A long-awaited study of autism prevalence in Korea came out today in the American Journal of Psychiatry. Most of the information we have about autism prevalence comes from the US, the UK and Europe, so many were looking at this as the “Korean Study”. It is that, and very much more.
The title of the study is Prevalence of Autism Spectrum Disorders in a Total Population Sample. I expect the study will be gathering quite a lot of press as the results are quite remarkable. For one thing, the autism prevalence is estimated at 2.64%. That’s right, over double the current estimates in the United States and the U.K.. For another thing, most of the prevalence is for autistic students who were previously unidentified and unsupported.
Unfortunately, I was unable to obtain permission to review the article pre-embargo for discussion here on the Left Brain/Right Brain blog. Instead, I wrote about this for the Autism Science Foundation as Prevalence of Autism Spectrum Disorders in a Total Population Sample. There you will find a more thorough review of the paper, complete with questions and answers with team member Roy Richard Grinker of George Washington University. The study was led by Dr. Young Shin Kim of Yale, and includes an international team:
Young Shin Kim, M.D., Ph.D., Bennett L. Leventhal, M.D., Yun-Joo Koh, Ph.D. , Eric Fombonne, M.D. Eugene Laska, Ph.D., Eun-Chung Lim, M.A., Keun-Ah Cheon, M.D., Ph.D. ,Soo-Jeong Kim, M.D., Young-Key Kim, M.D., HyunKyung Lee, M.A., Dong-Ho Song, M.D., Roy Richard Grinker, Ph.D.
Again, the full post can be found at the Autism Science Foundation blog.
Left Brain Right Brain 
I’ve thought the prevalence of autism was higher for quite some time. I just know too many people who are either diagnosed or should qualify for a diagnosis for the number to really be 1 in 100. There were two kids at the grade school I went to (+/-54 kids per grade) who weren’t diagnosed AFAIK, but in hindsight had extremely strong traits (one was my best friend and second cousin, her brother was more severely disabled and had a diagnosis of Aspergers). Both struggled quite a bit, I wonder if their lives would have been different had we known then what we know now. I don’t think they needed services in as so much that they needed understanding and tolerance.
I have a 6 year old brother diagnosed with PDD-NOS who I worry about- in many ways he is like them. He has a diagnosis, but people have difficulty accepting that a kid who is that bright, who can pass for typical so well really is on the spectrum. I know, in the past, he has been disciplined at school because of his differences, and I believe that even when we have these kids diagnosed we will continue to confront ignorance.
Wake me up when it’s 4O%. That’s what my family is…
A quick look at the World Health Organization schedule tables suggests that by three years of age children in South Korea receive about 1/3 fewer vaccinations than are suggested in the US and also fewer immunizations than are recommended in the UK.
Parents in the US and the UK may thus be reassured that the high apparent prevalence of ASD in South Korea could be due to “too few, too late.”
No no no. It’s cheap, imported Chinese kimchi pots. They are lined with MERCURY!
Most females either are not diagnosed or misdiagnosed. When those are accounted for, the numbers will go up.
Many countries are lacking in doctors, education and facilities to assess the population properly. When they can, the numbers will go up.
If they were to introduce another common disorder/syndrome to the umbrella of Autism (as they did with aspergers), the numbers would go up.
All of these things (and more) have been happening, continue to happen and will only happen more over time.
Sadly, all these factors do is fuel the ‘epidemic’ argument further as some people refuse to believe the facts.
Not sure but it seems like this study used a population-based methodology similar to the one used in the recent UK study that found a 1% rate in British adults. Is that right or am I mistaken?
Char, I think it’s similar.
Girls still were in the minority in this study (less so, but still). It’s likely in part because the socialization girls receive masks or helps compensate for an underlying ASD.
brian: thanks for looking up vaccine situation. Vaccine is, unfortunately, something that always comes up in this context.
I was actually thinking about kimchi too, since there are alternative medicine people who believe that autism can be helped/ cured/prevented by eating fermented foods or probiotics. They might not welcome this study. (I myself eat quite a loot fermented food, not because it affects my AS, but because I like to experiment with fermenting food at home, it’s one of mine geeky interest.)
Hi Sullivan –
On another thread, you said this:
As to skepticism towards the ASSQ: keep in mind that the ASSQ was not the diagnostic tool. It was a screening tool. The Kim study used actual diagnosis of the individuals.
Would this roughly be methodologically equivalent to the NHS study in adults; i.e., screen for likely candidates, diagnose a subset of those with high scores, and extrapolate out based on these findings?
Thanks.
– pD
pD,
I may have accidentally sent your long comment to spam. I’ll fish it out when u get to a real computer
Erm, oops, I just emptied the spam. Sorry pD.
