The ‘CDC Whistleblower’ and Vaccines: What the 2004 MMR Study Found, the Retracted Reanalysis, and the 2025 CDC Website Change

In 2014, a CDC scientist named William Thompson issued a statement saying that he regretted that he and his co-authors had omitted statistically significant information from a 2004 study on the timing of the measles, mumps, and rubella (MMR) vaccine and autism. Activists described him as a “CDC whistleblower,” and claims spread that documents showing a vaccine–autism link had been “thrown in the garbage.” The story became the basis of a 2016 documentary and remains central to debates about vaccines more than a decade later.

The debate took on new urgency in November 2025, when the CDC revised its own webpage on autism and vaccines in a way that medical organizations sharply criticized. This article explains what the 2004 study found, what Thompson said, why the most prominent reanalysis was retracted, what larger studies have shown since, and what changed at the CDC in 2025.

Was there a “CDC whistleblower” who proved vaccines cause autism? No. In 2014 a CDC scientist, William Thompson, said a 2004 study should have reported a subgroup finding differently, which anti-vaccine activists promoted as a cover-up. A reanalysis claiming a vaccine-autism link was later retracted for serious flaws. Many large, independent studies of over a million children have found no link between vaccines and autism. The scientific consensus is clear.

What Was the 2004 CDC Vaccine Study?

The 2004 study examined whether the MMR vaccine was linked to autism and found no overall association, one of many studies supporting vaccine safety. In the early 2000s, after Andrew Wakefield’s now-retracted 1998 paper suggested a link between the MMR vaccine and autism, researchers looked for evidence. One question was whether children who received MMR earlier were more likely to be diagnosed with autism. The CDC study by Frank DeStefano and colleagues, published in Pediatrics in February 2004, compared children with autism in metropolitan Atlanta with matched children without autism from the same schools.

The study found that the overall distribution of ages at MMR vaccination was similar in both groups: most children with autism (70.5 percent) and most controls (67.5 percent) were vaccinated between 12 and 17 months. There were no significant associations for vaccination before 18 or 24 months, including among children with developmental regression. Slightly more children with autism than controls had been vaccinated before 36 months (93.4 versus 90.6 percent), an odds ratio of 1.49 in the full sample. But in the subsample of children with Georgia birth certificate data, which allowed adjustment for more factors, the odds ratio was 1.23 and not statistically significant. The authors suggested the 36-month difference likely reflected vaccination requirements for children entering special education or early intervention programs, which children with developmental concerns often attend.

The study had limitations, including its case-control design and the fact that it only addressed timing, not whether vaccination itself was associated with autism. But its main finding was consistent with other research at the time.

What Did the “CDC Whistleblower” Actually Say?

In 2014, CDC scientist William Thompson said the 2004 study should have reported a subgroup analysis differently; he did not claim vaccines cause autism, and he has continued to support vaccination. William Thompson, a CDC psychologist and one of the 2004 study’s co-authors, began speaking privately in 2013 and 2014 with Brian Hooker, a biochemist and parent who believed vaccines caused his son’s autism. Some of their conversations were recorded without Thompson’s knowledge. In August 2014, Thompson released a public statement through his lawyers saying he regretted that he and his co-authors had omitted statistically significant information from the 2004 paper, specifically an analysis showing an association between earlier MMR vaccination and autism among African American boys in the full sample.

Thompson’s statement also said he believed vaccines had saved millions of lives and that he would never suggest any parent avoid vaccinating children. He did not claim that vaccines cause autism. Thompson provided documents to Congress, and his allegations have been repeated by vaccine critics, including Robert F. Kennedy Jr., now Secretary of Health and Human Services, who has described the episode as evidence that the CDC hid data.

The CDC responded at the time that the decision to rely on the birth certificate sample for race-specific analyses was made because race information and other covariates were more complete in that subsample, and that the overall conclusion was unchanged. Whether the subgroup analysis should have been reported remains a matter of dispute about research practice; the larger scientific question, whether MMR causes autism, has been addressed by much bigger studies.

What Happened to the Reanalysis Claiming a Link?

A 2014 reanalysis claiming a vaccine-autism link in a subgroup was retracted by the journal over undeclared conflicts of interest and serious methodological problems. Using data obtained from the CDC, Brian Hooker published a reanalysis in August 2014 in the journal Translational Neurodegeneration, reporting an increased risk of autism among African American boys vaccinated before 36 months. The paper was quickly criticized for its methods: it analyzed the case-control data as if it were a cohort study, did not account for the matching design, and focused on a small subgroup. In October 2014, the journal’s publisher retracted the article, citing undeclared competing interests and concerns about the validity of the statistical methods.

