Effects of EEG neurofeedback in adults with attention-deficit/hyperactivity disorder: a systematic review of randomised and non-randomised studies

Published July 22, 2026. This adult-only review found that randomized trials generally showed small, non-significant effects compared with controls. Larger improvements reported by naturalistic and personalized studies came from less controlled evidence. The review does not show that EEG neurofeedback reliably outperforms placebo for adult ADHD, but it identifies specific approaches that warrant better trials.

Rotkiewicz M, Yachou Y, Charef S, Bouaziz N, Januel D.
Journal of Neurodevelopmental Disorders · Publication date: 2026-07-22
Added to site: 2026-10-29
DOI: 10.1186/s11689-026-09720-1

Published July 22, 2026. Added to Neurofeedback.com October 29, 2026 after the adult-specific review surfaced in current indexing. This is a historical addition, not a newly published October 29 study.

Modalities

EEG neurofeedback, Theta/beta training, Slow cortical potential training, Personalized neurofeedback

Clinical focus

Adult ADHD, Symptoms and cognition, Protocol-specific evidence

Population

Adults age 18 or older with a formal ADHD diagnosis

Sample

Seven studies involving 390 adults: three randomized trials and four non-randomized studies

Intervention / method

One theta/beta study, one slow-cortical-potential study, one alpha-downregulation study, three personalized individual-alpha-peak-frequency studies and one QEEG-guided study

Comparator

Control conditions varied across the three randomized trials; four studies were non-randomized. The authors did not pool results because protocols and outcomes were too heterogeneous.

Clinician summary

Seven studies and 390 adults were included. Two of three randomized trials were rated high risk of bias and only one low risk; three of four non-randomized studies were rated high risk and one moderate. Cognitive gains were inconsistent and often compatible with practice effects, while electrophysiological findings did not converge across protocols.

How to interpret the findings

The review found only seven heterogeneous studies, including three randomized trials, and did not conduct a pooled meta-analysis, sensitivity analysis or formal publication-bias assessment. Most included studies were at high risk of bias. Safety evidence was incomplete: three randomized trials addressed adverse events and reported none related to treatment, but one deferred analysis was never reported, and the four non-randomized studies did not report safety protocols. Findings should not be generalized across every adult, ADHD presentation, device or neurofeedback protocol.

Equipment and disclosures

The review compared protocol families and did not establish superiority for any commercial system. No equipment is attributed to BrainMaster or another vendor.

The review reported no funding and no competing interests. Its authors disclosed using ChatGPT for language refinement and Gemini 3 Pro for visualisations, followed by author review and responsibility for the final article. No correction, retraction or expression-of-concern relation appeared in the checked PubMed and publisher records on October 29, 2026; this is a bounded check. AI-assisted source review for Neurofeedback.com, not independent human clinical review.

Related evidence

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