Treatment & Procedure
Neurofeedback (EEG biofeedback)
Biohacking
In neurofeedback, an EEG measures your brain waves, and a screen or tone tells you in real time whether a particular pattern is getting stronger. It is offered for concentration, sleep, stress reduction and peak performance, in practices and with EEG headbands for home use. That brain activity can be trained this way has been shown; that this does more for ADHD, sleep problems or attention than sham training has not.
In short
Neurofeedback is based on learning through reward: whoever produces the desired EEG pattern receives feedback and, over time, produces it more often. That works; frontal theta waves, for example, can be deliberately increased in studies. The decisive question, however, is whether the training works beyond general effects. In ADHD, a meta-analysis of 38 randomized trials found no improvement under blinded assessment, and in the large double-blind ICAN study, children with sham feedback did just as well. For insomnia and for attention in healthy people, the picture is similar. The most open question is sport, where the studies are positive but rarely sham-controlled.
What neurofeedback is and how it is supposed to work
Electrodes on the scalp record the electrical activity of the brain. Software breaks the signal down into frequency bands and rewards you as soon as a target pattern is reached, for example with a tone, a point or a film that keeps playing. Common protocols lower the ratio of theta to beta waves, increase the sensorimotor rhythm between 12 and 15 hertz or train alpha waves. The principle is conditioning: rewarded behavior occurs more often.
Neurofeedback does not deliver any current or magnetic field into the brain; it only measures. This fundamentally distinguishes it from tDCS and TMS, which stimulate the brain from outside: Neurostimulation (tDCS / TMS). A related approach is biofeedback via heart rate variability, in which the heart rhythm rather than brain waves is fed back: HRV measurement.
What is well supported
The first step is supported: people can learn to change their EEG activity. A meta-analysis on frontal theta rhythm in healthy people found a medium effect, with large variation between studies, and an increase in the target rhythm. It is also notable how well neurofeedback has been studied: there are several large, double-blind studies with sham feedback, more than for most biohacking practices.
In ADHD, a small effect appeared when only studies with established standard protocols were considered, along with a small improvement in processing speed. For sport motor tasks such as golf putting or shooting, a meta-analysis found a medium positive effect. These findings are the points where research would need to go further.
What the studies show
Westwood 2025 — ADHD, meta-analysis
38 randomized trials with 2,472 participants aged 5 to 40. The primary endpoint was overall ADHD symptoms, rated by people who probably did not know who received real training. Here there was no improvement (20 trials, 1,214 participants, standardized mean difference 0.04). Only in trials with established standard protocols was there a small effect of 0.21, and in the neuropsychological tests only for processing speed (0.35). The authors’ conclusion: no meaningful benefit at group level.
Neurofeedback Collaborative Group 2021 — ICAN study
144 children aged 7 to 10 with moderate to severe ADHD and an elevated theta-beta ratio, double-blind randomized to real training or identical-looking sham feedback using recorded EEGs. Blinding was successful. Both groups improved markedly, but for inattention the real training was superior neither at the end of treatment nor after 13 months (effect sizes 0.01 and 0.23, not significant). Response rates were 61 versus 54 percent, also without a significant difference. The neurofeedback group needed less medication at follow-up.
Schabus 2017 — insomnia, double-blind
25 people with primary insomnia each completed 12 sessions of real sensorimotor rhythm training and 12 sessions of sham feedback, with nine nights in the sleep laboratory. Subjectively, sleep improved equally under both conditions; objectively, in the sleep laboratory, it did not improve at all, and the sleep spindles that the training was supposed to strengthen also remained unchanged. The authors attribute the improvement to nonspecific factors such as attention and trust.
Recio-Rodriguez 2024 — sleep quality, meta-analysis
7 randomized trials, mostly with alpha training over 8 to 20 sessions. For sleep quality, the control conditions, including cognitive behavioral therapy or other forms of biofeedback, even performed better (5 trials); for insomnia severity there was no difference (3 trials).
Kimura 2024 — attention in healthy people, meta-analysis
41 randomized trials screened, 15 with 569 participants analyzed. Overall, attention improved slightly (standardized mean difference 0.27). In the eight trials with sham training as the comparison, no significant effect could be detected any longer.
Yu 2025 — sport, meta-analysis
25 studies screened, 21 analyzed, for example on golf putting, archery and rifle shooting or balance. Neurofeedback improved sport motor performance with a medium effect (Hedges g 0.78). The comparison was sham feedback in 9 studies and, in the rest, other protocols, normal training or rest; non-randomized studies were also included. The authors themselves call for more sham-controlled studies.
Where the data stop
The stricter the controls, the smaller the effect. Without sham control, neurofeedback often looks good, for example for attention in healthy people; with blinding and sham feedback, the advantage in ADHD, insomnia and attention largely disappears. The improvement itself is real; it simply also occurs under sham: attention, expectation and the regular sitting and concentrating play a part.