Subgroup findings in case-control studies are prone to chance results, particularly when many comparisons are made. That is why scientists look for confirmation in larger, independent datasets designed to answer the question directly.

Do Vaccines Cause Autism?

No. Numerous large, well-designed studies involving more than a million children across several countries have found no link between vaccines, including MMR, and autism. Since 2004, researchers in several countries have studied MMR and autism in populations far larger than the Atlanta study.

A 2014 meta-analysis in Vaccine by Luke Taylor and colleagues pooled cohort and case-control studies involving more than 1.2 million children and found no relationship between vaccination and autism, including for MMR specifically.

A 2015 study in JAMA by Anjali Jain and colleagues examined about 95,000 children with older siblings, including children whose older siblings had autism and who are therefore at higher genetic risk. MMR vaccination was not associated with increased autism risk, even among these high-risk children.

A 2019 Danish nationwide cohort study in the Annals of Internal Medicine by Anders Hviid and colleagues followed 657,461 children and found that MMR vaccination did not increase autism risk, did not trigger autism in susceptible children, and was not associated with clustering of autism cases after vaccination.

A 2021 Cochrane review of vaccines for measles, mumps, rubella, and varicella in children found no evidence of an association between MMR vaccination and autism.

The November 2025 CDC Website Change

On November 19, 2025, the CDC updated its webpage on autism and vaccines. As Contemporary Pediatrics reported, the page now states that the claim that vaccines do not cause autism is not evidence-based because studies have not ruled out the possibility that infant vaccines cause autism, and that this claim had historically been disseminated by federal health agencies to prevent vaccine hesitancy. The page kept its header reading “Vaccines do not cause autism.”

According to Columbia Law School’s Sabin Center, the CDC also said HHS had launched a comprehensive assessment of the causes of autism, including investigations of possible causal links. The American Academy of Pediatrics responded that the website had been changed to promote false information and urged the CDC to stop amplifying claims that sow doubt about routine immunizations. Public health experts noted that no new evidence was cited for the change, and that the scientific consensus, based on the studies described above, had not changed.

The change came amid broader shifts in federal vaccine policy under Secretary Kennedy, including the replacement of members of the CDC’s vaccine advisory committee in 2025, and during the largest U.S. measles outbreaks in decades. Our article on MMR vaccine safety and the Vaccine Injury Compensation Program describes the measles resurgence and how vaccine injury claims are handled.

Claim	What the record shows
“The CDC hid proof that MMR causes autism”	A subgroup result in the full sample was not reported; it was weaker in the subsample with better data. Thompson did not say vaccines cause autism.
“A reanalysis proved the risk”	The 2014 reanalysis was retracted for undeclared competing interests and statistical concerns.
“No one has studied this properly”	Studies involving more than a million children, including high-risk siblings, have found no MMR–autism link.
“The CDC now admits vaccines may cause autism”	The 2025 website change was made without citing new evidence and was criticized by major medical organizations.

Why Subgroup Results Need Caution

The heart of the Thompson controversy is a subgroup analysis, and understanding why scientists treat subgroups carefully helps explain the debate. When researchers divide a study population into many groups, by sex, race, age, and other factors, and test each for an association, some groups will show statistically significant results by chance alone. With a conventional significance threshold, roughly one in twenty comparisons can appear significant even when there is no real effect. That is why researchers typically plan subgroup analyses in advance, adjust for multiple comparisons, and look for confirmation in independent data before drawing conclusions.

In the 2004 study, the African American boys subgroup was small, the association appeared only at one age cutoff, and it was weaker in the subsample with more complete data. Even if the result had been published, it would have needed replication, and the large studies conducted since have not found an MMR–autism association in any group. Critics are right that transparency about all planned analyses is important, a principle that many journals now enforce through study registration and data-sharing requirements.

How Is Vaccine Safety Monitored?

Vaccine safety is tracked through multiple systems, including VAERS, the Vaccine Safety Datalink, and international monitoring, which continuously watch for rare adverse effects. The United States has several systems for monitoring vaccine safety. The Vaccine Adverse Event Reporting System (VAERS) collects reports from anyone, including patients and clinicians; it can detect unexpected patterns but cannot prove that a vaccine caused an event, because reports are unverified and include coincidental illnesses. The Vaccine Safety Datalink links health records from integrated health systems, allowing researchers to compare vaccinated and unvaccinated people. The Clinical Immunization Safety Assessment Project provides expert consultation on complex cases. The National Vaccine Injury Compensation Program provides no-fault compensation for certain injuries. These systems have identified real but rare side effects in the past, such as an increased risk of intussusception with an early rotavirus vaccine, which led to its withdrawal in 1999.

