Background Self-reported physical activity equivalent to at least 2 h a week is associated with reduction in the risk of hospital admission and death in COPD.1 However, whether self reported activity is accurate in COPD is uncertain. The gold standard objective measures of physical activity in daily life are direct observation and assessment of energy expenditure by calorimetry or the doubly labelled water method.2 We sought to test the validity of a home activity monitor (Actigraph)—for measurement of physical activity levels by comparing to calorimetry, the 6MWT and an activity questionnaire.
Methods 20 patients with COPD and five healthy elderly subjects underwent home activity monitoring over a period of 5 days, a 6MWT and were questioned twice regarding activity levels using the modified Baecke questionnaire (MBQ).3 Time spent non-sedentary according to the Actigraph was compared to perceived activity from the MBQ and 6MWT. A sub-group also underwent concurrent measurement of energy expenditure by indirect calorimetry and the Actigraph during a walk at steady state on a treadmill.
Results Actigraph output is summarised in Abstract P43 table 1. MBQ score did not relate to actual time active in patients (p=0.54), and was less reproducible on retest than in controls. Furthermore when patients were ranked according to activity level by MBQ and Actigraph, it was clear that patients both under and over-estimated their activity level. Actigraph activity levels related well to 6MWT (p=0.04). Accuracy of the Actigraph at measuring energy expenditure, compared to the gold standard of calorimetry, varied with walking speed, such that it was less accurate in those walking slowly.
Conclusions Patients self reported activity levels are likely to be inaccurate. Home activity monitoring is accurate at measuring exercise capacity, but not at measuring energy expenditure, due to slow speed of movement.