The Vaccine–Autism Myth: The Study That Was Withdrawn

This is a history, and it should be read as one. The events belong to the last years of the twentieth century and the first of the twenty-first; the scientific question they raised has been answered about as completely as science answers anything, and the person at the centre of the story was removed from the medical register more than a decade ago. What remains open is not the science. It is the far more human question of why the fear the science created has proved so much harder to retract than the paper that started it. That question is worth sitting with carefully, and with more sympathy than the subject usually receives.

A press conference in February 1998

On 26 February 1998, a group of doctors sat before the press at the Royal Free Hospital in north London to discuss a paper published that day in the Lancet, one of the oldest and most respected medical journals in the world. The lead author was a gastroenterologist named Andrew Wakefield. The paper described twelve children who had been referred to the hospital with intestinal symptoms and developmental problems, most of them a form of autism. It reported that in a number of cases, the parents recalled the developmental regression beginning shortly after the children received the MMR vaccine — the combined measles, mumps and rubella jab given to toddlers.

The paper itself, read strictly, was cautious. It stated plainly that it had not proved a link between the vaccine and the conditions described. It was a small case series, twelve children, hypothesis-generating at most. Had the matter rested on the printed words, it might have caused a ripple among specialists and nothing more.

It did not rest there. At the press conference, Wakefield went beyond the paper. He suggested that, as a precaution, parents might consider giving the three vaccines separately rather than as the combined MMR jab — a recommendation the science of the paper did not support and that his co-authors had not agreed to advance. That single spoken suggestion, from a doctor at a famous hospital, carried by a receptive press into every home with a small child, is where the fear began. The story that reached the public was not “twelve children, no link proven”. It was “doctor warns MMR may cause autism”. Those are very different sentences, and only one of them travelled.

What the investigation found

For several years the claim hung in the air, boosted by growing media coverage, while epidemiologists around the world tested it the only way it could be tested: by looking at very large numbers of children, far more than twelve, and comparing rates of autism among the vaccinated and unvaccinated. Study after study, across several countries and eventually millions of children, found no link. The science moved toward a firm conclusion that the MMR vaccine does not cause autism, and it has only hardened since.

But the science was one thread. The other was journalistic, and it was pursued by a reporter for the Sunday Times named Brian Deer, whose investigation, beginning in 2004, slowly took apart not the conclusion of the 1998 paper but its foundations.

Deer established several things that the Lancet’s readers in 1998 had not been told. The first concerned money. Wakefield had been receiving payments — eventually shown to total a substantial sum — from a firm of solicitors that was preparing litigation against vaccine manufacturers, seeking to prove exactly the kind of harm the paper appeared to support. This funding, and the conflict of interest it represented, had not been disclosed to the journal or to Wakefield’s co-authors. The second concerned the children. Rather than being a consecutive series of ordinary referrals, several of the twelve had connections to the anti-vaccine litigation; the sample was not what it appeared. The third, and gravest, concerned the data itself. Comparing the descriptions in the paper against the children’s actual medical and hospital records, Deer found discrepancies — the timings of symptom onset, the diagnoses, the reported findings — that could not be explained as innocent error. In some cases, records showing developmental concerns before the MMR jab had been described in the paper as beginning after it.

The institutions responded in sequence. In 2004, ten of the paper’s thirteen authors formally retracted its interpretation, withdrawing the suggestion of a causal link. In February 2010, after the longest hearing in its history, the General Medical Council found Wakefield guilty of serious professional misconduct, including dishonesty and the carrying out of invasive procedures on children that were not clinically justified. That same month the Lancet fully retracted the paper. In May 2010 Wakefield was struck off the medical register in the United Kingdom. In 2011 the British Medical Journal, in an editorial by its editor Fiona Godlee, went further than “flawed” and used the word “fraud”.

That is the documented kernel of this history, and it is unusually complete. A small paper, an unsupported spoken warning, an undisclosed financial interest, data that did not match the records, a full retraction, and a doctor removed from practice.

