Institutional members access full text with Ovid®

Share this article on:

Predicting Energy Expenditure from Accelerometry Counts in Adolescent Girls


Medicine & Science in Sports & Exercise: January 2005 - Volume 37 - Issue 1 - p 155-161
doi: 10.1249/01.MSS.0000150084.97823.F7
Applied Sciences: Physical Fitness and Performance

Purpose: Calibration of accelerometer counts against oxygen consumption to predict energy expenditure has not been conducted in middle school girls. We concurrently assessed energy expenditure and accelerometer counts during physical activities on adolescent girls to develop an equation to predict energy expenditure.

Methods: Seventy-four girls aged 13–14 yr performed 10 activities while wearing an Actigraph accelerometer and a portable metabolic measurement unit (Cosmed K4b2). The activities were resting, watching television, playing a computer game, sweeping, walking 2.5 and 3.5 mph, performing step aerobics, shooting a basketball, climbing stairs, and running 5 mph. Height and weight were also assessed. Mixed-model regression was used to develop an equation to predict energy expenditure (EE) (kJ·min−1) from accelerometer counts.

Results: Age (mean [SD] = 14 yr [0.34]) and body-weight–adjusted correlations of accelerometer counts with EE (kJ·min−1) for individual activities ranged from −0.14 to 0.59. Higher intensity activities with vertical motion were best correlated. A regression model that explained 85% of the variance of EE was developed: [EE (kJ·min−1) = 7.6628 + 0.1462 [(Actigraph counts per minute − 3000)/100] + 0.2371 (body weight in kilograms) − 0.00216 [(Actigraph counts per minute − 3000)/100]2 + 0.004077 [((Actigraph counts per minute − 3000)/100) × (body weight in kilograms)]. The MCCC = 0.85, with a standard error of estimate = 5.61 kJ·min−1.

Conclusions: We developed a prediction equation for kilojoules per minute of energy expenditure from Actigraph accelerometer counts. This equation may be most useful for predicting energy expenditure in groups of adolescent girls over a period of time that will include activities of broad-ranging intensity, and may be useful to intervention researchers interested in objective measures of physical activity.

1Division of Epidemiology, University of Minnesota, Minneapolis, MN; 2Center for Human Nutrition, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; 3Department of Exercise and Sport Medicine, University of North Carolina, Chapel Hill, NC; 4Department of Biostatistics, University of North Carolina, Chapel Hill, NC; and 5Department of Exercise Science, University of South Carolina, Columbia, SC

Address for correspondence: Kathryn H. Schmitz, Ph.D., MPH, Division of Epidemiology, University of Minnesota, 1300 South Second Street #300, Minneapolis, MN 55454; E-mail:

Submitted for publication January 2004.

Accepted for publication August 2004.

©2005The American College of Sports Medicine