Direct answer: No — vaccines do not cause autism. This is the most thoroughly tested claim in modern medicine, and it has failed every test. The idea traces back to a single 1998 paper about just 12 children, a paper that was retracted for fraud after its author was found to have been secretly paid £435,643 by lawyers suing vaccine makers. Since then, studies covering well over a million children — including one that tracked every child born in Denmark over eleven years — have found no link between vaccines and autism, no matter how the question is asked: MMR, thimerosal, aluminium, the number of shots, or children with autistic siblings. In December 2025 the World Health Organization's expert committee reviewed 31 more studies and reaffirmed the same conclusion for the fourth time. The debate survives anyway, because the timing looks convincing — the MMR shot is given at 12 to 15 months, exactly the age autism signs first appear — and because it touches the deepest fear a parent can have. The rest of this article shows both camps, fairly, with the numbers, and explains why the coincidence fools so many smart people.
Why this fight is louder in 2025 than ever
Most questions in this series have settled into a quiet consensus with a noisy fringe. This one has gone the other way: it is louder now than it was five years ago, and the noise is now coming from inside the institutions that used to settle it.
First, measles is back. The United States declared measles eliminated in 2000 — eliminated, not eradicated, meaning the virus no longer circulated inside the country. In 2025 that achievement unravelled. More than 1,300 measles cases were reported across more than 40 jurisdictions, with three confirmed deaths and 164 hospitalisations, and over 90% of cases in people who were unvaccinated or whose status was unknown. It became the third most active measles year since the vaccine was developed. Measles is not a mild rash: before vaccination, it killed millions of children a year worldwide, and it can cause pneumonia, brain swelling and a fatal brain disorder that appears years after the infection.
Second, the sceptics now hold high office. Robert F. Kennedy Jr., who built much of his public profile on claims linking vaccines to autism — including the thimerosal theory — now leads the US Department of Health and Human Services. In late 2025 the CDC's own vaccine-and-autism web page was quietly edited so that it no longer stated flatly that vaccines do not cause autism; the page now says the evidence does not establish that conclusion, without citing any new study. Fact-checkers and medical societies pointed out immediately that no new evidence had been presented — the wording changed, the science did not. Weeks later, in December 2025, the World Health Organization's Global Advisory Committee on Vaccine Safety reviewed 31 studies from 2010 to 2025 and reaffirmed, for the fourth time since 2002, that vaccines — including those containing thiomersal or aluminium — do not cause autism.
Third, the autism numbers themselves keep climbing, and the climb is real in the data even if its meaning is disputed. The CDC's April 2025 report found that about 1 in 31 eight-year-olds — 3.2% — had been identified with autism in 2022, up from 1 in 36 in the previous report and 1 in 150 when tracking began in 2000. Parents see that curve and ask the obvious question: something changed, so what was it? The vaccine theory offers a single, simple, blameable answer. The real answer — better detection, broader diagnostic criteria, and a condition that is largely genetic — is slower, less satisfying, and harder to argue about in a comment thread.
That is the shape of this debate: one side has the evidence, the other side has the better story. This article gives you both, and then the evidence.
Where the claim came from: twelve children and a press conference
Before 1998, almost nobody connected vaccines with autism. The idea was not a folk belief or an old suspicion. It was invented, in a specific place, on a specific date, by a specific person — and the paper that invented it is one of the most famous retracted papers in the history of science.
On 28 February 1998, The Lancet, one of the world's most respected medical journals, published a paper by Andrew Wakefield and twelve co-authors from the Royal Free Hospital in London. Its title was long and technical: "Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children." The paper described twelve children, aged three to ten, who had bowel problems and developmental disorders, and it suggested that the measles-mumps-rubella (MMR) vaccine might have triggered both — a new condition Wakefield called "autistic enterocolitis."
Read the paper carefully, and the first surprise is how modest its actual claim was. It did not prove a link. It said so itself: "We did not prove an association between measles, mumps, and rubella vaccine and the syndrome described." It was a case series — twelve children, hand-picked, with no comparison group — the weakest form of medical evidence, the kind of paper that normally generates a hypothesis for someone else to test, not a public-health scare.
The second surprise is what happened at the press conference. Wakefield did not wait for anyone to test the hypothesis. At the press conference launching the paper, he said parents should use single measles, mumps and rubella shots instead of the combined MMR — a recommendation that went far beyond anything in the paper itself, and one that no health authority anywhere supported. The newspapers did the rest. Within days, the story was everywhere: a doctor, a famous journal, a vaccine, and autism — and the careful wording of the paper itself was lost entirely.
