Update (8/29/22): All of this might be false, see here for more.

Some weird psychiatric trivia: no congenitally blind person ever gets schizophrenia (journal article, popular article). “Trivia” is exactly the right word for this fact; it’s undeniably interesting, but what do you do with it? So far nobody has done anything, other than remark “hmm, that’s funny”.

I was thinking about this recently in the context of the diametrical model of autism vs. schizophrenia. This is itself pretty close to psychiatric trivia - a lot of features of schizophrenia and autism seem to be opposites of each other. As I put it here:

Many of the genes that increase risk of autism decrease risk of schizophrenia, and vice versa. Autists have a smaller-than-normal corpus callosum; schizophrenics have a larger-than-normal one. Schizophrenics smoke so often that some researchers believe they have some kind of nicotine deficiency; autists have unusually low smoking rates. Schizophrenics are more susceptible to the rubber hand illusion and have weaker self-other boundaries in general; autists seem less susceptible and have stronger self-other boundaries. Autists can be pathologically rational but tend to be uncreative; schizophrenics can be pathologically creative but tend to be irrational. The list goes on.

In theory you could use this kind of thing to help figure out what causes both conditions. In practice, it’s a lot more complicated. Autism and schizophrenia also resemble each other in a lot of ways, a lot of the genes for one also increase risk for the other, and some people even get diagnosed with both. I try to make sense of this conflicting information by speculating that autism has at least two components, which are correlated and anticorrelated with schizophrenia respectively (see here for guesses about what these might be), but still - all speculation and trivia.

Anyway, here’s a study showing that congenitally blind people are about fifty times more likely to get autism than everyone else.

A full 50% of the blind children in the sample met criteria for an autism diagnosis, compared to about 1% of sighted children in the same area. And even blind children who don’t meet full autism criteria seem to have somewhat more autistic features than usual.

Maybe blindness makes children seem more autistic in some boring, mechanical sense? Like maybe if you can’t read other people’s body language or even consistently know where they are, it’s hard to be social and so you don’t interact with other people and that seems autistic to people trying to diagnose you? The researchers tried pretty hard to avoid that problem, getting really careful experts to do the diagnosis and asking about lots of different autism symptoms like mental disability, motor disability, and echolalia. Still, it’s easy to think of ways blindness could make you mentally disabled (eg you’re less likely to learn to read so you never exercise whatever cognitive functions are involved in reading) and motor disabled (I don’t know what’s up with the echolalia). Still, as far as we can tell that’s not it.

Maybe blindness something something interpersonal relationships? That’s the position of Hobson and Bishop, who use these kinds of studies to argue that autism is an “interpersonal disease” about how different people interact with each other, rather than being located in a single brain. It sounds a little too TED-talk-bait-y to me, and I don’t feel like I got a great grasp of their theory from the paper (maybe I should wait for the TED talk). Not even sure how I’d evaluate this.

I’m most sympathetic to the idea that some kind of neural hyperparameters are sensitive to the overall amount of sensation you’re getting, and if you lose a really important sense like vision, that shifts the hyperparameter values a bit. If some parameter goes from schizophrenia at one end to autism at the other, and blindness shifts it a bit towards the autism end, that would explain why no blind children get schizophrenia, lots of blind children get autism, and even the ones who seem neurotypical have unusually many subclinical autistic features.

I wish I could say more than “some hyperparameter or something”, but I’m not quite there yet. Probably the next step would be a very careful re-reading of Corlett, Frith, and Fletcher, who think they understand why sensory deprivation seems to treat schizophrenia (something something predictive coding), though before we get too excited we should remember that the sensory deprivation people also claim deprivation is helpful in autism.