Lewis Carroll On Mercury And Autism

About Seth Roberts

I am a professor of psychology at UC Berkeley and author of ... Full Bio

Seth
Roberts

From an article in Rolling Stone about mercury and autism:

The CDC “wants us to declare, well, that these things are pretty safe,” Dr. Marie McCormick, who chaired the [Institute of Medicine’s] Immunization Safety Review Committee, told her fellow researchers when they first met in January 2001. “We are not ever going to come down that [autism] is a true side effect” of thimerosal exposure. According to transcripts of the meeting, the committee’s chief staffer, Kathleen Stratton, predicted that the IOM would conclude that the evidence was “inadequate to accept or reject a causal relation” between thimerosal and autism. That, she added, was the result “Walt wants” — a reference to Dr. Walter Orenstein, director of the National Immunization Program for the CDC.

From Chapter 12 of Alice’s Adventures in Wonderland:

`No, no!’ said the Queen. `Sentence first–verdict afterwards.’

`Stuff and nonsense!’ said Alice loudly. `The idea of having the sentence first!’

`Hold your tongue!’ said the Queen, turning purple.

Eerily prophetic, no?

Thanks to Dev Rana.

Comments

Is that the extent of the context to be provided?

Did you notice the lengthy list of errors the publisher noted at the end of the article -- this is surprising in any science story, even one in a music magazine.

I have no particular stake in the issue and am not up on the literature, but surely it deserves better coverage than whatever this is meant to suggest.

T. Ryan Gregory | 02/24/08 | 22:28 PM

You can read more about the response to this article here:

http://www.rollingstone.com/news/story/7483530/kennedy_report_sparks_con...

Seth Roberts | 02/25/08 | 12:13 PM

pointed out that Kennedy is "not a scientist or a doctor" and dismissed his extensive evidence as nothing more than "a few scientific studies."

I'm often going to fall on the side of defending people who can discern fact from fiction, scientist or not, and I agree that there are cases where scientists have put advocacy ahead of data - but it's not all that common.

There are also plenty of instances where laypeople have done comprehensive reviews of science data and found crippling flaws but they were often statistical more than fundamental.

In the 2 years since the article was written and the almost 8 years since the conference that inspired it, this correlation has not gone unchecked. While I'm inclined to be of the Sherlock Holmes 'emiminate the impossible and consider everything else' mindset, there just doesn't seem to be anything here.

You've gone after the 'correlation does not equal causation' issue a lot more than I have so I assume you are not being snookered by it here. What in this research makes you think there is a concerted effort to, basically, allow a lot of kids to be damaged?

Hank Campbell | 02/25/08 | 13:51 PM

Hank, to answer your question: Nothing. I think the "concerted effort" is to evade saying something with personally unpleasant consequences to the person who says it. People don't like being attacked. People try to avoid being attacked. This is why certain people "wanted" certain conclusions. Self-interest.

There is substantial data that autism is caused by mercury poisoning. Not necessarily all autism -- at least some of it. I've described some of that evidence here:

http://www.blog.sethroberts.net/2008/01/27/does-mercury-cause-autism/

http://www.blog.sethroberts.net/2008/02/03/does-mercury-cause-autism-con...

When I posted that evidence I was attacked. A friend commented that I was "a glutton for punishment" -- meaning that I posted on the subject more than once.

Seth Roberts | 02/25/08 | 14:30 PM

Got it. This popped up out of the blue here so I didn't get the context. We had Gary Taubes and then suddenly girls in rabbit holes. That's so you. :)

Hank Campbell | 02/25/08 | 15:26 PM

True, the connection isn't obvious. The connection is this:

Gary Taubes is saying the establishment (e.g., most obesity researchers, writers like Jane Brody) has been terribly wrong, utterly wrong, about obesity.

The Rolling Stone article says the establishment -- such as the Institute of Medicine -- has been terribly wrong about what causes autism. They have too quickly dismissed the possibility that it is caused by mercury poisoning.

In both cases (Taubes and Rolling Stone) they make a good case. They have a lot of solid evidence on their side.

Seth Roberts | 02/25/08 | 16:00 PM

They have a lot of solid evidence on their side.

