What Can Dr. Jay Gordon, Pediatrician, Tell Us About Autism Epidemiology?
July 27, 2008 by Do'C 
Printer-Friendly Version
From the discussion section of a post over at Orac’s place at ScienceBlogs:
Hi Dr. Gordon,
On your website there’s a video ad for your “Vaccinations” book.
In the video, I noticed a slide that contained the following:Autism by the Numbers:
U.S. and Europe
USA - 1980′S: 1-2 per 10,000
USA - Late 1990’s: 1 in 500
USA - 2000: 1 in 250
USA - 2004: 1 in 166Given that the video is not much more than a short advertisement, if you mention the changes in diagnositic critera as well of the very definition of “autism,” it’s not evident in the clip.
Aside from diagnostic substitution, earlier diagnosis, increased awareness, and better recognition, etc., from a scientific perspective, to what degree do you think the following changes impact the numbers you have presented?
1. Autism was first added to the DSM in 1980.
2. PDD-NOS was added to the DSM in 1987.
3. The criteria for autism was changed (greatly broadened) in 1994.
4. Aspergers’s Syndrome was added to the DSM in 1994.In formulating a response, please keep in mind that the numbers you presented are for all autism spectrum disorders (which were not even all acknowledged in terms of diagnosis until well into the 1990’s. Also, please keep in mind that two-thirds of the estimated prevalence of “autism” is PDD-NOS and Asperger’s, according to current epidemiology.
Given that the numbers you present in this slide cover a wide range of changing diagnostic criteria, are they even meaningfully comparable?
Your best scientific response would be most appreciated.
Posted by: Do’C | July 25, 2008 2:05 AM
[..I'm sure that your comments, Ozzy and Do'C are correct: The revision of diagnostic criteria and registering of autism cases in various countries has made gathering data difficult and I use the numbers that suit my case. Are there sources that you consider more meaningful? Do you think that the "1/150" number is accurate or exaggerated?..]
Posted by: JayGordon | July 25, 2008 2:37 AM
Dr. Gordon wrote: “The revision of diagnostic criteria and registering of autism cases in various countries has made gathering data difficult and I use the numbers that suit my case. Are there sources that you consider more meaningful? Do you think that the “1/150″ number is accurate or exaggerated?”
How about a stab at the actual questions Dr. Gordon?
I’ll repeat it for you.Aside from diagnostic substitution, earlier diagnosis, increased awareness, and better recognition, etc., from a scientific perspective, to what degree do you think the following changes impact the numbers you have presented?
1. Autism was first added to the DSM in 1980.
2. PDD-NOS was added to the DSM in 1987.
3. The criteria for autism was changed (greatly broadened) in 1994.
4. Aspergers’s Syndrome was added to the DSM in 1994.In formulating a response, please keep in mind that the numbers you presented are for all autism spectrum disorders (which were not even all acknowledged in terms of diagnosis until well into the 1990’s. Also, please keep in mind that two-thirds of the estimated prevalence of “autism” is PDD-NOS and Asperger’s, according to current epidemiology.
Given that the numbers you present in this slide cover a wide range of changing diagnostic criteria, are they even meaningfully comparable?
Your best scientific response would be most appreciated.
Posted by: Do’C | July 25, 2008 4:50 PM
In an effort to be sensitive of your time, you could certainly answer the questions in the following way.
1. The changes in diagnostic criteria do have an impact, the changes in diagnostic criteria do not have an impact, I don’t know what impact the changes in criteria would have, or I don’t understand the question.
2. The data are meaningfully comparable, the data are not meaningfully comparable, I don’t know if the data are meaningfully comparable, or I don’t understand the question.
Posted by: Do’C | July 25, 2008 4:55 PM
To answer simply: Yes, the changes in diagnostic criteria do have an impact and make comparisons more difficult. However, I don’t agree that the data are not meaningfully criteria.
But, the wall is beginning to crack as this evening CBS News story began revealing the conflicts of interest in the vaccine promoters’ world. They are shameless.
[Link]
Jay
Posted by: jayGordon | July 25, 2008 11:35 PM
Dr. Gordon,
I appreciate that you’ve attemtped to answer my two questions. But, your answers are still a little vague (and the slide in your presentation is not).
when asked as simply as possible,
“to what degree do you think the following changes impact the numbers you have presented?”
followed by the actual DSM changes, you replied with:
“To answer simply: Yes, the changes in diagnostic criteria do have an impact and make comparisons more difficult.”
and
“However, I don’t agree that the data are not meaningfully criteria.”
I’ll assume you meant “comparable” instead of “criteria”.
I’ll also take the liberty of converting your negative phrasing to the positive. Please correct me if I am wrong.
You’re essentially saying that you would agree to a statement that the data may be meaningfully comparable.
Would you please expand on that. Feel free to draw upon your education, years of experience, logic and reason, and of course, actual descriptive epidemiology.
Can you explain, in light of the changes in diagnostic criteria, how the data in your slide could be meaningfully comparable?
Posted by: Do’C | July 27, 2008 12:59 AM
[..Sorry for the typos and the lack of clarity Do'C. What I meant to say was that the data can be compared meaningfully even though the diagnostic criteria have been changed. I think that we need new studies generating new data because I agree with you completely when you imply that changing diagnostic criteria over the years have made comparison much more difficult. Actually, you obviously are implying that the data are not meaningfully comparable. I think that things could be a whole lot better with new studies and new data but we have to work with what we have. I'll have to look at the slide you're referring to. ..]
Posted by: Jay Gordon | July 27, 2008 2:03 AM
Sorry for the typos and the lack of clarity Do’C. What I meant to say was that the data can be compared meaningfully even though the diagnostic criteria have been changed.
Can you explain, in light of the changes in diagnostic criteria, how the data presented in your slide could be meaningfully comparable?
Please be specific, and please provide supporting science for your explanation.
I think that we need new studies generating new data because I agree with you completely when you imply that changing diagnostic criteria over the years have made comparison much more difficult.
Dr. Gordon, Please don’t attribute a position to me that I do not hold. I don’t think the changes in diagnostic criteria, nor the changes in the very definition of autism (now considered a spectrum of disorders), make the comparison “much more difficult”. In fact, they make it very “easy” to make a completely invalid comparison.
Actually, you obviously are implying that the data are not meaningfully comparable.
That’s because they are not. Feel free to refute this position with actual science - epidemiological data.
I think that things could be a whole lot better with new studies and new data but we have to work with what we have.
Therein lies the apparent difference between you and a scientifically minded person. A scientifically minded person would recognize that the data cannot be used to draw certain conclusions (such as implied by your slide titled “Autism By The Numbers”), simply because the criteria and definition refer to very different things over the time period represented on your slide. A pseudo-scientific position frames erroneous hypotheses or draws erroneous conclusions from such data.
I’ll have to look at the slide you’re referring to.
No you won’t. I’ll assume that you are not purely lazy, or that you are not dodging this very simple question, rather, I’ll assume that you simply missed the fact that I already posted the information from your slide upthread. Again, I’ll attempt to save you some time and repost the information here.
Autism by the Numbers:
U.S. and Europe
USA - 1980′S: 1-2 per 10,000
USA - Late 1990’s: 1 in 500
USA - 2000: 1 in 250
USA - 2004: 1 in 166So, to repeat, here is the question I have for you:
Can you explain, in light of the changes in diagnostic criteria, how the data presented in your slide could be meaningfully comparable?
Please be specific, and PLEASE provide supporting science for your explanation.
Posted by: Do’C | July 27, 2008 6:32 PM
Also, Dr. Gordon,
Please try to resist temptation to respond to my statement that the criteria are not meaningfully comparable, with, “I disagree”. My statement is a fact, not an opinion. Your agreement or disagreement is irrelevant if you cannot provide evidence to substantiate your position.
The numbers presented in your slide are not meaningfully comparable due to the following reasons:
1. The definition of “autism” has changed over the referenced time period. It has gone from a singular disorder diagnosed with very strict criteria to a spectrum of disorders diagnosed by a wide range of criteria.
