Submaximal Exercise Testing in Cardiovascular Rehabilitation Settings (BEST Study)

This study compared changes in measured versus predicted peak aerobic power (V̇O ) following cardiovascular rehabilitation (CR). Peak cardiopulmonary exercise testing (CPET) results were compared to four V̇O estimation methods: the submaximal modified Bruce treadmill, Astrand-Ryhming cycle ergometer, and Chester step tests, and the Duke Activity Status Index (DASI). Adults with cardiovascular disease (CVD) who completed a 12-week CR program were assessed at baseline and 12 weeks follow-up. CPET, the DASI and three subsequent submaximal exercise tests were performed in a random order. Of the 50 adults (age: 57 ± 11 years) who participated, 46 completed the 12-week CR program and exercise tests. At baseline 69, 68, and 38% of the treadmill, step and cycle tests were successfully completed, respectively. At follow-up 67, 80, and 46% of the treadmill, step and cycle tests were successfully completed, respectively. No severe adverse events occurred. Significant improvements in V̇O were observed with CPET (3.6 ± 5.5 mL kg min , < 0.001) and the DASI (2.3 ± 4.2 mL kg min , < 0.001). Bland-Altman plots of the change in V̇O between CPET and the four V̇O estimation methods revealed the following: a proportional bias and heteroscedastic 95% limits of agreement (95% LoA) for the treadmill test, and for the cycle and step tests and DASI, mean bias' and 95% LoA of 1.0 mL kg min (21.3, -19.3), 1.4 mL kg min (15.0, -12.3) and 1.0 mL kg min (13.8, -11.8), respectively. Given the greater number of successful tests, no serious adverse events and acceptable mean bias, the step test appears to be a valid and safe method for assessing group-level mean changes in V̇O among patients in CR. The DASI also appears to be a valid and practical questionnaire. Wide limits of agreement, however, limit their use to predict individual-level changes.