About that autism and Alzheimer’s article

A Washington Post piece with the startling headline “Surprising links between autism, Alzheimer’s could change how we treat both” had a lot of people asking about its claims, including what the overblown headline implies. The piece avers that “Advances in brain imaging, DNA sequencing and molecular biology are revealing remarkable overlaps between autism and Alzheimer’s, scientists say — in genes, in neural circuitry, even in patterns of disease.” That word “disease” does, of course, firmly place the perspective on autism in this piece within the medicalized deficit-based framing that keeps redirecting research from what autistic people want and need. (you can google the headline if you’re dying to give WaPo a click; I am not, and accessed it through Apple News).

Here’s my tl;dr take on the piece: It's a flimsily constructed just-so story that presents half-truths as whole, especially when the whole truth contradicts the implication that autism and Alzheimer’s disease (AD) overlap in causes and effects, and it establishes a pattern of this kind of writing and reporting from this quarter.

The flimsiness
The flimsiness of the piece arises in part from its reliance on a slim selection of reviews or single studies, some that do not support the implication that autism and Alzheimer’s disease overlap in causes and effects. Linking to studies from a small carousel of not-great journals, the piece implies, despite a wave at some caveats, that one of these publications, for example, offers fresh findings of an association. But the publication is not an original research report with new findings and is clearly labeled as a “hypothesis and theory article.” It yields only a hypothesis that draws a tenuous line connecting the brain’s waste removal system to both AD and autism. It is not evidence of a link, although it may be evidence of a new angle on a grant application.

Readers can review the hypothesis paper for themselves to determine how valid they find the argumentation. Let’s just say it starts with a framing of “autism incidence is rising and so is the global incidence of AD” and builds a hypothesis on a large proportion of publications that themselves review existing studies and offer hypotheses. It is a sort of nesting doll of hypotheses, in fact, with nary a nugget of confirmation at the center. This is not a sufficiently sturdy foundation for the implications of the headline or the claims made in the piece itself.

The WaPo piece is written in five sections: (1) the risk associations of being autistic or of developing AD; (2) genetics; (3) “cellular housekeeping,” related to waste removal from the brain; (4) brain structures such as the amygdala and proteins that form them; and (5) a last wild, hail Mary pass to a protein directly implicated in AD. For all but one of these sections, the piece cites one scientific publication and offers maybe one researcher comment. With a couple of exceptions, the citations are from journal groups that are not widely respected in the scientific community, involve reviews or hypothesis papers that offer no new findings, or, in one case, reflect non-peer-reviewed conference proceedings. This foundation is beyond flimsy, and nowhere near the kind of solid, evidence-based examination needed to support an alarmist claim that would cause stress for millions of people and their families. And stress is among the many factors that do threaten the health of autistic people.

Half-truths
The half-truths are what stand out the most to me. For example, the WaPo piece states that people with AD often experience a loss of the sense of smell before other symptoms emerge, which is true. It then goes on to say that “smell issues have been reported in autism … but dismissed as a sensory quirk rather than as a possible clue to brain health.” That use of the word “issues” is redolent with the deceptive odor of a half-truth. The whole truth about those “issues” is that autistic people can experience either heightened and diminished sense of smell – hypersensitivity and hyposensitivity – and also may register smells just fine but not identify them in the same way as non-autistic people. Sensing the world in a different way is not the same as not sensing it at all. Plenty of people think cilantro tastes like soap and plenty don’t think that, but they can all taste cilantro. This pattern in autistic people is not like the lost sense of smell that can arise before the onset of better-recognized symptoms of dementia, including AD.

Another example of half the story being told is with the claims around tau protein, which has been linked robustly to AD and gained some dominance over falling-out-of-favor amyloid as a disease target. The WaPo article claims that tau and autism could be linked and cites a single study from 2020 as support. It is a mouse study at that. Studies examining tau levels in the blood of autistic people find low levels or typical levels or maybe high levels, or maybe it’s in ADHD instead. In other words, findings are all over the place — they are pattern-free.

Then there are the statements about genes. I could write 10,000 words about that, but I will spare you. The WaPo piece cites a study from a not-great journal group that reported “148 genes in common, many of them tied to the same fundamental processes that shape and sustain the brain over time.” [The research group reporting these findings is a team in Russia that makes a bit of a hobby of publishing in this journal about autism and a protein called mTOR, which isn’t hard to do with a pay-to-play journal.]

