THE INTERSECTION OF BRAIN SCIENCE & HUMAN POTENTIALPartner with us
neurofeedback.com
The newsletter
← Research & Evidence
Positive with limitationsTwo-stage study with a 16-person randomized sham-controlled trial followed by a 30-person single-arm multimodal cohort

Multimodal evidence associated with 20Hz rTMS in AD: neuroaxonal injury reduction, neural oscillatory rebalancing, and multilayer network integration

Published October 24, 2026. Cognitive scores improved in the active group of a very small sham-controlled stage and in a later single-arm cohort. Biomarker and EEG changes offer research clues about possible mechanisms, but the study is too small and structurally limited to establish clinical effectiveness or disease modification.

Wang T, Ji Y, Liu S, Wu H, Gu L, Guo M, Wang P, Xu G.
Behavioural Brain Research · Publication date: 2026-10-24
Added to site: 2026-10-25
DOI: 10.1016/j.bbr.2026.116536

Published October 24, 2026. Added to Neurofeedback.com October 25, 2026.

Original publication ↗
Modalities
rTMS, EEG, Blood biomarkers
Clinical focus
Alzheimer’s disease, Cognition, Neural oscillations and networks
Population
People with Alzheimer’s disease
Sample
16 participants in the randomized sham-controlled stage and 30 in the subsequent single-arm stage; possible cohort overlap was not resolved from the abstract
Intervention / method
Twenty-hertz rTMS targeting the dorsolateral prefrontal cortex, with cognitive testing in both stages and blood neurofilament light, EEG oscillations and multilayer network measures in the mechanistic cohort
Comparator
Sham rTMS in the 16-person first stage; pre-post single-arm comparison in the 30-person second stage

Clinician summary

In stage 1, active participants showed within-group MMSE and MoCA improvements while the sham group did not; the abstract does not report a direct group-by-time contrast or arm denominators. In stage 2, MMSE and MoCA improved alongside lower NfL, regional delta, beta and gamma changes, and network reorganization. Selected MoCA-network correlations ranged from R=0.389 to 0.478.

How to interpret the findings

The randomized stage included only 16 participants and the mechanistic stage was single-arm. Arm-specific denominators, a direct between-group treatment effect, confidence intervals, masking detail, adverse events, durability and cohort overlap were not available in the abstract. Multiple biomarker, EEG, network and correlation analyses raise multiplicity concerns. Imaging and biomarker associations do not show that network changes caused cognitive change or that neurodegeneration was reversed.

Equipment and disclosures

The abstract specifies 20 Hz dorsolateral-prefrontal rTMS and EEG but does not identify device manufacturers. No equipment is attributed to BrainMaster or another vendor.

Registered as ChiCTR2500095949 on January 15, 2025. The authors declared no competing interests. Funding detail was not available in the checked abstract or retained Crossref record. No indexed correction, retraction or expression-of-concern relation appeared in those records; this is a bounded check. AI-assisted source review.

Related evidence

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

Abstract checked · Checked 2026-10-25

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