The Documentary and the Debate

The Thompson story reached a wider audience through the 2016 documentary Vaxxed, directed by Andrew Wakefield, the former physician whose 1998 paper on MMR and autism was retracted and who lost his medical license in the United Kingdom. The film was withdrawn from the Tribeca Film Festival after criticism but was later released independently. Thompson himself did not participate in the film. The documentary helped turn a dispute about research reporting into a broader narrative of cover-up, which continues to circulate on social media.

Measles Returns

Debates over vaccine safety have real consequences when vaccination rates fall. Measles, declared eliminated from the United States in 2000, has returned in large outbreaks. The United States recorded more than 2,000 measles cases in 2025, the most in decades, along with deaths, mostly among unvaccinated children. Measles can cause pneumonia, brain swelling, and death, and it can weaken the immune system for years after infection. Two doses of MMR vaccine are about 97 percent effective at preventing measles.

What Research Shows About the Causes of Autism

Autism is a developmental condition with strong genetic contributions. A 2019 study in JAMA Psychiatry by Dan Bai and colleagues, using data on more than two million people from five countries, estimated the heritability of autism at about 80 percent, meaning most of the variation in autism risk in those populations was explained by genetic factors. Researchers have identified many genes associated with autism, and studies of brain development suggest that differences often begin before birth. Prenatal factors, such as advanced parental age, preterm birth, and certain prenatal exposures, have also been associated with risk.

The rise in autism diagnoses over recent decades has been driven in large part by broader diagnostic criteria, greater awareness, improved screening, and expanded services, though researchers continue to investigate whether true prevalence has also changed. Because signs of autism often become noticeable around the age when children receive routine vaccines, parents may understandably link the two in time, but timing alone does not show causation, and the large studies described above were designed precisely to test that question.

Why the Story Persists

The “CDC whistleblower” story has endured because it contains real elements: a government scientist did express regret about an unreported analysis, and research practices around subgroup reporting are legitimately debated. Distrust of institutions, the emotional weight of a child’s diagnosis, and a documentary that presented the story dramatically all amplified it. Addressing such concerns respectfully, by explaining what the data show and acknowledging legitimate questions about transparency, is more effective than dismissing them.

Questions Parents Can Ask

Parents with concerns about vaccines can ask their pediatrician about the specific evidence on each vaccine’s safety, what side effects are common and rare, how vaccine safety is monitored, and what the risks of the diseases themselves are. The federal Vaccine Injury Compensation Program exists for rare cases of vaccine injury. Discussing concerns openly with a trusted clinician, rather than relying on social media, helps families make informed decisions.

Thimerosal and Other Vaccine Concerns

The MMR controversy is one of several vaccine–autism hypotheses that researchers have investigated. Another concerned thimerosal, a mercury-containing preservative once used in several childhood vaccines. As a precaution, thimerosal was removed from or reduced to trace amounts in routine U.S. childhood vaccines by 2001. Autism diagnoses continued to rise afterward, and studies comparing children with different thimerosal exposures did not find links to autism. Questions about the number and timing of vaccines in the childhood schedule have also been studied, including a 2013 study that found no association between cumulative exposure to vaccine antigens and autism. Each of these questions has been investigated with large datasets, which is why scientific organizations consider the vaccine–autism question to have been extensively studied.

What Would Change the Scientific Consensus

Scientific consensus can change when new, high-quality evidence emerges. A credible challenge to current conclusions would need large, well-designed studies that account for confounding factors, preferably replicated in independent populations, and published with transparent methods and data. Statements of possibility, such as that studies have not ruled out every conceivable effect, are not themselves evidence, because no study can prove a complete absence of risk for every outcome. HHS has said it is conducting a comprehensive assessment of autism causes; if that work produces new data, it should be evaluated by the same standards applied to past studies.

Talking With Family Members Who Have Concerns

Many people encounter relatives or friends who worry about vaccines after hearing stories like the CDC whistleblower claims. Research on vaccine conversations suggests that listening to concerns, acknowledging uncertainty where it exists, sharing personal reasons for vaccinating, and pointing to trusted clinicians tends to work better than arguing or dismissing fears. Explaining the difference between an unreported subgroup analysis and evidence of harm, and pointing to the large studies that followed, can help address specific concerns.

Key Terms

Case-control study: A study comparing people with a condition (cases) to people without it (controls) to look for differences in past exposures.