MMR uptake in Britain fell. In some parts of London it dropped below 70%, far short of the roughly 95% needed to stop measles spreading. Measles cases, which had been rare, began to climb. The scare spread to the United States, Ireland and beyond. And the paper that started it all sat, unretracted, for twelve years.
The money, the fraud, and the twelve-year wait
If the story ended with a small, weak paper that was later corrected, there would be no article to write. It did not end there — because the paper was not an honest mistake. It was, according to the investigations that followed, a fraud with a paymaster.
The unravelling began with a journalist, not a scientist. Brian Deer of The Sunday Times spent years investigating Wakefield, and what he found was published in a series of articles in The BMJ in 2011 under the title "How the case against the MMR vaccine was fixed." The findings were devastating, and they have never been successfully challenged:
The children were not who the paper said they were. The paper described the twelve children as a consecutive series of ordinary referrals. In fact, several were recruited through anti-vaccine campaigners and lawyers, and some were already claimants in a lawsuit against the vaccine manufacturers. The paper also said the children were previously normal and then regressed after MMR; the hospital records showed that in several cases the developmental problems had been documented before the vaccine was given.
The data were altered. Deer compared the published paper with the children's actual hospital records and found the results had been changed to fit the theory: normal test results reported as abnormal, timelines shifted, diagnoses rewritten. The BMJ concluded the paper was "an elaborate fraud."
Wakefield was being paid by the other side. This is the detail that ends the argument about his motives. Two years before the paper was published, Wakefield had been hired by a lawyer building a lawsuit against the vaccine manufacturers, and he was paid £435,643 in fees plus £3,910 in expenses from public legal-aid money — about eight times his annual hospital salary at the time. The paper never disclosed this. He had also filed a patent application for a single measles vaccine of his own — the very product he was publicly recommending as an alternative to MMR. The children in the study were, in several cases, also his legal clients.
The procedures had no ethical approval. The children underwent invasive tests — colonoscopies, lumbar punctures, blood draws, for which Wakefield paid some children £5 each — without the proper ethical clearance. The General Medical Council, which regulates doctors in Britain, held the longest fitness-to-practise hearing in its history: 217 days spread across nearly three years. In May 2010 it found Wakefield guilty of serious professional misconduct, including four counts of dishonesty, and struck him off the medical register. He has never practised medicine in Britain since.
The Lancet fully retracted the paper in February 2010, with its editor calling it "utterly false" and saying the journal had been deceived. Ten of the twelve co-authors had already retracted the paper's interpretation back in 2004.
And here is the detail that explains why the scare outlived the paper: the retraction came twelve years after publication. In those twelve years, MMR uptake fell, measles returned, and millions of parents made a decision based on a study that was already falling apart. A retraction is a quiet correction in a journal. A press conference is a global event. The two were never going to reach the same number of people.
The studies that answered it: more than a million children
Here is the part of the story the debate usually skips: the Wakefield hypothesis was not ignored. It was tested — more thoroughly than almost any medical claim in history — by independent teams in several countries, using different methods, over two decades. Every test came back the same way.
Denmark, 2002 — 537,303 children. The first great test came from Denmark, where researchers tracked every child born between 1991 and 1998. Children who received MMR had the same rate of autism as children who did not: a relative risk of 0.92 for autistic disorder and 0.83 for other autism-spectrum disorders — in other words, if anything, slightly lower rates among the vaccinated, and statistically indistinguishable from no difference at all. The study was published in the New England Journal of Medicine.
Meta-analysis, 2014 — 1.2 million children. Taylor, Swerdfeger and Eslick pooled ten studies — five cohort studies and five case-control studies — covering more than 1.2 million children, and tested three separate claims at once: the MMR vaccine, thimerosal, and mercury exposure. None of them was associated with autism. Not as a trend, not in a subgroup, not in any analysis they ran. This is the study behind the "1.2 million children" figure, and it remains the most cited single answer to the question.
United States, 2015 — 95,727 children. Jain and colleagues studied American children with private health insurance, including the group the theory predicts should be most at risk: younger siblings of children with autism, who carry a much higher genetic risk themselves. MMR vaccination was not associated with autism in any group — not in the general population, and not in the high-risk siblings.
