David Brown: “Covered up” California data shows NO autism/thimerosal link

By David N. Brown

While the story of Poul Thorsen and possible misappropriation of funds continues to unwind, it is worthwhile to take a few steps back and place the claims by AoA and other anti-vaccine sites in a broader context. The major theme of their coverage has been that Thorsen's alleged misconduct somehow justifies much earlier claims against Danish autism/vaccine coverage. The Danes incurred the wrath of the anti-vaccine movement by reporting a simple finding: When they removed thimerosal from vaccines in the early 1990s, their country still experienced a rise in cases of autism. This has led to a long search for grounds on which to dispute these studies, typically alleging outright fraud as part of a conspiracy between Danish authorities and the US CDC. A particularly significant example is this statement from PutChildrenFirst , :

“In this email , we see data provided by Dr. Loring Dales from the California Dept of Health showing the relationship between the vaccination rates of DTP by second birthdays, and the number of autism cases in California. One of Dr. Simpson's colleagues mentions 'this looks like material for a graph.' The graph is created, page 3 of the email, and there is a clear, linear relationship between the increase in vaccination rates (from 50.9% to 75.7%) and the number of autism cases per year (from 176 to 1182, a 6.7x increase) between 1980-1994... “

Something which should be readily apparent is that the two things being measured are too different for direct and useful comparison: a percentage and a quantity. These are different enough to provide directly contradictory answers to the same question. For example, suppose you want to determine which country has the largest Jewish population. But, this could reasonably be measured two different ways: by percentage of the population, or by absolute numbers. If you measure by percentage, you will unequivocally reach the obvious conclusion: Israel. But if you measure by numbers, then it’s more or less a tossup between Israel and the US, with different estimates at different times favoring one or the other.

Another consideration which will be readily apparent to those familiar with the subject is that the vaccination rates given are preposterously low, for the population of California. For a point of reference, in the UK MMR scare, measles vaccination leveled off at 70%. Unsurprisingly, data from California conflict with these numbers, showing that in 1980 87.8% of children were vaccinated. At least one author has directly disputed the numbers given by anti-vaxxers. I think the more likely explanation is that we are looking at numbers for a small population, and it would obviously help very much if we had much more information than the (shall I say alleged?) email gives. I think we can get some idea of the size by comparison with Britain. A 2008 British report stated, “Fewer than half of all five-year-olds in London have had the MMR vaccine and booster jab… The Primary Care Trusts (PCT) with the highest percentage(s) are Kingston (76%), Hammersmith & Fulham (69%) and Sutton & Merton (68%). The PCTs that fare the worst are Newham (42%), Islington (43%), and Brent Teaching (43%).” Now, the 2007 population of London was 7.6M. The current population reportedly serviced by the Newham PCT is 247,700. Thus, the percentages given are consistent with a population as high as several million or as low as a few hundred thousand. I would be inclined to favor the latter, especially when compared to the autism numbers, which for a population of 100-200K would give a rate of 0.5-1%.

Then there is the data: To whatever extent it is useful to compare them, here are graphs of the numbers for vaccination rates (Graph 1) and autism diagnoses (Graph 2):

Needless to say, these graphs do not show anything remotely resembling a “linear relationship”, let alone a clear one. Vaccination rates first oscilated, then increased in an approximately “linear” fashion. Autism diagnoses, on the other hand, were already increasing in ca. 1980, and as early as 1984-85 the increase was accelerating in a manner consistent with an exponential trend. If the increase is quantified by percentage, a curious pattern is evident. From 1984-1992, increases were mostly oscilating in the range of 20-30%, peaking at 27% for 1989-90 and 1991-1992. Then in 1992, the rate of increase by percentage drops back to early 1980s levels.

The most significant thing about this, relative to the thimerosal issue, are the years 1986 and 1991. That year, the Hib vaccine, at that time containing thimerosal, was added to the recommended schedule, which together with the notorious whole-cell DTP vaccine made two thimerosal-containing vaccines on the schedule. What followed was the only recorded decrease in autism diagnoses during the 15-year period. In approximately 1991, acellular and mostly thimerosal-free DtaP vaccines were licensed to replace DTP. At about the same time (ca. 1991-1993), thimerosal-free Hib vaccines such as HibTiter, ActHIB and OmniHib were being licensced. (See MMWR 1993 report.) These developments could have marked the end of thimerosal-containing vaccine use in the US, if not for the 1991 addition of Hep B; as it was, there were half as many as there had been. This was accompanied first by the second-highest percentage increase, then by the lowest that was an increase, and then by a rebound to nothing more or less than the ca. 8% level prevailing in the early 1980s.

Thus, the data supposed to incriminate the CDC as having motive to conspire with Denmark to discredit an autism/thimerosal link instead show that US health agencies already had independent evidence of the same phenomenon the Danes reported: After exposure to thimerosal was reduced, cases of autism increased. But, these data also show, probably more clearly than the Danes’ data could have, that the “autism epidemic” was in evidence, for anyone looking for it, about a decade before it became a subject for alarm and speculation, and that its path was completely independent of changes in vaccination practices.


"Grendel" said...

if you analyse the entire period as a compound annual growth rate which which we find that vaccination rates have increased 3.1% per year over a 13 year period while at the same time autism diagnosis has increased at over 14% as a compounded rate - the only relationship between the two is that bot are numbers that increased. You may as well blame autism on inflation or global warming.

David said...

Or Computer-Generated Image special effects. Stop-motion rules!

Miss Sizer said...

Is it not more likely that during the 1980s, testing and early detection of children on the autistic spectrum was patchy at best, thus many of those affected were not detected by medical professionals until later in life. Many of these doctors received further training and support in the diagnosis of autism in children throughout their careers and so as education and detection of autistic spectrum disorders improved, the number of children recorded as having it, increased? Surely, some concept of improved medical practices must come into play as a factor?