We work in neurostimulation and sleep, and collaborate with Alzheimer's researchers.
The Amyloid hypothesis is not disproven, it is still ONE of the primary candidates for AD.
The problem with any Alzheimer's research is that the disease is still not well understood. It is likely that multiple diseases are being bundled in as a single disease. The tests for AD, are somewhat rudimentary. Beginning with psychological and neurological tests, the blood work to rule out other conditions, followed by a PET scan to look for brain atrophy, and CSF measures for amyloid and tau levels.
It seems almost like they're basically ruling out any disease we can actually measure for and then if it isn't one of those, it's AD.
Does this mean the Amyloid hypothesis is wrong? Unlikely. Is it incomplete? Absolutely!
But articles shouting that all the research should be thrown out are not helpful .
The AD community know that they don't understand the disease, and though therapeutics have been mostly focused on amyloid and tau, research into how the disease works continues.
> But articles shouting that all the research should be thrown out are not helpful .
Good thing OP isn't one, then. In fact it makes a pretty similar point: all the non-amyloid research also should not have been thrown out. Or rather, killed before it got that far; you can decide whether that's equivalent or worse.
Could the Amyloid/tau hypothesis be a cart before the horse situation? It is my understanding that the current hypothesis is that the buildup of these proteins causes Alzheimer's. Could it be that Alzheimer's causes these proteins to build up?
Its not the only disease this has happened to either. ME/CFS has been railroaded by European governments that funded only psychological research despite numerous pathological findings and better theories of the disease, this prejudiced treatment started in the 1970s and persists to this day including the corruption of the PACE trial results which researchers tried to hide the data of.
For a long time, fibromyalgia was only diagnosed by ruling out everything else. There was a lot of 'It's all in your head.' Last year they developed an actual blood test for it that detects the wonky immune system response.
One infuriating thing about PACE is that even the fraudulent results only showed a 22% recovery rate.
For a disease as serious as ME/CFS, a treatment with a 22% recovery rate is far from good enough. Even if PACE stood up to scrutiny it wouldn’t have made sense to give up on finding better treatments.
This argument gets invoked a lot when it comes to medical dishonesty, but I really don't think it applies in the case of ME/CFS. If we could find the pathology behind the condition, there is huge money to be made in pharmaceutical treatments. Just look at the enormous amount of money being made treating auto-immune illnesses with Humira/Skyrizi/Xeljanz/etc, treating diabetes with GLP-1 agonists and CGMs, and treating obesity with GLP-1 agonists (and depression before all that!). Sometimes treating the chronically ill is the most profitable option.
he was not in the trenches in 2003. in 2003 we were working in an Alzheimer's lab and everyone in our lab at least was expressing suspicion that there was something wrong with the hypothesis. pretty much every internal lab meeting started with "the amyloid hypothesis is... [this statement exists because our funding stream] though it is not conclusively proven, wink wink"
Granted, I have heard a scientist (in an unrelated field, I think it was some astronomical NPR interview) describe science as a bit like a supertanker...there usually is some prevailing thing that everyone believes in, but as contrary evidence piles up, the direction slowly turns.
I guess my own question is whether Alzheimer's/amyloid thinking was atypically stuck on one hypothesis, vs. is this just the slow pace of progress as usual for a given field? I mean...it's not like the amyloid deposition isn't there...
I only play a AD expert on TV (haha I jest...I like to say this because I had no intention of specializing in this when I was training but, hey, in the real world, you have to treat the "market" that rolls in the door....). I work more in the Parkinson's world, and while I would say there are cliques, which do affect who gets NIH (or used to get...I have no idea what's going on there now...), I can't say there's one prevailing "cabal" that's obsessed with any one direction. the bigger issue is that current Parkinson's research is a bit scattershot in too many directions.
My other pet peeve is somewhat unrelated, where the article mentions other directions like neuroinflammation and oxidative stress; the problem is these are also vague/broad topics, that have been thrown around like panaceas for every disease from head to toe; my own superstition is that when a new drug candidate comes out for "neuroinflammation" or "oxidative stress", I'd bet a healthy bunch of nickels it won't amount to much.
His blog has a category just for discussing Alzheimer's, and he's been talking about ever since he started blogging. So here's a post from 2002 where he points out (somewhat obliquely) that amyloid isn't a proven hypothesis: https://www.science.org/content/blog-post/alzheimer-s-vaccin...
