New Autism Study

California’s sevenfold increase in autism cannot be explained by changes in doctors’ diagnoses and most likely is due to environmental exposures, University of California scientists reported Thursday.

The scientists who authored the new study advocate a nationwide shift in autism research to focus on potential factors in the environment that babies and fetuses are exposed to, including pesticides, viruses and chemicals in household products.

Scientific American

Comments

  1. #1 stewart
    January 12, 2009

    Unfortunately, it’s a poor study. It resolutely refuses to consider the notion of diagnostic substitution, (the fact that the overall number fo children needing services is relatively consistent, although the labels change). Here’s a good paper from Psychological Science in the Public Interest that’s speaks directly to this point.

    http://www.psychologicalscience.org/pdf/cd/autism_epidemic.pdf

  2. #2 Orac
    January 12, 2009

    It’s a really, really crappy study you’re hawking here, I’m afraid:

    See:

    http://www.theness.com/neurologicablog/?p=454

  3. #4 David Lee
    January 12, 2009

    Posting is different from endorsing. Though Orac can do no wrong, since I read his blog.

  4. #5 Robert S.
    January 12, 2009

    At least the actual source article included some of the graphs. If the quotes from the researchers are accurate and indicative, they do not seem to be supported by the published research. The fact that the greatest rate of change occurred in the 2-3 yo. group is even more troubling to me. How the hell do you apply the DSM-IV criteria to a 2 year old. It seems nearly incomprehensible that a behavioral syndrome is being diagnosed at such an age. If someone suggested diagnosing 3 year old children with depression I should hope they would he soundly laughed out of whatever positions they held.

  5. #6 Harold L Doherty
    January 12, 2009

    Does Orac disagree with Dr. Bernardine Healy’s contention that the practice of thimerosal containing flu vaccines is a possible environmental inducement of autism that should be investigated?

    Does Orac disagree with the UK (Bromely, Mawer, Clayton-Smith, Baker) study showing that children born of mothers given the epilepsy drug valproate were seven times more likely to develop autism than children born to mothers who took no epilepsy drug while pregnant?

  6. #7 clinteas
    January 12, 2009

    Hahaha,oh Greg,

    did you realize what sort of can of worms you were opening LOL

    Without getting into the nitty-gritty of the trench wars on autism here,I ask everyone to just think about the diagnosis of ADHD and depression in children,and the incidence of these conditions since we started diagnosing them.Every kid that forgets to memorize a poem for class and instead goes riding his bike in the backyard is diagnosed with ADHD these days.
    Same goes for autism.
    And the nutters on both side of the trench are not helping to rationalize the dabate.

  7. #8 Azkyroth
    January 12, 2009

    Does Orac disagree with Dr. Bernardine Healy’s contention that the practice of thimerosal containing flu vaccines is a possible environmental inducement of autism that should be investigated?

    My understanding was that Orac disagrees with the kooks’ contention that the repeated investigations that have been undertaken thus far and failed to support a causal link should be thrown out and the investigations repeated until statistical noise produces a result that is to the kooks’ liking.

  8. #9 Greg Laden
    January 12, 2009

    Wow! I am truly astonished at the dismal quality of most of these comments. Come on people, we can do better than that. I know nothing about this issue, and a two minute inspection allows me to discover that most of you are speaking out of an orifice other than your mouth. Let’s review.

    stewart: Are we allowed to read only the first page of that paper? Anyway, it is simply not the case that the study refuses to consider changes in diagnostic rate. It IS, in large part, a study of changes in diagnostic rate. If you want to claim that changes in diagnostic rate against an otherwise stable background rate is the explanation for the documented trend, then pleas do so in a way other than waving a magic wand. The study makes explicit statements about this that you are free to mention and disagree with. That would be better than just saying it is not there. Right now you are sounding like a kook when you say that. Please revise.

    Orac: Hawking? Bite me. You know, sometimes I think it is really too bad for all those kids who should be getting vaccines that you are on their side. They deserve better. Thanks for the links to the blog posts, though, that was nice. Let me know if you have something substantive to say about this study, I’d love to hear it.

    Robert S.: That is not the source article you are linking to. You note: “they do not seem to be supported by the published research.” I just want to verify that you can’t say this without having seen the published research. You may be right, but you would have to look at the published research before actually knowing if you are right. Regarding the age and diagnosis issue, you’ve got nothing. The typical mildly or poorly informed member of the public often believes that early diagnosis of whatever (fill in blank) is impossible or too difficult because they can’t imagine how it is done. Until I see something more from you I’m going to assume that this is the case with your statements. Feel free to correct me if I’m wrong about that.

