latex gloved hands hold petri dish with a cure for autism symbolized as its contents, or perhaps genetic therapy gene therapy

Why is there not a cure for autism?6 min read

There is not a cure for autism, and not for want of trying—entire careers and many mil­lions of dol­lars have been spent on the search for a cure that has not appeared. This article is my view on why, despite a great deal of effort, sci­en­tists have not come any­where close to finding a cure.

We don’t know what autism is

The diag­nostic cri­teria are vague and behavior-based. The cur­rent Diagnostic and Statistical Manual (DSM) requires only two symp­toms to diag­nose autism: restricted/repetitive behavior or inter­ests, and dif­fi­cul­ties in social com­mu­ni­ca­tion, both of which must have been present since early child­hood. There is no bio­marker for autism. There is only the indi­vidual clinician’s opinion based on an inter­view with the autistic person and pos­sibly their par­ents or other family.

This means that the autism spec­trum is incred­ibly het­eroge­nous and likely biased by sex, race, and other demo­graphics (for example, com­pared to the DSM IV, the DSM V tends to miss females and people with high IQs). This is a problem because recruit­ment of autistic people into studies relies on their diag­nosis.

We don’t know whether autism has one fun­da­mental feature/cause or is a myriad of con­di­tions and comorbids—from anx­iety to intel­lec­tual dis­ability to gut problems—all grouped together by two common symp­toms. This does not mean that the diag­nostic cri­teria should be changed; this con­cept of autism may be the one that most closely carves nature at its joints. It does, how­ever, make autism dif­fi­cult to study sci­en­tif­i­cally.

Autism is genetic—but the genetics are highly complex

Autism is a genetic con­di­tion: about 80–90% of the vari­ance in autism devel­op­ment can be explained by genetics rather than envi­ron­mental influ­ences (Bai et al., 2019, Sandin et al., 2017). Some genetic con­di­tions can be treated with restric­tive diets, sup­ple­ments, or exper­i­mental treat­ments like gene therapy. However, autism is much more com­pli­cated.

Hundreds of autism “risk” genes have been iden­ti­fied; com­bined, they explain 10–20% of cases. Genes that affect the like­li­hood of devel­oping autism interact with each other affect the epi­ge­netic land­scape of the genome and are often dupli­cated or deleted. Some are common, while some are de novo—occur­ring only in the autistic person rather than being inher­ited.

Autistics tend to have higher muta­tional loads in the form of point muta­tions that sub­sti­tute one DNA base for another, or copy number vari­ants (dupli­ca­tions or deletions)—these are usu­ally scat­tered around the genome, so there is no single gene that is respon­sible.

There are sev­eral autism-associated syn­dromes that have known genetic causes, including Fragile X syn­drome, Rett syn­drome, 22q13 dele­tion syn­drome, cor­tical dysplasia-focal epilepsy, Angelman syn­drome, and tuberous scle­rosis. Gene therapy treat­ments for those are much more feasible—but syn­dromes caused by a single gene make up only 5% of autism cases com­bined. Even within that 5%, when you con­sider the extra­or­di­nary cost of gene therapy, it may not be eco­nom­i­cally fea­sible to develop one for each known cause.

Thus, going after autism at the root causes (genetic vari­ants) is unlikely to work. It is pos­sible that among the hun­dreds of dif­ferent genes that are asso­ci­ated with autism, sim­i­lar­i­ties could be found between them that point to a common process. Despite the many people who have tried to find such a process, there has been no simple answer.

Autism is dif­fi­cult to model

Biomedical sci­en­tists often use animal models to study dis­eases, under­stand causes, and test out poten­tial treat­ments. Autism is not suit­able to being studied this way (though researchers cer­tainly still try: animal research made up 44% of all autism research funding in the UK in 2016, according to Autistica).

Autism is hard to model due to its genetic het­ero­geneity; each model would have to uniquely match its human coun­ter­part. Also, autism is fun­da­men­tally about emer­gent func­tions of the mind that are uniquely human.

Animals that have been used to model autism include mice, rats, song­birds, rhesus mon­keys, and nema­tode worms, among others. These ani­mals are made “autistic” by mutating part of their genomes to some variant we know causes or increases the like­li­hood of some cases of autism, by removing oxy­tocin or opioid recep­tors, or by inflicting them or their preg­nant mothers with envi­ron­mental injuries with thalido­mide. Animal studies of autism often focus on one asso­ci­ated trait, such as repet­i­tive activ­i­ties like marble burying or self-grooming.

