Measuring physical activity in peripheral arterial disease: A comparison of two physical activity questionnaires with an accelerometer

Peripheral arterial disease (PAD)-related exertional leg pain may limit physical activity, thereby contributing to mobility loss and increasing cardiovascular morbidity and mortality in men and women with PAD. The objectives of this study were: (1) to compare objectively measured physical activity levels between patients with and without PAD, (2) to assess the validity of two physical activity questionnaires in patients with PAD. Twenty PAD patients from a noninvasive vascular laboratory and 21 patients without PAD from a general medicine practice wore an accelerometer continuously for 7 days to measure physical activity objectively. After 7 days, participants completed the leisure time physical activity questionnaire (LTPAQ), derived from the Health Interview Survey, and the Stanford 7-day physical activity recall questionnaire (PARQ). PAD participants had markedly lower physical activity levels than non-PAD participants as measured by accelerometer (803 kcal/week ±364 (range=284-2,000, median=708) vs 1,750 kcal/week ±1,296 (range=882-6,586, median=1,278), p<0.001). For the LTPAQ, physical activity levels in PAD and non-PAD participants were 609 kcal/week ±576 (range=0- 2,085, median=529) vs 832 kcal/week ±784 (range=53-2,820, median=623), p=0.128. For the PARQ, physical activity levels in PAD and non-PAD participants were 234 METS/week ±21 (range=214-301, median=229) vs 238 METS/week ±11 (range=225-268, median=234), p=0.454, respectively. Pearson's correlation coefficient for the association between the accelerometer and the log-transformed LTPAQ measure was 0.419 (p=0.006). Pearson's correlation coefficient was 0.348 for the association between the accelerometer and the log-transformed PARQ measure of physical activity (p=0.026). In conclusion, PAD patients have significantly lower physical activity levels than non-PAD patients. Two commonly used physical activity questionnaires were less sensitive than objective measurement to the association between PAD and inactivity.