I suspect your readers would be interested in a summary of this solid evidence rather than quips from a widely-critiqued Rolling Stone story from several years ago.

T. Ryan Gregory | 02/25/08 | 16:27 PM

For those who want to know what the impression is in the scientific community, you can check out the editorial in the journal Nature Neuroscience (10: 531, 2007) that appeared recently.

It is not openly available without a subscription, but here are the intro paragraphs:

Silencing debate over autism

Despite the lack of scientific evidence that childhood vaccines cause autism, extreme tactics used by those convinced that this hypothesis is correct have been increasingly successful in influencing public opinion and legislation.

The idea that autism is caused by vaccination is influencing public policy, even though rigorous studies do not support this hypothesis. Legislators are right to take into account the concerns of parent groups and others directly affected by autism, but policy decisions should be based on hard evidence rather than anxiety. More worryingly, some proponents have adopted tactics reminiscent of certain animal rights groups, which are aimed at shutting down the views of opponents.

The hypothesis is based on the observation that the number of autism cases increased in the 1980s, coinciding with a push for greater childhood vaccinations, which increased above recommended levels children's exposure to mercury in the vaccine preservative thimerosal. However, autism diagnosis continued to rise even after thimerosal was removed from US childhood vaccines in 2001. A review by the Institute of Medicine (http://www.nap.edu/catalog/10997.html) of over 200 studies concluded that that there was no causal link between thimerosal-containing vaccines and autism. Autism is no more common among vaccinated than unvaccinated children, and its incidence has not covaried with the presence of thimerosal in vaccines across different times and locations.

You can also read an alternative perspective at blogs such as Respectful Insolence and NeuroLogica.

T. Ryan Gregory | 02/25/08 | 16:24 PM

The author of the Nature Neuroscience editorial did not do his or her homework. At one point the editorial says:

In the end, these fears [that mercury causes autism] are driven by ideology rather than science.

Not so. The fears are at least partly driven by evidence. The editorial writer appears to be unaware that when mercurcy was removed from vaccines, at least some autism rates started to decline. See

http://www.blog.sethroberts.net/2008/02/03/does-mercury-cause-autism-con...

These declines do not prove that mercurcy causes autism, but they await a more convincing explanation.

Seth Roberts | 02/25/08 | 16:53 PM

You mean a study by the Geiers? Apart from the obvious problem with that regression in the paper you linked to in your blog, aren't the Geiers generally seen as somewhere out on the crazy fringe with people like Wakefield? No one takes their stuff seriously.

Ian Ramjohn | 02/26/08 | 02:31 AM

What's the "obvious problem with that regression" in the paper I link to? I'm unsure what you're talking about.

Seth Roberts | 02/26/08 | 02:38 AM

I'm not sure the problems are "obvious", though anyone with a modicum of statistical training would ask how the got their data and decided on transition points in the regression -- in fact, in one figure they use 2003 as the end of one line and 2002 as the beginning of the next, so were rates increasing (line 1) or decreasing (line 2) from 2002-2003?

More substantial criticism of such studies can be found here:

http://interverbal.blogspot.com/2006/03/review-of-early-downward-trends-in_15.html

http://autismnaturalvariation.blogspot.com/2006/03/cdds-data-101.html

I invite you to rebut the comments therein.

If you are actually interested in dealing with this topic scientifically, then you may also wish to discuss the more recent paper, using the same data base, that found no such relationship:

Schechter, R. and Grether, J.K. 2008. Continuing increases in autism reported to California's developmental services system: mercury in retrograde. Arch. Gen. Psychiatry 65: 19–24.

ABSTRACT

Context Previous analyses of autism client data reported to the California Department of Developmental Services (DDS) have been interpreted as supporting the hypothesis that autism is caused by exposure to the preservative thimerosal, which contains ethylmercury. The exclusion of thimerosal from childhood vaccines in the United States was accelerated from 1999 to 2001. The Immunization Safety Review Committee of the Institute of Medicine has recommended surveillance of trends in autism as exposure to thimerosal during early childhood has decreased.

Objective To determine whether trends in DDS autism client data support the hypothesis that thimerosal exposure is a primary cause of autism.

Design, Setting, and Patients Study of time trends in the prevalence by age and birth cohort of children with autism who were active status clients of the DDS from January 1, 1995, through March 31, 2007.

Main Outcome Measure Prevalence of autism among children with active status in the DDS.

Results The estimated prevalence of autism for children at each year of age from 3 to 12 years increased throughout the study period. The estimated prevalence of DDS clients aged 3 to 5 years with autism increased for each quarter from January 1995 through March 2007. Since 2004, the absolute increase and the rate of increase in DDS clients aged 3 to 5 years with autism were higher than those in DDS clients of the same ages with any eligible condition including autism.

Conclusions The DDS data do not show any recent decrease in autism in California despite the exclusion of more than trace levels of thimerosal from nearly all childhood vaccines. The DDS data do not support the hypothesis that exposure to thimerosal during childhood is a primary cause of autism.

T. Ryan Gregory | 02/26/08 | 07:54 AM

The more I look into this (which I will soon stop doing, as it is clearly a waste of time), the more dismayed I am that this is showing up on Scientific Blogging, and especially since it is taken almost as granted by Dr. Roberts that the data support a causal link. All I can do is direct readers of this post to other, more serious, blogs for a much more detailed assessment of the data.

http://scienceblogs.com/insolence/2006/03/the_geiers_go_dumpsterdiving_y_1.php

http://www.samefacts.com/archives/health_care_/2006/03/thimerosalautism_an_epidemiological_smoking_gun.php

T. Ryan Gregory | 02/26/08 | 08:38 AM

Partly driven by evidence? And what else?

T. Ryan Gregory | 02/25/08 | 17:03 PM

TR Gregory: I don't know.

Seth Roberts | 02/25/08 | 17:08 PM

I don't know much about the topic, and I would be interested in an unbiased overview of the evidence on both sides, which is what I think this site is really for. I do, however, object to not very subtle intimations of conspiracies in science. Back it up with data, and we'll go from there.

T. Ryan Gregory | 02/25/08 | 17:15 PM

No study has found any real evidence of a connection between mercury and autism. In fact, the autism "epidemic" has come as human exposure to mercury has declined. As for the chelation stuff - even if it worked (for which there's no evidence), is it really worth subjecting a child to sixty injections a month?

Kennedy's stuff has been widely debunked. It was never taken seriously in the scientific community, and I was amazed to see anyone discussing it still. Geier is out on the fringe - and getting paid quite nicely for being out there. And the activists have moved from blaming thimerosal to blaming power plants and fish.

As for the whole vaccination thing - from what I have seen, trained observers can pick out autistic children in first birthday video tapes, long before they were 'poisoned' by vaccinations.

Ian Ramjohn | 02/26/08 | 02:39 AM

Right -- I certainly had that impression from the Nature Neuroscience editorial and other blog posts, but my suggestion is that if Dr. Roberts wants to discuss the topic in any kind of reasonable way, he should present the evidence and not just vague hints that the scientific community is conspiring to cover up risks. I am willing to look at his arguments as objectively as possible, but I have no respect for not very subtle accusations without data.

T. Ryan Gregory | 02/26/08 | 07:22 AM

"The autism epidemic has come as human exposure to mercury has declined." This is misleading. The autism epidemic began at the same time that the exposure of young children to mercury greatly increased because the number of childhood vaccinations the average child received greatly increased. Each vaccination caused mercury exposure. How is that not "real" data?

And why is the data I discuss here

http://www.blog.sethroberts.net/2008/01/27/does-mercury-cause-autism/

not "real"?

Seth Roberts | 02/26/08 | 08:51 AM

Ah yes, cum hoc ergo propter hoc.

T. Ryan Gregory | 02/26/08 | 09:16 AM

TR Gregory, thanks for the links to critiques of the research I discuss. I had time to look at only one of them: the one at Interverbal. The writer makes 3 points. The first two are uninteresting. Point 1: all data have error. Point 2: in fact, it is reasonable to choose starting and stopping points of regression lines based on looking at data. Point 3 (shifting definition): This is interesting. But the crucial reference (Rollens, 2004) is not in the reference list.

