A Primer on Scientific Sleight of Hand, Based on a Recent Thimerosal Study

Over at The Boondocks, someone posted a recent ABC News story, Mercury-Containing Vaccine Vindicated. Since the toxicity of mercury is well established, I wondered what was behind that headline. Do I need to say that I was not impressed by what I found?

The article is long, but the basic information is presented in the first few paragraphs:

As federal health officials offer more evidence that the mercury-containing vaccine preservative thimerosal is safe, many vaccine experts say in retrospect that the U.S. Food and Drug Administration's decision to remove it from childhood vaccines may have done more harm than good by raising public fears. ....

A new study published in the New England Journal of Medicine concludes that early exposure to thimerosal does not cause any neurological problems. Thimerosal, used in vaccines since the 1930s, has been a topic of controversy since the FDA banned it in 1999.

Some claim that the additive causes autism and other brain development disorders in children. But the latest study joins a growing body of literature that shows thimerosal is safe and causes no long-term negative effects on children's health.

Knowing how research articles tend to be over-simplified in mainstream media reports, I searched for—and found—the original report: Early Thimerosal Exposure and Neuropsychological Outcomes at 7 to 10 Years. And what I read there has convinced me that USSA government “scientists” have—deliberately or inadvertently, I do not know—employed a scientific sleight of hand in the study. (Some accuse them of an ongoing refusal to face some uncomfortable facts. If true, given what is at stake for the medico-industrial complex, that comes as no surprise. Nor does the ADA’s huffiness over informed consent for amalgam fillings.) Because I don’t know how long the journal article will be publicly available, I’ll quote extensively from it.

But first, one other very important point: the study was funded by the CDC—a government agency. As such, it should come as no surprise that more often than not it supports the current nanny-ninnyisms, both directly and indirectly through the research it chooses to fund. Never forget, all government entities push an agenda, and you can be certain it isn’t one encouraging full disclosure of all relevant data and engaging in one’s own critical thinking. That said, one can fall into all sorts of conflict of interest traps. Anyway, on to the main event, so to speak.

From the article’s abstract, a warning flag is immediately raised:

We enrolled 1047 children between the ages of 7 and 10 years and administered standardized tests assessing 42 neuropsychological outcomes. (We did not assess autism-spectrum disorders.)

Given that autism is the primary neuropsychological problem alleged to develop from thimerosal exposure, it is highly suspicious that these disorders are not included in the study. I’ll return to this point. Here’s the complete description of the study’s participants:

We enrolled children from four HMOs that participate in the CDC's Vaccine Safety Datalink. Children from each HMO were eligible to participate if they were 7 to 10 years of age and had been enrolled in the HMO from birth through their first birthday. Birth dates ranged from January 1, 1993, to March 30, 1997; testing was conducted between June 1, 2003, and April 27, 2004. All parents provided written informed consent for their children to participate in the study. Children were excluded if they had certain conditions recorded in their medical records that could bias neuropsychological testing (e.g., encephalitis, meningitis, or hydrocephalus) or if their birth weight was less than 2500 g (Table A of the Supplementary Appendix, available with the full text of this article at www.nejm.org).

Notice that nowhere in there is any mention of the inclusion of non-normal children—were austistic or Asperger's children included in the study? Were any children with developmental delays—in particular cognitive challenges—included in the study? Not knowing these things about the sample studied is enormously limiting. I’ll return to this point too. On to how the researchers calculated mercury exposure:

We determined the mercury content of vaccines and immune globulins that the study children received when they were infants (1993–1998) from published data and the FDA (Table B of the Supplementary Appendix). We identified vaccines and immune globulins that children had received from HMO computerized immunization records, paper medical records, personal immunization records, and maternal interviews. Prenatal exposure to mercury included all known exposures of the mother to thimerosal-containing vaccines and immune globulins during pregnancy. We defined postnatal exposure as micrograms of mercury divided by the weight of the child in kilograms at the time of administration of each vaccine or immune globulin. Individual exposures were summed during the period of interest: birth to 1 month and birth to 7 months (1 to 214 days). We did not assess periods of thimerosal exposure after 214 days of age because we hypothesized that the potential effect of such exposure would be small. (Since most vaccines that are administered after 214 days would typically be given at 12 to 18 months of age, the dose per kilogram would be substantially lower.)