Yes, pD. Is there a better way to do it?
I knew prevalence estimated by studies would continue to increase with better case-finding, but 2.6% is not something I’d expected.
Hi Joseph –
Is there a better way to do it?
Well, I think there are different ways to do it, and from some previous conversations I’ve seen happen here, they should be more representative of reality. I can’t find the thread, but essentially the idea I saw proposed was that as you physically interviewed more people, you got more precise results and fewer false positives; the trade off being logistic difficulty and smaller sample sizes.
For example, you used to reference this study quite a bit: Brief report: autism and Asperger syndrome in seven-year-old children: a total population study
Wherein half of the population of school children, about 400 kids, were personally interviewed and assessed by the same individual. They reported a prevelance rate near 1%. I would tend to think that a study that involved randomized, personal evaluation of chidren would provide a clearer picture than screenings, refusals to participate, and weighted estimates. (?)
Obviously our problems with big shifts in terms of absolute numbers are an issue, but if Korea like values are real, they would have had to more than double their findings of cases. Do you think this is realistic?
I knew prevalence estimated by studies would continue to increase with better case-finding, but 2.6% is not something I’d expected.
I guess I’d be interested in your thoughts on if you think Kadesjo should have found ~ 2.5% based on the reports from the Korea study?
Similarly, isn’t this study evidence of an epidemic? After all, the NHS study in adults followed a very similar pattern, one you seem to acknowledge would result in better case finding, and found a very, very ‘low’ prevelance of 1% in adults.
The study from Brazil, Brief Report: Prevalence of Pervasive Developmental Disorder in Brazil: A Pilot Study which also used the ASQ as screening device and targeted follow up interviews found a prevalance rate of .3%
These studies are getting thrown around as if they are meaningful in isolation and grand conclusions are being drawn from them. I don’t think they are meaningful, except to illustrate what a total mess everything (still) is. Instead, it looks more like people are picking any study that seems to fit their meme and shouting it from the rooftops that it proves/disproves the epidemic.
This thread is great evidence of that; nobody seems bothered by a doubling or near tripling of ‘autism’ rates that includes an overwhelming majority of children with no obvious effect on school participation. Understanding why this doesn’t strike up skepticism in the individual lurker is difficult, but my gut feeling is that because this type of study tends to discount the notion of an epidemic, lots of people are sitting on their concerns about just how ridiculous these findings are. Whatever their reasons, it certaily isn’t because they have any evidence in the past that comes anywhere close to approaching these numbers, that’s for sure.
– pD
@pD: Note that Kadesjo did screen, except 50% of the children passed the screening. He could still have missed some children.
In that case, it was logistically possible to interview half the pertinent population. In a bigger population, you either screen and extrapolate, or you sample and interview. Sampling is non-trivial to get right statistically.
It’s not at all surprising that if the prevalence doubles, the proportion of mainstreamed children would also increase. Additionally, the proportion of children with mental retardation and epilepsy would drop. I’ve pointed this out about autistic children in CalDDS.
Clearly, it’s not straightforward to compare this prevalence finding to previous ones from Karlstad or Brazil. It’s not sufficient that they use the same screening tools and diagnostic criteria. The case-finding method is apparently unique, but I have doubts that even duplicating the case-finding method a different team of researchers would find the same prevalence. The researchers themselves are part of the “method” if you will.
Hi Joseph –
It’s not at all surprising that if the prevalence doubles, the proportion of mainstreamed children would also increase. Additionally, the proportion of children with mental retardation and epilepsy would drop. I’ve pointed this out about autistic children in CalDDS.
Hehe. This is what I mean.
What if, instead, they’d reported that prevalance had quadroupled, or increased by ten fold? Is there a number which might make you think, ‘You know what, these numbers might not be right, even though it fits the pattern of children affected even less making the cut onto the spectrum.’?
If someone were to report 5%, or 10% in Japanese children tomorrow, would you consider the findings suspect?
Would any reported prevalance rate give you cause for concern on the validity of the findings?
– pD
So what do they plan on doing with the info?
autism squeaks at least partially funded the study. Why didn’t they try this in America??
@pD: ‘Suspect’ is not the word. I’d just think that the researchers have a conception of autism that differs from most others.
I think it’s telling that Korea didn’t notice they had a 2.6% prevalence of autism. One official thought it would be more like 1 in 100,000. What this tells me is that the researchers think autism is one thing, and the locals think it’s another thing entirely.
>>It’s not at all surprising that if the prevalence doubles, the proportion of mainstreamed children would also increase. Additionally, the proportion of children with mental retardation and epilepsy would drop<<
@Joseph, @-pD
Anyone want to bet LD's go down?? I'm willing to lay down a c spot.