2. The criteria have broadened for autistic disorder alone to be much more inclusive.
(See the relevant versions of the DSM between 1979 and today if you need verification)
In order to show that your numbers are meaningfully comparable in reference to “autism,” you’ll need to provide evidence that numbers 1 and 2 are incorrect, or you’ll need to provide other scientific evidence that shows despite 1 and 2, the numbers are meaningfully comparable.
Posted by: Do’C | July 27, 2008 6:54 PM
I think it will be very interesting to see what Dr. Gordon comes up with in the way of actual science to support his position that, “the data can be compared meaningfully even though the diagnostic criteria have been changed”.
If he has time or inclination, perhaps he’ll tackle this logic problem too:
A logic problem for Dr. Jay Gordon
In 1980, you consumed one piece of fruit per day. The definition of “fruit” within your dietary preferences was very strict, and referred only to a red delicious apple.
Now in 2008, you consume one apple per day, one orange per day, and a banana. Good for you for making those diet modifications somewhere along the way! In addition, back in 1994 you expanded your apple preferences to include other varieties, including several red, some green and a yellow variety - again, good for you!
Based on the information provided, which statement is true?
1. There are more apples in the world today.
2. You consume more apples daily, than you did in the past.
3. You consume more fruits than you did in the past.
4. There is more fruit in the world today.
17 Comments »
RSS feed for comments on this post.
Important Comment Notes
Installed spam filter plugins may result in your comment being held in a moderation queue. Please consider making a copy of your comment "just in case". You can use many XHTML tags to mark up your comment if desired.
Additional information about comments and other contact can be found here
Comment by JayGordon — 27 July, 2008 @ 11:58 pm
James, if you don’t mind, I’ll post the email I just sent you in response to your email. If you object to this, please delete my post.
Thanks,
Jay
Thank you James.
I almost got to sleep early tonight, damn you!
I checked my email on my Treo just one last time and had to get out of bed because surfing is too tedious and slow as is typing any meaningful response.
But, thank you anyway.
I have always liked the phrase, “Don’t bring a knife to a gunfight,” and I sure made that mistake this time.
You couldn’t be more correct in pointing out that changing diagnostic criteria have made comparing ASD “then and now” difficult at best and impossible at worst.
I have only one really strong suit and it truly is a “knife” in Dave’s gun slinging crowd: My experience is really valuable. I can be attacked from every angle, scientific, statistical, logical, you name it. Defending my position in Orac’s group gets harder with each assault but I think that (syllogistically, paradoxically?) the reasons themselves are defensible: Adequate new studies haven’t been done to generate the data we need to argue rationally and even . . . pleasantly. I just can’t back down on that one weapon in my arsenal. I’ve seen a lot more vaccines, a lot more autism and related “disorders” and I’ve watched too many temporally related events to not adhere to my strong belief that in certain susceptible individuals our current vaccines trigger autism, diabetes and more.
Without that new data and lacking the time and training to interpret or massage the old data, I’m stuck saying “I know what I’ve seen,” and Dave’s crew—perhaps correctly—tell me to support my observations and experience with hard data.
My DVD and public pronouncements will be changed by the experience I’ve had in these past few days. Not because I believe with any less fervor that we are doing harm with our current vaccine formulations and schedules, but because I think that all the data I like is easily subject to interpretation that I don’t, like. (I didn’t say that well, it’s late, but I hope you know what I mean.)
What amazes me most about the crowd over there is the anger. I love children and have devoted my life to pediatric health care and relatively recently to delving as deeply as I can into the complex riddle of autism. I am one voice in the debate. I might not be as small a voice as I thought but I’m certainly not powerful enough to change national policy nor change Paul Offit’s mind. I might be able to change Amanda Peet’s somewhat more pliable mind . . .
You should be heard, I should be heard and the parents of children with autism who agree and disagree with me should be heard.
I agree that many, including me, jump from a reasonable hypothesis to a conclusion far too easily. I’ll work on that but I really would love to see studies ongoing. I reserve my greatest disdain for the official medical and scientific bodies who have promoted the “no further studies should be done” message.
