Medicine & Science in Sports & Exercise:
E-32 Free Communication/Poster - Objective Physical Activity Assessment: JUNE 3, 2011 7:30 AM - 12:30 PM: ROOM: Hall B
Horton, Magdalene; Kennard, Larry; Bassett, David R. Jr. FACSM; Crouter, Scott E. FACSM
1University of Massachusetts Boston, Boston, MA. 2University of Tennessee Knoxville, Knoxville, TN.
(No relationships reported)
PURPOSE: The purpose of this study was to examine the validity of 8 ActiGraph (AG) prediction equations for estimating EE in children.
METHODS: Fifty-three boys and 50 girls (mean ± sd; age, 11±1.7 yrs; BMI 21.4±5.5 kg·m-2) had their resting metabolic rate (RMR) measured and performed various activities ranging from sedentary behaviors to vigorous physical activity (PA). A total of 18 activities were measured and each participant performed six activities. Activity data was collected using an ActiGraph GT3X accelerometer, positioned on the right hip, and simultaneously EE was measured using a portable indirect calorimeter. The various equations predict EE in different units (i.e., METs, kcal·min-1, kJ·min-1, and net and gross kcal·kg-1·min-1), thus for the analysis all prediction equations were converted to METs, which were compared to measured METs (measured VO2/measured RMR).
RESULTS: Overall, for all activities and ages combined, all prediction equations significantly overestimated the measured METs (P<0.05). However, there were significant interactions by age and BMI (P<0.05). In general, the Trost equation worked best for 8-9 yr olds, while the Treuth, Puyau, and Mattocks equations worked best for those 10-11 yr olds, and the Freedson, and Puyau equations work best for those 12 and older. Regarding BMI, Trost equation worked best for normal weight children, all equations, except for the Schmitz and Pate equations worked well in overweight children, and the Freedson, Treuth, Mattocks, and Puyau equations were best for obese children. The Mattocks and Puyau equations were significantly different from the criterion for 5 and 6 activities, respectively. The Puyau equation tended to overestimate light intensity activities (<2.2 METs) (P<0.05), while the Mattocks equation significantly overestimated the EE of walking and running (P<0.05). All other equations significantly over- or under-estimated between 8 and 18 activities (P<0.05), with most of the differences occurring for walking, running, and PA <2.5 METs.
CONCLUSION: Overall, the Puyau and Mattocks equations provided the closest estimates of EE across a wide range of activities and ages. However, age and BMI appear to affect the validity of the ActiGraph prediction equations.
Supported by grant NIH 5R21HL093407-02