Neurofeedback and the Push for Drug-Free ADHD Technology

As digital and AI-driven tools move deeper into mental health care, one older technology is drawing renewed attention: neurofeedback, a brain-computer method now marketed widely as a drug-free option for attention deficit hyperactivity disorder (ADHD). It sits within a growing category of technology-based interventions being positioned alongside, and sometimes as an alternative to, stimulant medication. The interest is real, but so is the gap between how these tools are promoted and what the clinical evidence currently supports.

Neurofeedback is an EEG-based technique. Sensors measure electrical activity in the brain, and signals such as the theta and beta frequency bands are converted into a form the person can perceive, usually sound or visuals, with the aim of training more stable patterns linked to attention and impulse control. Sessions are non-invasive and require no medication, which is a large part of the appeal for parents and adults seeking a natural treatment for ADHD that avoids the side effects associated with stimulants. Several protocol types are in use, including theta-beta ratio, slow cortical potential and sensorimotor rhythm training. The technology itself is well established and has been studied for decades.

What remains unsettled is how well it works. The central difficulty was captured in a 2019 meta-analysis published in Frontiers in Psychiatry by Bussalb and colleagues, which observed that each meta-analysis published over the previous decade had contradicted the methods and results of the one before it, making a consensus on effectiveness hard to reach. That same 2019 analysis pointed to a recurring problem of who assesses the outcome, finding that improvements were significant when ADHD symptoms were rated by parents, who were not blinded, but not when rated by teachers, who probably were, a pattern suggesting expectation may account for part of the reported benefit.

The picture also differs by age. A 2022 pilot meta-analysis published in Scientific Reports by Fan and colleagues, pooling five randomised controlled trials, found no difference between EEG neurofeedback and comparison groups on inattention, hyperactivity and impulsivity, or total symptom scores in adolescents and adults. Results in children appear more favourable. A separate 2022 meta-analysis of 14 randomised controlled trials reported significant improvement in laboratory-measured sustained attention among ADHD participants who received EEG neurofeedback compared with controls, and a 2024 network meta-analysis published in Brain and Behavior by Wu and colleagues concluded that non-pharmacological interventions, and neurofeedback in particular, can significantly improve ADHD symptoms in children, while noting the evidence remained incomplete.

The fair summary is that neurofeedback is a legitimate, non-invasive technology under continuing study, but the evidence supporting it is not yet as strong or as consistent as that for established medical interventions such as stimulant medication and behavioural therapy, which remain the first-line treatments recommended by clinical guidelines. For now, it is best understood as a complementary or alternative option that some clinics offer rather than a proven replacement for standard care.

That distinction matters more as the wider field moves quickly. Regulators are now treating some software as medical treatment in its own right. In December 2025 the US Food and Drug Administration cleared LumosityRx, developed by Lumos Labs, a prescription digital therapeutic of 13 cognitive-training games intended to improve attention in adults aged 22 to 55 with ADHD. Its clearance was supported by the GAMES study, a randomised, double-blind, sham-controlled trial of 560 participants who used the app or a sham version for about 15 minutes a day over nine weeks. It followed earlier clearances such as EndeavorRx, a prescription video game (AKL-T01) that uses adaptive algorithms to train attention in children with ADHD. These products reached market through the kind of blinded, controlled evidence that parts of neurofeedback's record still lack.

In Australia, the regulatory scaffolding around all of this took shape over the same period. The Australian Government announced the Australian AI Safety Institute on 25 November 2025 to assess AI risks and harms, and released its National AI Plan on 2 December 2025, which stepped back from the mandatory high-risk guardrails it had been exploring in favour of relying on existing laws. On the disability side, industry reporting notes the NDIS has released a framework specifically for AI-enabled assistive technology, developed by CSIRO's Australian eHealth Research Centre, to encourage innovation while keeping tools safe and well-matched to individual needs.

The common thread is that technology-based options for ADHD and mental health are multiplying, from neurofeedback to prescription apps to AI assistants, but they sit at very different points on the evidence curve. For anyone weighing a drug-free approach, the practical takeaway is to treat the maturity of the evidence, not just the sophistication of the technology, as the thing that matters most.