Seth Roberts | 02/26/08 | 09:25 AM

Seth, that is an abysmal response. Point 1 is that the data are being misapplied. Point 2 is that to test a hypothesis, one needs to set the predictions first, which means determine the time when the expected cause starts and see if the effect begins then. Point 3 suggests that the criteria for inclusion in the dataset changed, and moreover that it did so in a fashion that would bias the dataset and perhaps account for the pattern. I think I am done discussing this with you, because I cannot have a serious conversation with someone so dismissive of basic scientific principles.

T. Ryan Gregory | 02/26/08 | 09:30 AM

the crucial reference (Rollens, 2004) is not in the reference list

A 60 second Google search turned up this:

http://autismdiva.blogspot.com/2005/07/autism-hysteria-history-october-2004.html

T. Ryan Gregory | 02/26/08 | 09:50 AM

Thanks for the additional info about Rollens (2004). That's very helpful.

Seth Roberts | 02/26/08 | 10:02 AM

This is way outside my field of expertise, but a quick Google search brought up two reviews articles,
http://www.ncbi.nlm.nih.gov/pubmed/17168158
and http://www.blackwell-synergy.com/doi/abs/10.1111/j.1365-2214.2006.00655.x , both of which state very clearly that they find absolutely no evidence for any causal link between vaccinations and autism. The second one moreover points out that

The recorded prevalence of autism has increased considerably in recent years. This reflects greater recognition, with changes in diagnostic practice associated with more trained diagnosticians; broadening of diagnostic criteria to include a spectrum of disorder; a greater willingness by parents and educationalists to accept the label (in part because of entitlement to services); and better recording systems, among other factors.

which I find an interesting consideration that has not been mentioned here so far.

Georg von Hippel | 02/26/08 | 10:08 AM

Thanks Georg. It is outside of my expertise also, but it doesn't take long at all to gauge the view of the scientific community based on the empirical evidence. And we have seen how proponents of the cover up hypothesis respond to that information.

T. Ryan Gregory | 02/26/08 | 10:18 AM

Yes, the rejection of the thimerosal-autism (or MMR-autism) link appears pretty much unanimous throughout the serious scientific literature.

And even if one was willing to assign a non-zero Bayesian prior to the cover-up hypothesis, its posterior probability still drops to zero pretty quickly when the vanishingly small likelihood of secret collusion between scientists from different continents, nations and disciplines is taken into account.

But I suspect that the real roots of those claims may well be psychological: many parents who are confronted with the problem of raising an autistic child may feel better if they have someone or something (other than genetics, for which the blame could be seen as falling back onto themselves) to blame for their predicament.

Georg von Hippel | 02/26/08 | 11:00 AM

That's an excellent point. Forgiving parents for not thinking scientifically is acceptable. We want to make sure scientists are thinking scientifically, though, and I like that about our community.

You go to other places and they somehow magically all agree with each other about everything. It's because they're handpicked for having common politics or ideology or beliefs. So they spend time linking to each others' articles and writing things like "X wrote an article on how Y (who naturally, does not write there) stinks and here's my take on why Y stinks.'

If we all agreed on everything, there wouldn't be a need for more than one of us. So I like that we have some debate.

That said, everyone's sort of dancing around the obvious question but I am a lousy dancer. Seth, do you think mercury causes autism? If so, what led you to that? (don't send me to your website - the whole point of spending 2000 hours creating this one was to not have to go to 50 websites to read great stuff)

The first time I heard about that correlation, I thought 'it's a common demoninator and has to be considered' but I think most people felt that way until later studies said that was not the case. Are they all flawed?

I don't think anyone here is a "circle the wagons, if one scientist is attacked we all lose credibility" mentality about this autism research but if we're going to be on record (and you're a featured writer here, so it is on record) endorsing Rolling Stone over the CDC on the matter, there should be a better reason than a hint that Walter Orenstein likes immunizing kids even if it means giving them autism.

Hank Campbell | 02/26/08 | 12:01 PM

Do I believe mercury causes autism? No. I am uncertain. I don't know what causes autism. Maybe mercury does, maybe not.

