Thimerosal: correlation, causation, and conspiracy

This is the first of two posts about thimerosal-containing vaccines and autism. The first post is about the scientific evidence for a connection between thimerosal-containing vaccines and autism. The second post will discuss whether pharmaceutical companies and federal regulators have responded appropriately to thimerosal concerns.

Until recently, I was agnostic about the safety of thimerosal. Then Orac stepped into the fray, repeatedly. I was thus inspired to investigate further.

As far as I can tell, the thimerosal/autism connection is totally unsupported by evidence. Autism diagnoses have been rising steadily over the past few decades. However, we shouldn't infer that autism incidence must therefore be increasing. Autism awareness has also risen steadily over the years, thanks to the hard work of activists, researchers, and clinicians. The diagnostic category has also broadened significantly since the 1940's to include an entire spectrum of developmental disorders.

Thimerosal made its debut as a vaccine preservative in the 1930s. The term "autism" had been used since 1911 to describe the symptom of extreme pathological introversion. However, autism wasn't recognized as a distinctive developmental disorder until 1943.

According to the World Health Organization's statement on thimerosal, concern over the safety of thimerosal was initially sparked by a confusion. After 70 seemingly uneventful years of widespread use, thimerosal came under scrutiny from the FDA in 1999. During a review of vaccination recommendations, some experts became concerned that the cumulative mercury exposure in the standard vaccine schedule exceeded U.S. exposure limits for mercury. As it turned out, the exposure limits were for set methyl mercury, whereas thimerosal is a derivative of ethyl mercury. The FDA had been treating ethyl and methyl mercury as equivalent, but we now know that two compounds have significantly different toxicological properties:

Expert consultation and data presented to the Global Advisory Committee on Vaccine Safety (GACVS) on 20-21 June 2002 indicate that the pharmacokinetic profile of ethyl mercury is substantially different from that of methyl mercury. The half-life of ethyl mercury is short (less than one week) compared to methyl mercury (1.5 months) making exposure to ethyl mercury in blood comparatively brief. Further, ethyl mercury is actively excreted via the gut unlike methyl mercury that accumulates in the body. Two independently-conducted epidemiological studies have been completed in the United Kingdom. These studies further support the safety of thiomersal-containing vaccines in infants in the amounts used in existing vaccines.

The FDA Center for Biologicals Evaluation and Research (CBER) offers a very comprehensive resource on thimerosal and vaccines.

In 2003, Geier and Geier claimed to have found a statistically significant association between thimerosal-containing vaccines and autism based on an analysis of data from the CDC's Vaccine Adverse Event Reporting System (VAERS). The American Academy of Pediatrics severely criticized the study for shoddy methodology and factual errors. According to the AAP:

The most important weakness of the article is the reliance on VAERS data to draw conclusions about adverse event associations or causality. VAERS is a passive surveillance system for reporting possible vaccine adverse events that depends on health care professionals, patients, and others to file reports. Health effects reported to VAERS as being associated with vaccines may represent true adverse events, coincidental occurrences, or mistakes in filing. Inherent limits of VAERS include incomplete reporting, lack of verification of diagnoses, and lack of data on people who were immunized and did not report problems.

Perhaps the best evidence against the thimerosal/autism hypothesis is the fact that banning thimerosal doesn't reduce autism rates. If thimerosal caused autism, you would expect autism rates to fall after the offending vaccines were stricken from the immunization schedule. Several nations including Denmark and Canada have already banned thimerosal-containing vaccines, but no declines in autism have been observed so far. In the United States, the FDA has worked with vaccine manufacturers to decrease thimerosal exposure in the standard infant immunization schedule by 95%. So far, decreased exposure hasn't translated into decreased autism incidence.

In my next post, I'll talk about the allegations of collusion, influence peddling, and data suppression. So far, I see no evidence of a conspiracy or a coverup. However, the evidence points to the usual combination public relations and influence peddling.

Comments

[Quisp]

I read the Kennedy article and was terrified. I read the Orac rebuttal and felt silly for being terrified. There is pretty much nothing scarier than the thought of actually causing the ruination of your own beautiful child. It's now at the top of my scary list, supplanting rabies.

[Lindsay Beyerstein]

The thimerosal issue bothers me because its putting parents off vaccination and/or making them feel guilty about doing what's objectively in their kids' best interests. That's just so irresponsible.

[Quisp]

It bothers me because I am exactly the kind of person who is vulnerable to those kinds of narratives (bad big business, greed, corrupt gov't, etc.). It helps to be a hypchondriac and to have a child. "Helps" in the sense of "palsies you with terror."

