THE INTERSECTION OF BRAIN SCIENCE & HUMAN POTENTIALPartner with us
neurofeedback.com
The newsletter
← Research & Evidence
Positive with limitationsSingle-center retrospective real-world cohort with multinomial propensity-score common-support trimming and adjusted three-group comparisons

Observation of the therapeutic effect of handheld induction therapy combined with electromyographic biofeedback on post-stroke dysphagia.

Published November 7, 2026. After four weeks, the group receiving conventional rehabilitation, EMG biofeedback and handheld electrical stimulation had better swallowing-test and oral-intake scores than the two comparison groups in this single-center retrospective cohort. Reported adverse-reaction rates did not differ significantly. Because treatment was not randomized and the combined group received two added components, the study cannot show that EMG biofeedback alone caused the difference or that the result will generalize to other settings.

Quan C, Ma T, Zhu X, Xu T.
Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association · Publication date: 2026-11-07
Added to site: 2026-11-12
DOI: 10.1016/j.jstrokecerebrovasdis.2026.108783

Published online November 7, 2026. Added to Neurofeedback.com November 12, 2026 after comparison with fresh and reserve candidates. The source publication date is separate from the site addition date.

Modalities
EMG biofeedback, Biofeedback, Electrical stimulation
Clinical focus
Post-stroke dysphagia, Swallowing rehabilitation, Combined rehabilitation
Population
Adults with first-onset stroke and dysphagia treated at one rehabilitation hospital from January 2023 through December 2024
Sample
105 eligible patients before trimming, with 96 patients in the final adjusted cohort, 32 in each of three treatment groups
Intervention / method
All groups received four weeks of conventional swallowing and eating training. Group B also received 20-minute surface-EMG biofeedback sessions five days per week. Group C received the same care plus 10- to 20-minute handheld intraoral low-frequency electrical-stimulation sessions five days per week. Outcomes included videofluoroscopic swallowing scores, MASA, FOIS, a composite effective-rate category and reported adverse reactions.
Comparator
Conventional rehabilitation alone and conventional rehabilitation plus EMG biofeedback; treatment was selected in routine care rather than randomly assigned

Clinician summary

After propensity-score common-support trimming, 96 patients remained, 32 per group. The combined-intervention group had higher post-treatment VFSS components, MASA 85.37 versus 73.25 and 73.31, and FOIS 6.05 versus 5.14 and 5.09, all reported p<0.001. Composite effective rates were 96.88%, 84.38% and 71.88% across combined, EMG-biofeedback and conventional groups, respectively, overall p=0.025. Adverse-reaction rates were 18.75%, 15.62% and 28.12%, p=0.440. The EMG-only versus conventional contrast was exploratory and did not show significant differences in the principal post-treatment scores.

How to interpret the findings

This was a nonrandomized retrospective cohort from one hospital with 32 analyzed patients per group and only four weeks of follow-up. Propensity trimming and regression can address measured covariates but cannot remove residual confounding or treatment-selection bias. The combined group added handheld electrical stimulation to EMG biofeedback, so the study cannot isolate the contribution of either component. The composite effective-rate threshold is less directly interpretable than validated patient-important outcomes. Ordinal VFSS and FOIS scores were analyzed as continuous variables, assessors were not reported as masked, and stroke type and lesion location were not stratified. The results are hypothesis-generating rather than practice-changing.

Equipment and disclosures

The EMG biofeedback group used a MyoDynamik surface-EMG system, and the combined group also used an LGT-2350A swallowing neuromuscular electrical stimulator. These are source-reported protocol details, not evidence of product superiority or attribution to BrainMaster.

The authors reported no external funding and declared no known competing financial interests or personal relationships. No correction, retraction or expression-of-concern notice appeared in the checked PubMed, publisher or bounded DOI and title records on November 12, 2026. The full text is CC BY-NC-ND 4.0. AI-assisted source review for Neurofeedback.com, not independent human clinical review.

Related evidence

Compare findings from the same clinical area, including mixed results.

Full text checked · Checked 2026-11-12

Summary prepared with AI assistance. Source verification does not imply an independent replication or a clinician's endorsement. Read our review method.