I didn’t mention this in my email, but I also object to bullying tactics in this discussion.
I look forward to talking with you at greater length and to perhaps meeting face to face someday soon.
Best,
Jay
Comment by M — 28 July, 2008 @ 12:42 am
Sorry, but Dr. Jay’s comment is just pathetic.
Autistic persons deserve better doctors than that. Not belief-based doctors, please!
Comment by mike stanton — 28 July, 2008 @ 1:21 pm
“I might be able to change Amanda Peet’s more pliable mind.” ?!?
Here in the UK that would imply that Amanda Peet is a weak minded individual who is easily manipulated. Is american usage significantly different?
Comment by Joseph — 28 July, 2008 @ 1:38 pm
BTW, Lotter (1967) found a prevalence of autism of 4.5 in 10,000. This was a screening at a time when the diagnosis was basically non-existent, and it used very early Kanner criteria which was operationalized according to Vic Lotter’s own ideas.
The question of how much of the observed increase in the diagnosis rate we can attribute to changes in criteria is interesting and intractable. I don’t think anyone can answer it, because there’s no mathematical model that says how much of an increase we can expect. All we know is that the prevalence of other disorders (mental retardation, speech disorders, etc.) is quite high, so practically anything is possible through diagnostic substitution. There is no doubt some excess labeling that is not due to substitution as well.
Comment by Pseudonym — 28 July, 2008 @ 2:58 pm
Dr. Gordon may very well believe he is seeing vaccinated kids becoming autistic, but it doesn’t explain why the majority of doctors are not seeing the same thing. He has fewer vaccinated kids in his practice (logically), and yet he is seeing more of these adverse events than others do?
I don’t think so. All one needs to have is a brain that is primed to see certain things and then will see certain things.
[moderated paragraph: topic relevance]
I am hoping that he is not offered any more opportunities to speak on national television, he has done much harm by making statements indicating the rate of autism has truly increased, saying that vaccines are “toxic” and need to be “greened” and especially by telling people that there is antifreeze in vaccines. I can’t imagine that he would sit still as a vaccinologist played back-seat driver and told him how he was running his pediatric practice all wrong. [.. remainder moderated: topic relevance]
Everyone has their own area of expertise. Dr. Gordon should realize that he’s in way over his head when it comes to autism and vaccines.
Gordon has “Deadly Immunity” on his website’s homepage. The amount of lies and nonsense in that mess is shocking, and to me it says a great deal about Jay Gordon’s lack of insight, lack of scientific training and lack of ethics.
He also is in a good location to explain why the autism epidemic, according to data from the California DDS, spreads out from Santa Monica and other wealthy areas of Southern California like a gradient. If autism is about vaccines, then what are they doing in Santa Monica, giving quadruple doses of each vaccine in the schedule?
Comment by Do'C — 28 July, 2008 @ 7:47 pm
Dr. Gordon, it is not “difficult at best”. Any such comparison is totally scientifically meaningless. There was no “ASD” diagnosed “then”. The concept of autism as a “spectrum” of disorders did not even exist in 1980!
Do you have any experience in descriptive epidemiology? If not, then your experience is not terribly valuable with respect to the information presented in your slide.
Have you read Shattuck, 2006? Schechter and Grether, 2008? How about Coo et al., 2008?
http://pediatrics.aappublications.org/cgi/content/full/117/4/1028
http://archpsyc.ama-assn.org/cgi/content/full/65/1/19
http://www.springerlink.com/content/wp0p251432g8677v/
I don’t need to tell you the correlation does not equal causation. You’ve seen a lot more cell phones, more of the internet, and more consumption of bottled water too.
Which you have yet to do.
“Belief with a fervor”. 100% science-free, ’nuff said.
Methinks you failed to delve very deep scientifically, so far. You seem to have taken a fairly faith-based position.
Why should you be heard? Do you have anything scientifically relevant to add?
Dr. Gordon, this is bullshit (and a straw man argument). There are studies ongoing.
