The Epidemiology of Autism Spectrum Disorders in Adulthood

Why does “epidemiology” matter? According to the World Health Organization (WHO: http://www.who.int/topics/epidemiology/en/) epidemiology is “the study of the distribution and determinants of health-related states or events (including disease).” As a series of reports published by the Centers for Disease Control, Atlanta (Centers for Disease Control and Prevention, 2012; Yeargin-Allsopp et al., 2003) over the past decade have shown, many children with autism^{Footnote 1} still go undiagnosed and unrecognized by services. As more children get recognized this number may be falling, which is progress. However, the differences in childhood recognition rates between the lowest and highest estimates are huge when comparing different areas within the USA. And what about adults? Who is working to recognize and diagnose them? All of this shows that there is no room for complacency about the number who remain undiagnosed.

Parents of children with autism wonder what will happen when the children have grown up and there is no one to care for them. It does not help that up to now everything we knew about autism prevalence (and much else about autism) related only to children. Yet most people alive now are no longer children. So what can epidemiologists offer? Epidemiologists are fortunate in that they can go where services and professional practitioners do not: they can “case find.” Case finding utilizes techniques for identifying people with physical and mental conditions whether or not they have been recognized or diagnosed by services. Only epidemiological methods can tell us the complete answer to questions such as: how many, who, and what are the characteristics of people with the condition. A complete answer to these questions is required so that the range of service provision needed can be adequately costed and planned for.

Professionals (doctors, teachers, psychologists, etc.), members of support groups (for people with the condition, for carers including family members) will all have their own perspective on these “how many” and “who” questions. And that perspective will depend partly on what we learn through people known to have the condition. But only a thorough collection and analysis of data systematically carried out across a defined population will give us the complete answer that includes those who are known to have autism and those who have autism but nobody knows about it. This chapter, as befits this book, casts light for the first time on that bigger population of adults with autism—both known and unknown. How many are there? What are their lives like? It tells us how many may be recognized or may not be. And therefore it can provide a voice for those who are not recognized. There is no advocacy group or organization primarily speaking for adults with autism who are undiagnosed—so who will speak up for them? What happens to individuals identified with autism in childhood when they grow up? Where do they go? This chapter is dedicated to them.

Before talking about what happened when our research group tried to find out how many adults there are with autism, it is worth reflecting on how little was generally known when we started. Before then no one had tried to find out how many adults have autism! But there are some sources of information that have provided some early clues. Studies following up adults who were diagnosed as children have shown that the condition does not go away (Howlin, Goode, Hutton, & Rutter, 2004). Worryingly, one set of studies suggested that adults with autism and intellectual disability may live shorter lives (Pickett, Paculdo, Shavelle, & Strauss, 2006), but this could be in part because adults with intellectual disability, particularly men and particularly those with epilepsy, have shorter lives.

Adults in Great Britain who have responded to postal and online surveys (National Autistic Society, 2008) stating that they have an autism spectrum disorder (ASD), are more often male (2:1), rarely aged 65 years or older and are rarely in full-time employment. They tend to be given a diagnosis of high functioning autism or Asperger syndrome, with 1 in 5 in receipt of psychological or psychiatric services. There are also many single case studies and indeed illuminating autobiographical accounts. But what about information on the lives of those undiagnosed? At best we have personal accounts from those diagnosed, looking back to life before their condition was recognized. In other words, quite a lot of hindsight, but little foresight.

Since completing our first study on the number of adults with autism in the population two other studies have been reported that add to our background of knowledge. An important new US study (Shattuck, Wagner, Narendorf, Sterzing, & Hensley, 2011) provides evidence to underpin the widespread concern that the transition from childhood to adulthood can result in unmet need for help with education and access to employment, even for children with recognized special education needs and autism. They used data collected in 2007–2008, from the National Longitudinal Transition Study 2 (NLTS2), a 10-year prospective study of youth receiving special education services, which included 680 youth in the autism category, 500 of whom were no longer in high school. For youth with an ASD, 35 % had attended college and 55 % had held paid employment during the first 6 years after high school. More than 50 % of youth who had left high school in the past 2 years had no participation in employment or education. Youth with an ASD had the lowest rates of participation in employment and the highest rates of no participation compared with youth in other disability categories. Similarly, a small study (Balfe & Tantam, 2010) described 45 teenage and adult individuals with Asperger syndrome or high-functioning autism who replied to an advertisement. It found that most were still living at home with parents, and had trouble understanding and responding to other people’s feelings, coping with life changes, and managing life skills such as cleaning and managing money.

Up to now, the answer to the question “how many” could only be estimated for children. In three recent large region-wide or national community surveys of children and young people in England (Baird et al., 2006; Baron-Cohen et al., 2009; Green, McGinnity, Meltzer, Ford, & Goodman, 2005) the prevalence of ASDs was approximately 10 per 1,000 children.^{Footnote 2} When a government National Audit Office research group (National Audit Office, 2009) asked all local health and social care service providers throughout England to say how many adults they knew with any form of autism, the answer seemed to be about 1 out of every 20 cases that should be recognized assuming a prevalence of 10 per 1,000. This would suggest that either most adults with autism were sufficiently well and independent not to need services as adults or it meant that a lot of adults with an autistic condition were “off the radar screen” and getting no help. One of the hopes of our study was to find out which (if either) of these possibilities was true.

There is evidence that individuals with an autistic condition are more likely to be diagnosed if they have another serious problem that brings them to attention such as another mental disorder, or difficulties with adaptation due to low levels of general intelligence (National Audit Office, 2009). In childhood, ASDs are associated with intellectual disability, male sex and an increased risk of epilepsy in older children. Among significantly intellectually disabled adults (less than 0.5 % of the overall adult population), a rate of autism of 75/1,000 was obtained from an intellectual disability population register (Cooper, Smiley, Morrison, Williamson, & Allan, 2007). The cases were identified from direct observation, detailed case records and interviews with carers, so it would be important to include adults with intellectual disability in any work on the number of adults with autism.

So, in summary, before the work described in this chapter began, we really had no idea how many adults were affected, what kinds of lives they were living and what factors were associated with having autism as an adult.

Previously, no one had ever done a survey to look at rates of autism among adults so to do this a research team needed to develop an approach. Surveys are going on all the time around us. They seem to be in the news every day—for example, opinion polls on who should be the next leader of the country—so we all know a bit about surveys and lots of us have been a respondent in one. If you are a survey specialist there will probably be no need to read the next few paragraphs and indeed you may prefer to read the technical scientific reports on our work (Brugha et al., 2009, 2011; Brugha, McManus, et al., 2012; McManus, Meltzer, Brugha, Bebbington, & Jenkins, 2009).

Any scientific or technical topic can be daunting and confusing so we are setting out a brief introduction to survey methods here. If you would like to read more, an experienced survey expert colleague and I have put much of what you need together in a recently published chapter in a public health textbook (Brugha & Meltzer, 2008). More knowledgeable readers can skip over much of the next section.

Let’s start with the two questions we were trying to answer with our survey. One, how many people have the condition and, two, what are the characteristics of people who have it (e.g., gender difference and likelihood of having been exposed to a possible environmental or genetic cause)? The ideal way to answer these kinds of questions is to ask them of everyone, as is done every 10 years in the UK Census. However, this is not a feasible solution for many more complex and detailed questions. So we do a survey—which means we select a much smaller group of people from the whole population and just ask our questions of them. But how can we be sure the answers for our group also apply to the whole population? We can’t, but the answers are more likely to be representative if everyone in the whole population, that means you and me, has a known chance of being in that group (we call the group a probability sample, or a purposive sample). We achieve this by taking a random sample: that means drawing up a list of “everyone,” for example an electoral roll, and choosing a group of people from that list on the basis of chance (using a computer program to speed things up considerably!). Compared to a convenience sample, such as persons responding to an advertisement, a randomly chosen group like that is far more likely to be representative, with about the same proportion of men and women, old and young, working and not at work, as that found in the population as a whole. (Indeed, in a survey, we compare the characteristics of the sample group with the census and where the two differ we make adjustments for those differences by weighting the responses).

