A Grand Unified Theory of Autism?

By Neuroskeptic | January 5, 2011 9:25 am

A physicist famously wanted to find the grand unifying equation behind the laws of nature, in a form that you could put on a t-shirt.

Neuroscientists Kamilla and Henry Markram have proposed a grand unifying theory of autism, and the key to it is in this picture. I wouldn’t want to be seen wearing it quite yet, but if the theory pans out, I’m sure we could come up with a more torso-friendly diagram.

So what does this mean? The Markrams call their idea the Intense World Theory. Essentially, they propose that all of the diverse symptoms of autism are direct or indirect consequences of the autistic brain’s being hyper-responsive to stimuli. (They published an earlier version of this theory in 2007).

Not the brain as a whole, and not each individual cell, either. Rather, they say that the abnormality lies in local microcircuits. The best known of these are the cortical columns and minicolumns. Neurons in any given microcircuit are connected both with their neighbors, and with more distant cells. A bit like a large company with offices in different cities: people within each office talk to each other, but they also phone and email the other branches.

The theory goes that the autistic brain has too many connections within any given microcircuit. So, when the circuit is activated, it reactivates itself too strongly, and shows a stronger, and longer, excitation. A bit like if the offices were open-plan, so everyone can overhear everyone else, and it all gets very noisy.

So what’s the evidence for this? There’s circumstantial support. It “makes sense”, if you’re willing to accept an analogy between hyperactive local neural circuits and hyper-intense psychological phenomena.

We know that a given cortical minicolumn responds to a particular type of stimulus, or aspect of a stimulus; there are minicolumns for horizontal lines, for lines at 10 degrees to the horizontal, and so on. People with autism are often fixated on little details. It’s a leap, but not an impossible one, to see these as related.

But the only really direct biological evidence is from rats. The story starts with valproate (VPA), an effective anticonvulsant also widely used in bipolar disorder. VPA has to be used with extreme caution in women because of the risk of birth defects.

Children whose mothers take VPA (and to various degrees other similar drugs) during pregnancy often suffer various physical and behavioural problems, the fetal anticonvulsant syndrome. Sadly, this happened quite a lot in the past, before the risks were appreciated. The key point is that autistic symptoms extremely common in children exposed to high-dose VPA.

Markram (and other people) have studied rats exposed to valproate in the womb. They found that, well, they’re weird. Proponents would say that they behave a lot like how an “autistic” rat would: they are less sociable, prone to repetitive behaviours, highly anxious, etc.

Can a rat “have autism”? That’s one to ponder. On the one hand, rats are surprisingly smart, sociable animals. For every human brain region, there’s a rat equivalent in roughly the same place, which does roughly the same thing. They have cortical columns and minicolumns like ours (we just have more of them). They even “laugh” when you tickle them. On the other hand… they’re rats. They run around gutters eating trash.

The t-shirt image at the top of this post is based on Markram and colleagues work on the cortical network properties of VPA-exposed rats (e.g. this and several other studies). These studies revealed hyper-connectivity within local microcircuits, and have also shown that circuits from VPA-exposed rats “learn” faster: they form new synaptic connections via the process of LTP at an accelerated rate, likely due to over-expression of NMDA glutamate receptors.

They admit that it’s a big leap from that to human autism. But it’s not an impossible leap. As they say:

This provided the potential cellular and circuit explanation for how an autistic brain could be easily trapped in a painfully intense world, potentially explaining a broad range of common autistic symptoms such as sensory sensitivity, withdrawal, repetitive behavior, idiosyncrasies, and even exceptional talents.

The major attraction of the theory is that it is a unified one: it seems to explain everything about autism, although maybe it’s just vague enough to be stretched to cover anything. For example, Markram attributes the social awkwardness of autistic people to an overactive amygdala, which makes them extremely anxious in social situations, especially when meeting people’s gaze; this, he says, means that they quickly learn to avoid other people in an attempt to cope with this Intense World.

