(Written in a hurry. I was almost going to title this “My poorly-researched pet theory of schizophrenia”. Hoping for feedback and pointers to relevant prior literature. I am very far from a schizophrenia expert. Really. I cannot emphasize this enough. Like, if I took an undergraduate psych test on schizophrenia right now, I might well flunk it.)
1. What’s my hypothesis?
My hypothesis is that the root cause of schizophrenia is (…drumroll…) a deficiency in medium- to long-range cortex-to-cortex connections. Some elaboration:
- When I say “deficiency”, I mean either “the connections aren’t there in their normal numbers” or “the connections are there, but for some reason they’re not accomplishing what they accomplish in neurotypical people”.
- When I say “cortex-to-cortex connections”, I think the main culprit is direct connections between Cortex Region A and Cortex Region B, but it’s also possible that the relevant thing is indirect connections between Cortex Region A and Cortex Region B, e.g. via the thalamus or cerebellum.
- When I say “medium- to long-range”, this definitely includes e.g. connections between different lobes, and it probably also includes connections across a few centimeters of cortex in humans. I haven’t really thought about what would happen if there was a deficiency in all connections of any length, including the very short ones, but I would weakly guess that this would present as schizophrenia as well.
2. How did I originally come up with that hypothesis?
I can pinpoint the exact moment: I was reading about visual processing abnormalities in schizophrenia, and more specifically the paper Weak Suppression of Visual Context in Chronic Schizophrenia (Dakin, Carlin, Hemsley 2005). They showed people pictures like this:
The task was to match the image contrast in the red circle to one of the circles on the left. The eye-popping results were:
- the schizophrenics did better than the control group,
- …with a p-value of 0.0000002!
- Indeed, “12 out of 15 [subjects with schizophrenia] were more accurate than the most-accurate control.”
- …i.e., there were only three subjects with schizophrenia who did not outperform each and every one of the 33 people in the control groups. And one of those three was later re-diagnosed as not schizophrenic after all!
Back to the task, the control group gets misled by the contrast level outside of the red circle, and the schizophrenic group didn’t. Here’s how I interpreted that:
Each part of the visual cortex is trying to make sense of (more specifically, predict) the sensory inputs that it’s getting. To do a good job at that task, in normal life, it’s helpful to take account of information happening elsewhere in the visual field. After all, distant parts of the visual field are conveying information about ambient light levels, textures and slopes, and things like that. When a neurotypical person grows up viewing naturalistic images, they form lots of predictive models that are near-optimally incorporating all that distant “context” information, pulled in from all across the visual cortex and maybe elsewhere in the brain too. And that screws them up in this particular task. The goal of this task is to ignore all context and look at the red circle contents in isolation. The visual system does not have a set of predictive models that are tailored for accomplishing this task, so the control groups tend to do poorly.
But schizophrenics are succeeding at the same task. Why? Well, presumably for them, the distant “context” information is not available in the first place. So their visual cortex is generally unable to leverage it for better predictive models. This makes their sensory predictions generally worse in a naturalistic setting. But it allows them to do better in this unnatural setting, where the contextual information is only there to trip them up.
3. Is this an existing hypothesis in the literature?
I assume by default that everything I have ever written about neuroscience is either unoriginal or wrong. (Sometimes both!)
I haven’t seen anyone write down a hypothesis superficially similar to the above, as far as I can recall. But I haven’t really gone looking for it either.
It does seem to have significant overlap with Dehaene’s theory that schizophrenia is a “disease of consciousness”. I’ll get back to that at the end. I would suggest that readers interested in the literature around these ideas should maybe consider starting with Dehaene’s schizophrenia discussions, and the papers that he cites, and papers that cite him, etc. Let me know what you find.
If you know of other relevant discussion, please share in the comments!
4. Does this hypothesis elegantly explain everything about schizophrenia?
Well, let’s try!
4.1 The onset of schizophrenia is typically in the late-teens-to-twenties
I propose to center this sub-story around synaptic pruning, which suggestively “continues into the late-20s” according to wikipedia. Maybe the symptoms of schizophrenia kick in when there is so little flow of information between different parts of the cortex that they fail to constrain each other—they get out sync, they come unmoored, they start spinning their wheels independently. And maybe there’s kind of a soft threshold of how much information flow between regions is needed to prevent that from happening.
