Tuesday, September 22, 2009

Has autism's hidden horde been found in England?

Here is a studythat has recently been done in England finding a prevalence rate of 1% of adults with autism, suggesting that there has never been an increase in prevalence of autism as many have suggested. I did not read the whole study in detail but did surface read it, and it seemed that they surveyed a variety of people all over England. There was about a 9:1 ratio of males to females rather than the usual 4:1 ratio and this may have been because of the sampling size, I am not sure. I tried to read it in detail, but my disability made it hard for me to comprehend the entire thing and have all of it sink in. I was debating whether or not to blog about it, but I have decided to anyhow in spite of my concentration problems. This may not be the survey or study to end all survey/studies and I am sure this won't convince the age of autism folks as well as others who believe in an "epidemic" of autism involving huge increases (I don't think anything will ever convince them), but it may give some evidence that there has always been a high and stable rate of autism and disprove that vaccines or something else in the environment caused an autism explosion in the late 1980's, early 1990s.

The environmental hypothesis has appeal among its adherents, however, due to the fact that it provides the possibility of finding a single cause for most ASD's and with the cause, the treatment or possibly even cure can be found.

One of the arguments that those who have this point of view is that there are no comparable numbers of adults with autism. Even though this survey may not prove that conclusively, I think it is the first published survey of its kind that has been done trying to get an assessment of what the prevalence of autism is in older populations. To date, as far as I know, the other side has never done such as study.

Mark Blaxill has given the arguments that a study done by Bard et. al. in a rural part of North Dakota in which they tried to find additional persons with autism diagnoses after the prevalence of a certain sample was obtained in the cohort's childhood, they failed to find any new cases. The problem with this one study was that it was done in a rural area where the prevalence rates for autism are low. If numbers of autistic persons has actually increased it has been in urban areas. For example in the 2007 CDC survey much higher rates were found in New Jersey than in Alabama. In California's regional centers, the autism prevalence is much higher in Los Angeles than in rural parts of Northern California. Some might argue that this is due to more pollutants being found in urban parts of california or a higher rate of pollution in New Jersey than in Alabama. Be that as it may, so far no one has found any pollutant in the environment that has caused this increase. If vaccination rates are much higher in New Jersey than in Alabama and much higher in Los Angeles than in rural parts of Northern California, I have yet to see evidence of this.

Another piece of evidence that Blaxill gave was a study done by Nylander in Sweden where the author searched for persons with autism in psychiatric hospitals and found low prevalences there. Aside from the fact that autistic adults may not be inclined to present in psychiatric hospitals in great numbers, particularly if they are pretty high functioning, we see some inconsistencies in Blaxill's reasoning. In some personal correspondence that he and I engaged in about six years ago he stated that the evidence for a true epidemic of autism was that the increases were geographically specific, only taking place in the USA and UK and not other parts of the world. So, it would seem strange that he would cite a study in an area he had deemed geographically irrelevant.

So, I suggest that the age of autism folks or one of their allies does their own study, refuting the recent British survey showing that the prevalence rate of autism in persons born in 1986 and earlier is in fact lower than in the younger cohorts.

I was skeptical that finding the autistic adults was feasible, since they are not likely to present to special education services and regional centers and such.

I am glad that studies are finally being undertaken to find the so-called hidden horde of autistic adults. I have written about the reasons previously

I apologize to readers of autism's gadfly that I was not able to read and understand the study in more detail. Hopefully more work will be done in this area and we can determine whether or not the prevalence rate is the same in older populations than in younger.


Jake Crosby said...

"I don't think anything will ever convince them."

-I know you did not make that line up but got that from somewhere else, and cannot believe people can say that thinking they even sound remotely intelligent. All such a quote does is display a tremendous bias on the part of whoever says it to disseminate information supportive of their preconceived views instead of finding the real answer to the problem. I hope you are aware of that, Jonathan.

Jake Crosby said...

Btw, I think I know why you are having difficulty reading that paper Jonathan - because it's uninterpretable. For instance, it lumps together only groups that have scored between a 0 and a 4 on the initial AQ survey, and people who scored from 5 to above on the AQ and completed the second phase as well.

