linking any vaccine or vaccine ingredient, or the childhood vaccine schedule, to the risk of developing autism or any neurodevelopmental disorder. But an alleged connection remains a persistent myth, propagated by the anti-vaccine movement. We have not written about this for a while so I thought I would check in on the data to see if there were anything new.
The exact claims of anti-vaxxers have shifted over the years, starting with the MMR vaccine, then thimerosal (a mercury-containing preservative), then just vague toxins, or “Too many, too soon” or perhaps it’s aluminum, or whatever – but it has to be something in the vaccines. The most prominent claim, however, that still is popular among antivaxxers is the notion that thimerosal is linked to autism.
The most compelling evidence against a link between thimerosal and autism is the fact that thimerosal was removed from the childhood vaccine schedule in the US in 2002, and this did not result in the predicted decline in autism diagnoses. Antivaxxers did make this prediction. David Kirby, who wrote Evidence of Harm, popularizing the thimerosal-autism claim, predicted that autism rates would plummet, which would prove the connection. He and others promoting this claim looked anxiously over the autism incidence rates as they came in, pointing excitedly at any downward fluctuations.
But the predicted decline never happened. I and others who thought the evidence did not support a connection made a very different prediction – that autism incidence would continue to rise until it reached its true level (I will explain this further below), and then would level off. That is essentially what is happening.
Antivaxxers not willing to give up on good propaganda started making increasingly absurd rationalizations. One I heard just a few years ago was that pediatricians were still using the thimerosal-containing vaccines they had in stock. However, thimerosal-containing vaccines in the schedule (it’s still in some flu vaccines) were phased out by 2002. Even if every pediatrician stocked up on these vaccines for the full shelf life of two years, that would have extended their use to 2004 (and to be clear, this is not what happened, but even if we take this extreme case).
As the years ticked by and autism rates did not plummet, the thimerosal hypothesis was looking worse and worse. Anti-vaxxers were persistent, however. They kept waiting for the predicted decline, like Linus waiting for the Great Pumpkin, but it never occurred. They tried to argue for more and more of a delay, but every year their strained arguments got less and less plausible.
This year the CDC published their , which include data through 2014. The rates ticked up a little bit again from the previous estimates covering through 2012. Incidence rose from 14.6/1000 to 16.8 (or 1 in 59 children).
So we now have data for 12 years following the complete removal of thimerosal from the vaccine schedule and no decline in autism diagnoses. There is also (not yet compiled nationally) showing no decline even through 2018 (that’s 16 years). If thimerosal were contributing to the number of cases of autism, we should definitely be seeing some effect by now.
In essence, by removing thimerosal from the vaccine schedule we conducted a natural experiment, with a very clear outcome. This experiment has been replicated in many countries around the world, and in fact world-wide use of thimerosal has been steadily declining for the last 20 years, without any correlated decrease in autism.
So – is there any one left who still clings to the thimerosal-autism claim despite this slam-dunk evidence for a lack of a correlation? Of course there is. In a 2017 interview, Robert F. Kennedy Jr. argued that thimerosal in flu vaccines was the culprit.
Healthline: It seems your main concern is the flu shot, which is given to children and pregnant woman. Is that your focus?
Kennedy: It’s being given to children at 6 months and it’s being given to pregnant women. So, today the exposure a child suffers is probably the greatest in history. If you give a flu shot to a woman who is one or two months pregnant, that child could be getting thousands of times of exposure levels to mercury.
Let’s look at the numbers. Total thimerosal dose to infants and young children peaked at 187.5 micrograms of thimerosal (not including flu vaccines, which are optional). The argument from antivaxxers was, as the thimerosal dose increased, so did autism diagnoses. Well, then, as the thimerosal dose decreased, so should autism diagnoses, which we now know did not happen. Multi-dose flu vaccines contain about 25 micrograms of thimerosal. So even if a child received a flu vaccine with thimerosal every year, they would not total the dose in the previous vaccine schedule for 7 years.
But remember – that 187.5 micrograms was in addition to any flu vaccines. No matter how you slice it, total thimerosal dose was decreased. There is also good evidence that mercury from thimerosal is cleared within days, and does not accumulate over years, so it doesn’t make sense to total the dose over years. Further, signs of autism are present as early as six months of age.
All of this is why Kennedy and others focus on the flu vaccines given to pregnant mothers. I don’t know what he is referring to with the “thousands of times” claim, or what he is comparing it to. In any case – this has been studied also, and there is .
Further still, only the fraction of flu shots that are given from multi-dose vials contain thimerosal. Shots given from single-dose vials and live virus vaccines do not contain thimerosal. The percentage of flu shots with thimerosal has been declining, and for this latest flu season contained thimerosal. Also, some states, like California, have banned thimerosal-containing flu vaccines, so the total dose of thimerosal for children in those states is zero.
So really it is completely absurd for Kennedy and others to claim that autism rates have not declined because of flu vaccines, and every year such claims get even more ridiculous.
So why, then, are autism rates increasing? There are a number of reasons, and we have explored them here before. It is a combination factors, including diagnostic substitution, increased surveillance, and broadening the diagnosis. The CDC attributes the recent increase to more diagnoses among minorities, who remain underdiagnosed compared to Caucasians. This is due to increased awareness and diagnostic efforts, often linked to services.
It is also possible that increasing maternal or paternal age is causing a real increase in autism incidence, but that is only part of the total increase.
We will keep tabs on the data, on any new significant published research, and keep reminding the public that the vaccine schedule is safe, it is closely studied and monitored, and there is no credible link to autism or any neurodevelopmental disorder.