Before saying more about the survey approach, let’s not rule out the idea of a census completely. For example, if you want to know how common autism is among children aged 8 years old in a city or a county, you could, with permission, examine all health and educational records on 8-year-old children for information suggesting that a child might have the condition of interest. And then you could ask a qualified health professional to examine those records more closely, or, more expensively, you could actually examine those children to find out which ones have the condition.^{Footnote 3} The CDC studies, referred to at the start of this chapter, do something quite like this. Child records are examined in defined geographical areas, although the children themselves are not re-examined. The problem with this approach is that it is only as good as the records—and indeed only as good as the services are at deciding which children should have a health check or an educational test that is accurately recorded. The fact that the number of children with autism found by the CDC researchers varies so much from one place to the next suggests that the quality of the records, or the services, is very variable between places. This puts some limitations on the value of statistics derived in this way, as the CDC authors have spelt out carefully and in detail.

So, returning to the survey sampling method and how to choose a group at random from the population—what are the advantages and drawbacks there? The big advantage is that the survey goes directly to representatives of the whole population—making no assumption about the quality of existing records (except for the quality of the list [sampling frame] from which the survey sample group is drawn). So when it came to studying adults, where there are known to be very poor records of who is affected by autism (National Audit Office, 2009), it was a “no brainer”: we had to choose the survey, with a direct assessment method. Since no one had done this before we had to invent specific methods for surveying autism in the adult population. Now that we’ve done it, lots of people have asked how we did it when no one else had managed to do it before. It’s a story worth telling—if only in the hope that others will now do the same in other populations and improve on our approach!

By now you will have realized that there is a survey cost issue. Examining people for a condition is a lot more expensive than checking records or asking people easy to answer questions about themselves. A medical or psychological examination for a complex condition like autism is very expensive so how can it be done at a reasonable cost? The answer is that we conduct the survey in two phases, and examine fewer people in the second phase. The first phase in our survey included screening questions for autism, and was based on interviews conducted face to face in people’s own homes. In the first phase we selected a large group of people and an interviewer went in person to their home and asked them many questions the answers to which might indicate that they have autism. But these questions are not enough to establish whether they definitely have autism: that requires a clinical assessment. But it is too expensive to assess everyone in that way as it takes a long time and can only be done by a specially trained interviewer.

By taking the survey sample method to a second phase, we resample from our first phase survey sample. Having taken a sample from a population and asked those willing to answer questions about their health in phase one, we then chose an even smaller subgroup to agree to a more detailed clinical examination, in a second phase, at a later date.^{Footnote 4} And to choose for the second phase a balanced or unbiased sample, we chose for the clinical examination some who look more likely to have the condition and some who do not. A tricky part of this is how to work out how likely someone is to have the condition. If this were the common cold or the flu you could just ask “in the last week have you had a cold or felt like you had the flu?” But most people would struggle with the same question if you replaced the words “cold” and “flu” with “autism” and “Asperger syndrome!” How we tackled this we’ll come back to later.

Now that we know roughly when and how to select a sample we need to consider how large the survey needs to be. For this we need a statistician who will turn this question around and ask you what exactly is the question that you are trying to answer? Let’s think of another example to illustrate this. In an example “leader of our nation” question a good answer is we want to know which candidate is ahead; but a better question might be to ask is: “is the most popular candidate far enough ahead in the voting population to win the election?” Thus, because it is an opinion poll of about 1,000 adults and not the election itself, we need to know if the front-runner’s lead also exists in the population at large or whether it just happens that lots more of his supporters were sampled by chance. A second poll on a different random set of 1,000 electors on the same day might show a 1 % difference but the other way around. 1,000 may not be enough if we are interested in looking at small differences.

