BACKGROUND

When data from California emerged in 1999 that suggested a dramatic increase in the rate of autism it was taken by some advocates as evidence for an epidemic. This was problematic because the scientific consensus is that autism is fundamentally genetic and you cannot have a genetic epidemic. Therefore there had to be some other explanation for the Californian data. The authors stated that their data provided no evidence for or against the idea that autism was increasing. But if more people were entering the system the obvious conclusion was that autism was on the increase. One idea was that the genetic “bullet” required an environmental “trigger” and perhaps there were more triggers now than before. About this time two ideas were suggested concerning vaccines. In the UK the MMR vaccine was implicated in the now infamous Lancet paper by Andrew Wakefield. in the USA Thimerosal, a mercury based preservative commonly used in vaccines, but never used in live vaccines like MMR, was proposed as a possible source in an article in Medical Hypotheses.

Both these ideas were tested and found wanting by medical researchers. But thanks to the internet both ideas had already spread amongst parent advocacy groups. These groups were used to battling with unsympathetic government agencies both for recognition of their children’s needs and for resources to meet those needs. So official denials of the vaccine link were met with predictable cynicism. The media picked up on the controversy and presented it as a human interest story: brave maverick doctors and campaigning parents taking on an unfeeling system that denied its culpability for the epidemic and refused to meet its obligations.

THE HIDDEN HORDE

So the idea of an autism epidemic caused by vaccines went mainstream. Following the public disgrace of Wakefield the vaccine link may no longer resonate with the media or the general public but the idea of an epidemic has survived. Those of us who question the notion of an epidemic of recent origin are challenged to explain the lack of autistic adults. Where is this “Hidden Horde?” There remains a devoted band of true believers in the vaccine link who frequent the comments sections of blogs and other social media sites. They often pose the question thus. This is a recent comment on Emily Willingham’s blog.

That British study found all the missing adults by asking survey questions. I don’t want to hear about a 50 year old man who lives alone and would rather stay home and read a book than go to a party. I don’t want to hear about a retired bus driver with a wife and a family who just discovered he has Asperger’s.

What I want to see are lots of adults like so many children that I know personally. I want to see the 40, 50, and 60 year olds who flap their hands, don’t speak or who have echolalia or who scream endlessly. I want to see adults who rock and spin and line things up like our kids do, along with ones who bang their heads on the walls endlessly and are still in diapers. I want to see the adults who have to be watched constantly because they’ll wander off at will. And I especially want to see middle aged and elderly people that are now called autistic and whose health histories include starting off as normally developing babies, but who suddenly and dramatically lost learned skills and regressed into autism. Finally, I want to see older people with autism who also have the health problems our children do, namely, bowel disease, seizures, and sleep disorders, to name only a few.

Unless and until you can show us adults with classic autism like we see in our children, this proves nothing.

Anne Dachel, Media editor: Age of Autism

Dachel and those like her are either ignorant of the facts or are misrepresenting them.

• Much of the increase in autism has been driven by changes in diagnostic criteria to include those whose cognitive and/or communication skills would have excluded them in the decades prior to the “epidemic.” I find it ironic that these same people are now complaining that fresh changes to the criteria will exclude more high functioning types and lead to the denial of services to people they have previously dismissed as not being truly autistic.
• The children Dachel describes are not the norm, even amongst those with so-called “classic autism.” I say this based on my experience of teaching these children for thirty years in a school for children with severe learning difficulties in the UK.
• A tiny minority remain non-verbal, prone to self harm and violent tantrums into adult life. But many more learn to talk, to manage their own behaviour and acquire sufficient self help skills to live with some independence in the community. I see these changes in children aged 2 to 19 years old.
• It is reasonable to expect that progress to continue into adult life, especially for those without significant impairments in cognitive and communication skills.

THE HIDDEN DATA HOARD

There is a problem regarding those who do have significant impairments in cognitive and communication skills. We have very little data on autism through the life span for these individuals, especially the older generation who may have missed out on an autism diagnosis when educational subnormality (UK) and mental retardation (USA) were more commonly diagnosed in this population. That is why I am grateful to Cos Michael, Autism & Ageing Project Manager at the National Autistic Society  for alerting me to some American research entitled “Prevalence of selected clinical problems in older adults with autism and intellectual disability.”

The study compares adults with autism (ASD) and Intellectual disability (ID) to those with ID alone. They took two survey samples from consecutive years, 2009/10 and 2010/11. They excluded those individuals with profound ID for whom any differential diagnosis is problematic as well as those whose level of ID was unspecified. Thus we had a group of individuals whose diagnosis with either ID or ID and ASD together could be ascertained with a reasonable degree of confidence. The primary focus of the study was on medical and  behavioural problems in older adults with ASD. They found no significant difference between the ASD/ID group and the ID group regarding physical/medical disabilities including epilepsy, but a significant increase in behavioural problems in the ASD/ID group, especially destructive or self injurious behaviour. Because the cut off point was age 59 there was little data on the impact of those diseases of old age like Dementia on the autistic ID population.

I was also interested in another aspect of the study. What light could it throw on the prevalence of ASD amongst  individuals with ID? ID affects around 2% of the population. Here we had figures recording the proportion of those with ID who also had ASD over a period of thirty years. First some caveats. This is not epidemiological data, any more than the Californian DDS data was. These are surveys. The 2010/11 survey has a much smaller sample than the 2009/2010 survey. We have no idea how the exclusion of subjects with profound or unspecified ID affected the results because we have no indication of how many of them were autistic. The authors acknowledge that the lack of unique subject identifiers means that that may have been some overlap over time or across states. A few individuals may have been counted twice. They are also aware that the potential for under diagnosis in older individuals means that some in the ID only group may have had an undiagnosed ASD.

So this study is indicative rather than conclusive. Bearing this in mind the raw data still suggests that there is a significant adult population with ASD and ID that exceeds the figures suggested by early prevalence studies.

Table 1 Demographic variables among the ASD group and the ID-only group in the 2009 to 2010 NCI sample of adults aged 30-59 years (adapted from Kats et al. Journal of Neurodevelopmental Disorders 2013, 5:27)

Table 2 Demographic variables among the ASD group and the ID-only group in the 2010 to 2011 NCI sample of adults aged 30-59 years (adapted from Kats et al. Journal of Neurodevelopmental Disorders 2013, 5:27)

Leaving aside those aged 30 to 39 in 2009/10/11 who are most likely to have been exposed to whatever environmental factors are supposed to have triggered the “epidemic” in the last quarter of the 20th Century, the proportion of those with an ID who also have an ASD suggests a prevalence in the range of 11 to 15 in 10,000 amongst adults with ID born in the 1950s and 1960s.

This is greater than figure of 4.5 in 10,000 found by Lotter (1966) but more in line with Wing and Gould’s finding of a prevalence of 20 in 10,000 for a population of children in special education in the Camberwell study. (1979) Gillberg reported similar findings in Copenhagen (1986) for children with special needs. (References for all studies mentioned in this paragraph can be found here.)

Perhaps more startling is the fact that Emerson and Hatton (2004)  estimate that less than a quarter of people with ID in the UK are known to health and social care agencies. The rest are managing in the community with no support. Presumably some of those are autistic as well. This all goes to suggest that there are already more ageing autistic adults than we know about and we are ill equipped to understand their needs, let alone provide for a dignified passage through old age.