> This is looking like one of the crazy ideas that just might work - stipulating, for the moment, that amyloid really is the cause of Alzheimer's. . .
Which is the only good take… if there wasn’t “something promising” with the idea, people wouldn’t have kept at it. There’s only so much time you’re going to waste on pure fantasy before you move on. The whole ordeal had me quite upset as I was supporting studies looking at this, and the patients are beyond desperate for any treatment. None of us got rich.
Every degenerative CNS disease is caused by death of non replicating cells. By definition, the cure of such diseases is equivalent to immorality. This already highlights the importance and difficulty of such achievement.
The amyloid hypothesis seems too simple and superficial to account for a decades long process (it is speculated that Alzheimer's starts up to 20 years before symptom onset). The ultimate problem is to find the underlining cause(s) and not correlations.
sunk cost, stockholm syndrome, refusal to change mental models, "it is hard to get a man to understand something when their salary depends on it not being true"
amyloid experiments, especially biophysics experiments, are really fucking hard and really fucking expensive. like, months of hundred hour weeks hard. then it usually fails. so when you burn years of this cycle to find a consistent result (artefact or not), its crazy hard to let go of it.
if you want a very rare example of someone who did the right thing:
45 citations in 10 years. how many amyloid researchers do you think are aware of this problem with plasticware being published. and probably most "know" it from lived experience, but they say "fuck it, i need to finish my phd/postdoc" and just scramble to push out any result, plastics being problematic be damned.
the only thing that could remotely stop the nonsense from continuing is if the NIH had the balls to take this result and issue guidance to halt all amyloid research that doesnt take plastic use into account. but 1) the program managers are not smart enough to do something like that and 2) such a deep challenge to the research agenda would shake the system way too much from the top AND the bottom. just easier to continue being a bureaucrat with a nice salary and a very "mid" approach to science.
There's a lot of researchers in a lot of roles and it turns out finding targets is just one of those roles, the rest focus on optimizing therapies against those targets and they like having well defined targets to work on.
Anti-science types keep holding this up as some kind of 'gotcha' or as a waste, but in the end it shows that the scientific method and the scientific establishment work efficiently.
Despite even intentional fabrication, the truth of it was found in a few years and the field marches on.
It was published and cited. 45 in 10 years is a reasonably okay citation count. What's the problem?
Science is slow and consensus based. Ideas are put forth, tested, and replicated. Eventually a consensus is formed. If enough new contrary evidence is collected a new consensus can form. The article you linked is not consensus changing, but it adds weight on the scale of change, and clearly it worked. Mission accomplished.
Science does not turn on a dime. It took many decades for Lynn Margulis's endosymbiosis theory to catch on, but it did. The evidence was eventually undeniable, and the consensus changed.
This AD story is just another slow consensus change. We still don't know exactly how AD works. We still know amyloid is involved somehow. We now know amyloid isn't everything.
The system works. It's just slower than you like it to be.
I guess it depends where you set the starting point. Was it a dead end exacerbated by fabricated data? Yes. Did the system correct itself (relatively) quickly. Yes.
HN (and Silicon Valley) has a contingent of people that want to attack the credibility of the scientific community so that they can present their own (usually very flawed) conclusions as being legitimate.
well that's a funny way to put it. i assure, there is a contingent of the scientific community that would like to attack the credibility of the scientific community to discredit the actually very flawed conclusions of the scientific community.
Something is very wrong and corrupt if a large field of science, an army of scientists spending billions of dollars, is ruled by one easily repeatable study/experiment, and yet nobody cares to repeat it.
Scientists do repeat experiments, they just don't publish the results because no one cares.
Specifically, results are replicated wherever you want to build on something someone has done. For example I'm making specific glass off of a Nature paper for a totally unrelated use.
If the glass doesn't work out will I publish my results? No, too much work to get rejected.
>Specifically, results are replicated wherever you want to build on something someone has done.
yes, you want to test your Alzheimer treatment, and you measure the amyloid in the patients before the treatment, and you don't find it in some patients... Are you not going to publish that?
