Despite a critical lack of supporting epidemiological evidence, there are those who cling to the belief that mercury found in the vaccine preservative thimerosal is an important cause of autism and is responsible for the recent epidemic increase in this diagnosis. Never mind that this epidemic is probably mostly or even entirely an artifact of increased efforts at diagnosis while simultaneously broadening the definition (reflected in the new terminology of “autism spectrum disorder” or ASD).

The argument was based on the assumption of causation from correlation: primarily in the 1990’s the number of routine childhood vaccines was significantly increased. During this same time the number of ASD diagnoses increased dramatically as well (from 1-3 to about 15 cases per 10,000, although the true incidence is probably between 30-60 per 10,000). It would be reasonable to draw from this the hypothesis that maybe there is a causal link between the two – but there are those who drew the conclusion that there must be a link – and that was their folly.

Subsequent to this there were a number of epidemiological studies to see test the correlation hypothesis – does exposure to thimerosal truly correlate to risk of developing ASD? After 8 years of studies the answer is a fairly definitive no. Multiple studies (for example), and multiple reviews of studies (here and here), have all concluded that there is no correlation.

The final test came when the US banned the addition of thimerosal to childhood vaccines in the US at the end of 2001, beginning of 2002. The removal of thimerosal should have led to a decrease in ASD. Well, now it is 5 years later and the incidence of the diagnosis is still increasing. Excuses abound – like for instance that vaccines in doctor’s offices were not recalled so thimerosal containing vaccine are perhaps still being used. Surveys, however, have shown that pediatric offices had very little thimerosal vaccine in storage, so the premise has not been validated. And despite this the relative (if not absolute) removal of thimerosal still should have resulted in a decrease in ASD. It hasn’t. The mercury believers are still holding out for a delayed effect, but their hopes are rapidly fading away.

But no one wants to abandon a beautiful theory just because the evidence doesn’t support it, so increasing the mercury believers are talking less about childhood vaccines and more about mercury from sources – such as environmental. Of course environmental mercury would not explain the apparent correlation between childhood vaccine and ASD, so in any case it would not support their original hypothesis.

One alternative target was pre-natal vaccines, which also contained thimerosal but now also is mostly free of this preservative. Good researchers like to dot every “i” and cross every “t” so now this hypothesis has been tested also. Judith Miles, who is the William S. Thomson Endowed Chair of Autism and professor of pediatrics and pathology in the Missouri-Columbia University School of Medicine, has just published an epidemiological study looking at the incidence of exposure to thimerosal pre-nataly in mothers who had children with ASD. A vaccine containing thimerosal is often given to prevent a reaction between maternal and fetal blood when the mother’s Rh factor does not match that of her child’s. Dr. Miles found no association – not with Rh incompatibility, with exposure to the vaccine or to thimerosal. None.

The mercury alarmists are not benign. It is OK to raise possible concerns about health risks, and even to lobby for research to answer important questions, but the mercury hypothesis proponents are raising often hysterical alarms about a risk that apparently is not real. Fear over mercury has also been adopted by the anti-vaccine crowd. The result is a decreasing compliance with vaccines and a resulting increase in otherwise preventable childhood infections. So there is a body count to be attached to their activity.

Further, while affluent countries have afford to use single-shot vaccines without preservatives, the poorer countries of the world rely upon cheaper multi-shot vaccines that require preservatives like thimerosal to prevent bacterial contamination. So poor countries may bear the brunt of negative health effects from unscientific fear mongering over vaccines and mercury.

The primary defense of the mercury hypothesis proponents is to argue that the epidemiological evidence should be trumped by toxicological evidence. Some still are trying to massage the epidemiology, but as that is a sinking ship most have already abandoned it for the SS toxicology.

Toxicological evidence basically looks at what mercury does to nervous tissue, what effect does it have on cells in general, how the body handles mercury, and how mercury affects physiology. The evidence from such research is clear – mercury is a terrible toxin, and it is specifically a neurotoxin. This is not in dispute. The “mercury militia” claims that such evidence should trump the epidemiological evidence – but this notion is flawed to the point of being naïve.

First, knowing what happens when you put mercury on top of neurons in a Petri dish is far from knowing what happens when mercury is ingested. Biological organisms are complex and it is hard to predict the net effect of any substance on a living organism. Ingested mercury may not get to neurons in sufficient doses to cause the same toxicity seen in the Petri dish, and the body may have mechanisms to mitigate toxicity again not seen in-vitro.

Second, the mercury proponents neglect the central issue of toxicology and that is that it is all about dose. They frequently argue that mercury is toxic at any dose, and therefore whatever exposure children are getting must be causing them harm. But this is simply not true. There are dose thresholds below which no meaningful toxicity takes place. In short they ignore the possibility that even though mercury is a neurotoxin, exposure to mercury from thimerosal was in a form a dose that was not sufficient to cause measurable biological harm. They simply rule out this possibility a-priori. Mercury is a toxin, they argue, therefore the epidemiology must be flawed.

In medicine in general, however, epidemiology trumps toxicology – if your question is: has this substance actually caused or is it causing harm? Toxicology is more theoretical –it shows that mercury may cause harm, but the epidemiology shows that it hasn’t.

The latest study adds one more epidemiological nail to the coffin of the thimerosal-autism hypothesis. RIP.