• home
• about
• new
• essays
• AI
• book

Frank Visser, graduated as a psychologist of culture and religion, founded IntegralWorld in 1997. He worked as production manager for various publishing houses and as service manager for various internet companies and lives in Amsterdam. Books: Ken Wilber: Thought as Passion (SUNY, 2003), and The Corona Conspiracy: Combatting Disinformation about the Coronavirus (Kindle, 2020).

Do Vaccines Really Cause Autism?

Two Substack Papers by Steve Kirsch Debunked

Frank Visser / Grok

"Two pieces of evidence that together show that vaccines cause autism"

Overview of the Paper

Steve Kirsch's article, published in April 2025 on his Substack, claims to provide "two pieces of evidence" proving that vaccines cause autism spectrum disorder (ASD). The core argument is that these data points demonstrate a causal link, shifting the burden of proof to the medical community to identify confounders. Kirsch asserts that AI analysis (via ChatGPT) supports this conclusion and criticizes the CDC and medical establishment for alleged cover-ups, including references to whistleblower William Thompson.

Key Claims

Parent Survey of 12,000 Respondents: Kirsch cites self-reported data from parents on the number of vaccines received and diseases in their children. He claims it shows a dose-response relationship: more vaccines correlate with higher autism rates. A graph (generated by ChatGPT) purportedly illustrates this linear trend. Data source: A publicly available Excel file on GitHub.

Pediatrician Doug Hulstedt's Cases: Hulstedt allegedly observed 44 instances of "rapid onset autism" in his practice, all occurring within 7 days of a vaccine. Kirsch calculates this as statistically impossible (probability of 1 in 1.7 x 10^13 by chance alone), implying causation.

Kirsch concludes that these findings make it "more likely than not" that vaccines cause autism and calls for replication by funded researchers.

Debunking the Claims

This paper relies on anecdotal, unverified, and methodologically flawed data, ignoring the overwhelming body of rigorous scientific evidence. Below, I address each claim specifically, followed by the broader scientific context. These critiques align with analyses from fact-checking sites, public health experts, and peer-reviewed literature, which have repeatedly dismantled similar anti-vaccine narratives.

1. The Parent Survey: Flawed Self-Reporting and Confounding Factors

Methodological Issues: The survey is voluntary and self-selected, likely drawing from anti-vaccine communities (e.g., promoted on platforms like Substack and X). This introduces selection bias—parents with vaccine-skeptical views or children with autism are overrepresented, skewing results. There's no verification of diagnoses or vaccine histories; parents self-report, which is prone to recall bias (e.g., misremembering doses or timelines). No control group, randomization, or blinding exists, making it unreliable for causation.

Confounders Ignored: Autism diagnoses often occur around ages 2-3, coinciding with routine vaccine schedules. Children receiving more vaccines may attend more medical visits, leading to earlier/better screening and higher reported rates—not causation. Socioeconomic factors (e.g., access to healthcare) also correlate with both vaccination rates and autism diagnoses. Kirsch dismisses confounders without evidence, but real-world studies control for these and find no link.

Statistical Manipulation: The "dose-response" graph is AI-generated from unadjusted data, not peer-reviewed analysis. Proper regression would account for variables like genetics (which explain 80-90% of ASD risk). Similar surveys by anti-vaccine groups (e.g., Hooker et al.) have been retracted or criticized for p-hacking.

Rebuttals: Science-Based Medicine and Snopes have critiqued Kirsch's prior surveys as "junk science" for these reasons. No independent replication exists because ethical researchers prioritize controlled studies over biased polls.

2. Hulstedt's 44 Cases: Anecdote Masquerading as Evidence

Lack of Verification: These are unverified case reports from one pediatrician, without medical records, peer review, or publication in a journal. "Rapid onset autism" isn't a recognized medical term; ASD symptoms emerge gradually, often retrospectively linked to vaccines due to temporal proximity (post hoc ergo propter hoc fallacy). Hulstedt's practice size, patient demographics, and full context are undisclosed, preventing scrutiny.

Statistical Fallacy: The 1-in-1.7e13 probability assumes random distribution and ignores base rates. Autism prevalence is ~1 in 36 (CDC data); in a busy practice vaccinating hundreds of children, 44 temporal associations aren't improbable. This is like claiming lottery wins "prove" a conspiracy—correlation isn't causation. Large epidemiological studies show no such clustering post-vaccination.

Cherry-Picking: Kirsch omits cases where autism onset predates or follows vaccines by months/years. Similar "temporal cluster" claims (e.g., from the discredited Wakefield study) have been debunked as coincidence.

Rebuttals: Pediatric experts, including those from the American Academy of Pediatrics, label such anecdotes as "dangerous misinformation." No broader data from Hulstedt's practice or similar clinicians supports this; it's isolated and unconfirmed.

Broader Context: The Myth's Origins and Scientific Consensus

Kirsch's paper recycles a debunked narrative stemming from Andrew Wakefield's 1998 fraudulent Lancet paper (retracted in 2010 for ethical violations and data falsification), which falsely linked MMR vaccine to autism in just 12 children. Despite this, the myth persists via misinformation networks.

Scientific Evidence Against a Link:

Large-Scale Studies: A 2015 JAMA study of 95,000+ children (including high-risk siblings) found no increased ASD risk from MMR. A 2019 Danish study of 657,000+ children showed vaccinated kids had lower ASD rates. Meta-analyses (e.g., Taylor et al., 2014 in Vaccine) of 1.2+ million children confirm no association.

Vaccine Ingredients: Thimerosal (mercury preservative) was removed from U.S. childhood vaccines by 2001; ASD rates rose afterward, disproving causation. Aluminum adjuvants and "vaccine overload" lack biological plausibility—infants' immune systems handle far more antigens daily from the environment.

Timing and Genetics: ASD is largely genetic (heritability >80%), with prenatal origins. Symptoms "emerge" post-vaccination due to developmental milestones, not vaccines.

Consensus Statements: CDC, WHO, AAP, IOM (2004 review), and Cochrane (2012) all state: Vaccines do not cause autism. A 2024 PMC review calls the hypothesis "effectively dismissed."

Why This Matters

Kirsch's unsubstantiated claims erode trust in vaccines, risking outbreaks (e.g., 2025 measles surges linked to hesitancy). Autism's rise stems from better diagnostics, not vaccines—prevalence was undercounted pre-1990s. Parents deserve evidence-based info: Vaccines save lives; autism research focuses on genetics and early intervention.

For reliable sources, consult CDC's vaccine safety page or the Autism Science Foundation. If you're a parent concerned about ASD, genetic counseling and screenings are far more actionable than vaccine myths.

"The link between vaccines and autism has been clear for over 25 years since the CDC did the VSD study"

Overview of the Paper

Steve Kirsch's article, titled "The link between vaccines and autism has been clear for over 25 years since the CDC did the VSD study," was published on Substack in September 2025. It reiterates and expands on Kirsch's longstanding anti-vaccine narrative, claiming definitive proof of a causal link between vaccines (especially thimerosal-containing ones) and autism spectrum disorder (ASD). The thesis: Government agencies like the CDC have suppressed evidence since 1999, and independent data (including Kirsch's surveys) confirm vaccines cause 71% or more of autism cases. He calls for accountability, replication of his work, and policy changes like mandatory vaccinated vs. unvaccinated studies.

The article recycles debunked claims from prior works (e.g., his 2023/2025 Substacks), citing historical emails, surveys, and selective studies while ignoring rigorous counter-evidence. Below, I debunk each major claim with scientific context, highlighting methodological flaws, misrepresentations, and contradictions with peer-reviewed literature.

Key Claims and Debunks

1999 CDC VSD Study by Thomas Verstraeten: Relative Risk of 11.35 for Mercury (>25 µg) at 1 Month and Autism

Kirsch's Claim: This "insanely high" risk (comparable to smoking-lung cancer) proves causation; Verstraeten spent 4 years "trying to make the signal disappear," per a 1999 email ("It just won't go away").

Debunk: The Vaccine Safety Datalink (VSD) is a robust CDC system, but Kirsch cherry-picks preliminary Phase I results from 1999, which were screening analyses, not final causal findings. Early signals for thimerosal (a mercury preservative) were confounded by factors like birth weight, socioeconomic status, and healthcare access—low-birth-weight infants get more thimerosal shots and have higher ASD risk independently. The finalized Phase II study (published 2003 in Pediatrics, Verstraeten et al.) analyzed 140,000+ children and found no significant association (adjusted relative risk ~1.0). The email refers to iterative data cleaning to address confounders, not a cover-up—standard epidemiological practice. No evidence of deliberate suppression; Verstraeten himself has publicly refuted misuse of his work.

Broader Context: Thimerosal was phased out of U.S. childhood vaccines by 2001; ASD rates continued to rise (CDC: 1 in 150 in 2000 to 1 in 36 in 2023), disproving causation. A 2010 IOM review of 12 studies rejected any link.

William Thompson/Frank DeStefano MMR Study: Evidence Destroyed to Hide Link

Kirsch's Claim: A CDC study found MMR-autism link in Black boys, but DeStefano ordered data destruction; Thompson's sealed congressional testimony (10 years ago) and "documents" prove it. Kirsch offered docs to press but was ignored.

Debunk: This stems from a 2014 misrepresentation by anti-vaccine activist Brian Hooker, who reanalyzed 2004 CDC data (Pediatrics, DeStefano et al., 657,000+ children) using flawed methods (e.g., arbitrary age cutoffs). The original study found no overall MMR-ASD link and no subgroup risks after proper adjustments. Thompson (a CDC co-author) expressed concerns about data presentation in a recorded call but did not claim causation or destruction— he clarified in 2014 that he questioned omission of a preliminary finding (a non-significant trend in a tiny subgroup: 2-5-year-old Black boys with private insurance), but full analysis showed no effect. No "sealed deposition" exists; Thompson testified voluntarily. Hooker's paper was retracted from Translational Neurodegeneration for invalid stats. Kirsch's "documents" are unverified and echo discredited narratives—no mainstream outlet pursues them due to lack of novelty or credibility.

Broader Context: The 2004 study has been validated by larger cohorts (e.g., 2019 Danish study of 657,000 children: MMR-vaccinated had 7% lower ASD risk).

Kirsch's 2023 Parent Survey (12,000+ Respondents): Odds Ratio of 3.5 for Autism with More Vaccines

Kirsch's Claim: Quick-to-design survey shows dose-response (OR=3.5 implies 71% of autism vaccine-caused); replicable via GitHub data, but ignored.

Debunk: As critiqued in my prior analysis of Kirsch's similar 2025 post, this is junk science: voluntary, self-selected from anti-vaccine circles (e.g., promoted on Substack/X), introducing severe selection and recall bias. No diagnosis verification, no controls for confounders (e.g., genetics explain 80-90% ASD risk; more-vaccinated kids get more screenings). The OR is unadjusted—proper multivariate analysis collapses. Ethical researchers avoid such polls; no peer-reviewed replication because they're unreliable. Science-Based Medicine (2023 review) called it "epidemiological malpractice."

Broader Context: Contradicts meta-analyses (e.g., 2014 Vaccine review of 1.2M+ kids: no vaccine dose-ASD link).

Control Group Study: Vaccine Harms, Including Vitamin K Shot and Autism Graphs.

Kirsch's Claim: Shows vaccination (even Vitamin K) causes harm; autism data in graphs.

Debunk: The "Control Group" is a 2023-2024 anecdotal survey by anti-vaccine group MAMA (Moms Across America), ~2,000 respondents (mostly unvaccinated). Like Kirsch's, it's biased (self-selected, no verification) and shows no causal proof—unvaccinated kids in the study had lower reported chronic issues, but this reflects healthy user bias (vaccine-hesitant parents monitor/report differently). Vitamin K shot (prevents hemorrhagic disease) has zero ASD link; a 2019 JAMA study of 1M+ kids confirmed safety. Graphs are descriptive, not adjusted—flawed like Kirsch's. No publication in reputable journals; dismissed by AAP as misinformation.

Broader Context: Vaccinated kids have fewer neurodevelopmental issues overall (2020 Annals of Internal Medicine).

Chris Exley on Aluminum: Highest Levels in Autistic Brains

Kirsch's Claim: Exley (aluminum expert) found unprecedented brain aluminum in ASD child; defunded due to pressure.

Debunk: Exley's 2017-2018 studies (J Trace Elem Med Biol) examined tiny samples (5-10 ASD brains) and reported high aluminum, but methods were criticized: small n, no controls for post-mortem contamination, and aluminum linked to all brains (not ASD-specific). No causation shown—aluminum adjuvants in vaccines are in trace amounts (safer than diet/environment). Exley resigned from Keele University in 2021 amid unrelated funding issues and misconduct probes, not a "cover-up." A 2022 Vaccine review found no aluminum-ASD link.

Broader Context: Infants ingest more aluminum from breast milk/formula than vaccines; no biological pathway to ASD.

100+ Studies Show Vaccine-Autism Link; No Vax vs. Unvax Studies Funded

Kirsch's Claim: Vaccinated vs. unvaccinated always shows unvax healthier; >100 pro-link studies vs. 30 no-link; Fauci blocked studies.

Debunk: False equivalence— "100+ studies" are low-quality (case reports, retracted papers like Wakefield's); the 30 "no-link" are gold-standard RCTs/epidemiology (e.g., IOM 2004: rejects causation). Vax-unvax studies exist (e.g., 2020 German KiGGS: unvax had higher disease rates) but are rare due to ethics—denying vaccines risks harm (e.g., measles deaths). NIH funds safety research; a 2013 bill failed for feasibility issues, not conspiracy. Fauci's quote is cherry-picked from 2001 (pre-major studies).

Broader Context: Global consensus (WHO, CDC, 2024 Cochrane): No link. ASD rise = better diagnosis/awareness.

Why This Matters

Kirsch's post perpetuates a dangerous myth, eroding vaccine confidence amid rising ASD diagnoses (now ~1 in 36, due to expanded criteria/genetics). Real ASD progress: Early screening/intervention, genetic therapies. Vaccines prevent 4-5M deaths yearly (WHO). For facts, see CDC's vaccine safety page or Autism Speaks' science summaries. If concerned, consult a pediatrician—myths delay help.