Nobody ever got tenure for repeating an existing experiment and getting the same results. It is a problem of incentives. Aside from doing something expensive like grant terms including potentially 'drafting' some scientists to attempt to replicate reproducible studies for funding as their next project I'm not sure what could be done to fix it.
Experiments are repeated if someone else is looking to expand on the work actually... it's just that if it doesn't work, no one wants to go to the effort of publishing that since it's a lot of effort and there'll be politics with the original author and there's the possibility that it was just a "push it through to satisfy the metrics" paper that every researcher "allows" every other researcher purely out of sympathy and there's the review process and deadlines and rebuttals and more deadlines and...
I suppose if you only value this entire end to end process then the experiment is not being repeated at all, but yeah, most things just get discarded when they don't work.
I mean it was caught and other scientists will work on the problem more and human knowledge will continue forward. Don’t trust cutting edge science or popsci articles because it’s steering a ship and course corrections happen. But the ship eventually gets on the right course.
The alternative is what? A wrong understanding of Miasma theory?
We work in neurostimulation and sleep, and collaborate with Alzheimer's researchers.
The Amyloid hypothesis is not disproven, it is still ONE of the primary candidates for AD.
The problem with any Alzheimer's research is that the disease is still not well understood. It is likely that multiple diseases are being bundled in as a single disease. The tests for AD, are somewhat rudimentary. Beginning with psychological and neurological tests, the blood work to rule out other conditions, followed by a PET scan to look for brain atrophy, and CSF measures for amyloid and tau levels.
It seems almost like they're basically ruling out any disease we can actually measure for and then if it isn't one of those, it's AD.
Does this mean the Amyloid hypothesis is wrong? Unlikely. Is it incomplete? Absolutely!
But articles shouting that all the research should be thrown out are not helpful .
The AD community know that they don't understand the disease, and though therapeutics have been mostly focused on amyloid and tau, research into how the disease works continues.
> But articles shouting that all the research should be thrown out are not helpful .
Good thing OP isn't one, then. In fact it makes a pretty similar point: all the non-amyloid research also should not have been thrown out. Or rather, killed before it got that far; you can decide whether that's equivalent or worse.
It seems involves some kind of metabolic disturbance, is there any explanation for this? https://academic.oup.com/metallomics/article/9/8/1106/601364...
Could the Amyloid/tau hypothesis be a cart before the horse situation? It is my understanding that the current hypothesis is that the buildup of these proteins causes Alzheimer's. Could it be that Alzheimer's causes these proteins to build up?
It could be. That's why we need more research. Well funded and not vilified.
Its not the only disease this has happened to either. ME/CFS has been railroaded by European governments that funded only psychological research despite numerous pathological findings and better theories of the disease, this prejudiced treatment started in the 1970s and persists to this day including the corruption of the PACE trial results which researchers tried to hide the data of.
Research fraud in medicine is alarmingly common.
For a long time, fibromyalgia was only diagnosed by ruling out everything else. There was a lot of 'It's all in your head.' Last year they developed an actual blood test for it that detects the wonky immune system response.
Out of interest, do you have a link to a paper?
One infuriating thing about PACE is that even the fraudulent results only showed a 22% recovery rate.
For a disease as serious as ME/CFS, a treatment with a 22% recovery rate is far from good enough. Even if PACE stood up to scrutiny it wouldn’t have made sense to give up on finding better treatments.
I mean, 22% sounds pretty damn good if there are no long lasting negative side effects of being part of the remaining 78%.
Like, sure, shoot for 200% cure rate, but even a success rate of 1% cured of a previously unrecoverable situation is insanely informative.
1 reply →
There’s money to be made; don’t let facts stand in the way of profits!
This argument gets invoked a lot when it comes to medical dishonesty, but I really don't think it applies in the case of ME/CFS. If we could find the pathology behind the condition, there is huge money to be made in pharmaceutical treatments. Just look at the enormous amount of money being made treating auto-immune illnesses with Humira/Skyrizi/Xeljanz/etc, treating diabetes with GLP-1 agonists and CGMs, and treating obesity with GLP-1 agonists (and depression before all that!). Sometimes treating the chronically ill is the most profitable option.