Denmark, 2019 — 657,461 children. The largest single study of MMR and autism ever done. Hviid and colleagues followed every child born in Denmark from 1999 to 2010, using the country's population registries to link vaccination records with autism diagnoses. Over five million person-years of follow-up, 6,517 children were diagnosed with autism. Comparing vaccinated with unvaccinated children, the fully adjusted hazard ratio was 0.93 — no increased risk — and the same result held in every subgroup they tested, including children with autistic siblings and children with other autism risk factors. The lead author summed it up in one sentence to NPR: "MMR does not cause autism."
Cochrane, 2012 — 14.7 million children. The Cochrane Library's systematic review pooled dozens of studies covering about 14.7 million children and found no credible evidence of any link between MMR and autism or Crohn's disease.
The "too many, too soon" test, 2013. When the MMR theory failed, the claim shifted: perhaps it is not one vaccine but the sheer number of shots in the first two years of life. DeStefano and colleagues tested that directly, measuring the total number of vaccine antigens children received in their first two years and comparing children with autism against children without it. There was no relationship. Children who received more antigens were no more likely to develop autism than children who received fewer.
Put together, this is what "the evidence" actually means in this debate: not one study, not one institution, not one country — but dozens of independent studies, several countries, both kinds of study design, and sample sizes running into the millions, all pointing the same way, with the largest studies pointing the most clearly. That is an unusual level of agreement in medicine. It is the level of agreement you get when a claim is simply false.
| Study | Children studied | What it tested | Result |
|---|---|---|---|
| Madsen et al., Denmark, 2002 (NEJM) | 537,303 | MMR and autism | No association (relative risk 0.92) |
| Taylor et al., meta-analysis, 2014 (Vaccine) | 1,256,407 | MMR, thimerosal and mercury | No association in any analysis |
| Jain et al., United States, 2015 (JAMA) | 95,727 | MMR, including high-risk siblings | No association in any group |
| Hviid et al., Denmark, 2019 (Ann Intern Med) | 657,461 | MMR and autism, all subgroups | No increased risk (hazard ratio 0.93) |
| Cochrane review, 2012 | ~14,700,000 | MMR, autism and Crohn's disease | No credible evidence of a link |
| DeStefano et al., United States, 2013 (J Pediatr) | 1,008 | Total vaccine antigens in first 2 years | No relationship with autism |
| WHO GACVS review, December 2025 | 31 studies, 2010–2025 | Vaccines, thiomersal, aluminium | No causal link — 4th confirmation |
Thimerosal: the mercury theory, and the experiment that killed it
When the MMR theory began to collapse, the argument moved to a different ingredient: thimerosal, a mercury-based preservative used in some multi-dose vaccine vials to prevent bacterial and fungal contamination. The claim was that mercury — a known neurotoxin — was damaging children's brains.
The claim had two problems from the start. First, thimerosal contains ethylmercury, which the body clears in days, not methylmercury, the form that accumulates in fish and causes the poisoning everyone pictures. The two are as different as ethanol in a glass of wine and methanol in windscreen fluid: same family, very different behaviour. Second, the MMR vaccine — the one at the centre of the original scare — never contained thimerosal at all. The two theories were about different vaccines, and they were often argued as if they were one.
But the thimerosal theory had something the MMR theory never had: a genuine natural experiment. In 1999, as a precaution, US health authorities asked manufacturers to remove thimerosal from routine childhood vaccines, and by 2001 it had been reduced to trace amounts in all of them (a few multi-dose flu vaccines were the exception). If thimerosal caused autism, autism rates should have fallen after 2001. That is what the theory predicted, clearly and publicly.
They did the opposite. Schechter and Grether, using California's statewide developmental-services registry, tracked autism rates from 1995 to 2007 and found that prevalence rose every single quarter of the study period — including every quarter after thimerosal was gone. The same pattern held in Denmark, Sweden and the UK, which removed thimerosal on different timetables: no decline followed removal anywhere. The largest natural experiment the theory ever faced, and it failed it in every country that ran it.
The aluminium argument — that the tiny amounts of aluminium salts used as adjuvants in some vaccines cause autism — met the same fate. The WHO's December 2025 review examined the aluminium evidence specifically, including ten randomised controlled trials and seven large cohort studies, and found no association between the trace amounts in vaccines and autism. A Danish nationwide study of children born between 1997 and 2018, reviewed in the same process, found no link either.