All of our genes are “in common,” so it’s no surprise that in looking at human brains, regardless of condition, there would be dozens categorized as being involved in brain-related processes.

A gene is an address for a sequence of DNA. A sequence at that address can have one use, like a single-family dwelling, or it can have many uses, each depending on which part of the sequence is used and how it’s regulated. The cited study involved a look at genes that correlate with the chances of developing AD and autism and found overlap. That’s not a surprise either. It would be odd if two human brain conditions did not overlap to some extent in associations with brain-related genes.

Not to get too wonky, but the biochemical pathway associated with some of these genes involves a very busy protein known as mTOR, which gets its figurative little protein hands on just about every single cellular process in every single tissue in the body. I mean, this thing is everywhere and also is associated with a ton of diseases, including several cancers. It has been linked to a lack of pruning of certain kinds of neurons in autistic brains, so that there are more connections among neurons than is typical, which is directly opposed to the AD pattern. mTOR also is linked to syndromes, such as tuberous sclerosis, that are associated with autism.

What studies like this do not show is specifically how the sequences at these locations are used in each condition. The WaPo piece says that “Many of them are deeply involved in how brain cells connect, signal and adapt over time,” but that “deep involvement” is a two way street or even a roundabout with 20 exits in terms of how the sequences at each of these 148 locations could be used – or misused – to result in different conditions. “Deeply involved” doesn’t mean “has the same role in two conditions.” If there is in fact a role in both autism and AD for any of these genes (and proteins they encode), it is entirely possible that the roles would be opposing. The protein mTOR, for example, can lead to overproduction or underproduction of proteins in the brain. Same gene – different outcomes.

Indeed, one of the genes mentioned specifically in the WaPo pieces, SHANK3, occurs in low levels in AD. It’s been implicated in a different way in autism. The work described (but not linked) in the piece suggests that mice with autism-related SHANK3 changes “are unusually resistant to developing full-blown dementia-like pathology.” One researcher noted: “So you have to double or triple down in introducing bad things into the mouse brain to even get something that looks like Alzheimer’s in a mouse.” That does not at all fit with the implications in the headline or what much of the rest of the piece tries to imply.

Even when fuller, more tempered pictures are presented, the part that argues against the premise of the piece is given almost as an afterthought. There is, for example, the arresting reference to shrinking of the amygdala in AD in a paragraph that seems to link this change to autism, as well. The amygdala is central to our socio-emotional processing, among other things. Although the piece says that the “amygdala is often enlarged” in autistic people, it lands on “findings vary by age and study design” as a way to keep autism in the “shrinking amygdala” zone with AD.

Those are the article’s “remarkable overlaps in genes, in neural circuitry, even in patterns of disease.” Perhaps my bar for remarkable is remarkably high, but none of this leaves an impression of “holy wow, these two things are remarkably related!”

If the reference to autism as a “disease” does not make the lens on this piece thoroughly clear, this should: “… the implications are profound: Both conditions remain mysterious and difficult to treat, and studying them together may open new paths for intervention.” And if the “ooooh, mysterious autism” and “disease to treat” tone doesn’t clarify the lens sufficiently enough for you, the WaPo reporter who wrote this piece also wrote, “Is autism preventable in certain cases after all? Some scientists say yes.” We dealt with that at the time for its eugenicist, thinly sourced attempt to “[stretch] these sources as far as it can to try to make the claim in its headline look substantiated, but readers, there is very much no there there.” This writer also is responsible for “breaking” the very old “news” about “four types of autism,” which I covered here (I wrote: “The WaPo  piece  claims the study was published with ‘little fanfare,’ although back in the summer, it was covered by little-known outlets such as  Scientific American,  Reuters,  The Daily Mail, and  New York Post, among  dozens of other publications. WaPo must have a very different definition from mine of ‘fanfare.’”)

Readers, I am sensing a pattern here, in keeping with the practices of my people.

