Comparison of accelerometers with oxygen consumption in older adults during exercise

Purpose: The purpose of this study was to compare two commercially available accelerometers with indirect calorimetry in a group of older adults (x +/- SD; 70.6 +/- 3.7 yr; N = 86, 44 males and 42 females). Methods: The accelerometers (Caltrac and Tritrac, Hemokinetics, Madison, WI) were worn while performing three submaximal, discontinuous (5 min exercise, 2 min recovery), progressive levels of treadmill walking and bench stepping. The treadmill exercise averaged 3.4 mph, at 0.4% grade, 3.0% grade, and 5.1% grade, while the stepping work rates (24 steps.min(-1)) were performed on 15.2-, 20.3-, and 25.4-cm steps. Estimated energy expenditure (EE) from the two accelerometers was compared with EE as measured by indirect calorimetry. Results: The Caltrac significantly (P < 0.05) overestimated EE at the three treadmill work rates (10-52% difference) and underestimated EE at the three stepping work rates (-19% to -28% difference). When comparing the changes in EE between work rates one, two and three, the Caltrac was not sensitive to the changes (increase in EE) that occurred during graded treadmill walking but did detect some changes in the stepping exercise. The Tritrac significantly (P < 0.05) underestimated EE for the three work rates of both the treadmill and stepping exercise when compared with indirect calorimetry but did detect differences in EE among work rates during stepping exercise (P < 0.05). Conclusions: These data indicate that the magnitude of the differences between measured and estimated EE is affected by exercise mode and intensity and that caution is warranted when using the acceleromerers in an attempt to quantify EE in older adults.