No, the autism prevalence did not go down in Denmark after the removal of thimerosal

25 Feb

Once there was an open question of whether the thimerosal containing vaccines, previously used for infants in the U.S. could be contributing to the increase in autism diagnoses being reported. Even with multiple studies showing no increased risk due to thimerosal exposure, there is still a group who not only believes in the mercury-induced epidemic, but they believe that there has been a concerted effort on behalf of the U.S. government and other groups to hide this “fact”.

An email obtained through a freedom of information act (FOIA) request is often cited in online discussions of not only the “fact” that the mercury-induced-epidemic is real, but that the government conspiracy is real. The email reportedly was made by one of the authors of a study from Denmark. The authors had used a sort of natural experiment: Denmark had phased out thimerosal from vaccines in the early 1990′s. They reported that the prevalence of autism continued to increase even with the phase out, thus indicating that thimerosal was not the driving force behind the increase.

The study, by Madsen, et al., was published in 2003. An email obtained by FOIA reportedly shows the authors removing data for the final years of the study period:

I need to tell you that the figures in the manuscript do not include the latest data from 2001….But the incidence and prevalence are still decreasing in 2001.

Sounds alarming, doesn’t it? Data which supposedly shows that autism rates actually dropped following the removal of thimerosal was removed from a paper. Well, I think they should have included the data, a possible explanation and done a follow-up study. That said, the statement really doesn’t bother me. Why? A few reasons but mainly because the data are clear that autism prevalence did continue to go up in Denmark. And much of that data are below.

First, what was the prevalence in the 1990s? Here’s a study from 2007. They were using data where from 2004 (followup through 2004):

Time trends in reported diagnoses of childhood neuropsychiatric disorders: a Danish cohort study. 2007
Birth cohort 1994-1995. Prevalence: 0.58%
Birth cohort 1995-1996. Prevalence: 0.47%
Birth cohort 1998-1999. Prevalence: 0.32%

Childhood autism:
Birth cohort 1994-1995. Prevalence: 0.18%
Birth cohort 1995-1996. Prevalence: 0.17%
Birth cohort 1998-1999. Prevalence: 0.18%

Yes, they are showing ASD prevalence dropping in the 1990′s, which should be an admission that the prevalence went down after the removal of thimerosal! I guess they forgot to censor that paper. While ASD prevalence seems to be going down, Childhood Autism is relatively flat. Why would that happen, do you think? Here’s one reason: average age of diagnosis for autistic disorder (childhood autism) is lower than that for the other ASD’s. While autism is not as obvious as many would have us believe, childhood autism *is* more obvious than the other ASD’s (and even with that it gets missed). Consider 1998-1999. That’s only 5 years or so until the followup date from the study: 2004. Average age of diagnosis for ASD in Denmark was about 5 years. So, about 1/2 of the ASD kids born 1998-1999 were not diagnosed at the time of the study. More of the ASD kids born in 1995-1996 and even more of those from 1994-1995. So, what looks like a declining prevalence is most likely just an artifact of how many years of follow-up there were from birth to study date.

I’d say it is even more than “looks like”. Consider the studies below. These are the papers I could easily find that give autism prevalence values for Denmark. I give the title, with a link to pubmed, the year of the study, the birth cohorts reported and the prevalence.

Autism prevalence in the 1990′s?

a href=”″>Association between thimerosal-containing vaccine and autism. 2003
Birth cohort 1990-1996. Prevalence 0.26%

Autism prevalence for cohorts including kids born after 2000? More like 1%. Consider this first paper:

Autism after infection, febrile episodes, and antibiotic use during pregnancy: an exploratory study. 2012
Birth cohorts 1997-2003. Prevalence: 1%

But there are more. The cohorts don’t always match between studies, and we don’t have individual years for kids born 2000 and later. But the prevalence is repeatedly reported as above the 0.26% found for the 1990′s.

In other words, the prevalence continued to go up.

