Another misinterpreted study. This time they are misusing Israel data

22 Feb

Kent Heckenlively from the Age of Autism blog has recent post up on a study out of Israel on autism. In it, he notes that the “incidence” of autism in Israel actually dropped for a few years

What’s curious, though, is how this population of medical professionals who were supposedly good at identifying and diagnosing autism in 1999, then had a drop of more than 40% by 2002. Did they lose their newly-acquired skills in a sophomore slump? If I’m not mistaken, those were also the years of the vaccine-autism “panic” which began shortly after the publication of Dr. Andrew Wakefield’s article in the Lancet, linking the MMR shot and autism. Was there a drop in vaccination rates in Israel after the Wakefield publication? Did vaccination rates then subsequently go up in later years?

Here’s the figure.

Figure 1 from Israel Autism Paper

Figure 1 from Israel Autism Paper

Let’s do this quickly. The interpretation is nonsense. Start from the fact that Google Ph.D.’s seem to rarely check simple facts using google.

Enter “Israel Vaccine Schedule” into google. The top link is a WHO site.

Take a look at how the MCV (Measles Containing Vaccine) coverage has varied with time.

MCV uptake was increasing steadily up until 2003, when it dropped for two years.

Year MCV Uptake Measles cases
1998 97% 8 (Year of Wakefield paper)
1999 97% 14
2000 97% 36 (year “incidence” starts to drop in Israel
2001 96% 19
2002 98% 2
2003 95% 124
2004 84% 116 (year MCV uptake drops)
2005 96% 2
2006 96% 9
2007 96% 539 (measles spike)

Obviously the drop in autism “incidence” in Israel isn’t related to measles vaccines–the drop in uptake happened after the drop in autism “incidence”. Note that I put “incidence” in quotes. The paper isn’t measuring incidence. I’ll get to that later.

In 2007, something else happened. Measles cases spiked. Yep, 3 years after vaccination rates dropped, there’s a big spike in measles. Are they connected? Possibly. I would want to see information like how many of the 539 measles cases were children 4-5 years old, for example.

Other issues with the paper.

1) they don’t give “incidence” or “prevalence”, really. They are giving the number of people who are getting services for autism. Sorta like the California DDS data. Why is this important? We don’t know how easy it is to qualify for services. We also don’t know how hard they are looking for people with autism. It isn’t the same thing as a measure of all the people with autism.

2) Note that 97.5% of the people with autism are Jewish. About 25% of the population in Israel is non Jewish (if Wikipedia is accurate). So, is being Jewish a “risk factor”, or is there an issue with access to services.

3) According to Mr. Heckenlively, the prevalence of Autism in Israel is 1 in 2,400. I guess low prevalence numbers invalidate studies when they are in Denmark, but not when they are in Israel?

By the way, consider that low “incidence” value for Israel. Now look at their vaccine schedule (I am only listing those for very young children)

DTaPHibIPV: 2,4,6,12 months
HepA: 18, 24 months
HepB: Birth, 1, 6 months
MMR: 1 year, 6 years

Hmmm. Remember how Israel has a low autism “incidence”?

Look at that first big combo vaccine–5 vaccines at once. I guess combo vaccines don’t cause autism, eh?

HepB is given at birth. I guess that isn’t a risk factor either, eh?

Separating out the combo vaccines, I count 28 different vaccines given by 24 months in that vaccine schedule. So much for “too much, too soon”, eh?

Of course, this is stretching the Israel data waaaay too far. Of course we can’t say that the Israel data prove that HepB, combo vaccines and too-many-too-soon are not risk factors. Just like we can’t use these data to support Dr. Wakefield’s hypotheses.

I’m glad to see autism studies come out of new countries. Let’s not use data with big limitations to support our preconceived ideas, shall we?

(note, I made corrections shortly after publishing. These did not change the content substantially. I did change the title to clarify that it is not the Israeli researchers who are misinterpreting the data)

9 Responses to “Another misinterpreted study. This time they are misusing Israel data”

  1. Catherina February 22, 2009 at 20:11 #

    Here is the breakdown of cases and ages:

    http://www.eurosurveillance.org/ViewArticle.aspx?ArticleId=8045

  2. Kev February 22, 2009 at 20:18 #

    It worries me that Heckenlively is employed as a science teacher.

  3. Sullivan February 22, 2009 at 20:28 #

    Catherina,

    thanks for that info on measles cases in Israel. There is a big spike in the 1-4 year old age bracket.

    Here is a quote from the paper.

    Those children should have been immunised against measles as part of the routine paediatric immunisation schedule in Israel. However, 96% were not vaccinated. Frequently, all the children in a large family were infected; two thirds of the cases belonged to family clusters of more than two patients per family. This was at least partly due to the fact that patients did not comply with the recommendations of timely post-exposure prophylaxis.

  4. Joseph February 22, 2009 at 21:38 #

    That’s a peculiar study. First thing to note is that it’s not possible to estimate the real incidence of autism. They are apparently looking at diagnosis incidence. I’m not referring to the fact that not every autistic person is diagnosed, but to the fact that date of onset is unknown.

    For example, if an 18 year-old teenager is diagnosed in 1999, that counts toward “incidence” in 1999.

    Now, the age of diagnosis is peculiar in Israel. It’s lower than in other parts of the world, and it has increased with time, not decreased.

    If you look at Figure 1, we see there’s a pretty big spike in “incidence” exactly in 1996, which continues to increase up to 1998, and then begins to drop.

    What’s a good explanation for this? The DSM-IV was published July 1, 1994. It probably took a year or two to catch on.

    So what happened in 1996 is probably not that some young kids developed autism. It’s more likely that a bunch of children of all ages got diagnosed with the recently released DSM-IV. As they ran out of older kids to diagnose, the “incidence” began to drop to where it should normally be.

    It would be interesting to see what the prevalence at something like age 5 would’ve been if a snapshot had been taken every year in the range. That would be much easier to understand than “incidence” and we could compare it to data we know.

  5. Cheryl October 5, 2014 at 22:11 #

    Would it be possible to use foot detox pads on children to help pull the toxins out and see if the autism reduces?

    • Chris October 5, 2014 at 23:01 #

      No, because those are bogus.

      • Sullivan (Matt Carey) October 6, 2014 at 01:30 #

        Check her link. Looks like a business effort.

      • Chris October 6, 2014 at 01:41 #

        Obviously one should avoid any business that is so clueless.

      • Chris October 6, 2014 at 01:42 #

        Just checked, and it is total loony tunes woo!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: