McGill researchers found the same blunted beta-band activity whether patients were processing sounds, sights or movements, and the patients furthest from normal had the most disorganised speech.

Schizophrenia presents as a list of apparently separate problems. Perception can be distorted, speech can become disorganised, movement and timing can be affected, and the relationship between those symptoms has never been obvious. Treating them as one disorder is a clinical convention more than a mechanistic claim.

A study in Science Advances from McGill University offers a candidate for something underneath them. Measuring brain activity with high-resolution magnetoencephalography, the team found a single oscillatory signal that was weaker and slower in patients, and that tracked several different symptoms at once.

“What struck us was that the same weakened brain signal kept showing up wherever we looked, whether patients were processing a sound, a sight or completing a movement,” said Hsi Wei, a postdoctoral scholar in McGill’s Department of Psychiatry and at the Douglas Research Centre, and the study’s lead author. “Seeing the same pattern across these different domains suggested there may be a common underlying process linking the symptoms.”

What beta-band activity does

The signal is in the beta band, the range of neural oscillations at roughly 13 to 30 cycles per second. It is one of the better-characterised rhythms in the brain, long associated with motor control, and more recently with something more abstract.

In the predictive-processing account that has become the dominant framework for thinking about perception, the brain is not passively receiving sensory input but continuously generating predictions about what it expects and comparing them against what arrives. Beta activity is widely interpreted as carrying the top-down part of that exchange — the maintenance of the current model — while faster gamma activity carries the mismatch signals that update it.

That interpretation is what makes the finding more than a correlation with a number. If beta activity reflects how firmly the brain holds and revises its own model of the world, a weakened beta signal is a weakened model, and the symptoms that follow would not be a list of separate faults but consequences of one.

Senior author Dr Lena Palaniyappan, a professor of psychiatry at McGill and director of the Centre of Excellence in Youth Mental Health at the Douglas Hospital, put it in those terms. “This gives us early but compelling evidence that measuring the brain’s operations, especially how it builds and revises its own model of the world, may help us appreciate that there is a common thread behind the confusing array of symptoms we see in disorders like schizophrenia,” he said.

Fifty people, one task, several measures

The study involved 25 people with schizophrenia and 25 healthy controls. Participants completed a task involving the timing of sights and sounds while their brain activity was recorded, and their speech was separately analysed for disorganisation.

Patients showed weaker and slower beta-band activity than controls. More informatively, the patients with the largest differences in that signal tended to have more trouble processing the timing of sights and sounds, more disorganised speech, and more severe clinical symptoms.

The multisensory timing task is a good choice of probe. Judging whether a sound and a flash occurred together is exactly the kind of judgment that requires the brain to construct a model of events in time, and it fails in schizophrenia in ways that are measurable and that patients do not have to report subjectively.

Linking the signal to speech disorganisation is the more unexpected connection, since language and audiovisual timing are not obviously the same function. A shared dependence on predictive machinery would explain it: speech production requires maintaining a plan across a sentence, and disorganised speech is what a failure of that maintenance would sound like.

The size of the study

The researchers are unusually direct about the limits, and the number is the reason.

“This is a first, modest-size study, and the associations with speech and clinical severity will need to be tested in larger independent samples before we can be confident in them,” Wei said.

Fifty participants is enough to detect a group difference in a strong physiological signal and thin for correlating that signal with three separate clinical measures. With 25 patients, a correlation has to be large to reach significance, and large correlations estimated from small samples are the ones most likely to shrink on replication. The release reports no correlation coefficients, so there is no way to judge how strong these associations are.

Nor can a cross-sectional study establish direction. Whether the weakened beta signal causes the symptoms, results from them, or reflects a third factor — including antipsychotic medication, which the release does not mention — is not something this design addresses.

What it would be useful for

Two applications are proposed, and they are at different distances.

The nearer one is identification. “The very first sign of schizophrenia begins in our adolescence, but this often gets missed for a long time,” Palaniyappan said. “If this finding is confirmed in larger samples, we may have a potentially useful method to identify who, among youth seeking help for mental health, needs extra support to prevent psychosis.”

The clinical problem he is describing is real and difficult. Most young people presenting with distress and unusual experiences do not go on to develop psychosis, and the existing tools for identifying who will are poor enough that services end up either over-treating or waiting. A physiological marker would change that calculation — though a marker validated on people who already have the diagnosis is not yet a marker that works prospectively on people who might.

The further application is therapeutic. Beyond identifying people at risk, Wei said, the brain signal may also be worth exploring as a future treatment target. Beta oscillations can in principle be modulated — by non-invasive brain stimulation, among other approaches — which is what makes a signal more appealing as a target than a structural finding.

The data were collected at the McConnell Brain Imaging Centre at the Montreal Neurological Institute-Hospital. Collaborating institutions include McGill University, the McConnell Brain Imaging Centre, the University of Montreal, the Research Centre of the University of Montreal Health Centre and the Douglas Research Centre; the release does not name the study’s other authors or state its funding sources.

---

Notes

1. Wei, H., Palaniyappan, L. “Neurophysiological dynamics related to perception, action, and language in schizophrenia.” Science Advances, 2026. DOI: 10.1126/sciadv.aeh5819.
2. McGill University. “Several schizophrenia symptoms linked to a single brain signal.” EurekAlert!, 22 October 2026. https://www.eurekalert.org/news-releases/1148865.