Subjects.

 This study was approved by the Medical Ethics Committee of the Catharina Hospital, Eindhoven, The Netherlands. Thirty consecutive patients with complaints of IC, confirmed by non-invasive testing (<0.9 Ankle-Brachial Index at rest or a fall in systolic ankle pressure by more than 20% after exercise), were recruited from the vascular outdoor clinic. Patients with comorbidity possibly limiting walking apart from IC (i.e. neurological disorders, severe COPD, congestive heart failure, orthopedic impairments) or with insufficient knowledge of the English or Dutch language were excluded. All participants provided verbal and written informed consent.

Study Protocol.

 Baseline characteristics, comorbidity, medical history and cardiovascular risk factors were recorded. All patients performed an open air distance measuring test and completed two treadmill protocols on one day. The order of the three tests was randomized using 'Randomizer' (Kwixo Designs, lite version) for Android smartphones. Three physiotherapy students were trained to standardize patients' encouragement.

Coupling of patient and physiotherapy student was also randomized using 'Randomizer'. Patients rested for 20 minutes (supine or seated) after each test. Primary measurement outcomes were maximal walking distance (MWD) as determined by the patient's own estimation, as obtained during outside walking and during the two treadmill tests. In addition, walking speed and the WIQ results were analyzed.

Self-reported MWD

 Patients were asked to estimate their MWD, defined as 'the maximum distance (meters) you can walk before you are forced to stop by leg pain'.

Outside walking

 To approximate a real 'daily life' maximal walking distance, patients were asked to walk outside on level ground in straight lines without interruption by traffic. They were instructed to continue until they were forced to stop by leg pain. MWD was recorded using both a measuring wheel (Stanley Black & Decker Inc., New Britain USA) and a GPS controlled device (iPhone 4s; Apple, Silicon Valley, USA). Walking time was recorded using the iPhone.

Treadmill tests

 Two different treadmill protocols were used. A graded incremental test (Gardner-Skinner Protocol, GSP) allows patients to walk at 3.2 kilometer per hour (km/h) with a 0% incline that increases by 2% every 2 minutes 13. The maximum test duration is 20 minutes. The non-graded treadmill protocol (NGTP) has a fixed 0% incline and allows the patient to walk with a favorite walking speed which was set during the first 30 seconds of the treadmill testing. Results are expressed as MWD.

 Walking Impairment Questionnaire

Patients were asked to complete the WIQ evaluating several components of daily walking ability including an estimation of walking distance, speed and stair-climbing ability (12). Patients were instructed to rank the degree of difficulty for each component using a 0 to 4 Likert scale. A validated Dutch version of the WIQ was used 14

 Walking Speed

Average walking speed of outside walking was calculated by dividing walking distance by time and was expressed in kilometer per hour (km/h).

Statistical analysis.

 It was assumed that outside walking reflected daily life walking most closely, and was therefore used as reference value. The Friedman two-way analysis of variance test determined differences between multiple assessments. A Wilcoxon Signed Rank test was used for comparison of two measurements. A Bonferroni method was used as post-hoc procedure for correction of multiple comparison testing.

 Various methods were used to assess study outcomes. Overall reliability was assessed by means of an Intraclass Correlation Coefficients (ICC) with 95% confidence intervals (CI). Variability in measurements was assessed using coefficient of variation (COV). The COV was calculated as the standard deviation of the absolute difference between two assessments (outside walking vs. patient's estimation, outside walking vs.

116
GSP, outside walking vs. NGTP) divided by the mean of the averages. Reproducibility was analyzed in Bland-Altman plots. These plots were used to visualize agreement between two measurements and were presented with 95% limits of agreement, calculated as the mean difference ± 1.96 standard deviation (SD). The mean of both measurements was depicted on the horizontal axis whereas the difference was illustrated on the vertical axis. Linear regression analysis was performed and mean difference (bias) was calculated. Spearman rank correlations were estimated to compare WIQ scores with walking distances and speed. A correlation coefficient was considered strong if = 0.7, moderate if between 0.3 and 0.7, and weak if = 0.3. P values less than 0.05 were considered statistically significant. Statistical analysis was performed using SPSS Statistics for Windows (version 20.0).