As for Rolling Stone vs. CDC, I'm sure they are both worth our attention and that neither is infallible. Until someone convinces me that this quote from McCormick

“[the CDC] wants us to declare, well, that these things are pretty safe,” Dr. Marie McCormick, who chaired the [Institute of Medicine’s] Immunization Safety Review Committee, told her fellow researchers when they first met in January 2001.

is wrong, then I will believe that someone at the CDC handled this situation very poorly.

Seth Roberts | 02/26/08 | 12:23 PM

It is another fallacy to confuse improper behaviour of an individual or political institution with determining what is or is not an objective scientific fact.

T. Ryan Gregory | 02/26/08 | 12:29 PM

I don't know what causes autism. Maybe mercury does, maybe not.

I do not know for certain either, but I know how we would find out: rigorous statistical analysis of large datasets from multiple sources published in peer-reviewed journals of high reputation. At present, this approach strongly challenges any link. You seem to have another approach in mind.

T. Ryan Gregory | 02/26/08 | 12:31 PM

"I know how we would find out: rigorous statistical analysis of large data sets from multiple sources published in peer-reviewed journals of high reputation." Poor Darwin: No peer review, no rigorous statistical analysis. He must have been mistaken. Poor Mendel: No peer review, no large data set. He must have been mistaken too. Poor Wegener: The entire scientific community disagreed with him about continental drift. He must have been wrong. Poor Ernst Wynder: No large data set. He must have been wrong that smoking causes lung cancer.

Seth Roberts | 02/26/08 | 18:56 PM

Wow, it's amazing that science has progressed at all, with so many scientists being evidence snobs! Are you suggesting maybe self-experimentation is a better test of the link between smoking and lung cancer?

Nobody is suggesting you need to put a p-value on everything - plate tectonics clearly is a subject that didn't need it. And you think Darwin didn't put his stuff up for the peer review typical of his time? He did what all good scientists do - subjected his claims to the scrutiny of his peers all over the world.

Michael White | 02/26/08 | 19:32 PM

You are aware how that makes you sound?

Georg von Hippel | 02/27/08 | 04:16 AM

Oh Seth. I am amazed you went there. Why not list Galileo and make it the official gambit?

Do you even know the history of evolutionary biology and genetics?

Do you know that both Darwin's theory of natural selection and Mendelian genetics were controversial in the scientific community, and that rigorous subsequent testing settled the issue? Same with continental drift. Data settled the debate.

Are you really arguing that large, rigorous, statistical datasets did not settle the question of smoking, but rather one person's opinion?

And by the way, are you stating that you do NOT think rigorous testing and peer review are necessary in science?

T. Ryan Gregory | 02/26/08 | 19:05 PM

There are many ways to "find out" the truth. That was my point. Wynder might be offended that you refer to his research as "opinion".

Seth Roberts | 02/26/08 | 19:15 PM

If he's a scientist of any merit, he would be more offended that you think his research neither demonstrates nor requires rigor.

T. Ryan Gregory | 02/26/08 | 19:18 PM

There are many ways to "find out" the truth.

Perhaps, but I'm concerned with the one called science.

T. Ryan Gregory | 02/26/08 | 19:21 PM

There are many ways to "find out" the truth. That was my point.

And my reply is that each of the examples you provided contradicts your point.

T. Ryan Gregory | 02/26/08 | 19:23 PM

What other way is there in the case of autism and vaccines? Just going by your gut feeling based on people you know with autism?

If everyone who got the vaccine got autism, and those who didn't get the vaccine didn't get autism, it would be simple. But biology is obviously not that simple. How do you suggest tackling this problem without statistics and lots of data?

Michael White | 02/26/08 | 19:47 PM

I'm not writing about how to do future research on autism. I'm writing about the research that has already been done.

The data pointing to a connection between mercury exposure and autism is not as worthless as is often claimed.

Seth Roberts | 02/27/08 | 07:21 AM

The data pointing to a connection between mercury exposure and autism is not as worthless as is often claimed.

Show us why, then. And then show us why the research reaching the opposite conclusion is not convincing. You may also want to just be clear in what your point of this post is: is it your claim that there is a cover up on this issue, for example.

