Comparative effectiveness of lower body positive pressure and traditional treadmill training on adults with mild balance impairment.

Background: Treadmill training and body-weight supported treadmill training are effective for improving gait and balance in various populations. Lower-body positive pressure (PP) treadmill training uses positive air pressure to support body weight, potentially offering advantages over traditional treadmill training by reducing joint impact and allowing longer sessions. However, no studies have directly compared PP treadmill training with traditional treadmill training in adults with mild balance impairment. Method: In this three-armed parallel design randomised controlled trial, 72 adults were randomly assigned to: i) PP treadmill training with 20% bodyweight support (PP-BWS), ii) PP treadmill training without bodyweight support (PP-noBWS), and iii) traditional treadmill training without bodyweight support (TT). Participants in all three groups completed 25 min of treadmill training, three times per week, for 8 weeks. Outcomes included the Berg Balance Scale (BBS), Timed Up and Go (TUG), Functional Reach Test (FRT), and postural sway and gait measured with smartphone accelerometry and force plates. Outcomes were collected at baseline, at the end of the 2nd, 4th, 6th, and 8th week, and follow-up data were collected in the 10th week. Data were analysed using linear mixed-effects models, with multiple-imputation sensitivity analyses. Results: All interventions resulted in significant within-group improvements in balance and mobility measured with the BBS, TUG and FRT. For balance measures, the primary analysis revealed a group by time interaction (p = 0.003) for the BBS, favouring traditional treadmill training and PP-noBWS at week 10, but no between-group differences for the FRT. TUG measures of functional mobility showed a significant group by time interaction (p = 0.028), initially favouring novel PP-BWS, but there were no between-group differences after week 4. This aligned with smartphone accelerometry outcomes, which showed no between-group differences for comfortable walking speed and gait symmetry. Between-group differences in standing postural sway did not consistently favour one group. Due to a large dropout rate at follow-up, a sensitivity analysis was completed; this confirmed the significant within-group effects on balance and mobility at week 10, but between-group differences in balance were no longer statistically significant. Conclusion: All treadmill interventions led to significant within-group improvements in balance and mobility over the 10-week period. The initial analysis suggested treadmill interventions without body weight support, traditional treadmill training and PP-noBWS, demonstrated larger improvements in balance at week 10, but between-group differences were not sustained after accounting for dropout rates in the sensitivity analysis. This may suggest that the altered gait mechanics and reduced sensory input during PP treadmill training with bodyweight support may limit the improvements in balance that accompany treadmill training.