Henry Markram is best known as the leader of the Blue Brain Project, which aims to simulate a brain using supercomputers. So he’s no stranger to big ideas. Whether this idea is as solid as it is big remains to be seen, but I think he’s to be applauded for at least having a crack at a unified account of autism, something which, as far as I know, no-one else has had the guts to try yet (Edit: But see the comments for a debate on that question)

ResearchBlogging.orgMarkram K, & Markram H (2010). The intense world theory – a unifying theory of the neurobiology of autism. Frontiers in human neuroscience, 4 PMID: 21191475

CATEGORIZED UNDER: animals, autism, papers, philosophy, science
  • http://www.blogger.com/profile/05660407099521700995 petrossa

    Being an autist myself and having studied the matter for the last 30 years intensely (due to my autism derived obsessional occupation) i'm more of the opinion it's a mostly corpus callosum issue. The wiring is just different, causing different integration of the hemispheres. From birth this causes a retardation of the forming of the usual circuits, or just causes parts of the hemispheres to develop more or less due to the surplus or lack of inter-hemispherical excitation.

    Which when being examined at a later stage is disguised as lateral underdeveloped modules. This explains in my view also the difference between aspergers (or other forms of high functioning autism) and hardcore autism with regard to their differences in right/left hemisphere formation/function.

    Its imo the sheer roll of the dice how the wiring is skewed that causes one to display varying levels of autism.

    But this piece fits in the puzzle as well. Could well be a contributing factor imo.

  • http://www.blogger.com/profile/11279402169161555639 Michelle Dawson

    There are a lot of unified accounts of autism.

    We have one (someone related it to the Markrams' here). I'm pretty sure Matthew Belmonte (who first put forward the local overconnectivity idea) has one. Marcel Just has one. Nancy Minshew has one.

    Michael Ullman has one. Uta Frith has one (last I heard, still the “absent self”). Simon Baron-Cohen has one. Armando Bertone (down the hall) has one. Francesca Happe has one (including that autism is fractionable). I'm pretty sure Jon Brock has one. Mark Bear has one, which has been a big trend. Michael Merzenich has one. And so on.

    The question really is who doesn't have one.

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    Michelle: I'm not familiar with all of those, but taking S B-C, is it true that he has a unified theory including both the biology and the psychology?

    Empathizing/Systematizing / Mindblindness is a theory about the mind. He's also promoted the idea of testosterone as a casual factor, but he hasn't tried to explain how it causes the symptoms afaik.

    There are plenty of other unified cognitive theories like weak central coherence, but as far as I can see, most of them are silent about the underlying molecular mechanisms. Correct me if I'm wrong, of course.

    The Markram's have at least attempted that, although as I said, it requires you to accept a kind of impressionistic analogy between neural microcircuits and mental phenomena. But that's not entirely silly.

  • http://www.blogger.com/profile/11279402169161555639 Michelle Dawson

    WCC stopped being a stand-alone unified account of autism some years ago (see FrithU, 2003), though it (maybe) did start out as one. It has been incorportated into other unified theories (Just et al., 2004).

    SBC's current model is presented as a grand unified account, as are many others.

    And I left out a bunch of others, including Ami Klin (whose world view includes the non-importance of some questions, including some aspects of cause), Manuel Casanova, Eric Courchesne, Thomas Bourgeron, on and on (here's one that popped up in 2009).

    Whatever else its problems, autism research has never suffered from a shortage of grand unified theories.

  • http://www.blogger.com/profile/14647896216499813443 Kapitano

    Just a thought, but if autistic brains are 'overwired', are some other brains 'underwired'?

    If the hypothesis is correct, should there be an undiagnosed Anti-autism condition? Might it even predict such a condition?

  • http://www.blogger.com/profile/08010555869208208621 The Neurocritic

    Their entire theory appears to rest on the VPA rat model. However, autism existed long before adult humans took valproic acid for therapeutic purposes. What other evidence exists that autism is a molecular syndrome? Do they have an alternate compound in mind? What is the mechanism if VPA isn't the catalyst? An “epigenetic insult”? That's rather vague…

  • http://www.blogger.com/profile/03449922526437153857 veri

    I want that shirt.

  • http://www.blogger.com/profile/05660407099521700995 petrossa


    Admittedly it's not necessary to have the condition you comment/study/theorize on but in my long years of contact with the professional community as a 'patient' it's either i met all the wrong professionals or none of them knew what they were talking about.

    Going by their behavior and the total lack of uniform opinion in the professional community i opt for the last.