    Harold: I am not a this moment familiar with the studies you (sort of) cite, but I can ask this: Do they look at interaction effects? Is there not a likely similar cause for the condition for which the mothers are treated and the autism in the children? You seem to be more engaged in a fight with Orac (which I can understand … he invokes those feelings with his lovely style) than a discussion over the cause of autism. The cause of atuism is more important than your ego or Oracs. Can we try to get on the same page here?

    clinteas: I’m not so sure. Kids these days ARE doin’ it wrong, as kids in EVERY generation have! Don’t you think?

    Azkyroth: That seems to be correct, but what about the possibility that the accepted methods are inadequate for all of the possible demand of research? For instance, isolating causes is at the base of almost all research designs, at some level, but this makes it difficult to study interactions.

  9. #10 qetzal
    January 12, 2009

    Anyway, it is simply not the case that the study refuses to consider changes in diagnostic rate. It IS, in large part, a study of changes in diagnostic rate.

    Did you read Orac’s second link?

    I wish I had access to the full text of the study, because the Natural Variation blog post suggests that the study didn’t directly evaluate diagnostic substitution at all. From the blog post:

    Finding [from the study]: Changes in diagnostic criteria have contributed a 2.2-fold increase in the rise of autism incidence in California.

    How is this determined? This being the most important area of the paper, my expectation was that I would find new data produced by H-P et al. showing that only a 2.2-fold increase may result from changes in diagnostic criteria. This is not the case at all. H-P et al. rely on prior work.

    Specifically, they rely on Kielinen et al. (2000), a study out of Finland based on data “collected from hospital records and the records of the central institutions for the intellectually disabled in the Provinces of Oulu and Lapland in 1996–1997.”

    How do the study authors justify their assumption that diagnostic substitution rates in California over a decade or more are numerically identical to those in two Finnish provinces over a 2 year period?

  10. #11 Orac
    January 12, 2009

    Bite me. You know, sometimes I think it is really too bad for all those kids who should be getting vaccines that you are on their side. They deserve better. Thanks for the links to the blog posts, though, that was nice. Let me know if you have something substantive to say about this study, I’d love to hear it.

    Give me a break with the self-righteous whining. You posted a link to a story on a study without comment, and the implication tends to be that you find it interesting and may even believe its results, especially given the excerpt you chose to post. You don’t think it’s worthwhile? Then write something–anything–about it, other than the “link and excerpt.”

    Come to think of it, you’re one to talk about not having something substantive to say about this study when you yourself didn’t bother to–oh, you know–actually say something about it. Pot, kettle, black, dude. At least I provided links to good skeptical discussions of the study to support my characterization that it’s a bad study. Indeed, I consider those discussions so good that, given that both bloggers beat me to discussing this study, I decided that I didn’t have much to add.

  11. #12 Orac
    January 12, 2009

    Does Orac disagree with Dr. Bernardine Healy’s contention that the practice of thimerosal containing flu vaccines is a possible environmental inducement of autism that should be investigated?

    Yes, I do disagree with Healy, who has turned into a bit of a crank over the last couple of years. Multiple large, well-designed studies have failed to find a detectable association between thimerosal-containing vaccines and autism. Scientifically, it’s a nonissue. In fact, even antivaccine activists realize it’s a losing issue, and that’s why they’ve changed to the “toxins” in vaccines gambit or the “too many too soon” mantra.

  12. #13 Greg Laden
    January 12, 2009

    Orac: Ignoring for the moment your irrelevant ranting about how and what I should blog, I have a serious question for you: Why would the shift from vaccines to “toxins” in vaccines have any wind at all? Isn’t the real killer study regarding the vaccines based on the idea that there has not been a downward shift in autism rates since the most yecky of the toxins was removed from vaccines? Is it the case that the autism = vaccines link is really mainly an anti-vaccine phenomenon that is exploiting autism rate increase (real or perceived)?

    BTW, I had not read the original paper because when I looked for it in the journal, it was not there. I had not realized it was published in the last full issue (I was expecting it in the forthcoming section.) I’ve got it and I’m reading it.

  13. #14 Orac
    January 12, 2009

    Ignoring for the moment your irrelevant ranting about how and what I should blog, I have a serious question for you: Why would the shift from vaccines to “toxins” in vaccines have any wind at all? Isn’t the real killer study regarding the vaccines based on the idea that there has not been a downward shift in autism rates since the most yecky of the toxins was removed from vaccines? Is it the case that the autism = vaccines link is really mainly an anti-vaccine phenomenon that is exploiting autism rate increase (real or perceived)?

