Monday, January 22, 2007

More Epidemiology From David Kirby

I realize I could be spending my time in more productive ways, particularly since the David Kirby show has clearly "jumped the shark". But I don't know if anyone else will feel it's worth it to take the time to verify David Kirby's numbers. Besides, I kind of like to analyze stats; it's a stim if you will.

The last time I posted, I looked into David Kirby's claims about drops in some IDEA numbers and I noted that there were discrepancies in the numbers reported by David Kirby compared to the only source he cited. I posted a request for clarification at, but my message was never approved apparently. I will try that again, adding a reference to this post.

On January 20, David Kirby reported (in EOHarm) that he scanned the IDEA data for drops in the autism category between 2004 and 2005. He found 4 states that show drops in the 3-5 administrative prevalence of autism between those two years: Hawaii, Iowa, Louisiana and Vermont. He adds:

The declines are modest, but noticeable, and the first ever decline in actual numbers that has ever been reported, as far as I can tell.

Well, he could not tell very far apparently. I went ahead and looked at state by state data between 2000 and 2005. After a quick glance at data for these four states only, I can see there was a big drop in the Iowa 3-5 caseload between 2000 and 2001; one in Louisiana between 2002 and 2003; and one in Vermont between 2002 and 2003. I am sure other drops can be found in other states, particularly those with small autism caseloads. David Kirby might have been close to realizing that when he said the following:

These four states have little in common and a from 4 completely different regions. The only trait they share is a relatively modest population rate. I don't know if that matters or not.

You see, the 3-5 caseload in California does not show any big fluctuations because California has thousands of children in that cohort. The four states Kirby found have less than 200 autistic children in that cohort. Fluctuations are expected statistically. I made a graph of caseload in these four states for 3 year olds only. Fluctuations are more obvious in that narrower cohort, but I chose 3 year olds because that would be the most telling cohort in terms of a drop in thimerosal exposure.

When you look at that graph, keep in mind that this is not a representative sample of states. This is a graph of four states handpicked by David Kirby based on whether their 3-5 prevalence has dropped between 2004 and 2005.


  1. I wonder if he'll figure out that he's not just playing to his adoring fans but that skeptics are watching and listening to his garbage, too? His evidence is getting thinner and thinner all the time. I guess the idea is to throw out a hundred easily disprovable hypotheses hoping that the mass of information will slow people down enough and impress them with the breadth of his knowlege or something.

    This autism is falling in 4 states thing is hysterical. He's back to the "autism is thimerosal poisoning and dose dependent" thing, I think, but what do you bet that you can't link these "drops" or the increases in other states to vaccine uptake or thimerosal exposure, one way or the other.

    Anyway, Kirby said that people were moving frrom other states to California to get our wonderful services... the ones Margaret Bauman compared to those of a "third world country." But it would be funny to find out if the numbers dropped in those states because parents moved to other states. :-)

  2. What strikes one very forcibly when you look at all the CSV data is the sheer number of numbers. Out of 50 states, only 4 show declines! This is just silly. As one astute person on EoH pointed out, maybe the decline in Louisiana is due to all the children moving away after the hurricane. Or maybe it's just a random fluctuation caused by those tiny case loads. Can you say under-diagnosis, Mr. Kirby?

  3. What does "jumped the shark" mean?

    I thought I was an obscurantist ranter but that phrase has escaped even my oculation and expository acutitionary eclectecceapocalypsis.

  4. Larry, look it up on Wikipedia. I linked to the wikipedia article when I first used the expression on my blog... I picked it up from someone else, it was new to me a month or so ago.

  5. His evidence is getting thinner and thinner all the time.

    I don't know what I'd compare it to. He's scanning all databases trying to find some small fluctuation that could remotely support his position. But you're right that maybe he's doing that just for the benefit of his fans at this point.

  6. Out of 50 states, only 4 show declines!

    Exactly. The way I'd explain it is that if autism prevalence were to level off in the US, you'd expect to see 25 states with drops and 25 with increases. We're pretty far from that yet. Next year we might see other 4 states (give or take) with drops.

  7. Is it also possible that the explanation for the drop for Louisiana's numbers between 2004 and 2005 are due to Hurricane Katrina? Remember, the Hurricane hit Louisiana on August 29, 2005 and people in the most populated area (New Orleans and surrounding suburbs) were evacuated, mostly out of state. Although I am sure numbers were established before the beginning of the school year there, it is possible that information may have been lost during the storm.

  8. Yes, that's very likely. Out of the 4 states, note that Louisiana is the one with the largest autism caseload. The 2005 drop is easily due to Hurricane Katrina. What I can't explain is the big spike in 2004.

  9. Hi Joseph,

    This might help explain the spike:

    Quoting from that site:

    "Dozens of medical diagnoses can be used to describe a child with some or all of the features of autism (e.g. autism, pervasive developmental delay, Asperger’s Syndrome, Rett Syndrome, Fragile X, multi-system developmental delay, etc.). The IDEA has only nine eligibility categories, only a few of which actually apply to children with autism. Although the educational assessment may suggest which types of educational assistance may help an individual child, the federal law (US Code, Title 20, Chapter 33, Section 1414) states that specific needs (not eligibility) should determine what services are offered. As a result, the educational assessment emphasizes eligibility rather than diagnostic accuracy. This wouldn’t be a problem except that the education “diagnoses” are reported to the U.S. Department of Education (USDE) and then used by others to support their positions on autism.

