SMR Neurofeedback: Sleep, Attention and the Evidence

If you are considering SMR training for sleep or attention, ask what the session actually rewards. SMR stands for sensorimotor rhythm. In this research, it refers to EEG activity around 12 to 15 hertz recorded over sensorimotor cortex. Hertz means cycles per second: the number describes how quickly the recorded rhythm oscillates.

In the Schabus insomnia study, the recording site was C3, a left-central scalp location. Participants watched a compass display and tried strategies to move its needle toward a target. During genuine feedback, the reward depended on increasing the selected 12-to-15-hertz signal. This makes the immediate task concrete: influence the recorded feature and see the display respond.

Whether that learning changes sleep is a further question. The study measured that outcome separately.

Keep the recording state in view

The sleep question arose partly because the trained frequency range overlaps with sleep-spindle activity observed during non-REM sleep. The proposed connection needed testing across two settings: practicing with feedback while awake and recording what happened during sleep afterward. Similar frequencies give researchers a question to investigate; they still need separate measurements of the two states.

The QEEG guide explains how frequency, recording location and context contribute to interpretation. That context keeps a rhythm label connected to what was actually measured.

When a protocol is proposed, ask what feature is being trained and what finding would show that the intended learning occurred. Then ask how the claimed benefit will be assessed separately.

The insomnia comparison is instructive

In Schabus and colleagues' double-blind study, people with primary insomnia received SMR and placebo-feedback conditions. Participants could increase SMR during genuine feedback. Subjective sleep improved, but the improvement was not specific to that condition, and objective measures failed to establish a corresponding specific treatment benefit.

That result distinguishes learning to influence a signal from improving the condition the training was meant to address. It is a finding about one studied protocol and sample; it cannot settle every possible SMR application. It does require restraint when claiming that SMR training reliably improves insomnia.

For persistent sleep problems, the sleep guide discusses assessment and established care. A training proposal should fit around that evaluation rather than substitute a brainwave explanation for it.

ADHD evidence needs its own reading

Different neurofeedback protocols have been studied in ADHD, often under different controls and rating conditions. A 2025 review of randomized trials found an overall null result on probably blinded total symptoms, with a small standard-protocol subgroup effect. It did not establish stand-alone treatment efficacy.

That synthesis should not be relabeled as proof that every SMR protocol works or fails. The ADHD guide examines the evidence and the distinction between protocol selection and demonstrated treatment benefit.

Seizure-related claims require a separate clinical discussion. A historical connection between SMR research and seizure control cannot justify self-directed training or changes in antiseizure treatment. Those decisions belong with the clinician managing the condition.

Define what you want the training to accomplish

For a proposed program, agree on the functional goal, the measures and the review point. Ask how recording quality is checked and how adverse changes are handled. A fixed session promise needs evidence for the particular outcome and population.

The alpha-rhythm article offers another example of why band names need context. A useful neurofeedback explanation tells you what is measured, what is rewarded and what result would count outside the session. Keeping those questions visible makes an interesting physiological method easier to evaluate honestly.