[Lindsay Beyerstein]

Happy Fathers Day, Quisp! Did you get breakfast in bed, or a ukulele serenade?

[rob helpychalk]

I'm with Quisp almost exactly, right down to the hypochondria. (I decided I had one of those inoperable tumors between your face and brain the other day, as the result of a stuffy nose.)

Although it is otherwise a mediocre movie, there is a great line in The Birdcage, where Robin Williams says "Parenthood makes cowards of us all."

Are there other hypothesizes causes of the increase in autism rates besides increased rates of diagnosis and the apparantly false thimerosal connection?

[Lindsay Beyerstein]

I'm a huge hypochondriac all by myself. I can't even imagine what it would be like to have kids to worry about.

I'm inclined to think that increasing autism diagnosis rates are mostly attributable to increased awareness and broader definition of the diagnostic category. Speaking of hypochondria, I wondered if I might have Asperger's syndrome after listening to a guy on NPR talking about people on the fringes of the autism spectrum!

Happy Fathers day to you, too, Rob. Thanks for the link.

[Quisp]

re asberger's: Do you do that rocking back and forth thing? (p.s. that's funny, but you don't have it.)

no breakfast in bed, though it is about to be GHEE-tah time.

"The Birdcage" is one of my favorite movies. Elaine May is a genius, I think. Hank Azaria is unbelievably great. "May I take your coat, as usual...or...for the first time?" Back before Calista Flockhart was scary. "It's like riding a psychotic horse toward a burning stable." etc. And it really nails the whole "moral majority" hypocricy bit, too. of course that's just gravy.

[thejtrain]

I haven't reviewed the data on this in a long time, since I don't do pediatrics, but when I read up on this a couple of years ago, my conclusion was the same as Lindsay's--there's just no juice in that orange.

Diagnoses of a condition go up when we get better at dealing with it. Since the SSRIs have come out, we diagnose a LOT more depression, because we actually have something we can do about it. Similarly, with autism, there has been a big push to recognize and deal with it over the years, so we get a lot more sensitive about picking it up. To twist the old analogy, if you're pounding nails in with a rock, you'll just worry about the worst ones sticking up, but if you go out and buy a nice hammer, there's no excuse not to get them all nice and flush.

A few years ago, the activists in this regard were pooh-poohing the epidemiological studies, saying that they wanted basic science research done on the subject before we could say for sure. That's not how it works--you can get a lot of things to happen in a lab that don't happen in the population, and we have good data to say this doesn't happen in the population. That's what we call "evidence based medicine" these days--we treat patients based on the way things do work, not on how they should work.

[Lindsay Beyerstein]

Similarly, with autism, there has been a big push to recognize and deal with it over the years, so we get a lot more sensitive about picking it up.

JT, do you get the impression that improved diagnosis for autism is improving outcomes? What kinds of interventions do kids get when they're recognized as belonging to the autism spectrum?

In my hypochondria, figured I must be somewhere in the infrared or ultraviolet of the autism spectrum--not visible to the naked eye, but somewhere along that continuum. I only started worrying after the guy on NPR started talking about how Asberger's was characterized monomaniacal fixations. I started thinking about the seemingly random parade of obsessions that defined my childhood: bears, forensic pathology, psittacines (especially cockatoos), Marxism...

[Andy]

Lindsay, will you marry me?

You combine being progressive and scientific in the best way. Too often, the folks on our side in the culture wars will swallow unscientific claims because of their loathing of all things corporate. You listen to the scientific evidence while being progressive. Sooo sexy.

[Cookie]

There's no doubt in my mind that not only me, but my entire family, are Asperger's. Although my obsessions aren't nearly so entertaining as yours, Lindsey. :=D

I really believe it's a normal part of being a parent of a child with a condition like autism to search desperately for causes. My family's done it - we have a niece with an odd ailment classified as "rare and expensive" - her kidneys are unable to process 4 aminos. And even though we know it's a autosomal homozygous (???) disorder, meaning it's for sure inherited, we spend hours and hours blaming GM foods, then wait! those El Charrito dinners she ate that one time ... no, wait! it was pop, we know it was the pop!

Etc. etc. etc.

This has been going on for years. I really do think it's a normal part of the wondering about a disease or disorder, even once acceptance has set in. But, we don't know anyone else who has this disorder or who has kids with it. So there's never been an opportunity to consolidate our wondering and make it official, so to speak. Instead, we're all stuck with what science has to say about it.

In any case, good post. I'm glad you made it. This is a topic of a lot of interest to me.