Cohort study: A study following groups of people over time to compare outcomes by exposure.

Subgroup analysis: An analysis of a subset of study participants, prone to chance findings if not planned and confirmed.

Retraction: Withdrawal of a published paper by a journal because of errors, misconduct, or other serious problems.

Heritability: The proportion of variation in a trait within a population attributed to genetic differences.

The Bottom Line

The “CDC whistleblower” story began with a real dispute about whether a subgroup result from a 2004 study should have been reported. It did not show that MMR causes autism, and the reanalysis claiming it did was retracted. Studies of more than a million children since then have found no link. The CDC’s November 2025 website change did not cite new evidence and was criticized by major medical organizations. Parents with questions deserve honest answers, and the weight of evidence continues to show that the MMR vaccine does not cause autism while protecting children from measles, mumps, and rubella.

Lessons About Research Transparency

Whatever one concludes about the 2004 study, the controversy illustrates why transparency in research matters. Practices that are now more common, such as registering study protocols and analysis plans before data are collected, publishing all prespecified analyses, sharing de-identified data for independent reanalysis, and disclosing conflicts of interest for all authors, including critics, help prevent suspicion that results were selectively reported. Public health agencies can strengthen trust by making their research data and decisions as open as privacy laws allow. Transparency serves both scientists and the public, because it allows claims of hidden findings to be checked directly.

The Human Side of the Debate

Behind the controversy are families seeking answers about their children’s development. Parents of autistic children often notice differences around the second year of life, when several vaccines are given, and some have described regression after a period of typical development. Their experiences are real, and the desire to understand why is understandable. Research has shown that early signs of autism are often present before parents notice them, and studies comparing vaccinated and unvaccinated children have not found that vaccination changes autism risk. Supporting autistic people and their families, through early diagnosis, services, education, and research into causes and support, deserves attention independent of the vaccine debate.

Why Independent Replication Matters

One of the most important features of the evidence on MMR and autism is that it comes from many independent research groups in different countries, using different data sources and methods. Studies from Denmark, Finland, the United Kingdom, the United States, Japan, and elsewhere have examined the question, some funded by governments and others by independent institutions. A cover-up at one agency could not explain consistent results across so many independent investigations. This is a key reason scientists remain confident in the conclusion, regardless of disputes about how a single study’s subgroup analysis was reported.

Common Misconceptions

“Thompson said vaccines cause autism.” His statement concerned omitted data and explicitly supported vaccination.

“The CDC destroyed all the evidence.” Thompson retained copies and gave documents to Congress; the data were reanalyzed by others.

“One study settled the question.” The conclusion rests on many independent studies across countries.

“Changing a website changes the science.” Scientific conclusions depend on evidence, not agency wording.

Key Facts at a Glance

The 2004 CDC study compared 624 children with autism to 1,824 controls and found similar ages at MMR vaccination. In 2014, co-author William Thompson said a significant subgroup result had been omitted, while affirming support for vaccination. A 2014 reanalysis was retracted for undeclared competing interests and statistical concerns. Studies including more than 1.2 million children in a meta-analysis, 657,461 children in Denmark, and about 95,000 children with older siblings found no MMR–autism association. In November 2025, the CDC changed its autism and vaccines webpage without citing new evidence, prompting criticism from the American Academy of Pediatrics.

A Balanced View

It is possible to take questions about research transparency seriously while also recognizing what the evidence shows. Scientists and agencies should report analyses fully, share data, and respond openly to questions. At the same time, the claim that MMR causes autism has been tested repeatedly in large, independent studies and not supported. Holding both ideas together, accountability for how research is conducted and confidence in what a large body of evidence shows, offers the most accurate understanding of the CDC whistleblower story.

Readers who encounter new claims about vaccines and autism can apply the same approach used in this article: identify the original source, check whether it has been peer reviewed or retracted, look at the size and design of the studies involved, and see how independent experts and professional organizations have responded.

Pediatricians are accustomed to discussing vaccine questions and generally welcome them. Bringing specific concerns, such as articles or videos that raised doubts, to an appointment allows a focused conversation about the evidence and about each family’s circumstances.

Clinicians can also explain how the recommended schedule is designed, why certain vaccines are given at specific ages, and what options exist for families with particular medical circumstances.

Asking questions is a normal part of informed decision-making, and good clinicians treat it that way.

Reliable Sources for Parents

Parents seeking information about vaccines can consult their pediatrician or family physician, the American Academy of Pediatrics, the Children’s Hospital of Philadelphia Vaccine Education Center, and peer-reviewed research summarized by organizations such as Cochrane. When government agency websites change in ways disputed by professional organizations, consulting multiple independent sources and discussing questions with a trusted clinician is especially important.

