The increase in autism diagnosis
According to one paper [1] there has been a 787% increase in autism diagnosis in the UK from 1998 to 2018. This present article discusses what could be driving this, with particular reference to a recent longitudinal genetic study. We conclude that one of the major factors behind this massive increase in diagnostic rates is that the term “autism” has undergone a mutation. It has expanded to include “symptoms” which previously were not “diagnosed” as autism. We question whether this is helpful for people who are ‘actually’ autistic. By ‘actually autistic’ we mean. very specifically, lacking a Shared Attention Module and Theory of Mind module.
The financial and social drivers
The usual explanation provided by people who work in the industry and their media outlets, (“journalists”), is that “we are getting better at diagnosing autism”. Presented as the sole explanation and an end of the discussion, this is obviously preposterous. It is a sign of the times that the media and anyone takes this at face value. I would suggest that some key drivers are:
- Changing social attitudes; to have a “special needs” child is now a social advantage, not an embarrassment
- Increasingly competitive and individualistic society; a special needs diagnosis brings a whole host of advantages including, but not limited to:
- large social security pay-outs *
- extra time in exams
- individualised support and special exemptions for the child at school
- An increasingly willingness of psychiatrists and clinical psychologists to hand out a “diagnosis”, reflecting a change in the role of professionals, from authoritative figures linked to the state, to agents providing a service to consumers.
No rational person could exclude the above as drivers for the massive increase in “diagnoses” of autism. Which is not to say, that, if “symptoms” could be quantified there is not some actual increase in autistic-like behaviours or, as we shall see, perhaps a change in what “autism” means.
Two “types” of autism?
In relation to other factors, a very interesting longitudinal study has just been published. [3] The study found that:
These findings indicate that earlier- and later-diagnosed autism have different developmental trajectories and genetic profiles. Our findings have important implications for how we conceptualize autism and provide a model to explain some of the diversity found in autism.
That is; people who are diagnosed in childhood have a different genetic profile, (statistically speaking of course), than those diagnosed later. Those diagnosed later have genetic profiles which correlate more closely with those characteristic of “ADHD” and mental illness than those diagnosed younger. Tentatively; this supports the thesis that the increase in diagnostic rates for autism is because the criteria have expanded and – that people who are basically, not autistic, are being diagnosed as autistic.
Theory of Mind Reading
At this point I should state my foundational position about “autism”. Currently, at least, I agree with the theory of “mind-reading” of Simon Baron-Cohen. [4] He posits a theory of brain development which is modular. Autism is the result of a specific developmental gap. The necessary “Shared attention module” and “Theory of Mind” modules have not developed. These modules are posited as functional and tied to specific areas of the brain. The former enables the ability to share attention on a common object with someone; to understand that they are looking at the same object you are and to understand they they know that you are. This is a prerequisite for the next module in a development and functional sequence: “Theory of Mind”, which enables you to posit and attribute mental beliefs and perceptions to other people, to make educated guesses about what they are thinking and feeling. ** These modules are necessary for “normal” functioning. Autism is a condition which results from a lack of these modules. The lack is genetically determined. This provides us with, at the theoretical level, a clear definition of autism; what is it is behaviourally, existentially for a person, and genetically. **
The hardware/software model
At this point, I would like to introduce a theoretical model which I find helpful. It is an analogy from computing science. In information technology, a “behaviour” of a system can be the result of either hardware of software. This applies equally to desired features of a system and to bugs. To take a simple example; if a computer monitor is flickering the problem can be either the result of a hardware issue, for example a circuit on a memory block is physically broken, or it could be the result of a software bug. The distinction has important consequences. If the issue is the at the hardware level it can only be fixed by replacing the broken hardware. However; if it is at the software level the problem is temporary and can be fixed, for example, by installing a new version of the software driver which contains a fix for the buggy code. (We should add that it might also, in some cases, be possible to get around a hardware issue with a software solution. NASA frequently has recourse to this option when dealing with malfunctions of its space probes. And, to apply the analogy to human beings. I am aware that one approach to treating people with brain injuries is about getting them to re-learn skills; using other, non-damaged, areas of the brain). In very general terms “mental illness”, especially “depression” might be more of a case of a software bug, a temporary and fixable condition, whereas autism in the sense of a lack of Shared Attention Model and Theory of Mind modules, is likely to be an unfixable hardware issue. (Though, following the model of software adaptations for hardware issues, Baron-Cohen has developed a training programme to help people without Theory of Mind develop some basic functioning in this area).
The possible cases: an increase in genetic malformations, an increase in software symptoms, or an increase in diagnostic rates?
It then remains to ask; are autistic “symptoms” actually, objectively increasing? If autism is understood specifically, by definition, as a genetic condition, and if symptoms are actually increasing, independent of any increase in the rate of diagnosis, then this would imply an objective increase in genetic malformation over a 25 year period. Is this possible? Unfortunately; I am not enough of geneticist to say much about this. Perhaps, one could explain such a rapid genetic effect by reference to environmental pollutants causing genetic malformation in babies in the womb, rather than generational mutation and inheritance. There is, of course, a theory, that people with autistic traits are likely to be drawn to other people with autistic traits and this could lead to an objective increase in autism via genetic inheritance, but I am not sure that this could explain a 787% increase in a 20 year period; presumably not. Even the environmental damage explanation doesn’t seem able to, even theoretically, account for the 787% increase over 20 years. So; an increase in actual genetic factors is unlikely, but not excluded altogether. It is not easy, however, to determine to what extent the increase in diagnostic rates is due to an increase in “symptoms” (without a hardware basis) or simply an increase in diagnostic rates.