And finally, there are smaller things about the article, such as making statements of fact that obscure greater nuance. The piece states that autism was “first formally recognized as a distinct diagnosis in 1980,” apparently to support the premise that few older autistic people have been available for studies of AD and aging. It’s a strange statement, as Leo Kanner defined the condition in 1953, a study in 1966 even quantified its prevalence, and it was listed in the DSM II (published in 1968). The 1980 marker is when it was separated completely from schizophrenia in the DSM III to stand on its own as a diagnosis. But, I mean, Temple Grandin was diagnosed as autistic in her teens because her mother used a diagnostic checklist, and Grandin is 78 years old.

The piece also falls into presenting opinion as facts. It frames “older” autistic people as largely absent from the data, citing a study from 2025 and saying that “just a tiny fraction of the more than 40,000 autism papers published between 1980 and 2021 included people over 50.” It does go on to say that “the number of studies about aging with autism is growing rapidly,” but it doesn’t quantify that statement, probably because there has been an almost 400% increase in research involving older autistic people since 2012 — which is 16 years of studies. And the WaPo piece omits the study authors’ own conclusion that:

many research areas … need addressing. Specifically, more research is needed on social isolation and the practicalities of living arrangements for older autistic people, as well as more studies including older autistic adults with intellectual disability.

For context, social isolation is a risk factor for dementia. I can’t help but wonder if this omission is a convenient one, as it does not serve the hand-waving narrative of the piece — and the sleight-of-hand away from some obvious factors involved in the higher rates of dementias among autistic people. Even the mention of the studies showing that association elides the fuller picture. Among autistic people without intellectual disability, those with co-occurring intellectual disability, and those with intellectual disability only, autistic people without intellectual disability had the lowest risk, and non-autistic people with intellectual disability had the highest risk. Clearly, there are factors in play here that aren’t related to autism.

In fact, reverting to my comment about unnecessarily raising stress levels for autistic people and their loved ones: So many things are risk factors for dementia and occur more often in autistic people, but they aren’t biologically baked in. They are social factors acting through human biology to affect health in autistic people and in people who develop dementia. These factors relate to stress, sleep, diet, metabolic condition, seizures, communication barriers, access to physical activity, social isolation, medication effects, and more. If we are going to have people digging into the “links” between autism and AD, what deserves a close look are these factors that are clearly associated with AD risk and more prevalent among autistic people – and then ways to mitigate these factors to improve the lives and health of autistic people.

News you can use

• TPGA editors Carol Greenburg and Shannon Rosa will be in Prague this week, covering INSAR. It is always ‘interesting’ to see the disconnect between what autistic people think autism research should be about, and what autism researchers who aren’t autistic or who don’t listen to autistic people think autism research should be about.
• Relatedly: TPGA will be doing AutINSAR, a Bluesky discussion about autistic priorities in research, again on Saturday, April 25. Here’s last year’s AutINSAR for context.
• Also interesting: Last year many researchers from outside the United States either didn’t want to come or were unable to secure funding to travel to Seattle for INSAR 2025. INSAR is not scheduled to return to the U.S., not through 2029 at least (look under “Future Annual Meetings”).
• Researchers with a great deal of investment in the idea that immunity and autism are related have published results of a small study of 36 autistic children and 18 non-autistic children indicating a difference between the two groups in a specific kind of immune cell known as a T regulatory cell. They report that some of these differences are in turn linked to gastrointestinal symptoms in the autistic group.
• The same very invested team in a separate study reported that in a rodent model, popping ‘adoptive’ versions of these cells into pregnant animals resulted in effects in their male but not their female offspring.

New at TPGA

Grief Changes an Autistic Person

A particular kind of grief follows me now that I am on a solitary path, a sorrow as spiky as my neurotype.

People you should know

• Here’s a Q&A with autistic advocate Jade McWilliams, who notes:

  An autistic child who grows up in a culture of acceptance, including in their family, has an easier time developing positive self-understanding, making friends, and developing resilience. As that child becomes an adult, they are better equipped to self-advocate.

• From the editors: Last week we featured AAC-using author Woody Brown. This week, we give another side-eye to the supposedly pro-communication advocates who claim to be “protecting” non-speakers and AAC users, while actually working against their interests — as once again happens in The Atlantic’s new piece on Brown. Do better, please.

Thanks for reading, and here’s to developing positive self-understanding, making friends, and developing resilience.

Got something autism-related to share with us? Send it along to editorial@thinkingautism.com.

Got a comment? We’d love to hear from you, so drop us a line below. Please note that comments are moderated per TPGA guidelines.