For a 10-year-old, relatively little synaptic pruning has happened so far, so maybe neurotypical people are way way way above that threshold, while future-schizophrenics are “merely” way above that threshold. Then by age 25, everybody has pruned lots of synapses, so the neurotypical people are somewhat above that threshold, while the schizophrenics have fallen below that threshold.
Good story? Yeah, I think it’s a pretty good story—I don’t think it’s special pleading / rationalization. In particular, I was guessing that this was the story before I knew the age range of synaptic pruning, and then when I looked that up, I was surprised and pleased at how good a match it was.
4.2 Positive symptoms—auditory hallucinations (hearing voices), “distortions of self-experience”, etc.
I figure, when a neurotypical person is subvocalizing, there’s communication between the motor cortex parts that are issuing the subvocalization commands (assuming that that’s how subvocalization works, I dunno), and the sensory cortex parts that are detecting the subvocalization which is now happening. Basically, the sensory cortex has ample warning that the subvocalization is coming. It’s not surprised when it arrives.
But in schizophrenia, different parts of the cortex can’t reliably talk to each other. So maybe sometimes the sensory cortex detects that a subvocalization is now happening, but hadn’t gotten any signal in advance that this subvocalization was about to be produced endogenously, by a different part of the same cortex. So when it arrives, it’s a surprise, and thus is interpreted as exogenous, i.e. it feels like it’s coming from the outside.
Wikipedia also suggests that psychotic episodes often include “feeling as if one's thoughts or feelings are not really one's own” and “believing that thoughts are being inserted into one's mind”. I think my hypothesis can explain them in a similar way.
Good story? Hmm, a priori, I think my hypothesis would more naturally predict that schizophrenics would experience these symptoms constantly, and not just during psychotic episodes. Not sure what to make of that. Hmmmm. Maybe I should hypothesize that different parts of the cortex are “completely unmoored from each other” only during psychotic episodes, and the rest of the time they’re merely “mediocre at communicating”? Or something like that? Still, I’m mostly pretty happy with how my hypothesis is pointing in a direction that’s consistent with the symptoms.
(EDITED TO ADD: For a better discussion, see my follow-up post, Model of psychosis, take 2.)
4.3 Negative symptoms
Wikipedia lists five common ones: “blunted affect – showing flat expressions or little emotion; alogia – a poverty of speech; anhedonia – an inability to feel pleasure; asociality – the lack of desire to form relationships, and avolition – a lack of motivation and apathy.”
What does my hypothesis predict for emotions in schizophrenia? I think it would predict that emotions should become generally less responsive to what you’re thinking, planning, expecting, seeing, hearing, etc. After all, suppose you see a friend in the crowd and feel an emotional response; this requires transferring information, via a long-range connection, from the face-detection part of the cortex (in the temporal lobe) to cortical areas involved in visceral reactions (probably amygdala, medial prefrontal cortex, and/or hippocampus). If the long-range connections are not there, or not effective, then seeing the friend won’t evoke any feelings at all.
So again, my hypothesis seems to predict that emotions / affects should be relatively unresponsive to what you’re thinking, seeing, etc. But my hypothesis does not seem to make any specific prediction about what those unresponsive emotions should be. If schizophrenics were deliriously happy all the time, regardless of what they were thinking, what was happening in the world, etc., that would also be consistent with my hypothesis, as far as I know.
I think my hypothesis could issue a more specific prediction, but it would require knowing a whole lot of nitty-gritty details about emotion-related circuitry in the amygdala and brainstem and so on. I have a major professional interest in that topic, but am nowhere near knowledgeable enough to make a first-principles prediction on this topic. I don’t think anyone can right now.
Good story? Yup, I’m pretty happy, I think this is a natural consequence of my hypothesis. I wish the hypothesis had made more specific predictions, but oh well, that shouldn’t really count against it.
4.4 Creativity
There’s a spectrum from “different parts of the cortex are all closely tied together into globally-coherent, self-consistent models” and “all the different parts of the cortex are totally unmoored from each other, each doing its own thing”, with schizophrenics closer to the latter end than neurotypical people. It seems very plausible to me that creativity would increase as you move forward along that spectrum. After all, if Cortex Region A and Cortex Region B are almost-independently generating (components of) thoughts, you’re sampling from a much broader space of A+B combinations than if the state of Region A is tightly correlated with the state of Region B at all times. Thus, in the former case, you’d wind up with more unexpected and “creative” A+B composite thoughts / analogies.