Did a 2017 Study Link Vaccines to Depression, Anxiety, and ADHD?

A study is often cited this way, but its findings were limited. A 2017 pilot case-control study in Frontiers in Psychiatry by Douglas Leslie and colleagues used insurance claims to compare children aged 6 to 15 newly diagnosed with conditions such as obsessive-compulsive disorder, anorexia nervosa, anxiety, tic disorders, ADHD, depression, and bipolar disorder with matched controls. Children newly diagnosed with anorexia were more likely to have had a vaccination in the previous three months, and flu vaccines were associated with new diagnoses of anorexia, OCD, and anxiety. The authors stated that the findings did not prove a causal role, noted the modest size of the associations compared with the clear benefits of vaccines, and encouraged families to keep vaccination schedules. Vaccine visits are also opportunities for doctors to notice and diagnose other conditions, which can create timing associations without cause.

Frequently Asked Questions

Was there really a “CDC whistleblower” about vaccines and autism?

In 2014, CDC scientist William Thompson said a 2004 study he co-authored should have reported one subgroup analysis differently. Anti-vaccine activists promoted this as proof of a cover-up, but Thompson did not say vaccines cause autism and has continued to support vaccination. His comments were about how to report data, not evidence that vaccines are unsafe.

Do vaccines cause autism?

No. This has been studied extensively in large, well-designed research involving more than a million children across multiple countries, and none has found a link between vaccines, including the MMR vaccine, and autism. The scientific and medical consensus is clear that vaccines do not cause autism, and the original claim behind this fear was based on fraudulent research.

What happened to the study that claimed a vaccine-autism link?

A 2014 reanalysis of the 2004 data, which claimed a link in a subgroup of children, was retracted by the journal that published it. The retraction cited undeclared conflicts of interest and serious methodological problems. Subgroup findings like the one it emphasized are statistically unreliable and prone to false positives, which is why they require great caution.

Why are subgroup results in studies treated cautiously?

When researchers slice data into many small subgroups, some differences appear by chance alone, producing false positives. A result in one subgroup that is not seen in the overall study, or replicated elsewhere, is considered weak evidence. This is why scientists rely on large studies and consistent findings across many studies rather than isolated subgroup results.

How do we know vaccines are safe?

Vaccines are tested in clinical trials before approval and then continuously monitored through systems such as the Vaccine Adverse Event Reporting System (VAERS), the Vaccine Safety Datalink, and international surveillance. These systems can detect even rare side effects. Decades of data from hundreds of millions of people support the safety of routine vaccines, with serious reactions being very rare.

References

1. DeStefano F, Bhasin TK, Thompson WW, Yeargin-Allsopp M, Boyle C. Age at first measles-mumps-rubella vaccination in children with autism and school-matched control subjects: a population-based study in metropolitan Atlanta. Pediatrics. 2004;113(2):259–266. PMID 14754936
2. Hooker BS. Measles-mumps-rubella vaccination timing and autism among young African American boys: a reanalysis of CDC data. Translational Neurodegeneration. 2014;3:16. PMID 25114790 (retracted)
3. Retraction note. Translational Neurodegeneration. 2014;3:22. PMID 25285211
4. Taylor LE, Swerdfeger AL, Eslick GD. Vaccines are not associated with autism: an evidence-based meta-analysis of case-control and cohort studies. Vaccine. 2014;32(29):3623–3629. PMID 24814559
5. Jain A, Marshall J, Buikema A, et al. Autism occurrence by MMR vaccine status among US children with older siblings with and without autism. JAMA. 2015;313(15):1534–1540. PMID 25898051
6. Hviid A, Hansen JV, Frisch M, Melbye M. Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study. Annals of Internal Medicine. 2019;170(8):513–520. PMID 30831578
7. Di Pietrantonj C, Rivetti A, Marchione P, et al. Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database of Systematic Reviews. 2021;11:CD004407. PMID 34806766
8. Bai D, Yip BHK, Windham GC, et al. Association of Genetic and Environmental Factors With Autism in a 5-Country Cohort. JAMA Psychiatry. 2019;76(10):1035–1043. PMID 31314057
9. Contemporary Pediatrics. CDC updates website, opens door for autism and vaccines link. November 20, 2025. contemporarypediatrics.com
10. Sabin Center for Climate Change Law, Columbia Law School. CDC Adds Debunked Link Between Autism and Vaccines to Website. climate.law.columbia.edu

Last updated: October 4, 2026