Physical activity is frequently a component of interventions designed to diminish weight gain in children. It is essential to determine whether the energy expenditure (EE) elicited by these interventions is sufficient to reduce the rate of weight gain.
Purpose: To quantify the EE of the Physical Activity across the Curriculum (PAAC) intervention. This intervention involved two 10-min physically active academic lessons per day, taught by classroom teachers.
Methods: We assessed EE of PAAC in 19 males and 19 females using both an indirect calorimeter (IC) (COSMED K4b2) and an accelerometer (ActiGraph) (AC). Independent t-tests were used to evaluate gender differences. Dependent t-tests were used to examine the difference between EE assessed by IC and AC. The agreement between EE measured by IC and estimated by AC was evaluated using a Bland-Altman plot. A Pearson correlation between EE measured by IC and estimated by AC was calculated.
Results: There were no significant gender differences for age, BMI, or EE; therefore, analyses by gender were not performed. The mean EE measured by IC was 3.1 ± 1.0 kcal·min−1 (3.4 METs). Mean EE estimated by AC (1.8 ± 0.9 kcal·min−1) was significantly lower (P < 0.05) than EE measured by IC (mean underestimation = 1.3 kcal·min−1). The Bland-Altman plot suggested increased underestimation with increased levels of EE. The 95% limits of agreement were large (−2.8 to +0.3 kcal·min−1). The correlation between EE measured by IC and estimated by AC was r = 0.68 (P < 0.001).
Conclusion: PAAC elicited a level of EE that may prevent excessive weight gain in children. AC significantly underestimated the EE of PAAC lessons and may not provide useful EE estimates in this context.
Center for Physical Activity and Weight Management, Energy Balance Laboratory, University of Kansas, Lawrence, KS
Address for correspondence: Joseph E. Donnelly, ED.D., Professor/Director, Energy Balance Laboratory, Center for Physical Activity & Weight Management, Schiefelbusch Institute for Life Span Studies, 1301 Sunnyside Ave, Rm 100, The University of Kansas, Lawrence, KS 66045; E-mail: firstname.lastname@example.org.
Submitted for publication July 2007.
Accepted for publication February 2008.