The Autism-Vaccine Myth

What was the impact of the alleged vaccine-autism connection?

What the public didn't know in 1998 was that the now-retracted study, which involved just 12 children, would turn out to have some serious flaws—and even to contain apparently falsified data. The 12 years between its publication and its retraction, however, left a lot of time for the unfounded and never-confirmed vaccine-autism link to take hold in the minds of worried parents—and thus for vaccination rates to suffer.

In 1997, the year before the paper was published, measles vaccination rates in the United Kingdom were over 91%. They started to fall in 1998 and in 2003-2004 reached a nadir of just 80%, although rates were even lower than that in specific areas. Only in recent years have MMR vaccination rates started climbing again in the U.K., reaching about 90% in 2013.

The vaccine-autism question: A timeline

Here are some highlights along the scientific journey the vaccine-autism hypothesis made from its 1998 publication in The Lancet to today. The American Academy of Pediatrics has compiled an expanded list of relevant studies.

1998-The Lancet publishes a paper by Wakefield et al. titled "Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children." A press conference and interview in which Wakefield says it is his "feeling that the…risk of this particular syndrome developing is related to the combined vaccine, the MMR, rather than the single vaccines" set off a media storm.

1999—A U.S. Food and Drug Administration review of the vaccine preservative thimerosal, which contains mercury, finds no evidence of its causing neurological harm, but the agency recommends the precautionary removal of the agent from vaccines administered to infants.

2001—Thimerosal is removed from childhood vaccines on the U.S. market.

—The Journal of the American Medical Association publishes "Time trends in autism and in MMR immunization coverage in California" by Dales et al. These authors looked for and found no association between MMR vaccination rates in young children and increased autism occurrence.

—Pediatrics publishes "No evidence for a new variant of measles-mumps-rubella-induced autism." by Fombonne et al. In this study of 96 children with pervasive developmental disorders, the authors found no evidence of a specific syndrome being related to whether or not they had received the MMR vaccine.

2002—The British Medical Journal publishes "Measles, mumps, and rubella vaccination and bowel problems or developmental regression in children with autism: population study" by Taylor et al. This study of 473 children with autism who were born between 1979 and 1998 offered "further evidence against a link between MMR and autism." The authors found no differences in regression or bowel symptoms in those born before and after the introduction of the MMR vaccine.

—Pediatrics publishes "Neurologic disorders after measles-mumps-rubella vaccination" by Makela et al. This study of 535,544 children in Finland found no link between MMR vaccination and autism.

—The New England Journal of Medicine publishes "A population-based study of measles, mumps, and rubella vaccination and autism" by Madsen et al. These authors describe "strong arguments" against the hypothesis that the MMR vaccine causes autism, based on an analysis of data from 537,303 children in Denmark, 82% of whom had received the MMR vaccine.

2003—The Journal of the American Medical Association publishes "Association between thimerosal-containing vaccine and autism" by Hviid et al. These authors, using data from 467,450 children born in Denmark, found no link between autism risk and vaccines containing thimerosal.

2004—Of the 13 authors on the 1998 Wakefield paper, ten formally retract its interpretation, stating a wish to make clear that the paper established "no causal link" between the MMR vaccine and autism.

—The Institute of Medicine of the United States' National Academies issues a report from its Immunization Safety Review Committee, which concludes that vaccines don't cause autism.

2007—The New England Journal of Medicine publishes "Early thimerosal exposure and neuropsychological outcomes at 7 to 10 years" by Thompson et al. These authors looked for links between early thimerosal exposure and neuropsychological outcomes in a group of 1,047 children aged seven to ten years and found no association.

2010—The Lancet, after an extensive investigation, including investigative work by journalist Brian Deer, issues a formal retraction of the 1998 Wakefield paper.

—The U.K,'s General Medical Council strikes Andrew Wakefield from the medical register (the U.K. equivalent of stripping a medical license), concluding, according to reports, that he had been "dishonest, irresponsible, and showed callous disregard for the distress and pain of children" in conducting the MMR-autism study.

—Pediatrics publishes "On-time vaccine receipt in the first year does not adversely affect neuropsychological outcomes" by Smith et al. This study of more than 1,000 children born between 1993 and 1997 found that on-time vaccination was associated with better performance on a number of neuropsychological tests.

—Pediatrics publishes "Prenatal and infant exposure to thimerosal from vaccines and immunoglobulins and risk of autism" by Price et al. This study of over 1,000 children found no increase in autism outcomes among children who had received thimerosal-containing vaccines, compared to those who had not.

2013—A major measles outbreak in Wales is linked to regional response to the 1998 Wakefield paper, which led to reduced measles vaccination rates in the region affected by the outbreak.

—The Journal of Pediatrics publishes "Increasing exposure to antibody-stimulating proteins and polysaccharides in vaccines is not associated with risk of autism" by DiStefano et al. This study of more than 1,000 children found no link between the number of antigens in vaccines and autism, and thus no support for the "too many, too soon" school of thought.

2014—Vaccine publishes "Vaccines are not associated with autism: An evidence-based meta-analysis of case-control and cohort studies" by Taylor et al. This meta-analysis of ten studies covering more than 1.2 million children finds no links between autism and MMR vaccination, mercury, thimerosal, or vaccination generally.

Why has the myth persisted?

Two fears powered the unfolding of these events. One was a fear of the unfamiliar ingredients in vaccines, including the mercury-containing preservative thimerosal, which as of mid-2014 was still used in some multidose flu shots but was otherwise phased out of routine childhood vaccines in the United States starting in 2001. The other was a fear of autism, an anxiety fostered by media stories pitting emotional appeals by high-profile anti-vaccine advocates against statistically based reports by medical researchers.

No link with autism—not to mercury, not to thimerosal, not to any vaccines, including the MMR—has been found.

These fears persisted even as evidence mounted that they were completely unfounded. Scientific verification relies on a process of testing and confirmation, not on a single observation. Researchers sincerely grappled with the question of a vaccine-autism link in numerous studies following publication of the 1998 Lancet paper. Some of these studies analyzed data from millions of people, in the quest to see if vaccines and autism might be linked. The overwhelming scientific consensus is that they are not. The Lancet paper specifically addressed a possible association between the MMR vaccine and autism, but later studies also looked at other vaccine-related factors, such as the mercury-containing preservative thimerosal, which is not used in the MMR vaccine. Thimerosal initially raised concerns because mercury is neurotoxic. However, no link with autism—not to mercury, not to thimerosal, not to any vaccines, including the MMR—has been found.

In some fraction of the American population, however, the belief in a link remains. One reason is a coincidence of timing: children are routinely vaccinated just as parents begin to observe signs of autism. Most vaccines are administered during the first years of life, which is also a period of rapid developmental changes. Many developmental conditions, including autism, don't become apparent until a child misses a milestone or loses an early skill, a change that in some cases can't help but be coincident with a recent vaccination.

Adding to such concern is the fact that, sometimes, vaccinations can lead a child to develop a high fever and accompanying febrile seizures. Such seizures are temperature-related and don't cause lasting damage. A tendency to experience febrile seizures runs in families, and about one in 20 young children will have one at some point. According to the U.S. Centers for Disease Control and Prevention, most febrile seizures happen when a child is sick rather than after a vaccination, though the MMR vaccine is associated with a slightly increased risk for febrile seizures. Children with and without autism have these febrile responses, but since their timing may coincide with emerging signs of autism, that can link the two incidents in a parent's mind, even though there is no causal relationship.

What do we know about what contributes to autism?

Largely ignored during this prolonged vaccine-autism controversy was clear evidence of a strong genetic influence on autism. For example, if one identical twin has been diagnosed with an autism spectrum disorder (ASD), the other twin is from 60% to 90% more likely to also be diagnosed with an ASD. Also known is that males are more likely than females to be diagnosed with autism and that the older a father is, the higher the likelihood of a diagnosis of autism in his offspring.

Nonetheless, autism has proven to be very difficult to pin down genetically in the way that single-gene disorders, such as cystic fibrosis and Huntington's disease, have been traced. The reasons for this difficulty have recently become clearer thanks to advances in DNA sequencing technology that have made it possible to compare the genomes of individuals with autism to those of individuals without autism, as well as to those of their parents. It is now evident that there are many genetic paths to autism, and that some mutations leading to autism are not inherited but arise spontaneously in reproductive cells or during development. These mutations help to explain how autism can appear in families that previously had no history of the condition.

Some mutations associated with autism are single changes in the DNA code that make up genes. Other kinds of mutations include extra copies or deletions of larger stretches of chromosomes that include multiple genes.

When investigators have homed in on specific genes, they have identified numerous candidates, many of them associated with the formation or function of the brain. These include genes that play roles in neural development and structuring, nerve signaling, and speech and language processing. Still, because of the rapid proliferation of genetic data regarding autism, scientists are still discovering the ways in which many other mutations lead to autism—a task that will likely take years.

The emerging scientific picture of autism is a condition that begins during fetal development, as a result of both genetic and environmental influences.

It is now very clear that vaccines are not among the environmental factors that cause autism. But scientists are eager to understand how genetic and environmental influences may interact and to explain how these factors play a role in the broad spectrum of symptoms manifest in different people with autism.