Autism is a very human con­di­tion, char­ac­ter­ized by dif­fer­ences in social­iza­tion and lan­guage as well as repet­i­tive behav­iors. Other ani­mals have dif­ferent cog­ni­tive capac­i­ties com­pared to humans, which is how researchers jus­tify using them in the first place.

They also have very dif­ferent social behav­iors; for example, mice are highly ter­ri­to­rial, so their behavior in a test of socia­bility may actu­ally be a test of con­fronta­tion. Some researchers are using pri­mate models instead to avoid that issue, but even apart from ethics, there are many traits that non-human ani­mals simply cannot replicate—in autism, those are the traits that matter.

The animal obser­va­tions actu­ally bear little rel­e­vance to the traits of actual human autis­tics. Mice do not face envi­ron­ments sim­ilar to those rel­e­vant for autistic humans, like get­ting through school and the work­place. The out­comes studied, such as marble-burying (repet­i­tive behav­iors) or pheromone emis­sions or ultra­sonic vocal­iza­tions (social inter­ac­tion), may be unre­lated to the effects people actu­ally want in a cure for autism.

They also fail to rep­re­sent the workarounds real autis­tics use in their daily lives, e.g. an autistic person using scripts to nav­i­gate con­ver­sa­tions instead of relying solely on their instincts and spon­ta­neous con­ver­sa­tion.

We don’t know what a cure for autism would look like in a person

What is a cure for autism? A person could be diag­nosed with autism for being in a high per­centile for each of the two diag­nostic traits — would a cure mean bringing those traits below the 95th per­centile? The 70th? All the way to the median? You could per­haps say it means changing the person’s brain so that they are per­ma­nently no longer neg­a­tively affected by their autistic traits, no matter their envi­ron­ment.

I would define a cure for autism as some med­ical inter­ven­tion given to an autistic person that per­ma­nently changes their brain so that they no longer pos­sess the under­lying incli­na­tions that led to diag­nos­able behav­iours. For example, a ‘pre­vi­ously autistic’ indi­vidual would no longer feel the desire to stim, rather than simply forcing them­selves to sup­press it (masking).

It is dif­fi­cult to imagine how the adult brain could be changed so dras­ti­cally, so the cure would have to be admin­is­tered in early child­hood or poten­tially in utero as a pre­ven­tion. Other pre­ven­tions may include selec­tively aborting or choosing not to implant embryos with genetics indi­cating likely autism (and remember, a vast majority of the genetic vari­ants are unknown).

Autism researchers are missing out on a huge source of information

That source is, of course, autistic people. There is a chasm between autistic people and autism pro­fes­sionals, with per­haps the most noto­rious example being the iden­tity lan­guage used: most autistic adults prefer “autistic people,” and autism pro­fes­sionals mostly call them “person with autism,” regard­less of pref­er­ence.

Autistic people’s knowl­edge of autism has been found to be more scientifically-based, which is not sur­prising con­sid­ering that many autistic people’s spe­cial interest is autism. If you’re looking to under­stand autistic minds—which, if you’re trying to under­stand autism, you should be—autistic people also have blogs all over the internet offering insights into their life expe­ri­ences and things that help them. You prob­ably need them more than you think: neu­rotyp­ical people tend to struggle to read autis­tics.

Of course, autism pro­fes­sionals should, as a matter of ethics, be lis­tening to autistic people anyway, but that’s not my point here—my point is that they are missing out on gen­uinely valu­able infor­ma­tion that will push the field of study for­ward.

All in your head: on a cure for autism as a human condition

Autism is defined by behavior—a dif­fer­ence in brain wiring. Brains are dif­fi­cult to target for drug delivery, gene therapy, and finding blood bio­markers. Brains are dif­fi­cult to reduce to indi­vidual com­po­nents, and in the fields of lan­guage and sociality, human brains are unique.

Autism is about the com­plex behav­iors involved in nav­i­gating the modern world, about all the inter­ac­tions within the brain and within society. With that in mind, it’s not sur­prising that autism is hard to pin down.

The good news is that it does not need to be pinned down. Instead of trying to erad­i­cate autistic people and replace them with neu­rotyp­i­cals inside previously-autistic bodies, we can focus our efforts on improving the quality of life for actual autistic people right now.

The failure of research to find a cure for autism may turn out to be a blessing in dis­guise.


Latest posts by Elle Loughran (see all)


    1. Author

      What makes you think I want to cure it?


  2. I do not need to be cured, I need to be accepted as I am.

  3. I don’t think autism is an exclu­sively human con­di­tion.

    That said, I think it might be pos­sible that some entire species (such as cats other than lions, the fox, cer­tain species of ham­ster, cer­tain birds of prey and one other bird species I saw which is social but extremely intro­verted, if you will) might have a neu­rology that cor­re­sponds closely to autism. Mind you, that doesn’t mean THEIR autism man­i­fests in the exact same way as a human, just that they could easily have sim­ilar ways of looking at the world, sim­ilar expres­sions and needs, and other such traits, shaped both by the autistic treats they share with autistic humans and the way their own species’ unique neu­rology fur­ther shapes that. And this would be fur­ther affected when for those species autism might be the typ­ical neu­rotype for their kind.

    And as for species that don’t typ­i­cally act autistic, I’ve seen my share of those too — in par­tic­ular, I’ve seen autistic dogs, three or four at least. And there’s this funny trait about two of the autistic dogs I met that doesn’t really overlap with human autism but which I wonder about as a pos­sible comor­bidity in canine autism specif­i­cally. That is, two of the autistic dogs I met had a neu­rotyp­ical lit­ter­mate with them, and in both cases the owner said that the neu­rotyp­ical dog is the one who will let you pet them and the autistic dog will not. But in both cases, the neu­rotyp­ical dog was the one that refused to let me pet them, while the “recal­ci­trant” (autistic) dog was the one who came right up to me, an autistic person, and let me pet them. Like, while dogs aren’t always so picky about people having a neu­rotype that matches theirs, I wonder if maybe autism in dogs could be comorbid with an unwill­ing­ness to interact with people or other non-conspecifics who don’t share one’s own neu­rotype. Like autistic dogs might be more likely to have neu­rotyp­ical sib­lings who only want to hang out with neu­rotyp­i­cals while they are more likely to only want to hang out with autis­tics — except for dogs.

    I don’t know if this is a pat­tern, of course, because it is just two dogs I saw, but if it is a pat­tern, that would be an autism symptom/comorbidity that dogs explic­itly do not share with humans, since humans do not have a cor­re­sponding fear of inter­acting with ani­mals who have “weird” neu­rolo­gies com­pared to them (and I mean explic­itly not shared — humans are far more like otters than dogs, animal-interaction-wise, in that they will pet or interact with just about any­thing so long as it’s safe to do so and some­times even when it isn’t safe).

  4. What I’m really won­dering is how we know for sure autism is a genetic con­di­tion. As the author asserts only 10 to 20 per­cent of the cases can be explained by the genes so far found. How do we know so sure that the other 80 to 90 per­cent also can be explained by genetics?

    1. Author

      Because geneti­cists don’t need to know the actual genes involved to know the per­centage of some­thing that is explained by genetics. Heritability is the pro­por­tion of vari­ance in a trait asso­ci­ated with genetic vari­ants rather than envi­ron­mental ones and there are many ways of cal­cu­lating it, but a simple one is the cor­re­la­tion between iden­tical twins raised apart. They have iden­tical genomes but dif­ferent envi­ron­ments, and you can cal­cu­late the her­i­tability based on the pro­por­tion of twins who are con­cor­dant for the trait (both have it/both don’t).

  5. Thank you.

    I wish researchers would under­stand this better and maybe work on things that were more fea­sible. I’d love to see cures or better treat­ments for epilepsy, anx­iety, sleep dis­or­ders, PTSD…

  6. If anyone cared about improving our quality of life, it wouldn’t cost any­thing. All they would have to do is to treat us with respect, dig­nity, and friend­ship. Give us a fair chance for employ­ment, edu­ca­tion, and to speak our minds. Stop shun­ning, crit­i­cizing, and rejecting us for things that cause no harm to anyone else. Recognize our strengths, our hard work, our accom­plish­ments, and, most impor­tantly, our poten­tial. Appreciate that every human being has weak­nesses and that we are no dif­ferent, and judge us fairly, instead of focusing on what we don’t do as well as others and for­get­ting that we also do some things better than others. Focus on making our expe­ri­ence of life better, rather than on making us do fewer things that make others feel uncom­fort­able. Understand that inside our minds are worlds and uni­verses that are not per­cep­tible from the out­side. In short, rec­og­nize that we are human beings with feel­ings, needs, and opin­ions and that we are just as valu­able to the world as any other person, instead of a problem that needs to be cleansed from exis­tence.

    In short, it’s the under­lying atti­tude that has to change, the one that has NTs con­vinced that we are infe­rior and broken and that we have nothing useful to say or to con­tribute. No amount of money is going to change that.

Talk to us... what are you thinking?