Comparing 6-minute walk versus treadmill walking distance as outcomes in randomized trials of peripheral artery disease

Abstract

BackgroundRandomized trials of people with peripheral artery disease (PAD) and intermittent claudication have traditionally used maximal treadmill walking distance as the primary outcome, but the 6-minute walk test is increasingly used as a primary outcome in randomized trials of PAD. This study compared relative changes in maximal treadmill walking distance versus 6-minute walk distance in response to a therapeutic intervention or control in randomized trials of participants with PAD. MethodsData from four randomized trials of therapeutic interventions in participants with PAD that measured both 6-minute walk and treadmill walking performance at baseline and the 6-month follow-up were combined. Two trials studied supervised treadmill exercise, one studied home-based walking exercise, and one studied resveratrol. ResultsOf 467 participants (mean age, 69.8; standard deviation, 9.7), the mean ankle-brachial index was 0.66 (standard deviation, 0.17). At the 6-month follow-up, participants with PAD randomized to control or placebo significantly declined in 6-minute walk distance (−10.2 m; 95% confidence interval, −18.2 to −2.2; P = .013), but improved maximal treadmill walking distance (+25.7 m; 95% CI, +6.0 to +45.3 m; P = .010; difference between change in 6-minute walk versus maximal treadmill walking distance: −37.3 m; 95% CI, −56.4 to −18.2; P < .001). Home-based exercise improved the 6-minute walk distance by 43.2 m (95% CI, +28.4 to +57.9), and supervised treadmill exercise improved the 6-minute walk distance by 25.0 m (95% CI, +14.7 to +35.2; mean difference, +18.2 m favoring home-based exercise [95% CI, +0.2 to +36.2 m; P = .048]). Among all participants, the presence (vs absence) of treadmill exercise training was associated with a 141.3-m greater improvement in maximal treadmill walking distance compared to 6-minute walk distance (95% CI, 88.2-194.4; P < .001), suggesting a benefit from treadmill training on the treadmill outcome. ConclusionsMaximal treadmill walking distance and the 6-minute walk distance are not interchangeable outcomes in participants with PAD. Participants with PAD randomized to control groups improved treadmill walking distance but simultaneously meaningfully declined in 6-minute walk distance. Supervised treadmill exercise training amplified improvement in treadmill walking distance because of a training to the outcome measure phenomenon.