Comparison of a subjective and an objective measure of physical activity in a population sample

Background: The aim of this study was to compare physical activity components in the long, self-administrated version of IPAQ with an accelerometer in a population sample. Methods: In total 980 subjects (18-65 years) wore an accelerometer (Actigraph) for 7 consecutive days and thereafter flled in the IPAQ. Measures of total physical activity, time spent in moderate and in vigorous activity as well as time spent sitting as assessed by the IPAQ and the Actigraph were compared. Results: The results showed signifcant low to moderate correlations (Rs = 0.07-0.36) between the 2 instruments and signifcantly (P < .001) higher values for sitting and vigorous intensity physical activity from the IPAQ compared with the Actigraph. The higher the values reported by the IPAQ the bigger differences were seen between the instruments. Comparison between the tertiles of total physical activity by the 2 instruments showed signifcant overall association with consistent agreement in the low and the high tertiles. Conclusion: The long form of IPAQ is a valid measure of physical activity in population research. However, the IPAQ likely overestimates actual physical activity as shown by its limited ability to classify adults into low and high categories of physical activity based on accelerometer data.