The increase is explained by a widening of the term “autism” to include people with software symptoms, who, previously, would not have been diagnosed as autistic
The genetic paper [3] identified two groups of autistic people; those with an early diagnosis were more closely associated with a typical autistic genetic profile and those with a later diagnosis were more closely correlated with a genetic profile typical of those with an ADHD or mental illness diagnosis. (Of course; all this is about statistical correlations, not causal pathways). The obvious temptation is to say that the increase in diagnostic rates is due to people in the second category who were previously not diagnosed at all now being diagnosed as having autism. In support of this, we can note that it is established that the increase in rates of ADHD “diagnosis” are linked to an increase in adult diagnosis. [5] If, this explanation is accepted, it would seem to relate to an expansion in the concept of “autism”.
This does not exclude the possibility of an actual increase in “software symptoms” but in fact, ‘actual autism’, that is lack of Shared Attention Module and Theory of Mind, is quite hard to fake, or to produce as a result of mental action/inaction. So; I would be inclined to look in the direction of an increase in “software symptoms”, or mental illness, as responsible for an increase in autism diagnostic rates. This thesis is supported by the thesis that ‘actual autism’ is genetic and genetic factors are unlikely to able to account for a 787% increase over 20 years.
Without taking the point too far, one could argue that adults, or parents of children, without serious, hardware deficits are “jumping on the bandwagon” to claim they have “autism”, in order to acquire certain benefits and exceptions. They may have real difficulties, but these are “software” difficulties. (Incidentally; if the government simply legalised cocaine or at least methylphenidate without prescription the rates of adult ADHD would drop rapidly as they would no longer need to get a diagnosis to get access to the stimulants).
Conclusion
We have concluded that it is unlikely that the increase in autism is due to an objective increase in genetic factors, though we have not completely ruled this out. Referencing a genetic study which indicated two groups of people with an autism diagnosis, we have put forward the hypothesis that the increase in autism diagnosis rates is likely to be due to an expansion in the term “autism” to include people who do not meet the criteria for what we are terming ‘actual autism’’; a lack of Shared Attention Module and Theory of Mind module caused by genetic factors. People with “mental illness” are now falling under an autism umbrella.
The umbrella is widening. This correlates with the other factors; a possibility that there really is a greater incidence of “mental illness” such as “depression”, and, of course, the driver that a diagnosis opens access to financial privileges and other benefits. We have not attempted to delineate between these two factors, which would be extremely hard to do. But we note that once the “autism” term is broadened to include “symptoms” such as those characteristic of ADHD and “mental illness” it becomes harder to distinguish between “real conditions” and people “faking it”. (The sympathetic view is, of course, to say that even if someone is exaggerating certain “symptoms” this still reflects a real difficulty in their life ***).
The implications for support provision
Clearly; resources for support are not unlimited. Special schools can increase their vigilance and only accept people with actual autism; those in the first category, with a real deficit in Theory or Mind module and Shared Attention module. However; mainstream schools are in danger of being overwhelmed with demands for support for young people “with autism” who, in fact, have, temporary, software conditions. With this in mind, it would be desirable to try to limit autism diagnoses to those in the first group.
* * PIP – a generous UK social security benefit paid to people regardless of whether they are in work or already on other benefits has seen an explosion in claims for psycho-social “disorders”. [2]
** In his book Mindblindness: An Essay on Autism and Theory of Mind, Baron-Cohen discusses a secondary form of autism. in which children have a limited ability to read the thoughts and beliefs of others; that is they have a limited Theory of Mind Module. These children, who form around 20-35% of autistic children, have developed further along the “normal” pathway before the disruption occurs. It follows, therefore, that parents and teachers are likely to notice that “something is wrong” with these children later. But, we can suggest the following; one reason for the increasing rate of diagnosis is that people in this subgroup, who, previously, just managed, are now falling under the diagnostic umbrella. This could be another explanation for the increase if diagnostic rates.
*** A very interesting thesis put forwards by an NHS psychologist in relation to this point is that increasing levels of “mental illness” really reflect unhappiness caused by living in a materially unequal and unjust society. https://substack.com/home/post/p-152792026 The Therapy Industry. Paul Maloney. Pluto Press. 2013
Notes
- https://acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.13505
- https://www.telegraph.co.uk/news/2025/07/05/mental-health-benefits-granted-531-people-a-day-last-year/
- https://www.nature.com/articles/s41586-025-09542-6
- https://thethinkingteacher.substack.com/p/mindblindness-an-essay-on-autism
- https://www.clinicaltrialsarena.com/analyst-comment/rising-adhd-diagnoses-7mm/?cf-view
[Image credit: https://commons.wikimedia.org/wiki/File:Autistic_Mind_5.png]
