A spinal fluid test that spots Parkinson's before the symptoms
A test that detects misfolded alpha-synuclein in spinal fluid identified Parkinson's in 88% of people who had it, and was positive in most people who had warning signs but no diagnosis. It is the closest the field has come to a biological test for the disease.
For sixty years, diagnosing has come down to a doctor watching how you move. There is no blood test. There is no scan that settles it. A neurologist takes a history, examines you, watches how you respond to medication, and makes a judgement — and in a meaningful share of cases, that judgement turns out to be wrong.
This study is a serious attempt to change that. Researchers ran a laboratory test on from more than a thousand people and asked a simple question: can we detect the disease itself, rather than the symptoms it causes?
What the researchers were trying to find out
Parkinson's is defined biologically by clumps of misfolded protein inside nerve cells. Until recently, the only way to confirm those clumps were present was to examine the brain after death.
A newer laboratory technique, the , offers a way around that. The principle is borrowed from COVID-era PCR testing. You take a sample of a person's spinal fluid, add a supply of normal alpha-synuclein protein, and wait. If even a trace of the misfolded form is present in the sample, it acts as a seed: it converts the added protein into the same wrong shape, over and over, until the clumping becomes detectable by an instrument. A tiny signal is amplified into an unmistakable one.
The question was whether this worked reliably enough, in enough people, to be useful.
What they did
The team used samples from the Parkinson's Progression Markers Initiative, a long-running international study that has been collecting fluid, scans, and clinical data from the same participants for years. That backdrop matters: these were not anonymous samples, but people whose diagnoses and symptoms were documented in detail.
They ran the assay on 1,123 participants, of whom 1,104 gave a clear result, across several groups:
- people with a clinical diagnosis of Parkinson's
- healthy volunteers of similar age
- people carrying Parkinson's-linked gene variants such as and , some with symptoms and some without
- people in the stage — no diagnosis, but warning signs such as or a lost sense of smell
Each sample was scored positive or negative, and the results were compared against what was already known about each person.
What they found
| Group tested | Share who tested positive |
|---|---|
| Diagnosed Parkinson's, overall | 87.7% |
| Diagnosed Parkinson's, with reduced sense of smell | 98.6% |
| Healthy volunteers | 3.7% |
| Carriers of LRRK2 variants with Parkinson's | 67.5% |
| Warning signs but no diagnosis | 86% (44 of 51) |
| Gene carriers with no symptoms at all | 8% (25 of 310) |
Three things stand out.
The test is accurate in typical Parkinson's. For people with the usual presentation — particularly those who had also lost their sense of smell — it was positive 98.6% of the time. Fewer than four in a hundred healthy volunteers tested positive.
It sees the disease before diagnosis, but only once warning signs have started. Of 51 people who had or a lost sense of smell without a diagnosis, 44 tested positive. Their brains already carried the biological signature.
That is different from simply carrying a risky gene. Among 310 people who had a Parkinson's-linked gene variant but no symptoms of any kind, only 25 — 8% — tested positive. The assay is not a crystal ball for anyone with a family history. It appears to turn positive somewhere between carrying risk and showing the first subtle signs, which makes when it flips the interesting question.
Genetic Parkinson's behaves differently. Roughly a third of people with LRRK2-related Parkinson's tested negative despite having the disease. That is not a flaw in the test so much as a finding in its own right: it suggests some people arrive at Parkinson's through a route that does not involve the same alpha-synuclein clumping. "Parkinson's disease" may turn out to be several biologically distinct conditions wearing similar clothes.
Why it matters
The most immediate consequence is for research rather than for the clinic.
Every trial of a faces the same problem: by the time someone is diagnosed, a large share of their -producing cells in the is already gone. Protecting cells that have already died is not possible. If a biological test can identify people years earlier, trials can start earlier, when there is more left to save.
It also helps with a quieter problem. Some people enrolled in Parkinson's trials do not actually have Parkinson's — they have a condition that resembles it. Those participants dilute the results and can make a working drug look useless. A biological entry test would tighten the enrolment.
What this doesn't tell us
This was a carried out at a single point in time, not a trial. Several limits are worth holding onto.
It does not measure severity or predict speed. A positive result says misfolded protein is present. It does not say how fast anyone will progress, or how badly they will be affected.
It does not prove the prodromal participants will all develop Parkinson's. They have the biological marker and known warning signs. Following them forward over years is the only way to learn what share actually convert, and how long it takes.
And that group was small. Fifty-one people is enough to be interesting and not enough to be certain: shift a handful of results either way and 86% becomes 78% or 94%. The headline number deserves less weight than the direction it points in.
And it does not yet work outside a research setting. The people running these assays were specialists using standardised samples. Reproducing that in ordinary hospital laboratories is a separate undertaking.
What to watch next
Three threads are worth following.
The first is the move away from lumbar punctures. Groups are working on versions of this assay that read skin biopsies, blood, and nasal swabs. A blood-based version would change the picture entirely.
The second is longitudinal follow-up. The prodromal participants who tested positive are being tracked. What happens to them over the next five to ten years is the real test of whether this predicts anything.
The third is how trials adapt. Expect to see new studies that require a positive assay to enrol — and expect some earlier failed trials to be reconsidered in light of who was actually in them.
The study behind this issue
Follow the links to read the original papers. Some journals charge for access; abstracts are usually free.