Yet, the probability a subject would even be selected for the second phase with scores from 5-9 on the AQ self-completion test for Asperger Syndrome were very small, ranging from .021 to .029 respectively. So the vast majority of people in this range, from scores of 5-9, would have been left out of the results calculating ASD based on a 10+ score on the ASOD because they would not have participated in phase II, which they would need in order to belong to the total sample used to calculate the final ASD prevalence.

jonathan said...

Jake: Is there any evidence that would convince you that vaccines don't cause autism and there has been no increase in autism prevalence?

john GO VOLS holla said...

The study is questionable because aren't the British an autistic lot anyway?

Jake Crosby said...

Jonathan, you apparently did not read my first comment, so I will say it again:

All such that question your asking does is display a tremendous bias on your part because it reveals an innate desire to find material supportive of preconceived views instead of figuring out the real answer to the problem, while ignoring my reasoning altogether. I hope I've made myself clear.

Jake Crosby said...

I never said you did, I only said what you convey by making that comment, big difference. Clearly you did not come up with such a loaded question, you probably heard it off an ND blog or from "Science"Blogs.

Anyway, that all doesn't matter. What you feel caused your ASD is your personal business and not something that is for me or anyone else to get involved in. All I can say is if you want to investigate the etiology of your disorder, you might want to look at sources of possible environmental prenatal and postnatal exposure you may have had. But that's just my recommendation.

BTW - Yeah, I saw, "Science"Blogs gave me a couple of plugs as well.

jonathan said...

I am not sure what caused my ASD though it is likely partially genetic due to the high heritably of autism and my family history of possible if not probable autism in a great uncle, depression, and ADHD and learning problems in some other relatives. I would sure like to know the real answers but probably won't.

Congrats on the plugs from "Orac" (Gorski) means you are coming up in the world

Jake Crosby said...

Thanks, he even called me a token autistic. This is coming from a guy who does not even have an ASD. Unbelievable.

Anonymous said...

Another piece of evidence that Blaxill gave was a study done by Nylander in Sweden where the author searched for persons with autism in psychiatric hospitals and found low prevalences there.

I always found that strange, because I draw the exact opposite conclusion from that same study.

Blaxill's mistaken assumption is that by simply looking for autistic adults among the out-patients of a psychiatric hospital somewhere, that's roughly equivalent to a whole population survey. That's nonsense. Autistic adults can't be all expected to be out-patients of psychiatric hospitals. Major fail.

Now, what Nylander & Gillberg did find which was of interest is that 17 of the 19 autistics they located at the hospital did not have a prior autism diagnosis. That is, 89.4% had been missed.

Anonymous said...

It seems Jake has trouble with the methodology they use in the paper to select phase 2 participants.

It took me a while to wrap my head around it as well, but in retrospect it's not that complicated, and it's kind of clever.

One thing to understand is that they don't use the AQ test as a screening tool, which would be the normal expectation.

It's reasonable to suppose that the researchers didn't trust the AQ test to not miss some autistics, probably intellectually disabled autistics, since the AQ test has mostly been tested with adults of normal intelligence.

They used it simply as one of several variables in the calculation of probabilities for phase 2 selection. Theoretically, they could've used any other variable, say, toss a coin - heads you have a 0.05 probability, tails you have a 0.15 probability of going forward. Of course, if they did this, they would only find about 6 autistics in phase 2, so the study loses statistical power.

The 19 autistics they found evidently must have been assigned a probability of passing to phase 2 which was about 3 times the probability assigned to a random person from phase 1. That's how the math works.

It's interesting and debatable why they didn't use a more conventional methodology. If they had used the AQ test as a normal screening tool, selecting everyone who scored 26+ for phase 2, then they might have found about 70 autistics, giving the study a lot more statistical power.

I'm sure they considered that, and they must have decided it's really not a good idea. This is the first near-whole population study of autism in adults, after all, so they couldn't make many assumptions about what works or not.