    But as as i also noticed the professional community hates criticism and even more from non-professionals. The appeal to authority most of them totally lack is somewhat childish.

    Each and every 'theory' i've read (and trust me i've had way more time to read then whichever professional and the intellectual capacity to understand what i read) is bogged down by overspecialization/hobby horse of the theorist.

    They are so confined by their education that their way of thinking can only progress along the lines of their education. Which going by the failure rate of the profession can't be that accurate. All overview is lost.

    Sure, peeling apart neurons works fine, neural networks well understood. All systems separately are studied in depth. And then one gets lost in the complexity of the whole thing, or get trapped in the pattern recognition system which causes theorist x to pose theory Y because he/she saw correlation Z.

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    Neurocritic: Well, given that VPA does cause ASD symptoms in humans, it must cause changes in the brain which are somewhat like those that underlie actual autism.

    It's not clear how VPA causes abnormalities. IMO it seems likely that it does so by blocking GABA breakdown and causing an overload of GABA which causes compensatory down-regulation of GABA stuff. VPA rodents lack parvalbumin+ GABA interneurons amongst other things. and we know that GABA is a neurodevelopmental signalling molecule.

    The abnormalities might be nothing to do with GABA. maybe VPA is just toxic for another reason. It seems unlikely, though, given that other anticonvulsants cause similar syndromes.

    there are various other rodent models of autism, based on genetic knockouts that cause autism in humans, and there is interesting converging evidence that they have deficient GABA interneurons.

    We know of a whole bunch of genes that when deleted cause autism (though usually not just autism, also other symptoms) in humans and a lot of them are involved in the formation of synapses e.g. SHANK. I guess in a sense everything in the brain is about the formation of synapses, but it's striking that they're not, say, receptors: Except for GABA receptors, interestingly.

    Ironically, the Markrams imply that deficient GABA probably isn't a major factor, but to my mind, their model makes more sense if it is, and I'm not sure if they've looked in the right places.

    So it's complicated but there's a story there if you want to believe it…

  • newgradstudent

    I think that push involving autism research (or research of any developmental disorder) must be interdisciplinary if we hope to understand the biological and psychological basis. How can you build a unified theory without expertise from different perspectives?

    Borrowing from schizophrenia research, recent work uses neuroendophenotypes (see neural signatures) to evaluate the connections between molecular/genetic causes and the phenotype of autism. I think this is a step in the right direction — using multiple levels of analysis to describe the brain differences.

    And @petrossa.. it's unfortunate that the professionals you have come across don't seem to know what they're talking about. Theory-development is daunting for the reason you mentioned (overly-specific narrow focus). I hope that someday I can contribute to theories of autism without excessive bias. It's my opinion that theory can only get us so far before it becomes cumbersome, unproductive, and drives us farther from helpful intervention strategies that enhance daily interactions.

  • http://latedx.wordpress.com/ latedx

    @ Kapitano
    I this this theory of over connectivity and under connectivity.
    MRI results of high functioning autistics shows there is an under connectivity in the frontal lobes and my own opinion is that the over activity is in the Somatic nervous system.

    I have autism and my mother didn't take any Valporate but I did have issues in the womb. It's my theory that there is no one cause of autism but it is a set of symptoms for brain abnormalities after trauma in the womb. I had RH disease and had an emergency C-section and blood transfusion.

  • http://www.blogger.com/profile/06647064768789308157 Neuroskeptic

    newgradstudent: Yes, the problem as I see it is that autism, more than just about any other disorder I can think of, spans every level of explanation.

    It has to start with molecules. We're not going to crack it without reference to genes.

    But then we'll need to work out how those genes affect the growth of the brain and that will mean lots of studies of cells and animals.

    But then we need to explain how that causes the symptoms, but before we can do that, we need to work out what the symptoms are and how they relate to each other psychologically e.g. is there a common cognitive thread linking all the symptoms, or are they just manifestations of abnormalities in different parts of the brain?

    Which is why I quite like the Markram's paper, there are a lot of holes but at least it tries to bridge the gap between molecules and symptoms. As I said I'm not aware of other attempts that ambitious but Michelle Dawson knows more about this than I do, so she's probably right that there are others.