Why it still feels true: the timing trap
None of the evidence above explains the most important thing about this debate: why do so many intelligent, careful parents remain convinced? The answer is not stupidity. It is timing — and timing, in this case, is genuinely cruel.
The MMR vaccine is given at 12 to 15 months of age. Autism's first clear signs — loss of words, loss of eye contact, withdrawal from social play — typically become visible between 12 and 18 months. Those two windows overlap almost perfectly. So every year, in every country, thousands of parents watch their child change in the weeks after a routine appointment, and the human brain does what human brains do: it connects the two events.
The connection is made stronger by the nature of autistic regression. In about a quarter to a third of autistic children, development looks typical and then skills are lost — words disappear, eye contact fades — often over a period of weeks. To a parent watching it happen, it does not look gradual. It looks sudden. And if the last memorable event before the change was a needle, the needle becomes the explanation. The fact that regression happens at the same age whether or not a child was recently vaccinated — and that it happens in unvaccinated children too — is almost impossible to see from inside one family's experience, because no family has a comparison group. They have one child, one timeline, and one story.
There is a second trap: the diagnosis itself arrives late. Autism is usually diagnosed between ages two and four, long after the MMR shot. So the parent's memory is not of a diagnosis following a vaccine — it is of a change following a vaccine, remembered years later, retold many times, and hardened by every retelling. Memory researchers call this reconstructive memory: each time we recall an event, we rebuild it, and the rebuild drifts towards the story we already believe. A parent who has spent years convinced the vaccine caused it will remember the change as sharper, closer to the appointment, and more sudden than it was.
None of this makes those parents dishonest. Their children really did change, really did lose skills, and really do need support. The error is not in what they saw. It is in what the seeing proves — and one family's timeline, however vivid, cannot prove a cause that millions of children's records disprove.
What actually causes autism
If vaccines do not cause autism, what does? The honest answer is that science does not have the full picture — but it has a much bigger piece of it than most people realise, and that piece points in a very different direction from the vaccine theory.
The strongest known factor is genetics. Twin studies consistently find autism highly heritable: when one identical twin is autistic, the other usually is too, and large meta-analyses put the heritability at roughly 80%. Hundreds of genes have been implicated, and the condition runs in families — the single biggest risk factor for an autistic child is having an autistic sibling or parent. This is also why the Danish studies checked the sibling subgroup so carefully: if vaccines triggered autism in genetically vulnerable children, that is exactly where the signal would show up. It did not.
Other confirmed contributors are prenatal: advanced parental age, certain pregnancy complications, and premature birth all raise the odds modestly. None of them is a single cause, and none of them involves a needle.
And the rising numbers? The CDC's 1-in-31 figure measures identification, not necessarily occurrence, and the agency says so itself. Since 2000, the diagnostic criteria have widened, screening has become routine, awareness has exploded, and groups that were historically missed — girls, minority children, children without intellectual disability — are now being counted. The rate varies wildly by place: 1 in 103 in Laredo, Texas, against 1 in 19 in California, in the same report, in the same year. A true environmental cause would not produce a fivefold gap between two American cities. A difference in how hard doctors look would.
None of this means the rise is entirely an artefact. Researchers are careful on this point: some of the increase may reflect a real rise in occurrence, and the environmental contributors are still being studied. But the one candidate that has been tested more than any other — vaccines — is the one that has been ruled out most completely.
At a glance: what vaccines do, and what they do not
| Claim | What the evidence says | Verdict |
|---|---|---|
| MMR causes autism | 657,461 Danish children (2019): hazard ratio 0.93; meta-analysis of 1.2m+ children: no link | False |
| Thimerosal (mercury) causes autism | Removed from routine childhood vaccines by 2001; autism rates kept rising in every country that removed it | False |
| Too many vaccines, too soon, overwhelm the immune system | 2013 study: total vaccine antigens in first 2 years had no relationship with autism | False |
| Aluminium in vaccines causes autism | WHO December 2025 review of trials and cohort studies: no association with trace aluminium | False |
| The 1998 Lancet paper proved a link | It said it did not prove one; it was retracted in 2010 after fraud was found; its author was struck off | False |
| Autism rates have risen sharply | CDC: 1 in 150 in 2000 to 1 in 31 in 2022 — but identification, criteria and awareness explain most of it | True as a count |
| Autism is largely genetic | Twin-study meta-analyses put heritability at roughly 80%; having an autistic sibling is the biggest risk factor | True |
| The timing coincidence is real | MMR at 12–15 months overlaps exactly with the age autism signs first appear — which is why the claim feels true | True — and misleading |
| Measles is harmless, so skipping the vaccine is low-risk | 2025: 1,300+ US cases, 3 deaths, 164 hospitalisations; measles can cause pneumonia, brain swelling and fatal late brain disease | False |
Claim vs evidence: the arguments, scored
| The argument you will hear | Evidence for | Evidence against | Score |
|---|---|---|---|
| "My child changed the week after the MMR" | Parents' accounts are sincere and consistent | Regression happens at the same age in unvaccinated children; millions of records show no difference in rates | Real experience, wrong cause |
| "The studies are all funded by pharma" | Drug companies do fund some research | The biggest studies were government-funded (Denmark, CDC); the WHO review covered 31 independent studies across many countries | Mostly false |
| "They removed thimerosal, so they must have known it was harmful" | It was removed as a precaution in 1999–2001 | Removal was precautionary, not a finding of harm — and autism rates rose afterwards, the opposite of the prediction | False inference |
| "The CDC changed its own page in 2025" | The CDC page wording did change in late 2025 | The edit cited no new study; medical societies objected; the WHO's December 2025 review of 31 studies reaffirmed no link weeks later | True about the edit |
| "Autism exploded — something must have caused it" | Identified prevalence rose from 1 in 150 to 1 in 31 | Criteria widened, screening and awareness rose, and rates vary fivefold between US cities — a pattern of detection, not a single cause | True rise, unproven cause |
| "Wakefield was silenced for telling the truth" | He lost his licence and his paper was retracted | He was paid £435,643 by lawsuit lawyers, the data were falsified, and the children were lawsuit claimants — the GMC found dishonesty | Debunked |
Popular opinions: what people actually say about vaccines and autism
This is the question that splits comment sections down the middle, and the split follows a pattern. Here are the positions you will actually meet online — paraphrased from real threads and debates, not invented — and how each one holds up.
"My child was fine, got the MMR, and a week later stopped talking. I know what I saw." This is the most powerful argument in the whole debate, and it is usually told with complete sincerity. Verdict: the experience is real, the conclusion does not follow. Regression at 12 to 18 months is a recognised pattern in autism, and it happens on the same schedule in children who were not recently vaccinated. With millions of children receiving MMR in exactly the months when autism first shows, thousands of coincidences are mathematically guaranteed — and each one feels, to that family, like proof.
"They say it's safe, but they also said opioids were safe and thalidomide was safe." Verdict: the distrust is earned, and the examples are real — pharmaceutical companies have hidden harms before, and regulators have failed. But this argument proves too much: it would also discredit seatbelts, because car companies have lied too. The question is not whether the industry has ever lied. It is whether this claim has been tested independently — and it has been, by government-funded teams in Denmark, by the Cochrane Collaboration, and by the WHO, across millions of children who had nothing to do with any drug company.
"Too many, too soon. Twelve shots before age two cannot be natural." Verdict: understandable, and directly tested. The 2013 study that counted every vaccine antigen children received in their first two years found no relationship between the total and autism. A child's immune system also handles far more antigens every day from ordinary life — a single cold exposes it to more immune stimulation than the entire childhood vaccine schedule. "Natural" is doing a lot of work in this argument: measles, the natural alternative, is a virus that can kill.
"My unvaccinated kids are fine, and my vaccinated nephew is autistic. That tells me everything." Verdict: it tells you about two families. Autism runs in families — the biggest risk factor is having an autistic relative — so a single family comparison is the weakest possible evidence, in either direction. The Danish studies exist precisely to replace this kind of comparison with one that covers hundreds of thousands of families at once. When you do that, the difference vanishes.
"Measles is just a rash. My parents all had it and they're fine." Verdict: survivorship bias in its purest form. Your parents survived measles because most children do — and the ones who did not are not here to argue. Before vaccination, measles killed millions of children a year worldwide, and even today it hospitalises about one in eight cases in the 2025 US outbreak. The parents who say "we all had it" are the ones whose siblings did not die of it.
"If there's no link, why did the CDC change its page in 2025?" Verdict: this is the sceptics' strongest new card, and it deserves a straight answer. The CDC's page wording changed under new political leadership, and the change cited no new study. The scientific response came weeks later from the WHO, which reviewed 31 studies from 2010 to 2025 and reaffirmed no causal link — the fourth such review since 2002. A website edit is a political act. A 31-study review is a scientific one. They are not the same kind of evidence.
"Autism was rare when I was a kid. Now every class has two autistic children. Something changed." Verdict: the observation is correct and the conclusion is premature. When you were a child, autistic children were usually labelled something else, or not labelled at all, and many were kept out of mainstream classes entirely. The criteria have widened, screening is routine, and awareness is incomparably higher. The fivefold gap between Laredo and California in the same CDC report is the giveaway: a real environmental cause does not respect city limits, but a difference in how hard doctors look does.
"I trust my paediatrician, who says vaccinate — and I trust the parent groups, who say don't. I don't know who to believe." Verdict: the most honest position in the debate, and the one this article is written for. The way out of it is not to pick a tribe but to ask what would change your mind: a single study, or a million children? A story, or a registry? The evidence for no link is not a press release. It is 657,461 Danish children, 1.2 million in a meta-analysis, 14.7 million in a Cochrane review — and a fraud at the origin of the claim, documented in the fraudster's own paperwork.
The pattern is worth noticing: the angriest voices in this debate are arguing about different things. One side is defending a child they watched change, the other is defending a mountain of data — and both are using the word "proof" to mean something the other side is not talking about. Strip away the identity, and the evidence leaves a much quieter answer: the shot and the change arrived together by coincidence of timing, and the tests that could have found a link, run on millions of children, found nothing.
What the argument costs
It is easy to treat this as an abstract debate about statistics. It is not abstract. The cost of the scare is measured in hospital beds.
When MMR uptake falls below about 95%, measles finds the gaps. Britain learned this after 1998, when uptake fell and measles — which had been rare — returned, including the first measles deaths in the country in over a decade. The United States learned it again in 2025: more than 1,300 cases, three deaths, 164 hospitalisations, and an outbreak centred on communities with low vaccination rates. Over 90% of cases were in people who were unvaccinated or whose status was unknown. Measles is one of the most contagious viruses known: in a room of unvaccinated people, one case infects roughly nine out of ten.
There is a cruel irony here. The parents who skip MMR are trying to protect their children from a risk they fear — and in doing so, they expose them to a risk that is real, documented, and currently circulating. Autism is not caused by the vaccine. Measles is caused by the absence of it.
None of this means vaccines are risk-free, and an honest article should say so plainly. Real side effects exist: sore arms, fevers, and, rarely, serious allergic reactions — about one in a million doses for MMR. Febrile seizures after MMR occur in roughly one in 3,000 to 4,000 doses in toddlers, and while frightening, they do not cause lasting harm. These risks are real, they are listed on the vaccine information sheets every parent is given, and they are dwarfed by the risks of the diseases themselves: measles hospitalises about one in five infected children in some outbreaks and kills roughly one to two in every thousand cases in high-income countries. That is the comparison that matters — not vaccine versus nothing, but vaccine versus the disease.
Frequently asked questions
Do vaccines cause autism?
No. This is the most studied question in modern medicine, and the answer has been the same in every large test: the 2014 meta-analysis of more than 1.2 million children, the 2019 Danish study of 657,461 children, the Cochrane review of about 14.7 million children, and the WHO's December 2025 review of 31 studies all found no causal link between vaccines — including MMR, thimerosal and aluminium — and autism.
What started the idea that vaccines cause autism?
A 1998 paper in The Lancet by Andrew Wakefield and colleagues, describing twelve children and suggesting the MMR vaccine might be linked to autism. The paper itself said it did not prove a link. It was later found to be fraudulent: the data had been altered, the children were partly recruited through lawyers suing vaccine makers, and Wakefield had been secretly paid £435,643 by those lawyers. The Lancet retracted the paper in 2010, and Wakefield was struck off the UK medical register.
Why do so many parents believe the link is real?
Because of timing. The MMR vaccine is given at 12 to 15 months, and autism's first signs usually appear between 12 and 18 months. Regression — the loss of words and eye contact — can look sudden, and when it follows a clinic visit by days or weeks, the connection feels obvious. It is a coincidence of timing, not a cause: regression happens on the same schedule in unvaccinated children, and population studies show no difference in autism rates between vaccinated and unvaccinated children.
Does thimerosal in vaccines cause autism?
No. Thimerosal contains ethylmercury, which the body clears quickly, unlike the methylmercury that accumulates from fish. It was removed from routine childhood vaccines in the US by 2001 as a precaution, and autism rates kept rising afterwards in every country that removed it — the opposite of what the theory predicted. The MMR vaccine never contained thimerosal at all.
Are vaccines completely risk-free?
No vaccine is risk-free, and no honest source claims otherwise. Common side effects are a sore arm and a short fever. Rare serious reactions exist: severe allergy occurs in about one in a million MMR doses, and febrile seizures in roughly one in 3,000 to 4,000 toddler doses — frightening but not causing lasting harm. These risks are far smaller than the risks of measles itself, which hospitalises about one in eight cases and can cause pneumonia, brain swelling and death.
If there is no link, why did the CDC change its vaccine-autism page in 2025?
The CDC's page wording was changed in late 2025 under new political leadership, and the change cited no new study — fact-checkers and medical societies noted this immediately. Weeks later, the WHO's Global Advisory Committee on Vaccine Safety reviewed 31 studies published between 2010 and 2025 and reaffirmed that vaccines do not cause autism, its fourth such conclusion since 2002. A website edit reflects a change of administration. A 31-study evidence review reflects a change in science — and no such change has occurred.
Sources
- Wakefield, A. J. et al. (1998). Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. The Lancet, 351, 637–641. Retracted 2010. — wikipedia.org
- Deer, B. (2011). How the case against the MMR vaccine was fixed. The BMJ, 342, c5347. — briandeer.com
- Deer, B. (2006). MMR doctor given legal aid thousands. The Sunday Times, 31 December 2006 — Wakefield paid £435,643 in fees plus £3,910 expenses. — briandeer.com
- Madsen, K. M. et al. (2002). A population-based study of measles, mumps, and rubella vaccination and autism. New England Journal of Medicine, 347, 1477–1482.
- Taylor, L. E., Swerdfeger, A. L. & Eslick, G. D. (2014). Vaccines are not associated with autism: an evidence-based meta-analysis of case-control and cohort studies. Vaccine, 32(29), 3623–3629.
- Jain, A. et al. (2015). Autism occurrence by MMR vaccine status among US children with older siblings with and without autism. JAMA, 313(15), 1534–1540.
- Hviid, A., Hansen, J. V., Frisch, M. & Melbye, M. (2019). Measles, mumps, rubella vaccination and autism: a nationwide cohort study. Annals of Internal Medicine, 170(8), 513–520. doi:10.7326/M18-2101
- Demicheli, V. et al. (2012). Vaccines for measles, mumps and rubella in children. Cochrane Database of Systematic Reviews — review of studies covering about 14.7 million children.
- DeStefano, F. et al. (2013). Increasing exposure to antibody-stimulating proteins and polysaccharides in vaccines is not associated with risk of autism. Journal of Pediatrics, 163(2), 561–567.
- Schechter, R. & Grether, J. K. (2008). Continuing increases in autism reported to California's Developmental Services System: mercury in retrograde. Archives of General Psychiatry, 65(1), 19–24.
- World Health Organization (2025). WHO expert group's new analysis reaffirms there is no link between vaccines and autism. 11 December 2025. — who.int
- Shaw, K. A. et al. (2025). Prevalence and early identification of autism spectrum disorder among children aged 4 and 8 years — ADDM Network, 16 sites, United States, 2022. MMWR Surveillance Summaries, 74(SS-2), 1–22. — cdc.gov
- Centers for Disease Control and Prevention (2025). Measles cases and outbreaks — 2025 US measles data: 1,300+ cases, 3 deaths, 164 hospitalisations.
- Medical Xpress (2025). Vaccines "don't cause autism": how scientists figured that out. — medicalxpress.com
Medical note: this article summarises published research for general information. It is not medical advice. If you have questions about your child's vaccinations, talk to your doctor or paediatrician — and if your child shows signs of developmental delay, ask for an assessment early, because early support makes a real difference whatever the cause.
Found this answer useful? This post is part of a series that takes questions many people search for but nobody has answered properly, and gives the real, detailed answer. Earlier in the series: whether breakfast is really the most important meal of the day, whether bottled water is safer than tap water, whether GMOs are safe to eat, whether eggs are bad for your cholesterol, whether vaping is safer than smoking, whether aspartame is safe, whether sugar makes kids hyperactive, whether red meat causes cancer, whether fluoride in tap water is safe, whether seed oils are toxic, whether organic food is actually healthier, the MSG food fight and whether expired poison becomes safe. If you have a question you want answered this way, reach out through asadwebs.com.
Be the first to share a useful question or thought.