    Ignoring for the moment your irrelevant complaining about my supposed ranting about how and what you should blog that you indulged in after your even more irrelevant ranting about how and what I should comment about after your post, see:

    http://www.sciencebasedmedicine.org/?p=14

    and

    http://scienceblogs.com/insolence/2007/11/cries_the_antivaccinationist_why_are_we.php

    It is exactly those studies that fail to find a drop in autism prevalence after thimerosal was removed from vaccines that I was referring to when I said that large, well-designed epidemiological studies have failed to find an association between thimerosal and autism.

    As for whether the “vaccines=autism” bit is mainly an antivaccine phenomenon, these days (and for at least the last three years), the answer is an emphatic yes. Note how antivaccine activists effortlessly move the goalposts. David Kirby, for instance, once said that, given that thimerosal was removed from nearly all childhood vaccines by the end of 2001, if autism rates didn’t fall by the end of 2005 it would be a “serious blow” to the mercury-autism hypothesis. Rates kept climbing through the end of 2005, and he then moved the goalposts to the end of 2007. 2007 came and went, and in 2008 he pushed the goalpost back to 2011.

    In the meantime, realizing that the mercury hypothesis was becoming increasingly untenable scientifically (it never was that tenable in the first place), antivaccination advocates switched to its being the vaccines themselves (which is what they thought all along; the thimerosal/mercury bit was a convenient bogeyman for them to point to hide their antivaccine views). That’s because it’s all about the vaccines, not any single ingredient in them:

    http://scienceblogs.com/insolence/2008/02/still_more_evidence_that_its_all_about_t.php

  14. #15 Joseph
    January 12, 2009

    I know nothing about this issue,

    Greg, this will sound crass, but evidently you don’t. Let’s see:

    stewart: Are we allowed to read only the first page of that paper? Anyway, it is simply not the case that the study refuses to consider changes in diagnostic rate.

    Stewart mentioned diagnostic substitution. The study does not look at diagnostic substitution at all. It looks at changes in diagnostic criteria, age of diagnosis and inclusion of “milder” cases. Is diagnostic substitution part of changes in diagnostic criteria? Not necessarily. I argue that in my critique.

    Of course, a more general way to explain what’s missing from the paper is that it does not consider awareness, as the researchers admit.

    Furthermore, the paper explains a 4.26-fold rise. It claims this means they cannot account for 6.85-fold rise. How so? Where’s the range of statistical confidence?

    Interestingly, the paper claims the impact of diagnostic criteria alone is precisely 2.2-fold. This comes from a single Finnish study, and the researchers use this in favor of the results of a meta-study claiming that the impact of criteria is 3.6-fold. If H-P et al. had used the 3.6 figure, they would’ve explained all of the rise (102% to be precise).

  15. #16 itzac
    January 12, 2009

    I’m about to talk mostly out of my ass, here goes:

    You know there is a third possible explanation for the rise besides statistical or environmental. That’s that the cause of autism is genetic, but that human mating habits have changed in recent decades to promote some genetic disorders. There are recent studies that demonstrate changes in partner selection by women taking birth control over women not taking the pill. These changes certainly have implications regarding the immune system. It possible there are other effects, as well.

    Before feminists jump on me, understand I’m totally with you, and I fully understand the implications of that statement. This is worth studying, and if turns out to be true, the proper response is to reevaluate how birth control is used, and to investigate non-hormonal forms of birth control for both men and women.

  16. #17 Jennifer
    January 12, 2009

    I would suggest that all interested people read the actual, published paper at:

    http://www.vetmed.ucdavis.edu/cceh/Epipaper1208.pdf

    The tone is completely different from that of the press releases.

  17. #18 CyberLizard
    January 12, 2009

    Oooh, Orac and Greg going at it!
    *sits back and munches popcorn*

  18. #19 Greg Laden
    January 12, 2009

    Jennifer: Yes, the tone is totally different. This paper is being slammed by people who, sadly, seem to know what they are talking about but will take this paper down because there is the possibility that it can be interpreted as opposed to their opinion. Which, in my opinion, is more interesting than the paper itself. The paper is OK, a step forward in some areas, good and flawed at the same time, but not as interesting as the debate itself, in this case!

  19. #20 Greg Laden
    January 12, 2009

    Joseph: The study at hand considers diagnostic substitution at a number of points. It is not the main focus.

  20. #21 Joseph
    January 12, 2009

    The paper is OK, a step forward in some areas, good and flawed at the same time, but not as interesting as the debate itself, in this case!

    I completely disagree. I think people have been too kind to the paper, perhaps because it’s not a paper by cranks. Frankly, I would’ve expected a paper like this from the likes of the Geiers.

    The only novel result from the paper is the 1.2-fold rise due to age of diagnosis. (I would caution this should not be interpreted as age of diagnosis not having changed much – that’s the change in the proportion of diagnoses by age 5).

    The other results are recycled from elsewhere.

    They use a single Finnish epidemiological study to come up with the figure on the impact of diagnostic criteria. It’s a complete stretch to suggest that the results of that study might be applicable to California DDS ascertainment between 1990 and 2006. Are we also to believe it’s precisely 2.2-fold and it can’t be a little higher?

    It’s true that the paper uses appropriately skeptical language at times. I mean, it has to. But you should see what the primary author said to the media.

    Honestly, if they wanted to raise funds for environmental research, there are probably better ways to achieve that. Calling autism an epidemic has many non-trivial moral and economic implications. False epidemics solicit false causes and waste research money.

  21. #22 Greg Laden
    January 12, 2009

    So, Joseph, your criticism is then mainly with the media side of this? There is a very different slant when comparing the paper and the UCD press report (go read the whole thing, don’t base anything on my little quote). And when we go one further step out into media land the fundamental nature of the paper is almost entirely lost, and it looks like a paper specifically asserting that there are specific environmental causes.

    But that happens to all papers all the time. No need to single this instance out as unique in this regard.

  22. #23 Kev
    January 12, 2009

    Greg, its quite obvious that Joseph *has* read the paper. Its equally obvious that when you posted this blog entry, you had not. I’m going to quote you something from the paper and then something from CDDS. See if you can spot the problem:

    H-P:

    The inclusion of milder cases has been suggested as an explanation for the increase in autism. Neither Asperger’s syndrome nor ‘pervasive developmental disorders not otherwise specified’ qualify under the category of autism in the DDS system.

    CDDS (private email)

    The current CDER was written in 1978 and updated in 1986, which is why the language is so out of date ( e.g., Residual Autism). California has clinicians in the field who are, of course, using modern criteria in their assessments but then they have to go backwards and try to fit those kids into the 1986 CDER. So you are going to have Aspergers kids, PDD-NOS kids in both categories 1 and 2. Categories 1 and 2 are called ‘Autism.’ But because there are so many clinicians, using lots of different techniques for evaluation, there is a lot of inconsistency and enrollment figures should not be misused as epidemiological data.

    Can you see the problem with this paper yet?

  23. #24 Joseph
    January 12, 2009

    So, Joseph, your criticism is then mainly with the media side of this?

    Obviously, it’s not mainly that. What I’m saying is that the paper should not be given a pass just because it uses some skeptical language, when the authors (and others) clearly want to communicate that this is a definitive study proving the “autism epidemic.” For reference, this is how Dr. H-P is quoted by WebMD.

    “When you put it all together, this doesn’t come close to explaining the increases in the last 10 years,” Hertz-Picciotto tells WebMD. “The more you whittle away at this increase, the more you have to say that what is left over is real. Given that autism cases keep going up, and can’t be fully explained by artifacts, environmental factors deserve serious consideration.”

    I took that from Autism Street, BTW, where you’ll find yet another post on the matter, primarily looking at diagnostic substitution in California, which incidentally is hard to deny once you look at an aggregate of all special education categories.

  24. #25 Robert S.
    January 21, 2009

    Greg: You are correct that I had not been able to find an accessible copy of the paper that the article commented on. However, if you read to the end of the article you linked to you would find that it was a reprint. After seeing that I found the original article (commenting on the paper in Epidemiology 2009;20:84-90) and found the reprint was without the graphs. I was unclear in my language regarding the press article vs the published paper which probably lead to my comment being misunderstood.

    Now that I have read the paper (thanks to Jennifer), I do have a problem with one of the conclusions made in the discussion section. They never seem to allow for early incidence not corresponding to the expected prevalence that might occur if the developmental delay was temporary. A very brief and shallow look at pubmed suggests that at least some children would no longer qualify under the diagnostic criteria if evaluated later (e.g. PMID: 17206522). Significant changes in the training of pre-k teachers in California has taken place since the last full birth cohort (1996). If nothing else it will be interesting to see if the conclusions drawn will correspond with the data as it develops over the next few years.

Current ye@r *