So what questions do we want to ask about autism in the adult population? If it is to ask how many adults have autism that could imply an interest in making a comparison with the frequency of this compared with other conditions; or it could be that an estimate is needed that is precise enough to inform the planning of services and to help to decide what they would cost to deliver. Or the question we might want to answer could be: “does autism affect as many older people as younger people?” For example if the true rate of autism is rising, as surveys of children over recent decades seem to suggest, you would expect fewer cases in older than in younger adults. This turned out to be the question that was of the greatest public interest when the results emerged, although we should emphasize that our study was not designed to answer this question specifically. Given the general interest in this question it proved fortunate that the adult survey began with over 7,000 willing respondents which provided some degree of precision to answer the question—“is autism associated with age?” Had our survey been much smaller we could not have done this.

So now you are probably wondering how we drew up our sampling list, and how we examined adults to decide which ones had autism? And how did we do this for adults of all levels of ability? Starting with the lists, we actually had to draw up two different lists because of the problem of range of abilities. Most adults live in what we call private households, such as an apartment, or just a room with shared bathroom and cooking facilities, or a house. But some adults with very low intellectual ability cannot care for themselves unaided and either live in a household together with a carer or in some kind of institution. We did this survey throughout England. The first of our two lists, covering adults able to manage without a carer, started out with a list of all postal delivery addresses (excluding large ones), known as the “small postcode address file.” In Britain, the Post Office has a list of all the addresses in the country to which a letter can be sent. This list is available to survey research organizations for selecting addresses for surveys. Such a list is not available in the USA. In England postal delivery workers are paid the equivalent of about 50 cents each time they update the list with new information—for example with a new house or apartment. Therefore the list is kept very up-to-date.

Addresses from a list of addresses throughout England were selected at random (to save on the time and travel costs of interviewers this was done in randomly chosen postal areas). Trained survey interviewers then visited the addresses and after excluding those that were found to be unoccupied (and overlooked errors in the list) they tried to make contact with an adult resident. If there was more than one adult living in the household, one was chosen at random and asked to take part. So we are confident that every adult living in a private household^{Footnote 5} in England when the survey was taking place had a known chance of being counted in the survey. About 59 % of those asked agreed to participate (we were able to run checks to compare those who did and did not agree to take part so that we could make adjustments to allow for differences, or “bias,” in the sample (Brugha et al., 2011)). For example, we found that younger men were least likely to cooperate when asked so we were able to make adjustments to our final results—the so-called “weighted prevalence.”

Fortunately for us, our second list already existed in the form of three population registers for adults with intellectual disability (as explained recently in a report on this second group (Brugha, Cooper, et al., 2012)). These registers cover about 1.5 million adults in three places in England, including rural, urban and large metropolitan areas. Our plan was to choose 500 adults from these three registers and examine them as we had done in the household study. But one potential pitfall from sampling from two lists is that someone could get counted twice, i.e., was in both lists (we also did some elaborate checks for adults who might have been left out of both lists, such as young adults living in college or in military establishments (Brugha, Cooper, et al., 2012)). We completed the study of the household sample in 2007. When we then began in 2010 to examine adults chosen (at random) from the intellectual disability registers we checked each one living in a private household to see if they could have answered the questions we asked in the 2007 sample and excluded those who could because otherwise someone of the same ability could have been included in both parts of the survey.

Autism is a developmental disorder and clinicians rely heavily on information on early development, such as the age at which the child begins to use language. In clinics, like the first author’s adult clinic, we try to interview a parent but our older adult patients often no longer have a living parent or one with good enough memory of their early childhood. Child surveys have had it real easy as there is nearly always an adult available but our survey included adults aged 60, 70, 80 and older! So the search was on for a method that could be applied directly to all adults including elderly adults who do not have a living parent or suitable informant. A further issue was how did we clinically examine adults across all ability levels to decide who were autistic in an equivalent way? This was not easy because no one had really developed and tested methods for adult surveys of autism.

