Kirby gets his maths wrong, and accidentally declares an end to the ‘autism epidemic’

6 Mar

Kirby’s crowing about a recent US government concession in a vaccine damage court case has got a fair bit of attention, and has been well-covered by several blogs – for a general appraisal, I’d start with posts on LBRB and Respectful Insolence, if you haven’t already. I was interested to see, though, that Kirby manages to get his basic maths wrong on the Huffington Post – meaning that his claims about autism, mitochondrial disorder and vaccine damage are left looking entirely ridiculous. Stephen Novella argues that “Among stiff competition, this is perhaps the most absurd and scientifically ignorant thing Kirby has every written.” It might be the most absurdly mathematically and statistically ignorant thing that Kirby has ever written, too.

For Kirby

mitochondrial disorders are rare in the general population, affecting some 2-per-10,000 people (or just 0.2%)

Actually, no – 2/1000 is 0.2%, 2/10,000 is 0.02%. As anyone with basic numeracy skills could tell you.

In itself, that doesn’t inspire much confidence in Kirby’s statistics . However, things get odder with Kirby’s numbers. He argues that

Mitochondrial disorders are now thought to be the most common disease associated with ASD. Some journal articles and other analyses have estimated that 10% to 20% of all autism cases may involve mitochondrial disorders, which would make them one thousand times more common among people with ASD than the general population

I can’t get these numbers to make sense. I will be as charitable to Kirby as possible: let’s say that all mitochondrial disease occurs in people with ASDs (it doesn’t, as Kirby appears to acknowledge in his post, but we’ll be kind here). Let’s take the lower end of Kirby’s claim that 10%-20% of all those with ASDs have mitochondrial disease. This would mean that 0.02% of the population constitute 10% of those with ASDs. Therefore, even being unrealistically charitable, Kirby’s figures could only allow for 0.2% of the population – only 1 person in every 500 – to be on the autistic spectrum. This is a remarkably low estimate of the percentage of people with ASDs, and taking a more realistic approach (for example, there are people with mitochondrial disorder who aren’t on the autistic spectrum) would reduce this low estimate further.

Maybe Kirby is declaring an end to the ‘autism epidemic’?

18 Responses to “Kirby gets his maths wrong, and accidentally declares an end to the ‘autism epidemic’”

  1. David N. Andrews M. Ed. (Distinction) March 6, 2008 at 13:29 #

    “Maybe Kirby is declaring an end to the ‘autism epidemic’?”

    Doubt it very much.

    He’s just shite at hard sums.

  2. Jon March 6, 2008 at 14:28 #

    Sorry David – that statement about the ‘autism epidemic’ was intended to be sarcastic, but this doesn’t come across well in print.

  3. Joeymom March 6, 2008 at 14:54 #

    Heck, he seems to be bad at EASY sums.

  4. Joseph March 6, 2008 at 15:19 #

    So what’s new?

    The 10% to 20% proportion must be wrong, or the prevalence of Mito disorders is wrong. For the former, the question to ask is how it was determined. Is it through some sort of “markers”? Were there controls in the study?

  5. Genevieve March 6, 2008 at 15:41 #

    Please join me in Blogging For Autism Awareness in April . http://momologue.blogspot.com/2008/03/join-me-in-blogging-for-autism.html

  6. Where's that horde? March 6, 2008 at 16:40 #

    He’s found the “mitochondrial hidden horde”!

  7. Leila March 6, 2008 at 18:40 #

    The quack labs and docs will make a killing testing and prescribing supplements for mito unnecessarily.

  8. livsparents March 6, 2008 at 19:23 #

    OK…now you’re just making fun of the mathmatically challenged. I hope you know that we are a coalition 10,000 strong with 50% in foreign countries and 47% here in the states. If you stretched us end to end we’d reach from NY to Chicago and back again. So next time you want to pick on us, you can be 110% sure we are going to mobilize our 100K membership…

  9. Ms. Clark March 6, 2008 at 20:22 #

    Livsparent,

    Kirby keeps trying to stretch his credibility from New York to Chicago and back again… All of Kirby’s twisted words, if laid end to end would reach from Brooklyn to Washington DC.

