Reply to: ‘Post-exercise criteria to diagnose lower extremity peripheral artery disease: Which one should I use in my practice?’ by Stivalet et al.

We thank Stivalet and colleagues for their letter. ¹ We agree that measurement of the post-exercise ankle–brachial index (ABI) in the symptomatic leg should occur first, as the time period immediately following peak exercise should correspond to the most significant arterial flow disturbances in the setting of obstructive lower extremity peripheral artery disease. We also agree that the artery used to measure the ABI should be consistent in both resting and exercise studies. In our own research study under discussion, switching to a different artery for the exercise ABI would have unnecessarily complicated direct comparison to the resting ABI. ²

Stivalet and colleagues are correct in pointing out the different exercise protocols used in the initial validation studies of the various exercise ABI thresholds as well as our own study, and this does complicate direct comparison of these studies. We may have seen a greater area under the curve (AUC) for each post-exercise ABI measure using a treadmill protocol with a greater speed or grade, although it is not clear if more vigorous exercise would favor one measure over the others. Instead, we suspect it would have improved the AUC for all measures. Additionally, if a large proportion of patients in our study already experienced lower extremity symptoms with the present protocol, it is less likely more vigorous exercise would increase the AUC. Although we suspect that not all patients developed claudication during the study, these details were not recorded.

Finally, to clarify the authors’ concern about the timing of our post-exercise ABI measurement, this occurred both immediately following exercise and at subsequent 2-minute intervals for either 6 minutes or until the ankle pressures returned to baseline. Therefore, because we collected ABI data in the symptomatic limb as soon as possible after peak exercise, we do not think this component of the protocol could be altered to significantly improve the AUC of each measure.