    Further clouding the issue, each state sets its own assessment standards and procedures, within the limits of the IDEA. Most states have followed the diagnostic guidelines in the American Psychiatric Association’s Diagnostic and Statistical Manual (4th edition)—the "DSM-IV"—but others have created assessment guidelines that are very different. However, many of the children assessed as autistic by the schools do not meet the DSM-IV criteria for autism."

  10. Also, I should note that there were revisions made to IDEA that were signed into law by Pres. Bush on December 3, 2004. The spike could also represent Louisiana's change in assessment criteria to meet those revisions.

  11. Thanks for that info, Anonymous.

  12. FYI - My second request at HuffPo for clarification on the discrepancies between David Kirby's numbers and those at has not been approved either apparently. I do not know on what basis the messages have been censored. They were not offensive messages in the least. They were just requests to clarify why the numbers don't match.

  13. Joseph wrote: "FYI - My second request at HuffPo for clarification on the discrepancies between David Kirby's numbers and those at has not been approved either apparently."

    Do you suppose that David Kirby approves those messages personally? That would be the only way that we could reconcile the HuffPo's apparent belief in free speech and the censorship that is occurring at that blog.

  14. Either David Kirby himself or a mercury activist would be my guess. I don't think anyone else would've censored messages like that.

  15. Who gives a shit about DSMIV? That is the whole hypocrisy that runs through these blogs,

    Because if you accept that DSMIV accurately describes what autism is you are all as guilty as Kirby of fabrication.

    DSMIV describes what a trade organisation has decided they would like autism to be for wholly professional reasons, and you think only the Geiers and the Bradstreets make things up, corruption is rife in Psychiatry and disorder is the name of the game, if you can lable it, then you can get paid for treating it.

    What really amazes me sometimes is this whole pretend game of objectivity of "science" against "science" you are all playing by the same rules without ever thinking whether the rules are valid or not.

    Why use a bad tool (like DSM in all its variants) to combat a bad argument?

    Think outside the box, I do.

  16. Larry
    i think the point is not that we accept these tools. Rather they rely on them to make their case for epidemics etc. When their weapon of choice lets them down (in this case IDEA data) it is legitimate to point this out.

  17. Larry,

    I think you mean IDEA, not DSM-IV, which was not the topic of the post. But ok.

    DSM-IV is a psychiatric definition, nothing more. It's not necessary to accept that they got it right or that they own it. But to argue the science that sorrounds certain definitions, you can't do it by discounting those definitions.

    You also can't argue science with non-science. That doesn't make sense to me. You can argue that the social model is a better model to the medical model, but that doesn't address questions of fact.

    The claim that there has been a DSM-IV autism epidemic is a scientific claim. You can only successfully argue that scientifically. Even if you argue that culture is what drives increasing recognition of autism, that is still a scientific proposition.

    I truly don't think an anti-science stance would do us any good.

  18. Any one who believes that DSMIV is useful for diagnosing autism is not on my side.

    Anyone who believes that autism belongs in the proper domain of psyciatry is not on my side.

    Until we change the way that autism is thought about these battles with Kirby are pointless because it all leads straight back to another medical model.

  19. I too have posted very innocuous comments at HuffPo and had them never show up. Mr. Bring-It-On, indeed.

  20. When you have a moment, check out what we are doing at I would love to hear your thoughts. I really enjoy your blog, thank you for what you are doing.


  21. Southeast and main Asian pandora jewelry countries have twisted rubies for centuries, cheap pandora bracelets but research as to where, and how to find more deposits is Pandora charms spare, and production has figured out how and mining companies,” Pandora beads Giuliani says, to look at exactly the right time and place.” pandora set Farther investigation of claret formation, based on tectonic scenery, cheap pandora geochemistry, fluid inclusions and isotopic ratios, allowed discount pandora Giuliani’s lineup to remodel a new prototype for the French Institute pandora 2010 of Research for Development (IRD) and the National Scientific pandora sale Center of Research, two government-sponsored knowledge Pandora Bangles and technology research institutes that aim to aid in the sustainable cheap pandora bracelets development of developing countries. Before the collision pandora bracelets prices of the Eurasian and Indian plates, lagoons or deltas sat in the regions where marble is giant, pandora bracelets and charms he says, “and there is the brains to expect that the new pandora bracelets sale thoughts should help development of the artless capital.” discount pandora bracelets Virginie Garnier, Gaston Giuliani and Daniel Pandora necklace Ohnenstetter urban the shape to do just that. They work for the garnet cheap pandora charms genesis. While studying the bedrock in Vietnam in 1998, the discount pandora charms French players found rubies, which detained traces of aluminum, chromium pandora charms sale and vanadium from universities, international corporations, governments pandora charms 2010 and why the rubies got there, and has created a paradigm Pandora beads to help these evaporites, Garnier says, when the Eurasian cheap pandora beads and Indian plates collided, raising the Himalaya Mountains.