Here are a few things anyone (even you) can quickly find on the CDC’s website:
http://www.cdc.gov/ncbddd/autism/seed.htm
Study to Explore Early Development (SEED)
SEED stands for the Study to Explore Early Development. It is a 5-year, multi-site collaborative study that will help identify what might put children at risk for autism spectrum disorders (ASDs) and other developmental disabilities.
http://www.cdc.gov/ncbddd/autism/research.htm
CDC-funded Projects
Thimerosal and Autism Study (September 2008)
Using the Vaccine Safety Datalink, CDC is comparing thimerosal exposure in children with and without autism. Certified specialists are using the most up-to-date, standardized diagnostic tests to evaluate children with autism. Review of medical records and interviews with parents are also included.
Italy Thimerosal Neurological Developmental Disorders Study (to be determined)
CDC is working with researchers in Italy to study children who were exposed to different amounts of thimerosal during infancy as part of a clinical trial of whooping cough vaccines. Some of the vaccines contained thimerosal, and some did not. The original clinical trial did not study neurological developmental disorders (NDDs), but researchers now are comparing their occurrence among participants to see how the level of thimerosal exposure affects NDDs.
Thimerosal Neurological Developmental Disorders Follow-up Study (to be determined)
This study looks at the relationship between thimerosal and neurological developmental disorders. Researchers are giving neuropsychological tests to children 7–10 years of age whose vaccinations in the first year of life could have contained thimerosal. The study will compare test results among children exposed to different quantities of thimerosal from vaccines and other exposures.
Comment by Do'C — 28 July, 2008 @ 10:07 pm
So true Pseudonym. The converse is also true - all one needs is to have a brain that is built to ignore certain things, and then certain things will be ignored.
Human perception is not necessarily always what many would like to think it is.
http://www.youtube.com/watch?v=voAntzB7EwE
http://www.youtube.com/watch?v=j779GMm6N5g
Comment by Do'C — 28 July, 2008 @ 10:12 pm
Indeed Joseph, thank you.
Dr. Gordon, where did you get the 1-2 in 10,000?
Comment by mike stanton — 29 July, 2008 @ 8:42 am
Regarding the effect of changing criteria on numbers, the Camberwell study by Wing and Gould (1979) is interesting. They studied the prevalence of autism in the special school population of Camberwell, effectively every child with an IQ < 70. When applying Lotter’s criteria (and they had the benefit of consulting with Lotter personally) they came up wit a similar figure of 4.9 in 10,000. But when they applied a broader set of criteria, the now famous “triad of impairments,” they identified a further 15 in 10,000.
It would be interesting to note how robust the combined figure of 1 in 500 for people with IQ 70 is what is driving the growth in overall numbers.
Comment by mike stanton — 29 July, 2008 @ 10:35 am
Sorry,
that last paragraph is garbled. It should read
“It would be interesting to note how robust the combined figure of 1 in 500 for people with IQ less than 70 is, and whether the growth in prevalence amonst people with IQ greater than 70 is what is driving the growth in overall numbers.”
Comment by Joseph — 29 July, 2008 @ 7:00 pm
@Mike: 20 in 10,000 is quite consistent with current findings on the prevalence of low functioning autism. See my post on the Amish.
Lorna Wing was a pioneer, way ahead of her time. She in fact suggested that prevalence could easily reach 1% if you combine those autistics with IQ less than 70 and those with IQ 70+ per Gillberg et al.
Comment by Orca — 4 August, 2008 @ 12:47 pm
See how many 78 year olds you can find with autism. It might help you understand the problem, DoC.
Comment by Do'C — 4 August, 2008 @ 8:35 pm
Orca, see page 10 of the following.
http://www.cdc.gov/nchs/data/nvsr/nvsr54/nvsr54_14.pdf
See the average life expectancy for males? What effect do you think this might have on the number of 78 year-olds who could be diagnosed as autistic? What do you know about the scientific study of utility of diagnosis?
Comment by mayfly — 9 August, 2008 @ 5:35 pm
In the past about 70% of autism suffers had I.Q.’s of < 70. Now that’s reversed about 70% of autistics have I.Q.’s greater than 70.
