Immunization Science

Review of Published Data on Thimerosal and Autism

The article

Thimerosal-Containing Vaccines and Autistic Spectrum Disorder: A Critical Review of Published Original Data. Parker SK, Schwartz B, Todd J, and Pickering LK. Pediatrics 2004;114:793-804.

The question

What is the quality of the evidence assessing a link between thimerosal-containing vaccines (TCV) and neurodevelopmental disorders (NDD), including autism and autism spectrum disorders (ASD)?

The study

This study sought to identify—from the National Library of Medicine’s Medline database (PubMed) and the Cochrane Library—all research publications reporting original data that were published between 1966 and 2004 that examined 1) for an association of TCV and autism, ASD, or NDD and 2) the human pharmokinetics (the accumulation and elimination from the body) of ethylmercury, the potential toxin in thimerosal.

The authors then evaluated each publication for the study design, the methods used, the types of analyses performed, and what interpretations could be made from the studies.

The findings

The authors identified 14 published studies, 10 epidemiologic and 4 pharmacokinetic. Each publication was critically reviewed.

The quality of the reviewed articles varied significantly:

  • The researchers found that 4 epidemiologic studies that purported to support an association between TCV exposure and autism—all by the same authors—used overlapping data sets and contained critical methodologic flaws. The flaws in these studies made it impossible to interpret the data.
  • None of the remaining retrospective and prospective cohort studies found an association between TCV and autism or ASD.
  • Two pharmacokinetic studies were excluded from further analysis because they did not include new original data. In one case, the study modeled theoretical estimates of mercury concentrations and the second because it used previously published extrapolated (calculated) data.
  • The two remaining pharmacokinetic studies suggested that ethylmercury is eliminated much more quickly than methylmercury, a known neurotoxin. Blood mercury levels after vaccination were not in the range of known toxicity. Both of these studies, however, were limited by their small sample sizes and differences in timing of specimen collection.

The relevance/bottom line

Of the published literature identified by this study, the majority of the epidemiological studies did not demonstrate an association between TCV and autism. A small number of epidemiological studies suggesting an association were found to be of poor quality with methodological flaws that made interpretation impossible.

What little data there are about the pharmacokinetics of thimerosal also suggest that an association is less likely.

NNii’s comment

The findings from an increasing number of studies should provide reassurance that there continues to be no evidence of an association between autism, ASD, or NDD and TCV. Many of the studies reviewed in this article have also been reviewed on this Web site.