[thejtrain]

JT, do you get the impression that improved diagnosis for autism is improving outcomes? What kinds of interventions do kids get when they're recognized as belonging to the autism spectrum?

I am just about the last person to ask, since I don't see kids at all. It's something I'll have to get more well-versed in, I guess, since a lot more kids will be carrying the diagnosis into adulthood (and to their internist), the way we've had to get more familiar with ADHD (which I was already, since I have it).

I might pose the question to some child-psych friends of mine.

[tr]

Vaccine manufacturers using thimerasol would be a juicy target for lawyers if a conncection to autism could be proven.

[Njorl]

" Autism diagnoses have been rising steadily over the past few decades. However, we shouldn't infer that autism incidence must therefore be increasing."

The state of California sponsored a very extensive study demonstrating that, while broadened diagnostic criteria and raised awareness accounted for some of the increase, it was no where near enough to account for current rate. That doesn't necessarily make thimerosol the culprit, but the culprit is almost something introduced by humans in the relevant time-period.

[Njorl]

Oh, just to cover all bases, CA also sponsored a follow up study to see if the increased diagnosis rate was due to people moving to CA to take advantage of their available resources.

[Njorl]

" Several nations including Denmark and Canada have already banned thimerosal-containing vaccines, but no declines in autism have been observed so far. "

This is not really relevant. Thimerisol exposure in Canada was always extremely low. Canada banned Thimerisol for everything but flu vaccines in 1960. After that, the US added more thimerisol-preserved vaccines to its regimen. Of course, if autism rates in Canada are about the same as in the US, or even trended the same (to compensate for systematic differences in diagnosis, and record-keeping) in this period, that would tend to exonerate thimerisol. I haven't seen any Canadian data, but if rates were much lower in Canada than the USA, I think it would be big news in the community.

Denmark also used little thimerisol compared to the USA. Denmark's autism rates were much lower than the USA's before it banned thimerisol. It is entirely possible that if thimerisol were causing autism in the USA, that it was not causing autism in Denmark. If so, banning thimerisol in Denmark would not result in a decrease in autism rates. Now, the only study I've seen of Danish autism rates shows a dramatic increase. You might think that this clears thimerisol, since something else is causing an increase in rates, but it is entirely possible that there are multiple culprits in the increase of autism.

[Njorl]

"In the United States, the FDA has worked with vaccine manufacturers to decrease thimerosal exposure in the standard infant immunization schedule by 95%. So far, decreased exposure hasn't translated into decreased autism incidence."

When did these changes happen?

You wouldn't see any effects of this until 4-5 years after the changes. Even when autism is successfully diagnosed, it is not usually officially diagnosed until a child is 5 years old. This allows parents greater flexibility in placing their child in special pre-school education. When a child hits 5, the benefits that states are required to give autistic children generally outweigh any advantages flexibility of placement might offer, so the diagnosis is made official.

[Njorl]

"Diagnoses of a condition go up when we get better at dealing with it. Since the SSRIs have come out, we diagnose a LOT more depression, because we actually have something we can do about it. Similarly, with autism, there has been a big push to recognize and deal with it over the years, so we get a lot more sensitive about picking it up. To twist the old analogy, if you're pounding nails in with a rock, you'll just worry about the worst ones sticking up, but if you go out and buy a nice hammer, there's no excuse not to get them all nice and flush."

This is not applicable. There have been no significant advances in the treatment of autism. It has become clear that the only significantly helpful treatment is therapy that is tremendously man-power intensive. This is just the kind of thing that dissuades diagnosis, except that virtually no insurance companies pay for any kind of useful therapy for autism.

[dwight Meredith]

You seemed to have missed the latest research on the differences betwen ethyl and methyl mercury.

The new study (pdf abstract here), was conducted at the University of Washington and funded by the NIH. It involved exposing infant monkeys to the different forms of mercury in qualtities that mimic those that were found in infant vaccines in the 1990s. The study concludes that ethyl mercury passes the blood brain barrier more easily and therefore remains in the brain longer. If you want the link to the full study please give me a hollar.

I was also suprised that you did not discuss the research of Dr. Richard Deth of Northeastern University. Dr. Deth's study found "there is an apparent link between exposure to certain neurodevelopmental toxins and an increased possibility of developing neurological disorders including autism." Dr. Deth's study also shows the mechanism by which mercury exposure may effect the development of the brain and shows that that effect may result in autism.

The press release of that work is here and the full study, in pdf is here.

The possibility of a link between mercury exposure and autism is a complicated issue. The epidemiological studies are not yet conclusive for a variety of reasons that would make this comment far too long.

Suffice it to say that a weekend of googling will not help anyone arrive at a well informed opinion. Over the last seven years, I have read and reread every study and every crtique of every study on this subject that has been published either on the net or in academic journals.

Anyone who thinks that there is now a final, definitive answer to the question of whether mercury exposure can cause autism simply has not scratched the surface of the issue.

BTW, I should mention that your blog is one of my favorites.

[Lindsay Beyerstein]

Thanks, Dwight. I completely missed that study. I will read it.

[R. Mildred]

This is not applicable. There have been no significant advances in the treatment of autism. It has become clear that the only significantly helpful treatment is therapy that is tremendously man-power intensive. This is just the kind of thing that dissuades diagnosis, except that virtually no insurance companies pay for any kind of useful therapy for autism.

What i think you're failing to see here is that early diagnosis is the best way to begin dealing with and treating Autism, The sooner a parent knows about having an autistic child the sooner they can begin dealing with them appropriatly

[Njorl]

I realize that early diagnosis and intervention is the best course, but that doesn't mean it will contribute to an increase in diagnosis. If such a course of action saved money for the entity doing the diagnosing, then it would. Early diagnosis of autism does not save medical insurers a dime. Consequently, most pediatricians don't know a damn thing about diagnosing autism.

Most states either have programs to detect autism, or mandate that their counties do so. Each diagnosis costs them money. Even in the long run, the earlier it is detected, the more money it costs them. The benefits of early detection are rarely monetary. They are only reflected in the well being of the autistic child and the child's care givers. Consequently, they have extremely low profiles.

[Alexis]

I agree that the evidence that thimerosal has caused an "epidemic" of autism is weak. However, it is not the case that concerns about the safety of thimerosal are based on confusion about the different biokinetics of ethylmercury (found in thimerosal) and methylmercury (found in fish). While ethylmercury has a shorter half-life in the blood than methlymercury, additional study has found that of the mercury remaining in the body, a higher percentage is in the brain, and of the mercury in the brain, a higher percentage is in the inorganic form. The implications of these differences for the relative toxicity of ethylmercury and methylmercury are uncertain (see http://ehp.niehs.nih.gov/members/2005/7712/7712.pdf). It may be the case that it is the inorganic form that does the actual damage to the brain (though it is the organic forms that are more readily absorbed by the body, and therefore that ethylmercury is more harmful than methylmercury.

In any case, the primary concern with mercury exposure, whether methylmercury or ethylmercury, is that it can may lead to small decreases in neurological performance. In the majority of cases (for methylmercury), these effects are "subclinical," falling short of what could be diagnosed as a disease, but that nonetheless are detectable when comparing highly exposed groups to less exposed groups.

The evidence that methylmercury exposure can be harmful is overwhelming; for ethylmercury, we can only guess. For that reason, even in the absence of strong evidence that thimerosal contributes to autism, taking thimerosal out of vaccines is prudent policy.

[Lindsay Beyerstein]

It was a wise decision to remove thimerosal from vaccines. There are other preservatives available and have been for some time. Thimerosal is the most active, but it's also a preventable source of injected mercury.

As a historic question, the current round of concern over thimerosal was sparked by investigators who were concerned about total cumulative mercury exposure and who did not differentiate between ethyl and methyl mercury when estimating whether the average infant was exceeding any of the three federally recognized sets of mercury exposure guidelines. If you assume that ethyl and methyl mercury are equivalently toxic, then it turned out that with the expanding infant vaccination schedule, infants were exceeding the EPA guidelines, but not the other two major sets of guidelines.

E- and m- mercury are clearly not equivalent. It doesn't necessarily follow that e-m is benign, let alone safe.

[Barbara]

I don't know whether thimerosol is safe as a preservative of vaccines. I tend to think that the link hasn't been proven. However, I also know that thimerosol is not necessary, and I know that mercury is definitely a heavy metal associated with neurological damage. I also know that many scientists who are responsible for making judgments on the safety of vaccines are so conflicted by their financial interests that they wouldn't be permitted to serve on a jury deciding whether any given manufacturer's product was safe. And yet, there they are, entrusted with decisions that affect the safety and health of millions. What is a pregnant mother of two to do? Well, for one thing, read the riot act to her children's pediatrician about using vaccines containing mercury.

There are so many contaminants in the modern world that may be contributing to autism that the chance of isolating a single compound in a systematic way is probably too much to ask for. That doesn't mean we should willy nilly be indifferent to each toxin that potentially increases the total load borne by each individual -- especially with respect to children whose neurological system is not complete at birth but continues to form throughout their childhood.