8 replies →
Related articles:
- The great brain clearance and dementia debate - https://www.nature.com/articles/d41586-025-00962-y (2025)
- The Devastating Legacy of Lies in Alzheimer’s Science - https://www.nytimes.com/2025/01/24/opinion/alzheimers-fraud-... (2025), referencing the book Doctored https://en.m.wikipedia.org/wiki/Doctored_(book)
- The maddening saga of how an Alzheimer’s ‘cabal’ thwarted progress toward a cure for decades - https://www.statnews.com/2019/06/25/alzheimers-cabal-thwarte... (2019, by the late Sharon Begley)
Derek Lowe's take:
https://www.science.org/content/blog-post/reaction-alzheimer...
he was not in the trenches in 2003. in 2003 we were working in an Alzheimer's lab and everyone in our lab at least was expressing suspicion that there was something wrong with the hypothesis. pretty much every internal lab meeting started with "the amyloid hypothesis is... [this statement exists because our funding stream] though it is not conclusively proven, wink wink"
Granted, I have heard a scientist (in an unrelated field, I think it was some astronomical NPR interview) describe science as a bit like a supertanker...there usually is some prevailing thing that everyone believes in, but as contrary evidence piles up, the direction slowly turns.
I guess my own question is whether Alzheimer's/amyloid thinking was atypically stuck on one hypothesis, vs. is this just the slow pace of progress as usual for a given field? I mean...it's not like the amyloid deposition isn't there...
I only play a AD expert on TV (haha I jest...I like to say this because I had no intention of specializing in this when I was training but, hey, in the real world, you have to treat the "market" that rolls in the door....). I work more in the Parkinson's world, and while I would say there are cliques, which do affect who gets NIH (or used to get...I have no idea what's going on there now...), I can't say there's one prevailing "cabal" that's obsessed with any one direction. the bigger issue is that current Parkinson's research is a bit scattershot in too many directions.
My other pet peeve is somewhat unrelated, where the article mentions other directions like neuroinflammation and oxidative stress; the problem is these are also vague/broad topics, that have been thrown around like panaceas for every disease from head to toe; my own superstition is that when a new drug candidate comes out for "neuroinflammation" or "oxidative stress", I'd bet a healthy bunch of nickels it won't amount to much.
2 replies →
His blog has a category just for discussing Alzheimer's, and he's been talking about ever since he started blogging. So here's a post from 2002 where he points out (somewhat obliquely) that amyloid isn't a proven hypothesis: https://www.science.org/content/blog-post/alzheimer-s-vaccin...
> This is looking like one of the crazy ideas that just might work - stipulating, for the moment, that amyloid really is the cause of Alzheimer's. . .
1 reply →
Which is the only good take… if there wasn’t “something promising” with the idea, people wouldn’t have kept at it. There’s only so much time you’re going to waste on pure fantasy before you move on. The whole ordeal had me quite upset as I was supporting studies looking at this, and the patients are beyond desperate for any treatment. None of us got rich.
Great and highly critical book on this general topic: “How Not to Study a Disease” by Karl Herrup from MIT Press:
https://direct.mit.edu/books/book/5216/How-Not-to-Study-a-Di...
The core problem is much older than stated in this focused review.
Every degenerative CNS disease is caused by death of non replicating cells. By definition, the cure of such diseases is equivalent to immorality. This already highlights the importance and difficulty of such achievement.
The amyloid hypothesis seems too simple and superficial to account for a decades long process (it is speculated that Alzheimer's starts up to 20 years before symptom onset). The ultimate problem is to find the underlining cause(s) and not correlations.
Halting degeneration would be a huge win in any of these disorders, and wouldn't be equivalent to immortality.
This is from 2022
So why do researchers is still hell bent on believing the beta amaloyd theory? I get voted down every time I ask this
sunk cost, stockholm syndrome, refusal to change mental models, "it is hard to get a man to understand something when their salary depends on it not being true"
amyloid experiments, especially biophysics experiments, are really fucking hard and really fucking expensive. like, months of hundred hour weeks hard. then it usually fails. so when you burn years of this cycle to find a consistent result (artefact or not), its crazy hard to let go of it.
if you want a very rare example of someone who did the right thing:
https://scholar.google.com/citations?view_op=view_citation&h...
