Skip to main content
ARS Home » Research » Publications at this Location » Publication #189205
Title: COMPARISON OF TWO DIFFERENT PHYSICAL ACTIVITY MONITORS

Author
item PAUL, DAVID - JOHNS HOPKINS
item Kramer, Matthew
item Moshfegh, Alanna
item Baer, David
item Rumpler, William

Submitted to: BioMed Central Developmental Biology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/25/2007
Publication Date: 6/25/2007
Citation: Paul, D.R., Kramer, M.H., Moshfegh, A.J., Baer, D.J., Rumpler, W.V. 2007. Comparison of two different physical activity monitors. BMC Medical Research Methodology. 7:(26). Available: http://www.biomedcentral.com/1471-2288/7/26.

Interpretive Summary: Activity monitoring is one method to measure the daily physical activity of humans. There are a number of different brands of activity monitors, so this study compared the results of two different ones. We found that the Actigraph brand produces significantly higher scores for physical activity than the Actical. We were able to produce a mathematical equation to inter-convert the data from the monitors to get good group estimates of physical activity, but it does not predict the physical activity from individual subjects very well.

Technical Abstract: A number of different activity monitors brands are available for investigators to use, but little is known about how they compare at different levels of physical activity in the field, nor if data inter-conversion between brands is possible. Over the course of 13 days, 56 women and men were fitted with two different activity monitors; Actigraph (Actigraph LLC; AGR) and the Actical (Mini-Mitter Co.; MM). Both monitors were fixed to an elasticized belt worn at the waist for the entire study period. The AGR detected a significantly greater amount of activity per day (216.2 106.2 vs. 188.0 101.1 counts/min, P < 0.0001). The average difference between monitors expressed as a coefficient of variation was 3.5 and 6.8% for log transformed and raw data, respectively. Inter-conversion of data results in a reduction in the bias and mean square error. Overall, inter-conversion of the data between brands produces excellent group mean estimates, but individual subject results must be interpreted with caution.