J Neuroeng Rehabil. 2026 Jun 25. doi: 10.1186/s12984-026-01952-5. Online ahead of print.
ABSTRACT
BACKGROUND: Transcutaneous auricular vagus nerve stimulation (taVNS) paired with task-specific training has shown promise for improving upper-limb motor recovery after stroke but its effects on gait and balance remain unclear. This study evaluated the feasibility and safety of taVNS paired with task-specific training and explored preliminary effects on walking and balance outcomes in adults with chronic stroke.
METHODS: A double-blinded, sham-controlled trial was conducted to evaluate active versus sham taVNS paired with task-specific training (15 one-hour sessions over six weeks). Feasibility metrics included recruitment, participant experience, adverse events (AEs), and blinding effectiveness. Clinical outcomes (walking speed, endurance, gait parameters, and balance) were assessed at baseline, post-intervention, and ≥ 30 days post-intervention, supplemented by qualitative interviews.
RESULTS: Twenty-five adults with chronic stroke completed this study (18 males, mean age 64.99 ± 13.45 years; 41.80 ± 38.00 months post-stroke, 14 received active taVNS). Active taVNS was considered feasible and safe, with a 93% consent rate; 96% of participants found the intervention useful/helpful, and 100% found the stimulation acceptable. There were no serious AEs, with 16% of participants reporting mild side effects. Both groups improved across clinical outcomes. No statistically significant between-group differences were observed. Between-group effect sizes were small for gait, balance, disability, and self-efficacy, and small-to-moderate for walking endurance; however, confidence intervals crossed zero for all outcomes.
CONCLUSION: Active taVNS paired with task-specific training was feasible and well tolerated in chronic stroke. No definitive between-group effects were detected. Observed effect sizes should inform the design and sample size calculations of future adequately powered trials using optimised taVNS protocols.
PMID:42351122 | DOI:10.1186/s12984-026-01952-5