We were fortunate in having advice and excellent training in the use of the Autism Diagnostic Observation Schedule (ADOS; (Lord, Rutter, DiLavore, & Risi, 2002)). The ADOS includes modules for all ability levels from the preverbal (child), module 1, to the older adolescent and adult, module 4. In the first part of the survey in private households we used module 4 and in the second part, in the intellectual disability registers, we mainly used module 1 (adapted for adults). For the more able adult in an institution who understood and could answer questions we used ADOS module 4. Thus we were able to combine the results of both modules to determine which adults met criteria for autism from the least able to the most able. The authors recommend a total score of 10 or more (“10+”) on module 4 and of 12 or more (“12+”) on module 1 in order to judge that a person meets criteria for autism. We checked these thresholds in our surveys in separate interviews with parents and carers using the Autism Diagnostic Interview-Revised (ADI-R (Lord, Rutter, & Le Couteur, 1994)) and a longer more detailed interview that also checks for other developmental disorders called the Diagnostic Interview for Social and Communication Disorders (DISCO (Wing, Leekam, Libby, Gould, & Larcombe, 2002)). We found that the same thresholds were appropriate for both parts of the survey (Brugha, Cooper, et al., 2012; Brugha, McManus, et al., 2012). For this extra work we purposively sampled individuals with a high risk of autism and a random sample of individuals with a low risk of autism.

Another important difference between the two parts of the survey was that in the intellectual disability sample, where we had originally wanted to examine 500 adults, we were able to use the appropriate ADOS module in all consenting cases. But our adult household survey covered over 7,000 adults across the whole of England. As mentioned earlier we had a two-phase design (a survey of a survey) but we needed some good questions in the first phase to help us decide who is more likely to be autistic and therefore more worthwhile choosing to undergo these more detailed clinical assessments. This is explained in greater detail elsewhere (Brugha et al., 2011) so here we will just concentrate on the headlines. Essentially, we needed a questionnaire an adult could fill out themselves that helps identify who is more likely to have autism. The obvious candidate was a very popular questionnaire that has also been available for anyone to complete online for many years (http://www.wired.com/wired/archive/9.12/aqtest.html) better known in the academic world as the Autism Spectrum Quotient (Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001) or “AQ.” We worked with the authors of the AQ to shorten this down from 50 to 20 questions and got permission to include them in the third adult mental health household survey of England that was in the field in 2006–2007 (McManus et al., 2009). That survey had a two-phase design: phase one was managed by one of us, SM, at NatCen (http://www.natcen.ac.uk/) and phase two by TB at the University of Leicester (http://www.le.ac.uk/). The phase two, clinical phase, examinations covered psychosis, anti-social and borderline personality disorder and, for the first time, autism. As explained elsewhere (Brugha et al., 2011) the phase two autism estimates from the clinical evaluations were weighted (back to phase one) to provide estimates for the household population. We also got funding to extend the survey to adults identified through the intellectual disability registers in England (Brugha, Cooper, et al., 2012) in order to obtain combined overall rates for adults of all ability levels.

In this chapter we are going to spare you the finer technicalities (available in our publications referred to already) and start looking at the main results and discussing what they mean. We start with the 2007 household survey because adults living in the community are less likely to be getting help and attention from services (Brugha et al., 2011). In the first phase of the household survey 7,403 adults completed the 20 question AQ and many other questionnaires relating to their health and life circumstances. We selected 849 for a clinical assessment and the ADOS-4 was completed on 618 adults.

