Validity of the International Physical Activity Questionnaire in a rural Hispanic Population

Abstract

<bold>INTRODUCTION:</bold> The primary purpose of this study was to examine the criterion validity of the IPAQ-long versus an accelerometer in terms of total physical activity (PA) among a group of Hispanics living in a rural area of Washington State. <bold>METHODS:</bold> A random sample of sixty participants (mean age: 49.8 ± 11 years, 81% female, 78% low acculturation) was taken from an existing intervention trial for diabetics. Participants wore an accelerometer on their hip for seven days and completed both the long (IPAQ-L) and short (IPAQ-S) versions of the IPAQ on the eighth day. Spearman's rank correlation coefficients and Bland-Altman plots were employed to examine the criterion validity and agreement between the two measures, respectively. Comparisons were done using both modified 10-minute bout activity as well as total minutes in activity >100 counts. <bold>RESULTS:</bold> Correlation coefficients for total PA in 10-minute bouts were modest for both the IPAQ-L (0.32; 95% CI: 0.07,0.56) and the IPAQ-S (0.33; 95% CI: 0.08, 0.59). The coefficients for total PA in minutes >100 counts were similar for the IPAQ-L (0.27; 95% CI: 0.03,0.52) and IPAQ-S (0.28; 95% CI: 0.02, 0.54). Results from the Bland-Altman plots suggest modest agreement between the IPAQ-L and the accelerometer in total measured PA, and that the IPAQ-long tended to overestimate minutes spent in total PA in bouts of at least 10-minutes. <bold>CONCLUSIONS:</bold> The results of this study suggest that both versions of the IPAQ have poor criterion validity in measuring PA and overestimate minutes spent active in a population of rural Hispanics.