Editorial by Katie Weisman, Canary Party, Executive Director and SafeMinds Board Member
Now that World Autism Awareness Day has passed, let’s assess where we truly stand. Last week, two studies were reported and spun by the CDC in the annual countdown to Autism Awareness Month, also known as “Let’s see if blue lights and awareness will keep my child from running into traffic” month. Both studies received a massive, coordinated and unforgivably uncritical airing in the press. I swear that none of the writers of any of the articles I read understands the concept of balanced journalism. I can guarantee that few of the writers or the scientists they quoted actually took the time to read the studies and think critically about them.
The first study, based on a phone survey of a nationally representative sample of parents reported autism spectrum disorders at the terrifying rate of 2% of school-aged American children (6-17). Let’s say that again – 1 in 50 children has autism. With 4 million children born per year in the United States that means approximately 80,000 children per year will get autism. At its most devastating, polio affected 57,628 people of all ages in 1952. Many died. Polio was a national health crisis demanding urgent federal response. Public health officials went into high gear. For this report of 2% of children with autism, we get, “We are just getting better at recognizing all these children with no language and zero social skills. They have always been here. You can relax”. Unfortunately, many of these children will die, too – from drowning, accidents and seizures associated with their autism – and, tragically, sometimes at the hands of those who are supposed to educate and care for them. With nothing personal intended against the speakers, here is the CDC’s event for Autism Awareness month – http://www.cdc.gov/ncbddd/autism/documents/2013-Autism-Awareness-Month-2013.pdf . You can judge for yourself if this is an appropriate response.
Here are the US autism prevalence numbers in the years they were reported (birth years and ages vary from study to study).
1970 1 in 14,857 (Wisconsin) 2009 1 in 235 (California DDS)
1987 1 in 8333 (North Dakota) 2009 1 in 110 (11 States, ASDs)
1989 1 in 2500 (Utah) 2009 1 in 91 (National Phone Survey, ASDs)
1999-2003 1 in 321 (California) 2010 1 in 124 (Salt Lake City – 8s, ASDs)
2001 1 in 150 (Brick, NJ) 2010 1 in 83 (Wisconsin Schools, ASDs)
2001 1 in 625 (Texas Schools) 2011 1 in 213 (San Franciso Bay Area, ASDs)
2002 1 in 671 (California) 2011 1 in 108 (EI in Massachusetts, ASDs)
2003 1 in 192 (Minnesota Schools ASDs) 2012 1 in 88 (14 States ASDs)
2007 1 in 150 (14 states ASDs) 2012 1 in 57 (Metro New Jersey 8s, ASDS)
2009 1 in 125 (South Carolina 4s ASDs) 2013 1 in 50 (National Phone Survey, ASDs)
Isn’t anyone worried yet?
While this new study does not have the most reliable methodology, how much more data do we need to tell that we are headed for a cliff? It is like a macabre game of chicken – how high can the autism numbers get before the country actually realizes there is a problem. I have decided that we need to rename our lead agency the CFCADEA – the Centers for Controlling Any Disease Except Autism.
The second study is worse. In the best tradition of tobacco science, this one has been publicized as proving that we are not over-vaccinating our children – that the number of vaccines a child gets has nothing to do with autism. In fact, the study didn’t even look at the total number of vaccines given – it only looked at the number of antigens. The problem is that they did a shoddy job answering a question that wasn’t really the question parents want answered. Parents want to know if vaccinated children (according to the regular schedule) have a higher rate of autism than unvaccinated kids. The only way to find that out is to actually do that study. And yes, it is over a decade since we first asked them to do the study. If vaccines are causing the autism epidemic, that is 800,000 kids…
Let’s be clear here:
1) It appears that there were no children in the study who were unvaccinated – therefore, no actual control group. If there actually were a handful who did not receive any vaccines, they are not mentioned anywhere in the text and they are rolled into the lowest exposure group, thereby mixing the placebo group with the exposure group.
2) Children with low antigen exposure were compared to children with high antigen exposure for their odds ratio of developing autism. This is like comparing 1 pack a day smokers to three pack a day smokers for lung cancer and saying that because they both developed lung cancer, the smoking had nothing to do with the cancer.
3) The study makes an assumption that all antigens are equally likely to cause problems and that the only thing that matters is the total number of antigens. This is like saying that it doesn’t matter what you are drinking, only the total ounces matter. Sure, throw some formaldehyde and aluminum in the blender. Don’t worry about any synergistic toxicity.
4) The study completely ignores all the other vaccine ingredients besides the antigens – except for thimerosal, which they studied, but then they didn’t bother to publish the data.