It turned out that the AQ was not very good at picking out potential autism cases: at best sensitivity was 74 % (the proportion of true autism cases picked up) and specificity was 62 % (the proportion of people who did not have autism and were correctly identified as not having autism). Therefore we found in the second phase very few people, 19 in fact, who met criteria for autism. In spite of this, we were still able to tell how many had autism overall and to look at factors that might be associated with having the condition (Brugha et al., 2011). We did not expect any of our measures to neatly divide the world into people with and without the condition. When compared with how often they occurred, the AQ scores (ranging from 0 to a maximum of 20) showed that the majority of people had a score somewhere in the middle of the scale—between 3 and 12—and very few had a low score 0–2 or a very high score, 12+. Thus there was no obvious extreme “bump” (peak) at the top or bottom of the AQ scores (for example IQ shows a bump at the lowest scores accounted for by people born with intellectual disability). The scores on the ADOS-4 were similar, with no obvious separation for the group with autism, but the distribution of scores differed from the AQ in that most of those examined scored zero.

Very few, about 1 % of adults, had a score of 10 or more, which is the score recommended for a diagnosis of autism (Lord et al., 2002) on module 4 of the ADOS. But almost 1.5 % of the population scored at a lower threshold of 7 or more; just over half a percent scored at a higher level of 13 or more (Brugha et al., 2011). This is how we concluded that autism affected 1 in 100 adults sufficiently “able” to live independently in private households and answer survey interview questions. All of this suggests that having autism is a question of degree and not an either/or finding. But planners still want a number for how many cases there are, in preference to an average score, and most of us prefer a number so that we can make comparisons between groups of people easy to understand, which is what we did next.

The most obvious comparison was between rates in adult men and women. We know in childhood that boys have been consistently more often affected (or identified) than girls. We found the same result in our general population household study: just under 2 % of men hit the 10+ threshold and only 2 per 1,000 (0.2 %) women, which is ten times fewer than for men. This gender difference is so great that we have to wonder whether the ADOS-4 is less good at picking up autism in women. In fact, there is growing interest in the research community in the possibility that there are many “missed” cases of autism in women because the focus for the last 50 years has been so heavily loaded to the male presentation of symptomatology. If we are to consider the historically presented male:female ratios of autism one sees that in individuals with autism coupled with intellectual impairment the rates are 4:1 males to females. However in individuals with autism who do not have clinical levels of intellectual impairment the rates are around 9:1. Is it really the case that intellect in women is a protective factor against having autism? Or are we missing a significant proportion of more able women with autism for other reasons? Whilst there is a popular theory currently that autism is an extreme form of the male brain, and biological factors explain the gender differences (Baron-Cohen et al., 2011) this does not seem to explain the intellect-based differences. Some authors suggest that women with autism are “masked” by other conditions, such as eating disorders (Rastam, Gillberg, & Wentz, 2003), depression or anxiety disorders, or borderline personality disorder (New, Triebwasser, & Charney, 2008). This may be the case, as adult psychiatry services are more likely on the balance of probability to assume a women presenting for assessment has a mood or personality disorder rather than a “male” condition such as autism. Similarly, as suggested above, existing diagnostic tools are developed to target the “typical” male presentation of symptoms and may therefore miss the female presentation, which research is indicating is different to that of males (Lai et al., 2011). Epidemiological studies in the future may need different tools and approaches to establish whether there is, in fact, a hidden population of women with autism.

The result that attracted most interest was the effect of age—or the apparent lack of an obvious age trend. As a clinician the first author was expecting to find lower rates of autism in older people because it is his impression, talking to affected adults, that increasing age and maturity brings some degree of improvement and because the descriptions given by parents seem to indicate that things were a lot worse in childhood. And of course the theory of an autism epidemic would predict that the proportion born with autism 50–80 years ago is far less than it is now. But we did not find that. In fact what we found when examining the survey data was that there was no obvious association between having autism and age. The likelihood of having autism appeared to go down very slightly as people got older. Expressed in more technical language: the predicted probability of being an autism case suggested a very gradual decrease with increasing years, although the trend was not significant; thus for every extra year of age the odds of being a case would decrease by 1 %.