    I shouldn’t laugh at his poor math skills, I didn’t catch Kirby’s error. But then I assume that 48.947% of what he says or writes is an outright lie, and 19.448% is twisted truth, the remaining 36.831% is correct, but it’s only correct because he hit on accurate statments randomly. :-]

  10. qchan63 March 6, 2008 at 20:51 #

    Just gotta say, i love this observation by Orac:

    “Let’s remember what the full title of Kirby’s book was: Evidence of Harm: Mercury in Vaccines and the Autism Epidemic, a Medical Controversy, not Evidence of Harm: Vaccines Aggravating a Rare Preexisting Genetic Mitochondrial Disorder and Causing a Condition That Mimics Autism.”

  11. Brian March 6, 2008 at 21:20 #

    I think we need to be generous. It has been shown that 72.4% of statistics are inaccurate.

  12. Jon March 6, 2008 at 21:57 #

    Livsparents- sure, some very bright people can’t do maths. Some of them do extremely good research. But – when maths is needed – they would generally work with those who are better at doing maths, or have their work checked.

    Certainly, if doing research in areas such as vaccination – where inaccurate information can lead to really bad health decisions, and thus to serious injuries and to deaths – one would want to have the figures right. Kirby, though, I suspect offers an example of how “incompetence robs [people] of the metacognitive ability to realize” their own incompetence [PDF article].

    Ms Clark- I wonder if Kirby’s credibility will prove to be as stretchy as his figures? Sadly, it looks like it may.

  13. Schwartz March 7, 2008 at 04:29 #

    I believe what Kirby was trying to say (pretty poorly) is:

    1) Mito Disfunction in ASD patients is up to 20% prevalence
    2) Mito Disfucntion in general pop has a prevalence of 0.02% (1 in 10000)

    Therefore Mito Disfunction is 1000% higher prevalence in ASD Patients than in the general population. (20% is 1000 times larger than 0.02%). I might be missing something, but the only error seems to be his initial 1/10000 = 0.2% which should read 0.02%.

    His 20% is loosely support by the retrospective study in the Case Study writeup by Poling-Zimmerman which showed SOME elevated markers in Autistic patient ranging from 22 to 47% (vs much lower results in controls that were suffering from neurological problems).

    I find all of this very interesting given that Ip et al., a bunch of highly respected researchers published a study in a peer-reviewed journal (let alone quoted in numerous other studies by experts), with huge calculation and measurement errors. None of the study population numbers in the analysis turned out to be even remotely correct.

  14. Ms. Clark March 7, 2008 at 05:40 #

    Isn’t it amazing that so many autistic kids like Hannah are out there, gosh, even 20% of autistic kids…, maybe (the ones who regress tragically and immediately following a vaccine)
    and not one DAN! quack has pointed out that about 20% of his patients have this one kind of biochemical problem that shows up in labs like x, y and z.

    Isn’t it amazing that not one of the biomuddled quacks has applied to get research funding to identify this subset?

    And isn’t it amazing the kind of money certain fat-headed wealthy individuals have put out on nonsense trying to call autism as another name for mercury poisoning, (you know like the fat heads at SAFE MINDS and elsewhere) hasn’t gone to look at allllll this mito disease in autism? Instead they funded phone surveys and monkey studies and mouse studies that ended up to be fraudulent… that sort of thing. Now they are calling for research into this? Hmmmm.

    Well, all I can say is it’s about time. And I also think that they won’t find any more like Hannah at all, but I could be wrong about that so they should look. Starting with the other clients that the antivax sharks have had under their greedy noses all this time.

  15. David N. Andrews M. Ed. (Distinction) March 7, 2008 at 09:52 #

    Hi Jon… I probably did miss the sarcasm in your post, but nonetheless it probably doesn’t affect the outcome too seriously 🙂

    Like Joeymum intimates… Kirby’s pretty ropey at lots of things and lots of levels… 😛

  16. Joseph March 7, 2008 at 18:07 #

    I’m surprised Swartz missed what the error is. The numbers just don’t add up, Swartz. It’s impossible for Mito disease to have a 0.02% prevalence and at the same time affect 20% of autistic children. You see that?

  17. Jon March 7, 2008 at 22:16 #

    Yes, as far as I can see Kirby’s figures aren’t consistent with current estimates of the percentage of the population who are on the autistic spectrum.

  18. Schwartz March 8, 2008 at 02:42 #

    Joseph,

    Now I read it properly. Thanks for correcting me.

    I looked over those numbers several times, and completely missed the point. 😦

Comments are closed.

%d bloggers like this: