Mapping the entire field of autism causation studies in one article
(Editor’s note: This article is too long for many email systems so please click the headline to read the full article on the Substack site.)
I. INTRODUCTION
It seems to me that the proper way to understand the autism epidemic is to read everything that has been written on autism causation, throw out any studies that are characterized by a financial conflict of interest or fatally flawed study design, and see what patterns emerge from the papers that are left. During my doctoral thesis I reviewed about 80 of the top studies in autism epidemiology and toxicology. That was groundbreaking at the time because the vast majority of mainstream scholars don’t have the courage to discuss any papers that threaten the profits of powerful industries.
As I’ve continued to work in this space over the last six years I now realize that there are over 800 autism causation studies in the English language focused on the U.S. It’s daunting to think about trying to wrap one’s head around a field that large. So most public health officials just grab a favorite study here or there to justify their biases and that is exactly the wrong way to approach this topic. There has to be a better way of working through the available knowledge on this issue.
Now I believe that I’ve figured out how to map the entire field of autism causation studies (about 850 papers in all) in one article. If you sat down to read each article individually, it would likely take you several years. But as I will show below, you don’t necessarily have to do that. There is a way to move through all of the literature at a meta level that I believe leads to the right answer and a viable plan for how to stop the autism epidemic.
Let’s start with a quick introduction and then get into the different types of studies.
In the early 1980s, vaccines were so harmful that vaccine manufacturers routinely lost in court. They lobbied the U.S. Congress to pass the 1986 National Childhood Vaccine Injury Act to give themselves liability protection. And they pinky-swore to make vaccines safer but there was no legal mechanism in the bill to enforce that promise so they never did.
Pharmaceutical companies proceeded to add as many vaccines as possible to the schedule. Prior to 1986, there were 3 routine vaccines totaling 7 injections. Today the CDC’s Maternal and Child & Adolescent vaccine schedules include 19 vaccines requiring 76 injections with 94 total doses of antigen (I’m actually less worried about the antigens than the other ingredients in the shots).
No one in a position of authority bothered to measure the impact of the growing vaccine schedule on the health of children. Most regulators were auditioning for a job with Pharma because that’s where the money is. Politicians depend on Pharma donations for their re-election campaigns. The mainstream news media get most of their revenue from Pharma advertising so they were never going to bite the hand that feeds them. Pharma invested heavily in public relations to lay siege to any remaining pockets of resistance.
Mercury (thimerosal) was grandfathered in as “Generally Recognized As Safe” because it’s easier to do that than actual safety testing. Aluminum adjuvants were allowed with only minimal safety testing — 1 man, 3 rabbits, and ever-moving goal posts (chapter 9 of my thesis covers the regulatory history of aluminum adjuvants). The gold rush was on so vaccine manufacturers were free to add whatever they wanted to vaccines and they would all be approved because the regulators and the medical industry were captured mind, body, and spirit by Pharma.
The autism rate skyrocketed in the 1990s and has continued to increase ever since. Rates of ADHD, life threatening allergies, autoimmune disorders, asthma, childhood cancers, diabetes, and epilepsy soared too and those are probably vaccine injuries as well. But autism spectrum disorder (ASD) is more costly than those other conditions because it’s a lifelong disability with no known effective treatment (some parents have been able to recover their children through holistic and alternative therapies but the percentage who are successful in doing so is still in the single digits).
At that point, the people who created the autism epidemic had to pretend to look for the cause. But they had to make sure to never find the actual cause because then the flow of research funding would stop and lots of these doctors and scientists would go to jail or be hung from lampposts by angry parents of injured kids. So an entire industry was created to cover up the autism epidemic.
II. TWENTY-TWO STUDIES IN THE VACCINE COVER-UP
Since 2000, more than twenty scientific studies have concluded that there is no association between vaccines and autism. The most widely cited studies are:
- Fombonne & Chakrabarti, 2001;
- Madsen et al., 2002;
- Mäkelä, Nuorti, & Peltola, 2002;
- Pichichero, Cernichiari, Lopreiato, & Treanor, 2002;
- Hviid, Stellfeld, Wohlfahrt, & Melbye, 2003;
- Madsen et al., 2003;
- Nelson & Bauman, 2003;
- Stehr-Green, Tull, Stellfeld, Mortenson, & Simpson, 2003;
- Verstraeten et al., 2003;
- Wilson, Mills, Ross, McGowan, & Jadad, 2003;
- Andrews et al., 2004;
- Heron & Golding, 2004;
- Smeeth et al., 2004;
- Honda, Shimizu, & Rutter, 2005;
- Fombonne et al., 2006;
- Miles & Takahashi, 2007;
- Thompson et al., 2007;
- Baird et al., 2008;
- Hornig et al. 2008;
- Schechter & Grether, 2008; and
- Tozzi et al., 2009.
Most of these are studies that claim no association between MMR or thimerosal-containing vaccines and autism, which is odd because the CDC’s own internal research shows that both of these types of vaccines do indeed cause autism (see 2014 statement from William Thompson and 2014 SafeMinds analysis of FOIA documents obtained from former CDC researcher turned GSK executive Thomas Verstraeten).
J.B. Handley also documents the conflicts of interests and fatal flaws in study design for most of these papers on a brilliant website called 14studies.com.
More recently, vaccine supporters have made a last stand with Hviid et al. (2019) but that study is also fatally flawed (for example the autism rate in their sample was more than 65% lower than in the general Danish population; see analysis in Hammond, Varia, & Hooker, 2025 and James Lyons-Weiler, 2019).
Furthermore even though randomized, double-blind, placebo-controlled trials are the gold standard of biomedicine, none of the studies listed above has a proper control group of unvaccinated children (the Informed Consent Action Network provides the details here). The failure to conduct proper double-blind RCTs renders all of these studies scientifically invalid.
And just like that we’ve demolished the entire basis for the claim that vaccines do not cause autism.
III. FIVE LARGE AUTISM GENETICS STUDIES
In the 1990s, the Human Genome Project captured the public’s imagination and the government’s scientific spending. Claiming that autism is genetic was a win-win because it offered the hope that autism might be curable through genetic engineering.
The federal government then sank more than $2 billion into searching for the gene(s) for autism… and found nothing that explained more than 1% of cases.
Not to be outdone by the federal government, private foundations also sought to prove that autism is genetic and failed categorically.
The genetic explanation for autism has always been problematic because there is no such thing as a genetic epidemic — the human genome just doesn’t change that fast.
The Autism Genetic Resource Exchange (AGRE) was established in 1997 by the Cure Autism Now (CAN) foundation, a predecessor organization to Autism Speaks (which later merged with CAN in 2007). AGRE collected genetic (DNA) and phenotypic (clinical, behavioral) data from 2,000 families with at least one member diagnosed with ASD and made the data freely available to qualified researchers globally. This led to the production of 169 scientific journal articles but no major breakthroughs that get us any closer to understanding autism causation or treating autism symptoms. Below I’ll explain more why and how all of these gene studies fail in a similar fashion.
As readers of this Substack will remember, Jim Simons (1938 – 2024) was a billionaire hedge fund manager with a daughter with autism. He wanted to invest some of his wealth in addressing autism and many of the top scientists in the country took advantage of him by telling him that autism was likely genetic. Jim set up the Simons Foundation and proceeded to spend over $300 million searching for the gene(s) for autism. The Simons Foundation Autism Research Initiative (SFARI) launched a project called the Simons Simplex Collection (SSC) in 2007 that gathered genetic, clinical, and behavioral information from approximately 2,600 “simplex” families — those with one child diagnosed with ASD, unaffected parents, and typically one unaffected sibling. SSC has produced 132 peer-reviewed publications and identified “102 risk genes.” But it has produced no major breakthroughs that get us any closer to understanding autism causation or treating autism symptoms.
In 2010, the Autism Sequencing Consortium (ASC) was founded by Joseph Buxbaum at the Icahn School of Medicine at Mount Sinai, New York and supported by the Broad Institute and NIH. Like other multi-million dollar health studies, ASC launched with a breathless promotional article in a major journal. Rather than focusing on the whole genome, ASC focuses on sequencing the exome which is “the portion of the genome that contains all the exons, which are the protein-coding regions of DNA.” The claim is that the exome “represents a small percentage of the total genome, about 1-2%, but it contains a majority of the known disease-related genetic variations.”
To date, the ASC has sequenced approximately 50,000 exomes from ASD cases, unaffected siblings, and parents. A search of PubMed shows 22 peer-reviewed publications associated with ASC. In 2020 they published a paper highlighting the role of 102 genes in autism and in 2022 they identified 72 more. These sorts of studies produce excited headlines in the mainstream media but no breakthroughs that get us any closer to understanding autism causation or treating autism symptoms.
IN 2011, A COMPREHENSIVE STUDY OF TWINS AND AUTISM SHOWED THAT AUTISM IS NOT PRIMARILY A GENETIC DISORDER… AND THIS MADE NO DIFFERENCE IN THE TRAJECTORY OF THE INDUSTRY
In the early 2000s, as the autism rate soared, political leaders in California wanted to better understand what was happening. So California contracted with sixteen of the best geneticists in the U.S. and gave them access to all birth records in the state. They produced a study titled “Genetic heritability and shared environmental factors among twin pairs with autism” (Hallmayer et al., 2011) which is the most comprehensive study of twins and autism to date.
They found that genetic heritability explains at most 38% of ASD cases; in two places they explain that this is likely an overestimate. So at least 62% of autism cases (and likely significantly more) are caused by something other than genes. But the search for the gene(s) for autism had already become a large and very profitable industry and this study showing that autism is NOT primarily genetic did little to slow the growth of this field.
As the cost of genetic sequencing went down, Autism Speaks launched the MSSNG study in 2014. MSSNG is not an acronym, the leaders of the study just liked the way it sounded (it’s pronounced “missing”). They’ve sequenced the genomes of 13,801 individuals belonging to what they call family “trios” (two parents and one affected child) or “quads” (two parents and two affected children). To date, MSSNG has produced 138 peer reviewed publications. They claim to have identified 134 “genes associated with autism” but again produced no major breakthroughs that get us any closer to understanding autism causation or treating autism symptoms.
Undeterred by the failure of all genetic research projects to date, the Simons Foundation massively expanded their genetic research portfolio with a new project in 2016 — the Simons Foundation Powering Autism Research for Knowledge (SPARK). As of 2025, SPARK has enrolled over 100,000 individuals with ASD and 250,000 total participants (including family members) across the U.S. Recruitment is facilitated by 31 clinical sites (mostly major pediatric research hospitals). To date, SPARK has produced over 40 peer-reviewed publications. Thus far they’ve identified “ten new autism-risk genes” but no major breakthroughs that get us any closer to understanding autism causation or treating autism symptoms.
STRAIGHT UP CENSORSHIP
As the failures of the Simons Foundation genetic research efforts mounted, rather than change course they hired the editor of Retraction Watch, Ivan Oransky, to push for retraction of studies that question the genetic narrative in connection with autism research. Given that there is an entire multibillion dollar industry built up around gene and autism studies, scientific journals are more than happy to accede to Oransky’s requests to censor the narrative on behalf of their patrons.
WHY STUDIES OF GENES AND AUTISM FAIL (THIS WAS KNOWABLE IN THE EARLY 2000S BUT MOSTLY IGNORED BECAUSE THERE WAS SO MUCH MONEY TO BE MADE)
The human genome contains 3.1 to 3.2 billion base pairs. When one feeds thousands of human genomes with several billion base pairs each into a computer and asks it to look for an association, it will certainly find many based on chance alone. But it’s the classic problem of “correlation is not causation.”
One of the world’s foremost epidemiologists, John Ioannidis, points out in “Why Most Published Research Findings are False” (2005) that only about 1/10th of 1% of these sorts of fishing expeditions (“discovery oriented exploratory research with massive testing” — usually nutrition and genetic studies with large numbers of variables) are reproducible.
As Sheldon and Gruber show in their book Genetic Explanations: Sense and Nonsense (2013) the entire theory of the case that single (or even multiple) genes code for a particular disease has unraveled in recent years.
Generally speaking, the Mendelian understanding of genes has been replaced in recent years by a completely different paradigm. British philosopher of science John Dupré at the University of Exeter argues in his book Processes of Life: Essays in the Philosophy of Biology (2012) that DNA is neither a blueprint nor a computer code for biological outcomes but rather a sort of warehouse that the body can draw upon for a range of different purposes:
The assumption that identifiable bits of DNA sequence are even “genes” for particular proteins has turned out not to be generally true. Alternative splicing of fragments of particular sequences, alternative reading frames, and post-transcriptional editing — some of the things that happen [naturally] between the transcription of DNA and the formatting of a final protein product — are among the processes the discovery of which has led to a radically different view of the genome…
. Coding sequences in the genome are therefore better seen as resources that are used in diverse ways in a variety of molecular processes and that can be involved in the production of many different cellular molecules than as some kind of representation of even a molecular outcome, let alone a phenotypic one (pp. 264–265).
The people who actually study genetics know that, at least when it comes to autism, genetic determinism is dead. But there is a fortune to be made from pretending otherwise. So the story that is sold to government and private foundations is that the “genes for autism” are out there somewhere just waiting to be found if only they will keep the research money flowing.
Government plays along with this ruse because funding genetics research keeps scientists away from studying toxicants which might threaten powerful interests. The result is an entire multibillion dollar research industry that produces hundreds and hundreds of peer-reviewed articles that never get us any closer to understanding autism causation or providing a cure.
As the search for an “autism gene” repeatedly failed, geneticists came up with a placeholder theory that they call “genetic dark matter” patterned after the dark matter in astrophysics that is said to make up most of the universe — that astrophysicists cannot explain or measure. The idea is that a gene for autism must surely exist but that they do not have the tools to detect it yet. This has kept the grant money going for now. But the entire scheme is untenable.
For more on the boondoggle of the mythical search for the “gene(s) for autism” please see my article, “Nearly everything that we’ve been told about genes and autism is wrong” (2025).
IV. FOUR LARGE EPIGENETIC STUDIES
The University of California, Davis launched the Childhood Autism Risks from Genetics and the Environment (CHARGE) study in 2003 to investigate environmental causes and risk factors for autism and developmental delay. It is led by one of the most respected and widely-published environmental epidemiologists in the world, Irva Hertz-Picciotto. CHARGE is a case-control study where researchers identify children age 2 to 5 with autism and compare them with similarly matched children without the diagnosis of autism. They have enrolled more than 2,000 autism families in their studies and produced foundational reports on the effects of:
- air pollution (e.g., particulate matter, nitrogen dioxide, ozone)
- pesticides (e.g., organophosphates, pyrethroids, carbamates)
- heavy metals (e.g., mercury, lead, cadmium)
- per- and polyfluoroalkyl substances (PFAS)
- polychlorinated biphenyls (PCBs)
- nutritional factors (e.g., folic acid, vitamin D)
- flame retardants (e.g., polybrominated diphenyl ethers – PBDEs)
- maternal metabolic conditions (e.g., obesity, diabetes) and
- volatile organic compounds (VOCs).
To date, CHARGE has generated 144 peer reviewed publications. But I recently discovered that none of their studies controls for vaccines (vaccinated vs. unvaccinated, number of vaccines, timing of vaccines, etc.) as a possible confounding factor — even though in many cases that information is available to them. The failure to control for vaccine exposures renders all of the CHARGE studies unreliable.
To be clear, all of the toxicants they study are a problem, can likely cause autism, and should be better regulated or banned. What I’m saying though is that one cannot measure the relative impact of each of these chemicals without including a variable for the potentially confounding effect of vaccines.
So for example, a brilliant CHARGE study, Shelton et al. (2014) found that mothers who lived within 1.5 km (less than 1 mile) of agricultural fields sprayed with various pesticides had elevated risks of autism in their offspring. But who is most likely to live that close to the fields? Farmworkers and other low income residents. So it is also possible that the children born to women who live closest to agricultural fields get lower quality vaccines through the Vaccines for Children Program and this explains the higher autism risk. Or perhaps these children were not vaccinated at all and the increased autism risk is entirely from pesticides. But we’ll never know the relative risk of each factor because Shelton et al. (2014) did not control for vaccination status.
Or take another example. Lots of CHARGE studies claim that supplementation with folic acid during the first month of pregnancy reduces autism risk. But vaccines and other toxicants can cause dysregulated folate metabolism. And for some of these women, supplementing with folic acid increases autism risk in their offspring because their bodies cannot convert folic acid to folate (see Raghavan et al. 2018). By failing to control for the number of vaccines taken by the mother before and during pregnancy we are unable to unravel the relative effects of genetic mutations, vitamin supplementation, vaccines, and pesticides.
Why would some of the best epidemiologists in the world spend so much time, money, and effort and then make a mistake this basic? The answer is pretty straightforward — the field of autism research is so polarized and politicized that everyone involved with these studies knows that if they include vaccines as a variable they would instantly lose all of their research funding and be blacklisted from future research funding. That one, principled, and scientifically necessary decision would immediately and permanently end their careers. So they avoid the variable that shall not be named even though this omission renders all of their work unreliable.
I would just add that all of these mainstream autism causation studies fail in a similar way — they engage in circular reasoning (the logical fallacy in which the premise of an argument assumes the conclusion to be true).
- The vaccine studies assume that vaccines are safe and effective so they never bother with a proper placebo group that might prove otherwise.
- The gene studies assume that genes are the cause so they just gather trillions of data points until they can find a spurious association (the gene studies don’t control for vaccination status even though the possible mutagenic effects of vaccine ingredients on DNA is an ongoing concern).
- And the epigenetic studies assume vaccines could not be a factor so they don’t control for them (in spite of the fact that some of the toxicants they are studying in the environment are the same toxicants being injected directly into children’s bodies).
CHARGE (and other epigenetic studies that I describe below) are following standard practice in epidemiology that typically does not consider vaccination status a confounding variable in examining environmental risk factors for autism. But that’s precisely the problem — standard practice in each of these research fields assumes away the question of vaccines rather than studying it.
The political economy of autism causation research is such that these scholars will likely never fully understand the autism epidemic because they are prohibited from stepping outside the constraints of circular reasoning (not because they are bad people per se but because assuming away politically explosive problems is how these professions survive in the face of overwhelming corporate power).
In 2006, the UC Davis MIND Institute launched the Markers of Autism Risk in Babies – Learning Early Signs (MARBLES) study. MARBLES is a prospective longitudinal study for pregnant women who already have a biological child with autism. Information about each participant’s genetics and environment is collected through a number of sources, including:
- Blood, urine, hair, saliva, and breast milk, as well as through home dust samples, in order to obtain a comprehensive picture of the environment surrounding each pregnancy.
- They also conduct interviews with the mother and access medical records in order to uncover more information about any behavioral aspects or trends that may contribute to the development of autism.
- Mothers maintain detailed diaries tracking health symptoms, diet, and product use during and after pregnancy.
- They also conduct standardized assessments of the child’s development up to 36 months of age.
See more here Substack
Please Donate Below To Support Our Ongoing Work To Defend The Scientific Method
PRINCIPIA SCIENTIFIC INTERNATIONAL, legally registered in the UK as a company incorporated for charitable purposes. Head Office: 27 Old Gloucester Street, London WC1N 3AX.
Trackback from your site.