Current research suggests that the causes of autism are strongly genetic but that there remain some environmental influences. If there were major environmental effects that had begun or increased in recent decades you would expect those to be associated with lower autism rates among older people. Instead it seems more likely now that the possible environmental causes are not actually new causes, which suggests that they are factors that have been in our environment not just in the last 20 or 30 years but actually with us in the last 50–80 years and probably longer. This finding means that scientists need to be broader in their search for environmental causes of autism: it cannot be all down to new factors like the spread of cellular telephones, computer games, use of the internet or the introduction of new medicines and vaccination programs (although our study cannot rule out small hard to detect effects for any of these factors).

The results of the survey also confirmed two other important factors long thought to be associated with autism, low IQ (Brugha et al., 2011) and epilepsy (Rai et al., 2012). Adults with no educational qualifications were twice as likely as those with a school leaving qualification to have autism. Adults with a university degree only had a five times lower risk compared to those with just a school leaving qualification. Verbal IQ was clearly reduced in our autism cases—indeed none of our cases in the community had a verbal IQ score greater than 100. Most of my clinical colleagues and I know adults with autism who are highly intelligent—and for example, have completed a PhD. But the survey results suggest that they are rare and exceptional. What could make sense is that having a Ph.D. means that you are more likely to stand out if you are autistic and more likely to be encouraged to obtain a clinical assessment (and smart enough to ensure that you make it to an assessment clinic). This finding also made us realize how important it was for us to extend the survey, as we have done since, to include adults at the lowest levels of intellectual ability.

What about the lives of the people we found with autism? We compared the people we found to have autism with the remainder of the population in our sample. Here too there were surprises. We were expecting to find hardly any in a marriage or cohabitating, and most to be unemployed. Neither factor stood out quite as strongly as that. We found that adults with autism were four times more likely than people without autism to be single than in a long term relationship. More surprisingly, there was no association with being unemployed. The important policy message here is that our focus should not just be about getting adults with autism into work but taking a closer look at whether autism places them at a disadvantage in the workplace. We also need to understand how so many have already managed to hold down a job and to look at the quality of their working lives for signs of autism-related vulnerability and exploitation (which we did not ask about in this survey) as well as looking for good news messages about success in the work place.

Housing is also very important in our lives. What stood out was that adults with autism, compared to adults who do not have autism, are far are less likely to be buying their home or renting from a commercial (private) owner of the property they live in and are more likely to be relying on local government support for housing. It also appears that they are somewhat more likely than people without autism to be living in a low income, more deprived neighborhood. Taken together, the social circumstance of adults with autism is relatively poor compared to the rest of the adult population.

And finally, what about the services that adults with autism receive—diagnostic recognition and support services? In England, like in most western European countries, no one has to pay for health care when they go looking for it (in Great Britain they will have already paid for it through their taxes especially if they are in paid employment). The most astonishing finding of all was that none of the cases of autism we found in our community survey had been given an official diagnosis. They were entirely unrecognized; only in one case was a family member thinking of raising with the doctor the possibility that her relative might have a condition such as Asperger syndrome. Bear in mind that the household survey in 2007 did not include adults with significant intellectual disability (who we have gone on to study since). We also found it quite difficult to work out if adults with autism in the community were using services or had welfare entitlements. What stood out as significant was that they did not know how to answer questions about welfare entitlements (which may not come as a surprise to anyone working with adults with autism because typically they struggle to manage money sensibly and to budget and pay off debts on time). In all our other surveys since 1993, looking at adults with mental disorders, in each case we have found that those with a mental disorder told us they were more likely to be getting health care specifically for mental health reasons either from their family doctor or from community and hospital services than people without that disorder. The autism group is the first and only group that this does not seem to apply to. Being undiagnosed goes with being untreated as well. This is important and needs to be studied further and, of course, acted upon.

Adults with Intellectual Disability

We will now look at the second part of our project. Research on children with ASD suggests that up to a half will also have difficulties in learning. Research on adults who had learning difficulties while growing up shows that most manage to live independent lives as adults without having to rely on someone else to care for them. This also appears to apply to our 2007 adult autism survey findings in the household population where those with autism appear to be managing to live independently, although at a lower than average level of ability. What about adults clearly recognized as having more severe and significant intellectual disability seen in our latest 2010 survey?