It remains open whether individual subgroups benefit and whether standard protocols do have a small specific effect. In sport, sufficiently blinded studies are lacking. The studies above come mostly from laboratories and practices; whether their results apply to home headbands advertised for focus, meditation or sleep is therefore not answered.
Status, approval and legal
Neurofeedback is not a medicine. In the US, biofeedback devices that feed back, for example, the brain’s alpha activity fall into a separate device class; for prescription battery-powered devices for relaxation training, premarket notification is not required there. This does not imply approval for the treatment of ADHD or insomnia.
In Germany, neurofeedback does not appear in the G-BA’s Remedies Directive (Heilmittel-Richtlinie). In ADHD it does not replace guideline-based treatment; the studies above argue against a specific benefit at group level.
Safety
Neurofeedback is non-invasive and does not deliver any energy into the brain. In the ICAN study, side effects were evenly distributed between real training and sham feedback. However, most studies do not record side effects systematically, so the safety data are less good than the low risk would suggest.
The greater practical risk is an indirect one: time and money for many sessions, and that an effective treatment, for example cognitive behavioral therapy in ADHD or insomnia, is postponed.
BK-Score Well studied – effect not confirmed
| Human evidence | 6 | |
|---|---|---|
| Mechanism | 4 | |
| Safety data | 6 | |
| Hype gap | 3 | |
| Track record of use | 7 |
An unusually large number of blinded studies for a biohacking practice, and the clean ones are negative: a meta-analysis of 38 randomized trials with 2,472 participants (JAMA Psychiatry 2025) found no improvement in ADHD in blinded assessments, and only a small effect with standard protocols; the double-blind ICAN study with 142 children saw no difference from sham feedback either at the end of treatment or after 13 months. In insomnia, neurofeedback was no better than sham (Brain 2017), and in healthy people the attention effect disappeared in comparison with sham training. That EEG patterns can be deliberately changed has been shown; how this is supposed to lead to symptom relief has not. In the large study, side effects were equally distributed in both groups; they are rarely recorded systematically. Focus, sleep and peak performance are what is advertised – the most open question is sport, where studies are positive but mostly without sham training.
The score rates the state of knowledge, not the substance. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)
Frequently asked questions about neurofeedback (EEG biofeedback)
What does neurofeedback do?
With it, people can learn to change their brain wave patterns. Whether that relieves complaints is another question: in ADHD, insomnia and attention, real training was no better than sham feedback in blinded studies.
Does neurofeedback help with ADHD?
A meta-analysis of 38 randomized trials found no improvement in core symptoms under blinded assessment, and only a small effect with standard protocols. In the double-blind ICAN study with 142 children analyzed, real training was no better than sham either at the end of treatment or after 13 months.
Does neurofeedback help with falling asleep?
In a double-blind study with 25 people with insomnia, sleep improved subjectively to the same extent under real and sham feedback, and objectively, in the sleep laboratory, not at all. A meta-analysis found no advantage over other treatments.
What is the difference between neurofeedback and neurostimulation?
Neurofeedback only measures brain waves and feeds them back; you have to do the training yourself. Neurostimulation such as tDCS or TMS delivers current or magnetic pulses into the brain from outside. Results from one cannot be transferred to the other.
Does neurofeedback improve concentration in healthy people?
Across all studies there was a small effect on attention. In the eight studies that compared it with sham training, it could no longer be detected.
Is neurofeedback dangerous?
It is non-invasive and does not deliver any energy into the brain. In the large ICAN study, side effects were equally distributed in both groups. The main risk is that an effective treatment is postponed.
Related
- Related topicNeurostimulation (tDCS / TMS)
- Related topicPEMF (pulsed electromagnetic fields)
- Related topicHeart rate variability (HRV)
- Same sectionBioadaptive impulse application
- Same sectionVagus nerve stimulation (active)
- Same sectionFloating (isolation tank)
Sources
- Westwood et al., JAMA Psychiatry 2025 (ADHD, meta-analysis)
- Neurofeedback Collaborative Group, J Am Acad Child Adolesc Psychiatry 2021 (ICAN study, ADHD, double-blind)
- Schabus et al., Brain 2017 (insomnia, double-blind)
- Recio-Rodriguez et al., Front Neurosci 2024 (sleep quality, meta-analysis)
- Kimura et al., Imaging Neurosci 2024 (attention in healthy people, meta-analysis)
- Pfeiffer et al., Neurosci Biobehav Rev 2024 (frontal theta, meta-analysis)
- Yu et al., Scand J Med Sci Sports 2025 (sport, meta-analysis)
- US Code of Federal Regulations, 21 CFR 882.5050 (biofeedback devices)
- G-BA, Remedies Directive (Heilmittel-Richtlinie), version of October 29, 2024
Open in the database – with search, filters and comparison (German app)
Information only, not medical advice and not a usage or dosage recommendation. Prescription-only and unapproved substances belong in the hands of a physician. Last updated: 2026-10-05.