Anatomy of a Scare

The charges against Wakefield were the least of what was undermining the vaccine theory of autism. What would eventually become an overwhelming body of evidence showing that childhood vaccines did not increase the risk of autism began to pile up. In 2002 scientists led by Brent Taylor of the Royal Free reported that their study of 473 children had found no difference in the rates of autism between those who had received the MMR and those who had not, providing "further evidence against involvement of MMR vaccine in the initiation of autism," they wrote. Scientists in Finland, studying 2 million children, reached the same conclusion in a 2000 paper. So did scientists at Boston University, studying the medical records of 3 million children, in 2001. In 2004 a study of the medical records of 14,000 children in Britain found that the more thimerosal the children had been exposed to through vaccines, the less likely they were to have neurological problems. Also that year, the Institute of Medicine (IOM) in the United States, having reviewed 200-plus studies, rejected the vaccine-autism hypothesis. Not only did it find no evidence of a link—and, indeed, evidence against the existence of a link—but it took aim at the original 1998 Lancet paper by Wakefield's group. Because autism symptoms typically appear at the same age that children get the MMR, the panel said, it was inevitable that some children would first show symptoms of autism soon after being vaccinated. Coincidence is not causality.

If the IOM panel thought that would be the end of it, they were naive. From the halls of Congress to the airwaves to the pages of leading newspapers, true believers went at the vaccine-autism link more passionately than ever. After the IOM released its report, Rep. Dave Weldon of Florida, a physician, took to the House floor to denounce the CDC for its vaccine-autism research. The agency, Weldon charged, was guilty of "selective use of the data to make the associations [between vaccines and autism] disappear" and had engaged in "a public-relations campaign [on behalf of vaccines] rather than sound science."

The following year, a story by environmental lawyer Robert F. Kennedy Jr. called "Deadly Immunity" made the case against thimerosal in Rolling Stone. Activists used large chunks of the money they were raising from terrified parents to spread their message. On June 8, a full-page ad for Generation Rescue, which had been founded the month before to push the thimerosal theory, ran in The New York Times, proclaiming, "Mercury Poisoning and Autism: It Isn't a Coincidence." It included quotes from several politicians, with Burton stressing research that "indicated a direct link between exposure to mercury and autism," and Sen. John Kerry saying "mercury has been linked to autism." Later that year, a book titled "Evidence of Harm: Mercury in Vaccines and the Autism Epidemic," by journalist David Kirby, got huge attention in the media. Kirby appeared on "Imus in the Morning" several times, as did politicians supporting his thesis. Sen. Joseph Lieberman, citing the growing incidence of autism coupled with the increase in the number of required vaccines, said, "Make sure your kids are getting vaccines without thimerosal."

Throughout this saga, the "vaccines cause autism" side could claim more powerful persuaders than the dry-as-dust scientific papers and even drier scientists trying to reassure parents. On Sept. 18, 2007, model and actress Jenny McCarthy appeared on "The Oprah Winfrey Show" to promote her new book, "Louder Than Words," in which she describes curing her son Evan's autism—which she blames on the MMR—with diet and chelation, a process that chemically binds heavy compounds in the body so they can be excreted. Asked about the CDC statement that science does not link vaccines to autism, McCarthy said, "My science is Evan." Researchers were dumbfounded that so many parents rejected the conclusions of the CDC, the Institute of Medicine and the American Academy of Pediatrics (which after its thimerosal debacle had put itself foursquare behind childhood vaccines). "The issue for people like Jenny McCarthy isn't that doctors and scientists and public-health officials haven't listened to parents," writes Paul Offit in his 2008 book "Autism's False Prophets: Bad Science, Risky Medicine and the Search for a Cure." "It's that they've been unable to find any evidence to validate parents' concern."

The anti-vaccine campaign was having an effect. As parents postponed vaccinating their children, or refused vaccines entirely, children were suddenly catching preventable diseases, and some were dying. The number of measles cases in the United States reached 131 in 2008, the highest in decades. Last month five children in Minnesota became infected with Hib. Four developed serious complications; the fifth child died. Other parents, believing that yanking mercury out would cure a child's autism, opted for chelation. Unfortunately, it can pull out vital metals such as iron and calcium as well as toxic mercury and lead.

An overwhelming majority of vaccine and autism experts were convinced that parents were putting their children at real risk over a phantom fear. But perhaps no one understood that the MMR theory, in particular, was a house of cards better than molecular biologist Stephen Bustin of the University of London. In 2004 the U.K. High Court asked him to inspect the Dublin lab that had reported measles genes in the guts of autistic children right after they received the MMR, an important confirmation of Wakefield's theory. It was an uncomfortable situation, Bustin recalls, playing cop at another scientist's lab. But he discovered a number of problems. The genetic material the lab had found was DNA, but measles genes are made of RNA. The equipment was so poorly calibrated that its results depended on where in the machine the sample was placed. Wakefield defends his collaborator, saying that a later test confirmed "the fidelity and high quality of [the Dublin lab's] methods … The original results that found measles virus genetic material in intestinal biopsies in 75 percent of the autistic children compared with 6 percent of the nonautistic controls still stand."