At the time of writing, the findings of the more recent work in 2010 on adults with intellectual disability (Brugha, Cooper, et al., 2012) have not yet appeared in a peer refereed scientific journal and we are still studying them, so that these can only be described in summary form here. We assessed 276 adults in this ability range. A high proportion of the adults we found through the intellectual disability registers had not developed any receptive language skills and the assessment required the cooperation of carers (who have a very difficult role and may in part explain why some chose not to take part in the research). But adults in this very low ability range are a very small part of the overall population—less than half of 1 %. We also had to go through quite elaborate consent procedures. Also, as expected, rates of autism were much higher in this very low ability group. And when we added the results of the two surveys together (2007 and 2010) the overall rate of 1.0 % in the 2007 household survey hardly increased, coming close to 1.1 %^{Footnote 6} in the combined populations. As in the household survey the autism rate was higher in males, but that gender difference was not as great. The rate of autism was also highest in those with the very lowest ability levels. The presence of ASD appeared to be unrelated to ethnic grouping. And finally there was also little to suggest an association with age—rates were similar in older as in younger intellectually disabled adults. Taking the two surveys together the clearest association throughout is that between autism and ability level—the less able are clearly more likely to have ASD and more likely to need care.

Lots of people have asked about what “types” of ASD we found in this project: how many cases had Asperger syndrome, autism, high functioning autism, and so forth? We don’t know. This is because the adult survey approach does not give us the fine-grained information on early development of the type that can be collected by specialist autism teams in clinics. And in line with the changes expected in DSM-5 (http://www.dsm5.org/ProposedRevisions/Pages/proposedrevision.aspx?rid=94), we agree that those distinctions about subtype do not stand out in the evidence we have about variability in autism. Nor do they seem to help with treatment recommendations. What we think matters most in deciding on care and treatment, taking account of the care the adult already has, is the person’s overall ability level and the presence of other conditions like epilepsy, and problems with use of drugs and alcohol, risk of self harm and of suicide, all of which we plan to look at in the coming months and years using our survey data.

The work described in this chapter shows that it is possible to study autism among the adult population using similar methods to those used to study other mental disorders. It is vital that others undertake similar work elsewhere. There is no previous literature with which to compare our findings. For many the most surprising and concerning finding is that there are so many adults with autism in the community without any recognition or diagnosis, even in a country with health care that is free when needed for everyone. Only those adults with significant intellectual disability were known to also have ASD.

Under-recognition of autism in adults poses an enormous public health challenge. Just a few countries are beginning to address this. In England an autism-specific Act of Parliament is being followed up with a national strategy to improve understanding, recognition and support for adults with ASD (Department of Health, 2010). A key component of this is training professionals who are in caring roles to be aware of how autism manifests itself in adults and supporting adults through recognition, where appropriate, with diagnosis and through understanding and adapting to the kinds of disability that affect adults on the autism spectrum. Outside of the United Kingdom there appears to be relatively little evidence of this kind of approach although the authors are aware of positive initiatives in the Netherlands and Sweden, within Europe.

We would argue that there is a need for training, not just research. We know enough to be taking action and not just waiting for answers from research. The finding in our survey of under diagnosis and of lack of treatment needs to be thought of in terms of availability of support and treatment that is seen as effective. At the time of writing, the UK National Institute for Health and Clinical Excellence (NICE) is finalizing a clinical guideline for adults with autism (http://guidance.nice.org.uk/CG/Wave23/1#keydocs). This includes detailed systematic reviews of screening tests, diagnostic methods, medical and psychological approaches to treatment, and information on the kinds of service structures that need to be developed, albeit within the UK National Health Services, for health and social care.

Discovering how many adults in the community have autism should be about more than counting how many. It should be about opening up to providing help and care. The potential benefits are only just becoming apparent.