45 citations in 10 years. how many amyloid researchers do you think are aware of this problem with plasticware being published. and probably most "know" it from lived experience, but they say "fuck it, i need to finish my phd/postdoc" and just scramble to push out any result, plastics being problematic be damned.
the only thing that could remotely stop the nonsense from continuing is if the NIH had the balls to take this result and issue guidance to halt all amyloid research that doesnt take plastic use into account. but 1) the program managers are not smart enough to do something like that and 2) such a deep challenge to the research agenda would shake the system way too much from the top AND the bottom. just easier to continue being a bureaucrat with a nice salary and a very "mid" approach to science.
There's a lot of researchers in a lot of roles and it turns out finding targets is just one of those roles, the rest focus on optimizing therapies against those targets and they like having well defined targets to work on.
Old and outdated.
How so?
Anti-science types keep holding this up as some kind of 'gotcha' or as a waste, but in the end it shows that the scientific method and the scientific establishment work efficiently.
Despite even intentional fabrication, the truth of it was found in a few years and the field marches on.
https://scholar.google.com/citations?view_op=view_citation&h...
45 citations in 10 years for a paper pointing out a major systematic flaw in almost all foundations of amyloid reserch. the system is not working.
It was published and cited. 45 in 10 years is a reasonably okay citation count. What's the problem?
Science is slow and consensus based. Ideas are put forth, tested, and replicated. Eventually a consensus is formed. If enough new contrary evidence is collected a new consensus can form. The article you linked is not consensus changing, but it adds weight on the scale of change, and clearly it worked. Mission accomplished.
Science does not turn on a dime. It took many decades for Lynn Margulis's endosymbiosis theory to catch on, but it did. The evidence was eventually undeniable, and the consensus changed.
This AD story is just another slow consensus change. We still don't know exactly how AD works. We still know amyloid is involved somehow. We now know amyloid isn't everything.
The system works. It's just slower than you like it to be.
4 replies →
> it was found in a few years and the field marches on
I agree but 16 years is still significant. It represents 5% of the modern medicine era.
I guess it depends where you set the starting point. Was it a dead end exacerbated by fabricated data? Yes. Did the system correct itself (relatively) quickly. Yes.
Working as intended.
HN (and Silicon Valley) has a contingent of people that want to attack the credibility of the scientific community so that they can present their own (usually very flawed) conclusions as being legitimate.
I noticed it’s very prevalent here. Lots and lots of academia is dead, useless, fraudulent, etc comments.
well that's a funny way to put it. i assure, there is a contingent of the scientific community that would like to attack the credibility of the scientific community to discredit the actually very flawed conclusions of the scientific community.
[dead]
Something is very wrong and corrupt if a large field of science, an army of scientists spending billions of dollars, is ruled by one easily repeatable study/experiment, and yet nobody cares to repeat it.
Scientists do repeat experiments, they just don't publish the results because no one cares.
Specifically, results are replicated wherever you want to build on something someone has done. For example I'm making specific glass off of a Nature paper for a totally unrelated use.
If the glass doesn't work out will I publish my results? No, too much work to get rejected.
>Specifically, results are replicated wherever you want to build on something someone has done.
yes, you want to test your Alzheimer treatment, and you measure the amyloid in the patients before the treatment, and you don't find it in some patients... Are you not going to publish that?
1 reply →
Nobody ever got tenure for repeating an existing experiment and getting the same results. It is a problem of incentives. Aside from doing something expensive like grant terms including potentially 'drafting' some scientists to attempt to replicate reproducible studies for funding as their next project I'm not sure what could be done to fix it.
Experiments are repeated if someone else is looking to expand on the work actually... it's just that if it doesn't work, no one wants to go to the effort of publishing that since it's a lot of effort and there'll be politics with the original author and there's the possibility that it was just a "push it through to satisfy the metrics" paper that every researcher "allows" every other researcher purely out of sympathy and there's the review process and deadlines and rebuttals and more deadlines and...
I suppose if you only value this entire end to end process then the experiment is not being repeated at all, but yeah, most things just get discarded when they don't work.
gotta trust that science yo
I mean it was caught and other scientists will work on the problem more and human knowledge will continue forward. Don’t trust cutting edge science or popsci articles because it’s steering a ship and course corrections happen. But the ship eventually gets on the right course.
The alternative is what? A wrong understanding of Miasma theory?
No no no.
Trust the Scientist! For they are pure of heart and unswayed by pride, greed, lust, avarice, or envy
Cash rules everything around me.