This is where most of the increase is coming from. Note this does not mean that children who would have been given an MR diagnosis in the past are not now called autistic, but that increase is dwarfed by the rise in HFA and Aspergers.
Comment by JayGordon — 10 August, 2008 @ 2:00 pm
Everyone intentionally misses the point, ALL of the previous studies and data are subject to a variety of interpretations. No matter how loudly people shout at one another, honest people will still disagree. I offer only one thing to the discussion: Thirty years of observation in a clinical, reasonably busy pediatrics practice. Nobody else involved here, and not Dr. Offit or any of the other experts involved can contribute this perspective. You may consider my perspective and my experience anecdotal and/or worthless, but I still offer it to you. I don’t deny having an agenda. My agenda is to ask everyone involved to look harder at vaccine formulations and schedules and change the way we do things in the best interests of children.
Jay
Comment by Do'C — 10 August, 2008 @ 5:21 pm
Dr. Gordon,
The “data” (and I use the term loosely) presented in the slide I originally asked you about, suggest an interpretation that is completely without scientific merit, or logical basis, and is probably best represented by the term, “credulous”.
Perhaps you’d like to give that little logic problem a shot?
Your point about previous studies and data being subject to a variety of interpretations was not “intentionally” missed by me. I see your point, but find it completely without merit to support your argument so far.
Honesty and disagreement are irrelevant to scientific reality. One may be honest, and one may disagree, but neither guarantees a clue about autism epidemiology.
No offense Dr. Gordon, but this is one of the weakest appeals to authority I have ever seen. Appealing to your “exclusive” status (based on number of years worked) and indirectly including yourself as an “expert” is about as irrelevant an argument that can be made, unless you can show that your experience is scientifically relevant.
What scientfic knowledge has been uncovered and documented as a direct result of your experience? Please try to be as specific to autism as possible, and provide appropriate supporting references/citations.
I’m willing to give you the benefit of the doubt, without declaring a priori worthlessness.
What scientfic knowledge have you uncovered and documented (or are in the process of documenting), and submitted for peer-review, as a direct result of your experience? Please be specific to autism, and provide appropriate citations for the readers.
Do you think it is in the best interest of autistic children to suggest that they are part of an “epidemic” without actual scientific data to support such a claim?
I won’t leave it readers to glean such a perspective from my opinion about the numbers from your slide, I’ll also refer them to the following written by you.
Linked Article By Dr. Gordon
If you have actual scientific data that shows an autism “epidemic”, please feel free to present it. If you don’t, I’d suggest that a very core premise of your argument with respect to vaccines and autism is fatally flawed.
Not to worry though Dr. Gordon, this is easily remedied with actual evidence or errata.
Comment by Prometheus — 11 August, 2008 @ 11:28 am
According to Dr. Gordon:
Now, I suppose that Dr. Gordon has some data to support his assertion that “everyone” (himself included, correct?) is intentionally missing the point? I’ll be waiting for that
I, for one, have been intentionally making a similar point: that the educational and social services “data” on autism prevalence are not sufficiently rigorous or consistent to make a valid comparison between this year and last year, let alone between this year and twenty years ago.
And I’ve been giving Dr. Gordon the benefit of the doubt by assuming that he was missing the point unintentionally.
“People” can be absolutely “honest” and still be dead wrong. I assume that the people involved in these “debates” are “honest” - to the extent that they are telling what they believe to be the truth. What I have said - over and over - is that telling “the truth” (as you perceive it) does not make you automatically correct.
Nobody else can offer Dr. Gordon’s perspective, true enough, but some can offer as many years of pediatric practice and others - such as Dr. Offit - can offer as many years of specialty pediatric practice and research into the science of immunity. That is a perspective Dr. Gordon can’t offer.
Dr. Gordon’s perspectives and experiences are anecdotal, since he has not - so far as he has reported - gone back through his “experiences” and systematically looked for the data. If he had, I assume that he would have brought forth the data rather than his “interpretation” of his “experiences”.
Finally, I don’t consider Dr. Gordon’s perspective or experiences “worthless” - I just don’t value them as highly as he does.
Prometheus