The fork: the science reversed, the belief hardened

Here is the strange turn, and it is the reason this history matters beyond itself. Everything that could go right in the correction of a scientific error went right. The paper was investigated, discredited, retracted. Its author was sanctioned by the highest professional authority in his country. Enormous studies confirmed the vaccine’s safety. By every measure of how science is supposed to police itself, the system worked.

And the fear grew anyway. In the years after the 1998 press conference, MMR vaccination rates in Britain fell as frightened parents held their children back. Measles, a disease that had been all but eliminated, returned; outbreaks followed, and with them the serious complications measles can bring. The belief that the vaccine was dangerous spread across the English-speaking world and outlived every stage of the paper’s collapse. The retraction did not undo it. The striking-off did not undo it. The belief had become detached from the study that spawned it, and once detached, it could not be reached by anything that happened to the study.

Why? Not because the parents who held it were foolish or careless. This is the part that demands care.

Why frightened parents were caught by it

Autism is, for most families, diagnosed in the second or third year of a child’s life. The MMR vaccine is given at around the same age. A parent watches a bright, responsive toddler and then, over some months, sees the child change — withdraw, lose words, retreat somewhere the parent cannot follow. It is one of the most frightening experiences a parent can have, and it arrives without warning and, for a long time, without explanation. And it arrives close in time to a needle the parent chose to allow into their child’s arm.

The mind cannot help but draw the line. This happened, and then that happened, so this caused that. It is one of the oldest reasoning errors we have, and it is not a mark of stupidity; it is the default setting of a frightened animal looking for a cause. Add to it a real and agonising need: a parent facing a lifelong condition in their child wants, desperately, two things — a reason it happened, and a moment when it happened. The vaccine offered both. It gave the shapeless catastrophe a date and an author. It converted “we do not know why, and there was nothing you could have done” — which is unbearable — into “it was the jab”, which is terrible but at least holds still and points somewhere.

That is the coincidence trap, and it is worth naming plainly because it is what makes the myth humane rather than contemptible. The parents caught in it were doing the thing love does: refusing to accept that a child’s suffering had no cause and no culprit. Wakefield’s suggestion did not create that need. It walked into a need that was already there, aching, and gave it a target.

What the history teaches

There is a wider pattern here, and it is why medical mistrust, once seeded, is so slow to lift. Trust in medicine is not a given; it is earned or squandered by conduct, and it has a long memory. When the record shows that medical institutions have genuinely betrayed the people in their care — as it does, catastrophically, in the Tuskegee syphilis study, where a real US government programme let real men go untreated for decades — that betrayal does not stay contained to its own case. It becomes a general permission to doubt, and every later medical reassurance has to be offered into an atmosphere already thick with earned suspicion. The bitter irony of the MMR affair is that the mistrust it fed on was not entirely irrational, even though this particular claim was false. People had reasons to wonder whether doctors always told them the truth. Wakefield’s story arrived into that atmosphere and borrowed its credibility.

The same reversal appears elsewhere on this desk, in the story of thalidomide, where a genuine pharmaceutical catastrophe and cover-up gave the public a real, evidenced reason to fear that drug companies would lie about safety — a fear that then, in less careful hands, hardened into the belief that all medicine is poison. The MMR myth sits in that same territory. It took a legitimate instinct — check what you put in your child; do not simply trust — and attached it to a specific claim that the evidence does not support and never did.

The honest close is not a scoreboard. The science is settled and has been for years, and repeating the verdict does nothing for the parents still caught in the fear, because the fear was never really answered by the verdict. What the history teaches is gentler and harder. A false belief that grips this many careful, loving people is almost never held out of ignorance. It is held because it does a job that the truth, in its first raw form, cannot do — it gives a reason, and a moment, and a place to put the grief. Meeting that belief with contempt guarantees it will endure, because contempt confirms the mistrust it grew from. Meeting it with the truth alone is not enough either, because the truth on its own does not carry a frightened parent across the thing the myth was built to cross. What the record leaves us is a need to understand — the need of a mother or father, staring at a changed child, unwilling to believe that no one is to blame, and no one, therefore, to have stopped it. That the answer they were handed was false does not make the question they were asking anything less than the most human one there is.