Self-paced and fixed speed treadmill walking yield similar energetics and biomechanics across different speeds

Abstract

Background: Treadmill assessments are often performed at a fixed speed. Feedback-controlled algorithms allow users to adjust the treadmill speed, hereby potentially better resembling natural self-paced locomotion. However, it is currently unknown whether the energetics and biomechanics of self-paced differ from fixed-paced treadmill walking. Such information is important for clinicians and researchers using self-paced locomotion for assessing gait. Research question: To investigate whether energy cost and biomechanics are different between self-paced and matched-speed fixed-paced locomotion. Methods: 18 healthy participants (9 males/9 females, mean +/- standard deviation age 24.8 +/- 3.3 years, height 1.71 +/- 0.81 m, weight 65.9 +/- 8.1 kg) walked at four different self-paced speeds (comfortable, slow, very slow, fast) in randomized order on an instrumented treadmill while three-dimensional motion capture and gas exchange were measured continuously. The average walking speed during the last 2 min of the self-paced trials was used to match the speed in fixed-paced conditions. Linear mixed models were used to assess differences in mean values and within-subject variations between conditions (self-paced and fixed-paced) and speeds. Statistical Parametric Mapping was used to assess differences in kinematics of the lower limb between conditions. Results: Although self-paced walking consistently resulted in a 4-6% higher net cost of walking, there were no significant differences in the net cost of walking between conditions. Further, there were also no differences of clinical relevance in spatiotemporal outcomes and sagittal-plane lower-limb kinematics between the self-paced and fixed-paced conditions. Within-trial variability was also not significantly different between conditions. Significance: Self-paced and fixed-paced treadmill walking yield similar energetics and kinematics in healthy young individuals when mean values or linear measures of variation are of interest.