Maria’s mother has dementia. She forgets names. She repeats questions. But lately, she’s also withdrawn, angry, and refuses to eat. Her doctor isn’t sure if it’s the dementia getting worse or something else.
Millions face this confusion. Over half of people with cognitive impairment develop neuropsychiatric symptoms like depression, anxiety, or agitation. These are hard to diagnose. Patients often can’t explain how they feel. Families struggle to tell what’s what.
Right now, doctors rely on observation and guesswork.
But brain activity might hold the answer.
A signal hiding in plain sight
For years, scientists have studied brain waves using EEG. These tests track electrical activity across the brain. Most focus on seizures or sleep. But a quiet signal called “microstates” has been overlooked.
Think of microstates as the brain’s default radio stations. They switch rapidly, like channels flicking every fraction of a second. Each one controls a different mental task. One handles attention. Another manages emotion.
In healthy brains, these stations switch smoothly. In dementia, the signal gets fuzzy. Some stations stay on too long. Others vanish.
Now, researchers have found four key microstate patterns that act like fingerprints for mental health symptoms in dementia.
Four brain wave clues
The study looked at 78 older adults with cognitive impairment. Some had mood or behavior symptoms. Others did not. All had EEGs.
Scientists analyzed 36 different microstate traits. Four stood out.
The most important was the duration of Class C microstate. It stayed active longer in patients with neuropsychiatric symptoms. The other three were transition patterns between brain states—how the brain jumps from one station to another.
When these four markers were combined, they formed a powerful signal.
The model worked almost perfectly
A computer model used these four traits to tell who had mental health symptoms.
It was right 90% of the time.
Even better, it never missed a case. Sensitivity was 100%. That means every patient with symptoms was correctly flagged. Specificity was 77%, meaning a few people without symptoms were flagged by mistake.
That’s far better than current tools, which often miss these issues or confuse them with dementia itself.
This doesn't mean this treatment is available yet.
But there’s a catch.
The study was small. Just 78 people. And it was a snapshot in time, not a long-term test. The results need to be confirmed in larger, more diverse groups.
Also, the model hasn’t been tested outside the lab. EEGs are common, but analyzing microstates requires special software and training. Most clinics don’t have that.
Still, experts say this opens a new path.
A new window into the mind
Right now, we treat the brain like a black box in dementia care. We see behavior and guess what’s inside.
This method lets us peek inside without scans or spinal taps.
It could help doctors tell if a patient is depressed or just more forgetful. That changes treatment. Depression can be managed with therapy or meds. Agitation might need different support.
Right now, many go untreated because symptoms are mistaken for dementia progression.
Not ready for your doctor’s office
You can’t get this test today. It’s still in research labs.
But the tools are close. EEG machines are already in many hospitals. The next step is building user-friendly software that can run this analysis automatically.
Researchers plan to test the model in bigger studies, including people with mild cognitive impairment and different types of dementia.
A step toward personalized care
Dementia is not one disease. It affects people differently. Some lose memory first. Others lose mood control.
This approach could help tailor care to the individual.
Instead of guessing, doctors could use brain wave data to guide treatment.
That day is still a few years away. But for families like Maria’s, it offers hope.
One day, a simple EEG might answer the question: Is it the dementia—or something else?
And that could change how we care for millions.