5) The study implies that the reduction in the number of antigens in the DPT vaccine should have made it safer. Yet, in order to reduce the amount of antigen in a typical vaccine (which also saves the manufacturers money since the antigen is the most expensive ingredient), an adjuvant is usually added to stimulate increased immune response to the decreased amount of antigen. The adjuvant ASO3 has recently been linked to narcolepsy in children who received swine flu vaccine in Europe. While this particular adjuvant is not found in US vaccines, others are, and it begs the question of what happens when you reduce the amount of antigens but have to increase the amount of adjuvants. There is no evidence that “fewer antigens with more adjuvants” produces a safer schedule, but this is what was implied by this study.
6) Due to the large amount of antigens in the old DTP vaccines, the study is really a comparison of the antigen load between the old DTP and new DTaP vaccines. The overwhelming number of antigens in the DTP would make any impact of any other vaccines irrelevant in terms of categorizing the children as low vs. high exposure. The authors also knew that the DTP was being phased out starting in the early 1990’s so antigen load was decreasing even as autism prevalence was going up. There was no reason to do this study since these children were born in 1994-1999 and the trends were going in opposite directions.
7) In addition, the authors included 186 children among the 752 controls who, while they did not meet the cutoff criteria for Autism Spectrum Disorder, did meet diagnoses of speech delay, learning disability, ADHD and so on. This is like doing a study of diabetes cases and intentionally putting people with pre-diabetes in the control group.
8) To cap this all off, the study design was so badly flawed to begin with, that any results it reports are invalidated. It reused a dataset that Price et al. used in 2010 to look at thimerosal exposure and autism. In both studies, negative findings were reported, which isn’t surprising given the study designs. In a nutshell, when you design a case-control study, matching is often used to efficiently control for confounding variables that are associated with the outcome (in this case, a diagnosis of autism spectrum disorder). However, the researchers must ensure that the confounders that they match on are not associated with the exposure of interest (antigens or thimerosal) or they will reduce the power of the study to detect an effect. An example of this would be studying the effect of high cholesterol on heart disease and matching on the basis of “high-fat diet”. Since a high-fat diet is strongly associated with high cholesterol, you have eliminated relevant differences between the cases and controls in the design of the study. Similarly in this case, the researchers matched cases of autism and control children on the variables of MCO (managed care organization) and year of birth. Since MCOs typically order vaccines in bulk lots from a single manufacturer and since year of birth is associated with which vaccines are on the schedule at the time, both of these were strongly associated with the number of antigens and the amount of thimerosal that the children received – thereby eliminating the differences between cases and controls in the study design. For a complete discussion of the problems with this dataset, here is the link to a recently published rebuttal, funded by SafeMinds, of the 2010 paper: http://cdn.intechopen.com/pdfs/41866/InTechVaccine_safety_study_as_an_interesting_case_of_over_matching_.pdf
For those who haven’t looked at the US vaccine schedule lately, I strongly suggest that you go to the following link http://www.cdc.gov/vaccines/schedules/hcp/imz/child-adolescent.html and look at what our kids are getting. Ask yourself, would I, as an adult, be OK with getting 9 shots for a total of 13 diseases in one doctor’s visit. That is what we are giving our 15 month olds.
Would you let your doctor give you any 9 drugs at one time?
Would it make you feel safer if I told you that this combination of shots had never been tested against a placebo for adverse events?
Would it make you feel better if I told you that if you do have a reaction to the vaccines, none of the manufacturers has any financial liability whatsoever?
We seriously need to consider whether this protection from infectious disease is ultimately producing the healthiest children. Vaccines are the only medical product given routinely to healthy infants, unlike most drugs which are given only when a disease already exists. The standard of safety for these products should be the highest we can achieve.
I would like to point out that the last three major studies purporting to disprove a link between vaccines and autism have all been done by the same group of CDC-affiliated researchers using the same dataset from the same MCO’s . The data had limitations to begin with in terms of participation rate and the exclusion of preemies and multiples, and it has not improved with successive iterations. To assume that the CDC has no conflict of interest in this study is ludicrous. With its responsibility for the vaccine schedule and control of infectious disease, along with its holding of vaccine patents with manufacturers, the CDC is not an unbiased source of research on vaccine safety. And don’t get me started on the part where the authors thank Paul Offit, MD, for his help. It would sure be nice if some of the reporters out there would recognize that they are being spoon-fed spin. For those who want to go back and read the studies, here are the links.
Happy Day After World Autism Awareness Day. Isn’t it time to start the real conversation about autism?