  • http://www.blogger.com/profile/03449922526437153857 veri

    ..that must've been humbling. I was thinking that Kandinsky design wouldn't match my complexion but maybe a monochromatic baby tee will. Wouldn't it be cool if they had a luxury brand using science symbols. Like cufflinks with a symbol of your personalized DNA, embroiderd peptides, equations lightly printed on the back of a collar. Something subtle and charismatic, not geeky and loud.

  • http://www.blogger.com/profile/03449922526437153857 veri

    Van Huesen? had computer shirts. Looked really sexy. The pattern and texture reminded me of computer chips but softer and humane.. His and Hers EEG towels, my ECG across his chest. My gosh I think like a capitalist how disurbing.

  • http://www.blogger.com/profile/15107067137612954306 GamesWithWords

    @Neuroskeptic: I had the same response as @Michelle Dawson. There are lots of grand unifying theories out there.

    If we assume all behavior arises from activity of the brain, then talk about psychological mechanism and talk about biological mechanism is the same talk using different vocabulary. Since most of us are interested in autistic *behavior*, IMHO the vocabulary of behavior is the right one. But in any case, it's a red herring since lots of theories say *something* about biological mechanism.

    I wasn't sure, from your description, whether the Markmans have that much to say about behavior. The explanation of eye-gaze avoidance seems straight-forward, but how does their theory explain theory of mind deficits? Language deficits? Obsessive fixations?

    It seems like our knowledge of both behavior and its underlying biological mechanisms is so primitive, any grand theory is going to have to be pretty hand-wavy at this point. I'd love a really good theory of *one* aspect of *one* subtype of autism.

  • http://www.blogger.com/profile/05660407099521700995 petrossa

    This has been proposed all over the place (and ridiculed) but i see autism as an environmental driven genetic adaptation due to the way society is shaping up over the last 6 millenia.

    To my mind it's the brain adapting itself to another environment, were oldtime strategies don't work anymore or are mostly counterproductive (emotion driven behavior is nowadays a negative survival factor) plus the selection for intelligence.

    As with all adaptations the first 'models' are defective, the really hardcore deep autists. But as time goes on these nonviable mutations will get less and the viable ones will get more pronounced.

    By now it's generally acknowledged that gifted children are overrepresented amongst autists. High functioning autists come close already to being a more functional mutation.

    As soon as the point has been reached were the limbic system has been relegated to it's proper importance (somewhere along with the appendix) the human mind is freed from its restraints.

    I feel free. I would never ever want to be different. Emotions don't register much on my radar, leaving me free of the restraints they impose.

    The only real problems my 'afflication' gives me are caused by the intolerance of NT people. For the rest it only gives me positive things.

    A free thinking mind, an extremely retentive memory, a singular drive to reach a goal and and overall higher intelligence.

    Try telling that to your 'professional' You immediately get this big stamp on your forehead: Narcissistic personality with illusions of grandeur.

    After no matter what you say it goes in one ear and comes out the other with the person opposite you nodding 'wisely' going 'uhuh, scribble, uhuh, scribble'.

    Autism isn't a disease, it's an adaptation to the surroundings evolving.

  • http://www.blogger.com/profile/15225859145004971487 Jon Brock

    Interesting that you mention GABA. I've just blogged on a new study looking at gamma brain oscillations in a VPA mouse model of autism


    From my limited knowledge, there's quite a connection between GABA and gamma. The authors tried to 'rescue' the autistic mice using MPEP, which targets the glutamate system, although results in terms of gamma oscillations were a bit messy. Would be interesting to know if drugs directly targeting GABA would have an effect.

    Re grand unifying theories, there are a few out there (thanks Michelle for putting me in such illustrious company) although you're right in that they don't all try to link across so many levels of explanation.

    The problem for me is knowing what such a theory should actually try and explain. For example, lots of people with autism have high levels of anxiety. And there's pretty compelling evidence that lots have good visuo-spatial skills. But we don't know whether individual people who have high anxiety also have good spatial skills. So a unifying theory might be trying to explain the co-occurrence of two phenomena that don't actually co-occur!

  • http://www.blogger.com/profile/00187465138890222167 LokaSamasta

    The cannibal hypothesis of autism pretty much explains everything too. Try as one might, one just can't find fault with it!

    Apart from it simply not being allowable as an explanation for autism, that is.



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar