You know the rule. The answer is, “No.” But the assertion has gone viral on social media thanks to the zombie-like resurrection of a long-told, oft-debunked story that the US Centers for Disease Control (CDC) is hiding its own data linking autism and mercury in vaccines. If you see such assertions in your timelines and newsfeeds (sample headline: “CDC Caught Hiding Data Showing Mercury in Vaccines Linked to Autism”), send the disseminators here. Why? Read on.
In 1999, four authors affiliated with the CDC submitted an abstract [donotlink linked here**] for a conference … a CDC conference for fellows of its Epidemic Intelligence Service (EIS). The EIS, by the way, serves as the Interpol of infectious disease, tracking down elusive perpetrators worldwide and stopping them before they can harm again. In other words, they are people who dedicate their lives to saving lives. Every year, the program also sponsors a conference. And 1999 was no exception.
In 1999, one Thomas Verstraeten and three colleagues submitted an abstract for the EIS conference, presenting results from the early stages of a two-phase study*. It was preliminary, as many, many such submitted abstracts are. They indicated ‘no strong preference for a poster presentation’, which means that they were OK with getting up in front of the conference attendees and discussing their findings on the record. The zombie-like story making the rounds would have you believe that submitting the abstract “required the approval of top CDC officials prior to its presentation at the Epidemic Intelligence Service (EIS) conference,” but all conference abstracts require approval from the people running the conference–which, in this case, was the CDC.
The authors reported using raw data from the Vaccine Safety Datalink and HMOs in the Pacific northwest to identify mathematical relationships between thimerosal-containing vaccines and developing neurological and renal impairment (thimerosal is a preservative that contains ethylmercury and prevents dangerous contamination of large volumes of vaccine. It currently is present in multidose vials of flu vaccine). In their comparison of what they call the “highest exposure group” to an “unexposed group,” they reported an increased risk for nondegenerative neurological disorders.
Conference abstracts and the accompanying data are almost always preliminary. In fact, the likelihood that conference material and what finally appears in a peer-reviewed journal will differ is quite high. Much conference material never appears in a full, peer-reviewed article at all because completion of the study yields the much-dreaded “negative results.”
The year of the EIS conference submission*, 1999, was a turning point for thimerosal in vaccines. And then in 2000, the Simpsonwood Conference took place, an assemblage of experts from inside and outside the CDC to discuss the thimerosal issue. The entire transcript of that conference is available here. Verstraeten was in attendance and presented on the data related to the not-even-remotely concealed results from his two-phase study. One thing he noted–and this issue probably is one of many that drives differences between a conference abstract and a final publication–was variability related to the HMOs gathering the data. Those differences mattered to the outcomes and had nothing to do with thimerosal.
In his presentation at the Simpsonwood conference, Verstraeten noted,
This is the result for autism, in which we don’t see much of a trend except for a slight, but not significant, increase for the highest exposure. The overall test for trend is statistically not significant.
Later in the presentation, we learn that phase I of the study looked only at raw numbers from a database while phase II involved chart examination to confirm diagnoses and added in an HMO. The second phase involved new data following on the study described in the 1999 abstract. This chart review matters. As one of the other authors on the 1999 abstract notes in the Simpsonwood presentation:
Now with autism, if we limit it to children with exposure at either one month or three months of age… there is a relative risk that is no different than one and that is replicated whether we limit it to children with a diagnosis mentioned in the chart where the child was referred to a specialist, or the child was confirmed by a specialist.
In other words, the chart review refined the original raw data and effaced any finding of increased risk. ETA: If you’re feeling wonkish, writer Lindsay Beyerstein elaborated at length in 2005 on the Simpsonwood conference and the limitations of the presented study.
In spite of the openness of this process and the adherence to an original two-phase plan for the study, Verstraeten found himself (and continues to find himself, it seems) the target of accusations of manipulating or hiding data, particularly when the peer-reviewed paper from this study was published in 2003 (abstract here). His having gone to work for “Big Pharma”–in this case, GlaxoSmithKline GSK +0.93%–following completion of his appointment at the CDC brought further accusations of complicity in a coverup. Indeed, the accusations were so hot that Verstraeten responded to them in a 2004 commentary published in Pediatrics, recounting the history. Bottom line was, he was a foreign citizen whose fellowship with the CDC was ending, and he sought and obtained employment in his home country, in his field.
In his 2004 commentary (which is behind a paywall), Verstraeten says, “Did the CDC water down the original results? It did not.”
He goes on to write
The CDC screening study of thimerosal-containing vaccines was perceived at first as a positive study that found an association between thimerosal and some neurodevelopmental outcomes. This was the perception both independent scientists and antivaccine lobbyists had at the conclusion of the first phase of the study. It was foreseen from the very start that any positive outcome would lead to a second phase.
In other words, when you dig into raw numbers and find some mathematical relationships, then you have a reason to move to the second planned phase of examining the charts. If you don’t find anything, phase II is a non-starter.
He then notes
Because the findings of the first phase were not replicated in the second phase, the perception of the study changed from a positive to a neutral study. Surprisingly, however, the study is being interpreted now as negative by many, including the antivaccine lobbyists. The article does not state that we found evidence against an association, as a negative study would. It does state, on the contrary, that additional study is recommended, which is the conclusion to which a neutral study must come.
Perhaps you don’t want to take Verstraeten’s word for it because he went to work for Big Pharma. Those who oppose vaccines often rely on the US Congress, for better or for worse, to make their arguments. So, here’s a link to the findings of the Senate Committee on Health, Education, Labor and Pensions from their 2007 investigation into allegations that the CDC used Simpsonwood to cover up a thimerosal-autism link and that Verstraeten manipulated data.
Here’s what the Senate committee concluded regarding allegations against Verstraeten:
Allegation # 2: The Centers for Disease Control (CDC) convened the Simpsonwood Conference to cover up the finding that thimerosal causes autism.
Findings: The allegation is not substantiated. … Instead of hiding the data or restricting access to it, CDC distributed it, often to individuals who had never seen it before, and solicited outside opinion regarding how to interpret it. The transcript of these discussions was made available to the public. The data was also discussed at the Advisory Committee on Immunization Practices, a public forum held on June 21 and 22, 2000. Simpsonwood participants generally agreed that the VSD data set was weak, it was difficult to assess causality, and further study and investigation were warranted.
Not exactly the behavior of government scientists bent on a cover-up.
The committee also found that
Allegation # 3: Dr. Thomas Verstraeten, MD, MSc, was pressured into changing his research position regarding a causal link between thimerosal and autism.
Finding: The allegation is not substantiated. … HELP Committee staff interviewed Dr. Verstraeten with regard to his findings and his participation in the Simpsonwood Conference. …Review of the phases of Dr. Verstraeten’s study, “Safety of Thimerosal-Containing Vaccines: A Two-Phased Study of Computerized Health Maintenance Organization Databases,” and examination of his voluntary response to Committee questions during his interview reflect that his intention was always to conduct a two-phase study. … there is no evidence that GlaxoSmithKline hired Dr. Verstraeten for the purpose of pressuring him to manipulate his data on a causal link between thimerosal and autism. … Dr. Verstraeten was working in the United States at CDC on a temporary visa. Near the completion of his tenure with CDC, he began searching for employment in his native country and found employment with GlaxoSmithKline where he continues to be employed.
In spite of the neutral findings from Verstraeten’s study, the US Public Health Service (USPHS) and the American Academy of Pediatrics (AAP) jointly recommended in 1999 that thimerosal should be phased out of use in the handful of childhood vaccines that included it. In the wake of considerable further study showing no link between thimerosal and developmental disorders, that recommendation was retired in 2002. Now, say AAP doctors in a 2012 commentary:
Had the AAP (and, we suspect, the USPHS) known what research has revealed in the intervening 14 years, it is inconceivable to us that these organizations would have made the joint statement of July 7, 1999. The World Health Organization recommendation to delete the ban on thimerosal must be heeded or it will cause tremendous damage to current programs to protect all children from death and disability caused by vaccine-preventable diseases.
*phrase added for clarity. I also refer readers to this comment offering some more information about resources discussing the 1999 abstract and the Simpsonwood conference, etc. **Someone alerted me to a way to link without driving traffic to a specific site, so I have now added a link to the 1999 abstract submission itself, although it is given almost in full in the second link of the post, and clarifed in this sentence that 1999 was its submission year–which was already noted throughout the remainder of the post.