Most of this sounds pretty good to me, but I wonder why vaccines given after 214 days weren’t included. Their “dose per kilogram” argument isn’t particularly convincing to me, since newborns undergo the greatest growth spurt humans have. Is there a big jump in number of vaccines given sometime after 214 days that the researchers want to ignore? I know some cutoff needs to be used, but this one seems to be too early to be a thorough examination of thimerosal’s influence, as we know mercury accumulates in the body with repeated exposure, particularly in the central nervous system. Continuing to the neuropsychological assessment:

Each child was assessed on 42 neuropsychological outcomes selected on the basis of findings from the CDC's screening study, previous studies of methyl mercury, and the recommendations of an external panel of independent consultants. Most of the measures were collected during a 3-hour neuropsychological assessment performed by trained evaluators. The outcome measures included speech and language indexes, verbal memory, achievement, fine motor coordination, visuospatial ability, attention and executive-functioning tasks, behavior regulation, tics, and general intellectual functioning (Table C of the Supplementary Appendix). Measures of attention, hyperactivity, and executive functioning were based on reports from parents and teachers. We evaluated motor tics, phonic tics, and stuttering on the basis of a combination of ratings by evaluators and reports by parents and teachers (Table D of the Supplementary Appendix). Personnel who were administering the neuropsychological battery were not aware of the children's exposure to mercury or medical history. Mothers were asked to refrain from giving children selected prescription medicines for attention deficit–hyperactivity disorder (ADHD) the day before testing. Since the CDC is conducting a separate case–control study of autism in relation to mercury exposure, a measure of autism was not included in the test battery.

This sounds pretty straightforward. However, that last sentence really drops a bombshell—the CDC is conducting a separate study of autism and mercury exposure? Before, upon reading that sentence about not assessing autism-spectrum disorders, I suspected that the researchers were fudging the sample somehow. Now I know that happened—maybe not intentionally, but that doesn’t matter. What matters is this is being reported as if it were good science, and it isn’t. (It was very disappointing, while researching the study, to find Medical News Today's story noncritically summarizing the study and its results).

The researchers have artificially and post hoc split the population of interest. To try to accurately assess the impact of thimerosal in vaccines on cognitive functioning in children, the target population is “all children who have been vaccinated”. With a large enough sample size, that would necessarily include some severely autistic children. I know there will be protests to that: trying to give cognitively impaired children the same test as non-impaired children presents problems. Without them, though, the complete range of possibilities is not represented in the data. The data are clipped, rather than being truly representative, and therefore may not give an accurate picture.

Doesn’t the fact that the CDC is doing a separate study mean that comparisons can still be done? Well, yes ... sort of. Comparing results from two separate studies is never as good as making comparisons within one study, though. Differences in the tests used, the researchers who conducted the study, the experts who assessed the children, and other methodological differences become confounding variables that can make it difficult, if not impossible, to attribute statistically significant effects to the variable of interest—mercury from thimerosal—with the degree of certainty suggested by the p value (speaking somewhat statistically sloppily). That is to say, it’s impossible to factor out unwanted contributions to differences in scores from that of the mercury. Further, while comparing significant effects across studies can be done, it isn’t as methodologically strong as conducting one well-designed study. An essential component of a well-designed study is proper identification of the population, and making one’s best effort to get a representative sample. By separating the real population into two separate populations, the researchers have failed at this. And where’s the mention of a genuine control group—that is, children who had no thimerosal exposure during the same time periods? There is none. There appears to have been no effort to try to even recruit such a group, which admittedly would have been difficult. But the inclusion of those participants would have resulted in a much stronger and more informative study.

By not describing the participants in this study very well, it’s very difficult to know how much credence to give the results. Are these essentially normal kids? It seems so, but we just don’t know. How many children tested were able to complete all the tasks? What was the median time of completion for each kind of test? How many of the children in the sample were on ADHD meds?

How does a study that is claimed to have tested the relationship between thimerosal and “neuropsychological deficits” actually do this, when the sample apparently does not include any children with the most severe neuropsychological deficits? Yes, including them would make the testing much more difficult to carry out, and might possibly introduce confoundings; however, without data from such individuals this study is severely flawed. It strongly appears not to be a representative sample of the entire population of children that they're claiming it is. Whether it was intentional or not, this design is an unconscionable experimental fuckup.

And, aside from whatever one wants to say about the study, all of this kind of research has a fundamental flaw that I think the nanny-ninnies and medicrats will never grok: individuals are tremendously varied in all kinds of ways. “Massed science” like this might inform us as to the larger trends, but there are always outliers. Some individuals’ biochemistry makes them more susceptible to mercury, while others are much hardier. Following the general trend would pose no problem for the latter individuals, but could well be fatal for the former. As we don’t currently have the ability to test for such things (thanks in part to regulatory barriers and interference in research), let alone doing so at birth, relying on this kind of science is a real crapshoot for the individual. Yet the nanny-ninnies have no problem mandating that all get vaccinated, or that trans fats be banned, et cetera and ad nauseam. Given that thimerosal is almost 50% mercury by weight, I don’t see how any scientist with a conscience can continue to assert that injecting lots of it into infants and young children is a good thing.