T. Ryan Gregory | 02/27/08 | 07:27 AM

Aside from the debate the other 7000 articles here can add some insight. Autism: Study Says Thimerosal Mercury Doesn't Have Time To Accrete In Baby Bodies and even

Node search:autism

Hank Campbell | 02/28/08 | 19:28 PM

There are actually two related articles listed above that provide new data suggesting no mercury link, and the irony has not been lost on all of us.

T. Ryan Gregory | 02/28/08 | 21:14 PM

The Rolling Stone article explains at least part of the reason the data suggesting a link between mercury and autism is not as worthless as claimed: Because there was pressure to deny such a link. This caused biassed evaluations of the evidence.

Seth Roberts | 02/28/08 | 19:41 PM

part of the reason the data suggesting a link between mercury and autism is not as worthless as claimed: Because there was pressure to deny such a link.

Premise 1 - The data have been rejected.

Premise 2 - Some people rejected the data because of invalid reasons.

Conclusion - Therefore, the data are valid.

This logical fallacy has a name, and is in addition to the previously used cum hoc ergo propter hoc fallacy. I'll open it to readers to name this one, plus any others in earlier posts that I may have overlooked also.

T. Ryan Gregory | 02/28/08 | 20:39 PM

It's always interesting to see certain common features of back and forths like these:

1. Derogatory descriptions. Example: "quips".

2. Exaggerated negative comments. "No one takes their work seriously"

3. Statements asserting superiority. "I'm interested in the one called science." "I'll open it to readers..." "You are aware how that makes you sound?" "I am amazed you went there." "Do you even know...?" "I cannot have a serious conversation with someone so ..."

4. Magic words. "Rigorous." "Science." "Peer-reviewed." "Scientific."

Seth Roberts | 03/02/08 | 01:36 AM

For readers keeping track, that's Cum hoc ergo propter hoc, Non sequitur (perhaps argumentum ad logicam), and now the style over substance fallacy.

And I'm sorry if you find the words "rigorous", "science", and "peer-reviewed" offensive.

T. Ryan Gregory | 03/02/08 | 09:04 AM

Emmy winning actor Jeremy Piven of Broadway play, "Speed the Plow," by David Mamet had to leave the production due to a high mercury count in his system. The director has called the ailment, which can actually turn deadly, bogus, despite orders from a doctor to stop working and seek treatment immediately. Jeremy experienced mercury poisoning after eating too much sushi. He is an avid sushi eater, regularly eating sushi twice a day. Mercury, though almost always present in trace levels in the human body, is a very toxic element and it has deadly side effects from exposure to even small amounts.

Derrick j (not verified) | 01/05/09 | 06:06 AM

I coudn't do worse than refer you to this

http://leftbrainrightbrain.co.uk/?p=2077

"In a single document, we now have an expert on both mercury toxicology
and autism. Not faux experts, or worse, businesspeople and public
relations people, but an actual, bone fide expert in both fields. I.e.
we have a good document to give to people who are being snowed under by
the misinformation campaign promoting autism as mercury poisoning."

And ....

Dr. Rodier notes that the comparison that autism and mercury poisoning appear similar isn’t even close.

In other words, because the symptoms of methylmercury poisoning
are not similar to those of autism, the authors have tried to construct
a new, hypothetical kind of mercury poisoning from symptoms of toxicity
of other mercury species and symptoms never reported for any kind of
mercury exposure. The hypothesis is not based of facts; instead, the
facts are being selected, manipulated, and shaped to fit the
hypothesis. The hypothesis is then offered as evidence. But hypotheses
are not evidence."

Laurence Arnold | 04/16/09 | 18:56 PM

More Seth Roberts articles

About Seth Roberts

Comments

Add a comment

Know Science And Want To Write?

What People Are Saying

Books By Writers Here

Not a Scientific Blogging member yet?

Register Now for a Free Scientificblogging.com Account

Already a Scientific Blogging member?

Sign In Now

More Seth Roberts articles

About Seth Roberts

Related Articles Here On Scientific Blogging

Comments

Add a comment

Know Science And Want To Write?

What People Are Saying

Books By Writers Here