THE HYPE INDEX · JULY 30, 2026
A plastic spoon in your brain is a very good sentence. It is vivid, it is frightening, and it is the reason most people have heard of this study. It is also not in the study.
THE CLAIM
There is a plastic spoon's worth of microplastic in your brain, and it is causing dementia.
Everywhere since February 2025. On podcasts, in wellness marketing, in headlines. Usually delivered as two facts. It is one measurement, and one thing the researchers specifically refused to conclude.
VERDICT
Half True
Something real is being described. It is just being stretched much further than the evidence goes.
HYPE INDEX
53 / 100
The higher the score, the further the claim has travelled from what the evidence supports.
The plastic is really there and the study is real. The spoon came from an interview, and the study's own words rule out the dementia claim built on top of it.
THE NUMBER
0
Times the words “plastic spoon” appear in the published study. The comparison was made by one of the authors, to reporters, on the day it came out.
The spoon is not in the study
This is the fact everything else rests on, and it takes one search of the text to check.
The study measures how much plastic is in a gram of brain tissue. It never multiplies that by the weight of a whole brain. It never arrives at a number of grams, or an object, or a spoon. Search the paper for the words and you get nothing.
The spoon came from the study's senior author, talking to journalists on publication day. That is the version that travelled. It does not appear in the university's own press release either.
Turning a laboratory measurement into a household object is a real step with real assumptions behind it, and it is the step that made this study famous. It is worth knowing that step happened in an interview rather than in the reviewed research, because the review process covered one and not the other.
What was actually measured
Take the spoon away and there is a specific set of numbers underneath.
Researchers examined brain tissue from people who had died, comparing samples from 2016 with samples from 2024. Roughly 20 to 28 people at each point. They found about 50% more plastic in the 2024 samples than the 2016 ones, and far more in brain tissue than in liver or kidney.
Those are the findings. They are genuinely interesting. They are not a spoon.
The study also carries a correction from March 2025, covering duplicated images and a missing step in the methods. Corrections happen and this one was properly published. We mention it because we would want it mentioned about us.
The dementia part, and the sentence printed right underneath it
This is where the claim goes wrong, and the study says so itself.
Twelve brains from people diagnosed with dementia showed far more plastic than any of the others. That is a big difference, and it is why this became a dementia story.
Immediately after reporting it, the authors write that dementia damages the brain's ability to clear things out, and that they would expect that damage to raise these readings on its own. Then they state plainly that no causation is assumed. The conclusion repeats it.
Read that again, because it is unusual. The researchers are not just declining to claim that plastic causes dementia. They are naming the opposite explanation and saying it is the one they would expect: a brain already damaged by dementia clears plastic less well, so more builds up. Same finding, arrow pointing the other way.
There is also no matched comparison group for those twelve brains. They are measured against samples collected differently. That is a limitation, not a scandal, but it is why twelve people are carrying a great deal of weight here.
Other scientists dispute the measurement itself
This is not a fringe objection. It was published in the same journal.
In November 2025, nine researchers across five countries published a formal challenge in Nature Medicine, arguing the study lacked adequate contamination controls and validation.
The technical worry is worth understanding, because it is simple. Brain tissue is very fatty. The test used here heats the sample and identifies what comes off. Fat breaking down can produce fragments that look a lot like the specific plastic the study reports finding most of. So the reading might be picking up fat rather than plastic. The authors say their preparation method reduces that risk.
The authors replied the same day, describing the exchange as being about the limits of even the best current methods. That is a fair description, and it is also an admission that the headline number is not settled.
One more thing. The check often described as independent confirmation was five samples re-run by people who are co-authors on the same paper, using the same technique. That tests whether the method repeats. It does not test whether the method is measuring the right thing.
What holds up, and it is not nothing
The verdict is Half True because half of it is true.
There is microplastic in human brain tissue. That is not in doubt. In 2026 a completely separate group in China found it in the large majority of samples they looked at.
The rise between 2016 and 2024 matters a great deal if it holds up. And the dementia link is worth chasing hard, whichever direction it eventually points.
None of that needs a spoon, and none of it needs the word “causing.” The honest version is this: we are accumulating plastic in our bodies, nobody yet knows what it does to us, and the one team that looked at dementia said plainly that they were not claiming it causes anything.
What would change our mind
Said in advance, so we can be held to it.
If an independent laboratory measures the same thing using a different method, with proper controls for the fat problem, and gets a similar number, this score drops.
And a study designed to separate the two directions, by measuring people before dementia begins, would settle the part that actually matters. Until one exists, anyone telling you plastic causes dementia is going further than the people who did the work.
How it scored
We score every claim on the same five things, worth 20 points each. They add up to the number at the top. If you disagree, you can point at the exact line you think is wrong, which is the entire point.
What we check | Score | Why |
|---|---|---|
Source quality | 3 | Peer reviewed in a leading medical journal and free to read. The research is not the problem here. |
Sample and method | 15 | Around 52 brain samples in total, from people who had died, across two years only. The dementia group is twelve people with no matched comparison group. |
Independence | 3 | University researchers with nothing to sell. Nobody here profits from the answer. |
Replication | 14 | No independent lab has reproduced the amount using a different method. The check often cited as confirmation was done by co-authors using the same technique. |
Drift | 18 | The study measured how much plastic is present. The claim asserts what the plastic is doing, which the researchers specifically refused to conclude in the same paragraph. |
Hype Index | 53 | Half True. Something real is being described, stretched much further than the evidence goes. |
WHO SCORED THIS
Five different AI models, built by five different companies, read the sources listed below and scored this claim separately. They did not see each other's answers, and they did not see ours. We ran it three times.
THE FIVE MODELS SAID
53 / 100
THE EDITOR SAID
58 / 100
Five points apart. We publish the models' number and keep the editor's on the record, so that when this call resolves in January you can see which one was closer.
All three runs came back at exactly 53, and the five models landed within 7 points of each other. That is unusually tight. Five machines agreeing is not the same thing as five machines being right, so we show you the spread either way.
The strongest objection, printed word for word. One model argued the score should be lower, because some reputable coverage does describe microplastics as a possible risk factor rather than a proven cause, which puts the popular version closer to the real scientific conversation than to pure sensationalism.
Every model's individual score and reasoning is on the edition page. The code that runs the scoring is public at github.com/marklynd/quorum.
OUR CALL · WE FIND OUT IN JANUARY
By 31 January 2027, no independent laboratory will have published a measurement of plastic in human brain tissue anywhere close to the amount this study reported.
We are right if no such published figure exists from a team unconnected to the original authors by that date. We are wrong if one is published, whatever it shows, because the point of the call is whether anyone else can measure this at all. Reports that simply find plastic present, without measuring how much, do not count either way.
It gets marked right or wrong on that date either way, and it stays on the record. See every call we have made.
What to do with this
If someone tells you plastic is causing dementia, ask whether the study said that. It says the opposite, in the same paragraph that reports the finding.
Keep two questions apart. Is plastic in there? Yes. Is it hurting you? Nobody knows yet, and the researchers say so themselves.
When a lab measurement arrives as a household object, ask who did the conversion. Here it was an interview, not the study.
Where the numbers came from
The study itself, free to read Nature Medicine, February 3, 2025
The published correction Nature Medicine, March 31, 2025
Nine researchers formally challenging the method Nature Medicine, November 13, 2025
The authors' reply to that challenge Nature Medicine, November 13, 2025
Independent experts reacting on publication day, including the fat problem Science Media Centre, February 3, 2025
Reporting that traces the spoon back to author interviews The Transmitter, February 25, 2025
The university's own announcement, which does not mention a spoon University of New Mexico, February 28, 2025
A separate group finding microplastics in brain tissue Nature Health, April 20, 2026
Every claim we have scored, with its full breakdown and its call, is at thehypeindex.com/record. Think one of the five scores above is wrong? Tell us which one and why. Every challenge gets a published answer, including the ones we turn down.
Mark Lynd, Editor