Testing the “Extreme Female Brain” Theory of Psychosis

Source: Christopher Badcock

Up until recently, the imprinted brain theory (above) had only one serious rival: the so-called "extreme male brain" theory of autism (EMB). And that was only half a rival in the sense that it applied only to autistic spectrum disorder (ASD), not psychotic spectrum disorder (PSD). But now a new study by Felicity Larson, Meng-Chuan Lai, Adam Wagner, the MRC AIMS Consortium, Simon Baron-Cohen, and Anthony Holland has remedied the situation and added what they call the "extreme female brain" theory of psychosis (EFB). This formidable authorship (30-odd in all) also chooses to designate the imprinted brain theory “the diametrical ASD-P model,” where “P” stands for psychosis. This is because the theory is based on the diametric model of cognition, which proposes that ASD represents an extreme deficit in mentalism (hypo-mentalism), while PSD corresponds to what it calls hyper-mentalism (also diagrammed above).

According to these authors, “Brosnan et al. [*] tested if psychotic illness more broadly might represent the EFB, in the same way ASD represents the EMB,” adding that “This would provide directly measured behavioral-cognitive evidence supporting the model.” In one of the most widely read of all these posts, I commented on that claim, and will not repeat the point here that, despite using Baron-Cohen's methods, Brosnan et al. effectively endorsed the Crespi-Badcock theory. These authors go on to assert that “Crespi and Badcock […] proposed that if the EMB leads to autism, because of ‘mentalizing’ deficits, the EFB may lead to psychotic illness and paranoia, because of excess and inaccurate mentalizing.” Furthemore, they argue that

In the context of mixed genetic and clinical evidence, a more parsimonious explanation is that in at least some cases, ASD and psychosis share causal factors and are thus related in a fundamental biological sense. In turn, this suggested there may be shared behavioral features between the conditions, rather than placing them as opposites on a spectrum.

Specifically, the authors comment that

It is unclear why, therefore, the diametrical ASD-P model compares a stable, lifelong condition that develops early in life (ASD) with what could be considered an extreme and acute manifestation of schizotypy (psychosis).

But there is nothing unclear if you think about it for a moment. The relevant difference between PSD and ASD—and one, furthermore, that the imprinted brain theory is unique in readily explaining—is that ASD is an early-onset condition often diagnosed in childhood but that PSD is a late-onset one that usually does not show itself until after puberty simply because mentalism takes years to master. Furthermore, if ASD represents arrested mentalistic development, PSD represents pathological over-development—hyper-mentalism—which will inevitably look like “an extreme and acute manifestation of schizotypy.” 

“In order to address some of the issues with the EFB” the authors “investigated empathizing and systemizing in individuals with a dual diagnosis of ASD and psychotic illness.” As these authors rightly note, the diametric model implies that “any case of apparent co-morbid psychosis and ASD is likely to be a misdiagnosis, rather than a true co-occurrence of two distinct conditions.” Indeed, had the authors been able to bring themselves to recognize the fact, they could have added that this is also the conclusion strikingly implied by what has been by far the biggest and most searching test of the imprinted brain theory so far. This was a Danish study of a population of 1.7 million which revealed that risk of ASD co-varies inversely with that of PSD as evidenced by birth size interpreted as a proxy for the expression of imprinted genes: increased birth size increases risk of ASD but reduces that of PSD, and decreased birth size does the opposite, just as the theory and its diametric model of mental illness predicts. And clearly, if this is correct, the ostensible test of the theory we are discussing—which is miniscule by comparison—is cast into doubt. The implication must be that most or all of this study's 65 "co-morbid ASD and psychotic illness" cases may indeed have been misdiagnosed—at least if you define "co-morbid" as suffering from serious symptoms of both types of disorder at the same time. And of course, choosing such a group to test the diametric model sounds like playing the game with loaded dice!

However, a more subtle and insightful interpretation of “co-occurring autistic and psychotic traits" is possible, as I pointed out in a recent post. Ahmad Abu-Akel and his colleagues argue that such co-occurence can exert opposing influences, "producing a normalizing effect possibly by way of their diametrical effects on socio-cognitive abilities." The implication is that "some individuals may, to some extent, be buffered against developing either illness or present fewer symptoms due to a balanced expression of autistic and psychosis liability.” Indeed, according to the diametrical model's more humane, modern view of mental illness, you could see normality as such a state of "co-morbidity," with autistic and psychotic trends not just in balance, but cancelling each other out! 

But however all that may be, what of the results? In the words of the authors,

There were overall differences in the distribution of cognitive style. Adults with ASD who had experienced psychosis were more likely to show an empathizing bias than adults with ASD who had no history of psychosis.

This is hardly surprising, indeed, it is exactly what—making due allowances for the contentious issue of co-morbidity—the diametric model would predict. The authors add that “This was modulated by IQ, and the group-difference was driven mainly by individuals with above-average IQ.” Nor is this surprising if you note that, according to my diametric interpretation of IQ, there are two types of intelligence—mentalistic and mechanistic—with the latter being stressed in current tests of the kind used in this study. Inevitably, this would reveal group differences between populations at different ends of the mentalistic continuum because autistics can be gifted in many mechanistic skills while relative superiority in verbal as opposed to spatial skills is a feature of PSD, as I pointed out in a previous post.  Finally, and again as the imprinted brain theory would predict if you regard mania as a psychotic disorder and empathizing as an aspect of mentalism influenced by maternal and female genes (top), “In women with ASD and psychosis, the link between mania/hypomania and an empathizing bias was greater than in men with ASD.” 

* Brosnan M, Ashwin C, Walker I, Donaghue J. Can an "Extreme Female Brain" be characterised in terms of psychosis? Pers Indiv Differ. 2010; 49: 738–42.

(With thanks to Ahmad Abu-Akel for bringing this to my attention.)