Good story? Yeah, I think so. To keep myself honest, I pretended for a moment that schizophrenics were less creative, and tried in good faith to “explain” why that is exactly what my hypothesis predicted all along. I couldn’t come up with anything. Good!
4.5 Anticorrelation with autism
(This section needs to be read in conjunction with my earlier post The Intense World Theory of Autism.)
Scott Alexander wrote a 2018 blog post called Diametrical Model Of Autism And Schizophrenia.
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.
I’ve previously been skeptical of this kind of thinking because there are many things that autists and schizophrenics have in common, many autistics who seem a bit schizophrenic, many schizophrenics who seem a bit autistic, and many risk factors shared by both conditions. …
As far as I can understand, our task is to explain two things: (1) autism and schizophrenia seem generally anticorrelated (as in the first paragraph), and (2) there is definitely such a thing as “a person who has both autism and schizophrenia” (example ref) (as in the second paragraph). I claim that my hypotheses predict both of those things.
As in my autism blog post, I claim that the diagnostic criteria for autism are all downstream from “hypersensitivity”—things that would be pleasantly stimulating (or even barely noticeable) for a neurotypical person, are overwhelming and aversive for someone with autism.
One possible cause of hypersensitivity is “there are unusually many (or unusually effective) cortex-to-cortex connections”.
If that’s the cause, and if many of those connections are medium- or long-range, then this cause of autism would simultaneously be a protective factor against schizophrenia. And I suspect that this is the thing that very often happens.
However, you could get both schizophrenia and autism if the autism had a different cause from the one I mentioned just above. For example:
- Maybe someone could have an extreme excess of local cortical connections (thus causing autism) while also lacking longer-range cortical connections (thus causing schizophrenia).
- Maybe the long-range cortical connections could be messed up in a way that makes them fire way more than usual (thus causing autism) while mostly failing to communicate useful information (thus causing schizophrenia).
Stuff like that.
OK, now let’s look at the sentences in the above excerpt one-by-one:
- For the first sentence (on genes): Consider a gene that says “Hey neurons! When in doubt, make more synapses! Grow more axons! Make bigger dendritic trees!” This gene would probably be protective against schizophrenia and a risk factor for autism, for reasons discussed just above. And vice-versa for the opposite kind of gene. Nice!
- For the second sentence (on the corpus callosum), my hypotheses straightforwardly predict the opposite (unusually large corpus callosum in autism, unusually small in schizophrenia). Oops! I was very weirded out by this, and in my first draft I had a bunch of text about how my hypothesis had made a flagrantly wrong prediction. (It was peppered with words like “Dang” and “Hmm”.) But then I checked google, and e.g. this says that the corpus callosum is unusually big in autistic kids, and this finds “a relative absence of crossing fibers” in the corpus callosum in schizophrenia, which are both consistent with my hypotheses. So maybe Scott accidentally typed it in backwards? Granted, it does seem to be the case that lacking a corpus callosum altogether (“corpus callosum agenesis”) puts one at unusually high risk of autism, which is counter to my prediction. But on the other hand, this paper (IIUC) says that corpus callosum agenesis puts one at an even higher risk of schizophrenia, which matches my prediction! Anyway, I’m mostly inclined to ignore corpus callosum agenesis in the first place. It’s a pretty extreme and unusual case, and presumably induces other downstream changes, and I’d bet that corpus callosum agenesis patients have an unusual combination of symptoms that might not fit well into our usual classification schemes. So anyway, I think my hypotheses are basically fine in regards to the corpus collosum.
- For the third sentence (nicotine), it seems a natural consequence of nicotine creating strong feelings, which would be appealing to schizophrenics who have blunted affect in general (see discussion of “Negative symptoms” above), and aversive to autistic people who are feeling overstimulated in general (see my autism post). [UPDATE: "Slimepriestess" in the comments section says I’m super-wrong about this bullet point. Oops! I’ll look into it more carefully when I get a chance.]
- For the fourth sentence (rubber hand illusion), see the discussion under “Auditory hallucinations” above for the schizophrenia side. On the autism side, this is probably related to most (not all) people with autism having unusually strong long-range connections, as discussed above.
- For the fifth sentence (creativity), see previous section and above.
Good story? I think everything hangs together remarkably well, unless I’m mistaken about the corpus callosum thing.
4.6 Relation to myelination
My hypothesis would seem to predict that if someone’s brain is producing defective myelin, such that their long-range axons were lousy at transmitting information, this person would probably be highly prone to schizophrenia. So I google-searched for “schizophrenia myelination”, and found a bunch of results that appear to confirm this prediction (e.g. 1,2). But, this might be less impressive than it sounds; I kinda assume that if you search for any pair of neuroscience words you’ll probably find a bunch of papers about how they’re deeply related ¯\_(ツ)_/¯
4.7 Schizophrenia and blindness
There’s a story that blindness prevents schizophrenia. This is counter to my hypothesis; my hypothesis predicts that schizophrenia and blindness should have more-or-less nothing to do with each other. So then I was questioning whether the “blindness prevents schizophrenia” story was true in the first place. I spent a few hours looking into it last summer, and wound up feeling like it probably wasn’t. I wrote that up as a short blog post: I’m mildly skeptical that blindness prevents schizophrenia.
5. Dehaene’s related discussion of schizophrenia as a “disorder of consciousness”
See Dehaene’s book “Consciousness and the Brain” chapter 7 (I had read the book a long time ago and completely forgot about that section, then was reminded of it here).
I mostly wrote this post before (re)reading that book chapter, and it does seem like some of the things I wrote in this post are rehashing ideas in the literature. Yay! That’s very reassuring! For example, it seems like Fletcher and Frith have written this with a qualitatively similar idea to my Section 4.2.
Also, more direct evidence:
Diffusion tensor imaging reveals massive anomalies of the long-distance bundles of axons that link cortical regions. The fibers of the corpus callosum, which interconnect the two hemispheres, are particularly impaired, as are the connections that link the prefrontal cortex with distant regions of the cortex, hippocampus, and thalamus [59]. The outcome is a severe disruption of resting-state connectivity: during quiet rest, in schizophrenic patients the prefrontal cortex loses its status as a major interconnected hub, and activations are much less integrated into a functional whole than in normal controls [60].
At a more microscopic level, the huge pyramidal cells in the dorsolateral prefrontal cortex (layers 2 and 3), with their extensive dendrites capable of receiving thousands of synaptic connections, are much smaller in schizophrenic patients. They exhibit fewer spines, the terminal sites of excitatory synapses whose enormous density is characteristic of the human brain. This loss of connectivity may well play a major causal role in schizophrenia.
What does any of this have to do with “consciousness” though? Well, Dehaene is the leading advocate of “global workspace theory”, which I don’t want to get into, but which centrally involves (a subset of) long-range cortex-to-cortex connections.
I think calling schizophrenia a “disease of consciousness” is slightly misleading. Some long-range cortex-to-cortex connections are related to the “global workspace”, others aren’t, but I think they’re all generally impacted by schizophrenia. For example, I think the visual perception example that I opened with in Section 2 is probably (although not necessarily) related to schizophrenia’s impact on few-cm-long connections within visual cortex. If so, that kind of neuronal connection is almost definitely not part of the “global workspace” or “consciousness”, I would argue. But it’s still part of schizophrenia.
That said, I’m kinda nitpicking, and really if someone describes this post as “an amateurish restatement / endorsement of Dehaene’s theory of schizophrenia”, I wouldn’t be particularly upset.
Hm, I had a vague memory that contrast detection relies on something like lateral inhibition but when I thought about it a bit more it doesn't really make sense and I guess I conflated it with edge detection in the retina.
Regarding cerebellum in hearing voices: If I understand your model correctly, it goes something like this. Region S (sender) "generates voices" and region R (receiver) "hears voices" generated by S. R expects to receive those signals from S (or maybe even just expects to receive these kinds of signals in general, without specifying where they come from). R gets surprised when it receives unexpected signals and interprets them as "not mine". R would expect to receive them, if it first got a message "hey, S is soon going to send some voice-signals to you". Isn't this exactly the role of the cerebellum, to learn that, e.g. if S activates in this particular way (about to "generate voices"), then R will soon activate in the other way ("hears voices") and therefore it would make sense to preempt R, so that it can expect to get that particular signal from S and act accordingly even before receiving that signal?