"Long-held assumptions"?
« previous post | next post »
"New Autism Data Challenge Long-Held Assumptions", MedPage Today 10/3/2025:
Autism diagnosed during early childhood had a distinct genetic and developmental profile compared with autism diagnosed later, a large analysis of multiple cohorts showed.
[…]
The results challenge a long-held assumption that autism has a unified underlying cause.
The cited research paper is Xinhe Zhang et al., "Polygenic and developmental profiles of autism differ by age at diagnosis", Nature 10/1/2025. The abstract ends
These findings indicate that earlier- and later-diagnosed autism have different developmental trajectories and genetic profiles. Our findings have important implications for how we conceptualize autism and provide a model to explain some of the diversity found in autism.
The paper's evidence and its conclusions seem solid. But the MedPage article's reference to "a long-held assumption that autism has a unified underlying cause" is misleading, starting with the implication that "autism" is a category with well-defined, stable, and generally-accepted boundaries.
Several years ago, I quoted from a paper by Laurent Mottron, "A radical change in our autism research strategy is needed: Back to prototypes", Autism Research 6/2/2021, whose abstract starts this way:
The evolution of autism diagnosis, from its discovery to its current delineation using standardized instruments, has been paralleled by a steady increase in its prevalence and heterogeneity. In clinical settings, the diagnosis of autism is now too vague to specify the type of support required by the concerned individuals. In research, the inclusion of individuals categorically defined by over-inclusive, polythetic criteria in autism cohorts results in a population whose heterogeneity runs contrary to the advancement of scientific progress. Investigating individuals sharing only a trivial resemblance produces a large-scale type-2 error (not finding differences between autistic and dominant population) rather than detecting mechanistic differences to explain their phenotypic divergences.
And it's easy to find similar concerns expressed across previous decades.
A few relevant past posts:
"Translating 'phenotypically diverse'", 5/12/2020
"'Reliability is confused with truth'", 6/26/2021
"Intonation in 'human emulation mode'", 5/9/2021
"Grouping-think", 6/9/2022
"RFK Jr on Autism", 4/18/2025
A relevant recent skeet:
The mystery of medical diagnosis!
— Richard Kadrey (@richardkadrey.bsky.social) October 3, 2025 at 2:25 PM
Of course,"stars" are actually a natural kind, though some of the things we see in telescopes turn out to be galaxies and so on.
wgj said,
October 4, 2025 @ 8:23 am
Indeed it would be more correct to describe it as a "never-existed assumption", at least in terms of scientific consensus.
J.W. Brewer said,
October 4, 2025 @ 10:25 am
The phrase "population whose heterogeneity runs contrary to the advancement of scientific progress" has a weird, and I assume unintended, victim-blaming implicature. It's not like most of the individuals involved were trying to obstruct scientific progress! They by and large* aren't the ones responsible for their not being given more fine-grained diagnoses.
*There may be some countervailing tendency where "activist" types interested in promoting a sense of commonality and "community" among affected individuals are incentivized to promote a lumper rather than splitter approach, because that inherently yields a single larger "community" rather than a range of smaller but more precisely defined "communities."
Mark Liberman said,
October 4, 2025 @ 10:35 am
@J.W. Brewer "They by and large* aren't the ones responsible for their not being given more fine-grained diagnoses.":
The phrase "more fine-grained diagnoses" has an unfortunate, and I assume unintended, implicature — namely that the the overall category is a scientifically appropriate "natural kind", just needing to be further sub-divided.
In this case, that may be analogous to a belief that the four classical elements are a well-founded place for physical chemistry to start, just needing further subdivision…
J.W. Brewer said,
October 4, 2025 @ 11:29 am
That's a fair point and a different word than "fine-grained" might have been usefully vaguer. I suppose I'm not even entirely sure what "natural kind" would mean in this sort of taxonomy, since I suspect a grouping-things-by-similar-symptoms approach may yield different categories than a grouping-things-by-similar-underlying-causes approach, yet both approaches may have value.
Mark Liberman said,
October 4, 2025 @ 11:42 am
@J.W. Brewer "a grouping-things-by-similar-symptoms approach may yield different categories than a grouping-things-by-similar-underlying-causes approach, yet both approaches may have value."
That's clearly true when the similar-symptoms case is something like fever or unconsiousness. The problem is that "autism" doesn't correspond to any single symptom (which might well have multiple causes), but instead is defined (in a different way in each edition of the DSM) as "at least N of the following M symptoms", leading to a situation where (for example) some category members are totally non-verbal while others are hyper-verbal, and others are neither; and similarly for all of the other listed symptoms.
Brett said,
October 4, 2025 @ 12:22 pm
All three of my children have relatively mild autism spectrum disorder. Their official diagnoses are based on rather different combinations of symptoms, as the condition manifests itself in somewhat different ways in each of them. However, while the strongest diagnostic indicators are different for each of them, there is nonetheless a very clear commonality to their conditions. Their shyness under similar circumstances, obsession with certain kinds of minutia, and other characteristics are unmistakably of a common origin, in spite of superficial differences.
That is not to say that everyone with autism shares these traits—although I still think there is generally more commonality that one might infer just from reading diagnostic reports. Autism represents a broad basin of attraction in mental functioning, into which it is relatively "easy" for a brain with any of various neurological issues to fall. It is a constellation of symptoms—most of which are present to some degree in most people with autism spectrum disorder, but which vary a lot in intensity between different individual presentations—that can arise out of a much wider variety of underlying neurological phenomena.
J.W. Brewer said,
October 4, 2025 @ 12:56 pm
Brett (like the DSM) is perhaps talking about a category like what Wittgenstein called a Familienähnlichkeit, which is difficult to reconcile with the notion that "nature" objectively has identifiable "joints" at which a proper taxonomy can and should subdivide it. Whether that latter notion can be made to work with a given set of phenomena probably depends on the details of the phenomena. It's hard enough to make it work in historical linguistics …