Cardiopulmonary exercise testing in severe osteoarthritis: a crossover comparison of four exercise modalities*

Cardiopulmonary exercise testing is performed increasingly for cardiorespiratory fitness assessment and pre-operative risk stratification. Lower limb osteoarthritis is a common comorbidity in surgical patients, meaning traditional cycle ergometry-based cardiopulmonary exercise testing is difficult. The purpose of this study was to compare cardiopulmonary exercise testing variables and subjective responses in four different exercise modalities. In this crossover study, 15 patients with osteoarthritis scheduled for total hip or knee arthroplasty (mean (SD) age 68 (7) years; body mass index 31.4 (4.1) kg.m^-2) completed cardiopulmonary exercise testing on a treadmill, elliptical cross-trainer, cycle and arm ergometer. Mean (SD) peak oxygen consumption was 20-30% greater on the lower limb modalities (treadmill 21.5 (4.6) (p < 0.001); elliptical cross-trainer (21.2 (4.1) (p < 0.001); and cycle ergometer (19.4 (4.2) ml.min^?1.kg^?1 (p = 0.001), respectively) than on the arm ergometer (15.7 (3.7) ml.min^-1.kg^-1). Anaerobic threshold was 25-50% greater on the lower limb modalities (treadmill 13.5 (3.1) (p < 0.001); elliptical cross-trainer 14.6 (3.0) (p < 0.001); and cycle ergometer 10.7 (2.9) (p = 0.003)) compared with the arm ergometer (8.4 (1.7) ml.min^?1.kg^?1). The median (95%CI) difference between pre-exercise and peak-exercise pain scores was greater for tests on the treadmill (2.0 (0.0-5.0) (p = 0.001); elliptical cross-trainer (3.0 (2.0-4.0) (p = 0.001); and cycle ergometer (3.0 (1.0-5.0) (p = 0.001)), compared with the arm ergometer (0.0 (0.0-1.0) (p = 0.406)). Despite greater peak exercise pain, cardiopulmonary exercise testing modalities utilising the lower limbs affected by osteoarthritis elicited higher peak oxygen consumption and anaerobic threshold values compared with arm ergometry.