wore the Actigraph concurrently during waking hours, for 3 of 7 days (pre-schoolers) or 7 of 10 days (school-age children). We tested two prediction equations based on accelerometry. Agreement between predicted and measured TEE was assessed using the 'Bland Altman' approach. Results. Mean TEE measured by doubly-labeled water was 5.8 MJ/d (SD 1.6). Mean error for the Ekelund equation was +0.3 MJ/d (limits of agreement -3.7 to +4.3), and for that of Puyau et al. was -0.3 MJ/d (limits of agreement +3.2 to -3.8). Conclusions. Simple approaches using the Actigraph appear to be inadequate for the estimation of free-living TEE in young children at present.