Using maternally reported data to investigate the association between early childhood infection and autism spectrum disorder: the importance of data source. 2012
Birth cohorts 1997-2003. Prevalence: 1%

Parental age and autism spectrum disorders. 2012
Birth cohorts 1980-2003. Prevalence: 0.73%

The rising prevalence of autism: a prospective longitudinal study in the Faroe Islands. 2012
Birth cohort 2002. Prevalence: 0.56%
Birth cohort 2009. Prevalence:0.94%

A comparison of autism prevalence trends in Denmark and Western Australia. 2011
Birth cohort not in abstract. Prevalence (ASD) 0.685%

Risk of autism spectrum disorders in children born after assisted conception: a population-based follow-up study. 2011
Birth cohort 1995-2003. Prevalence: 0.61%

Association of hospitalization for infection in childhood with diagnosis of autism spectrum disorders: a Danish cohort study. 2010
Birth cohort 1980-2002. Prevalence 0.52%

Association of family history of autoimmune diseases and autism spectrum disorders. 2009
Birth Cohort 1993-2004. Prevalence: 0.48%

The incidence and prevalence of pervasive developmental disorders: a Danish population-based studyBirth cohort 1971-2000. Prevalence 34.4 (“corrected”)

Association between thimerosal-containing vaccine and autism. 2003
Birth cohort 1990-1996. Prevalence 0.26%

A population-based study of measles, mumps, and rubella vaccination and autism 2002
Birth cohort 1991-1998. Prevalence: 0.14%

The more early birth cohorts a given study uses, the lower the average prevalence. The more years included post 2000, the closer the prevalence is to 1%.

Sure, it would be great if we had data for prevalence by individual birth year going to 2000 and beyond. But there’s enough data above to make it clear that the autism prevalence in Denmark did not go down after 2000. Quite the contrary. Similarly, measures of autism prevalence in my home state, California, have continued to go up even after thimerosal was removed.

The idea that thimerosal caused an autism-epidemic is just not consistent with the facts. The same data people used over and over to make the argument that thimerosal causes autism–the various autism prevalence estimates–don’t support the idea any more. Thimerosal went away, autism prevalence continued to climb.

So I’m just not swayed when I read people write about how the autism prevalence went down in Denmark with the removal of thimerosal. It didn’t go down. It is sad that people are being misled to think there is still some substance to the mercury-epidemic idea. Cherry pick one email out of how many FOIA’d documents, present it out of context and play it up as something real–while ignoring the evidence that is right in front of you. This is not treating the autism community with respect. This is misleading people.

By Matt Carey

About these ads

11 Responses to “No, the autism prevalence did not go down in Denmark after the removal of thimerosal”

  1. Daniel Comin February 25, 2013 at 09:51 #

    Reblogged this on Reflexiones y pensamientos sociales.

  2. Marc February 25, 2013 at 15:55 #

    Why not just make it up as you go? Seems to be the norm on this blog. DSM V soon to be released and then what?!

    • Sullivan (Matt Carey) February 25, 2013 at 16:08 #

      You can follow the links to the studied cited above. The autism prevalence in Denmark is higher after the removal of thimerosal. People are giving out selective facts to create the image this isn’t the case.

      Rates of autism diagnoses in California are also still going up, long after the removal of thimerosal.

      The thimerosal epidemic idea wasn’t strong to begin with. If people want to deny or hide the data that shows it isn’t true, shame on them.

      • brian February 26, 2013 at 00:36 #

        Thanks for pointing out much of the information that is ignored by conspiracy theorists who focus on a redacted e-mail message related to the 2003 Madsen study.

        It would have been better if Masden et al. had discussed the problems of incomplete case ascertainment in recent data, since that phenomenon should have been familiar to vaccine-phobic parents knew of Mark Blaxill’s 2001 presentation to the Institute of Medicine. Perhaps Age of Autism readers are pleased that, as suggested by Blaxill’s chart titled “Vaccine Mercury Burden and Autism Risk: United States, the prevalence of ASD fell as exposure to thimerosal decreased, in 1995, but, unfortunately, that seems to have been due to incomplete case ascertainment: the prevalence of ASD appearedto decrease simply because many of the cases in the most recent birth cohort were yet to be diagnosed.

        Vaccine-phobic parents also seem to ignore work related to ASD prevalence and thimerosal exposure Canada. The authors of the most recent study (linked below) concluded: “As in other studies where the effect of the discontinuation of thimerosal in childhood vaccines was examined, no change in the underlying population trends for PDD rates could be observed in relation to thimerosal discontinuation.” In fact, the prevalence of PDD in each individual birth cohort born after the use of thimerosal in pediatric vaccines was discontinued in 1996 was consistently higher than that in cohorts born prior to 1996, even in children born as much as six years after the ban. (Notably, the prevalence of PDD was slightly lower in the most recent cohort (surveilled in kindergarten) than in the previous year–most likely due to incomplete case ascertainment.)

  3. Grant February 25, 2013 at 21:20 #

    Sully, up until 1995 they only counted autistic inpatients in the study. Starting in 1995 they included autistic outpatients also. In another study done by this group on the MMR-autism link, they state that autistic outpatients outnumber inpatients by a factor of 13:1 in Denmark. If you just compare the original cohort, autistic inpatients, rates of autism dropped by 10 fold in Denmark. Not to mention that the current rate of autism in Denmark, Finland and Iceland is only 1.2:1000 compared to 1:88 here. It’s like monitoring Malaria rates in Canada and using the data in Africa.

    • Sullivan (Matt Carey) February 25, 2013 at 21:43 #


      Don’t fall for the “comparing autism prevalence around the world” trick. JB Handley tried it with a pseudo paper, and Chris Shaw’s group did it again recently. They tend to compare, for example, Iceland a long time ago to the US today.

      Consider the data in “Do aluminum vaccine adjuvants contribute to the rising prevalence of autism?”

      They quote Iceland prevalence of 12.4/10,000. Follow the citation and you get this abstract:

      This clinic-based study estimated the prevalence of autism in Iceland in two consecutive birth cohorts, subjects born in 1974-1983 and in 1984-1993. In the older cohort classification was based on the ICD-9 in 72% of cases while in the younger cohort 89% of cases were classified according to the ICD-10. Estimated prevalence rates for Infantile autism/Childhood autism were 3.8 per 10,000 in the older cohort and 8.6 per 10,000 in the younger cohort. The characteristics of the autistic groups are presented in terms of level of intelligence, male:female ratio, and age at diagnosis. For the younger cohort scores on the Autism Diagnostic Interview-Revised and the Childhood Autism Rating Scale are reported as well. Results are compared with a previous Icelandic study and recent population-based studies in other countries based on the ICD-10 classification system. Methodological issues are discussed as well as implications for future research and service delivery.

      Besides the fact that they are using cohorts that are 20-40 years old, they somehow added the prevalences from the two cohorts (8.6+3.8=12.4). Which is, frankly, nonsense.

      You can find an expanded version of the Finland report online. Here’s a bit of the abstract:

      The data were 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. The age-specific
      prevalences obtained in this study showed the prevalence to be lowest, i.e. 6.1 per 10 000, in the oldest age group of 15- to 18-year-old adolescents and highest, i.e. 20.7 per 10 000, in the age group of 5-to 7-year-old children, when the criteria of ICD-10 and DSM-IV were used.

      So, we are looking at prevalences for birth cohorts of 5-7 and 15-18 year olds in 1997–so birth years 1978-1982 and 1989-1992 from that same period. Interestingly enough, the overall prevalence is 13.9 (at least in the report I found), not the 12 something Shaw reports (The total prevalence (± 5% CI) of AD in Northern Finland was 13.9 (12.0–15.7)/10 000.). Minor detail, maybe the published report says something different. The quote is from the followup study. But, wait, the author of the Finnish study is saying the prevalence of AD–that’s autistic disorder, not ASD’s. So, another false comparison on the part of Shaw.

      Comparing a prevalence for cohorts from the 1970-early 1990s to the prevalence in Denmark in 2000 (which multiple reports show as being 1% or more) is dishonest. Comparing apples to airplanes, as one researcher puts it.

      Stay tuned for the California data tomorrow. Administrative autism prevalence continues to climb, a decade after the removal of thimerosal.

    • Sullivan (Matt Carey) February 25, 2013 at 22:20 #

      “Starting in 1995 they included autistic outpatients also…”

      Why isn’t the prevalence 13 times higher in 1996 than 1995? (take a look, the prevalence by one report went down 1994–>1995–>1998)

      OK, compare:

      Birth cohort 1995-1996. Prevalence: 0.47%

      with the study:

      Birth cohorts 1997-2003. Prevalence: 1%

      Both are in the “outpatient” time period. Prevalence doubled.


  1. No, the autism prevalence did not go down in Denmark after the removal of … - February 25, 2013

    [...] Read more at Left Brain Right Brain. [...]

  2. No, the autism “rate” in California did not go down after removing thimerosal from vaccines | Left Brain Right Brain - February 26, 2013

    [...] all said, the talk made me dive back into looking at autism prevalence. I decided to finally write about the fact that the autism prevalence in Denmark is higher post thimerosal than while [...]

  3. Commulink - March 22, 2013

    [...] No, the autism prevalence did not go down in Denmark after the removal of thimerosal ( [...]

  4. Autism in America | Enlightened Lotus Wellness - April 1, 2013

    [...] No, the autism prevalence did not go down in Denmark after the removal of thimerosal ( [...]

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s


Get every new post delivered to your Inbox.

Join 952 other followers

%d bloggers like this: