Association of autistic spectrum disorder and the measles, mumps, and rubella vaccine. A systematic review of current epidemiological evidence.

Wilson K, Mills E, Ross C, McGowan J, Jadad A.

Archives of Pediatric and Adolescent Medicine 157:628-634. 2003

Question: What is the evidence, pro and con, showing that autism spectrum disorders are related to use of the MMR vaccine?

Background: Many people are convinced that the measles, mumps, and rubella (MMR) vaccine is responsible for their children's autism. Some studies reported in the media have supported the idea that the vaccine causes a "variant" form of autism, while other studies have shown no relationship at all. In this review the authors posed various hypotheses or theories and looked at the evidence for each of them. The theories were that 1) the rates of autism in people who were vaccinated are higher than in people who were not vaccinated; 2) the perceived increase in rates of autism is the result of the use of MMR vaccine; 3) the development of autism occurs around the same time as the MMR vaccination; and 4) a new type of autism (a variant form) is related to the use of the vaccine.

Data Sources: The authors searched CINAHL, PsychINFO, MEDLINE, PreMedline, TOXLINE, Biological Abstracts, the Cochrane Library, and HealthSTAR for English-language and non-English language articles. They used the terms "measles, mumps, rubella vaccine" and "autistic disorder" and "autism spectrum disorder" and included variations of the search terms. The database searches were supplemented with examination of reference lists of articles retrieved and by contacting the authors of published studies to see if they were involved in other research on the topic.

Study Selection: The studies found were included if there were original epidemiological studies, if the studies used clearly defined populations, how the subjects had been diagnosed, and if the studies attempted to find an association between the MMR vaccine and ASD.

Data Extraction: Two independent rates evaluated the studies and recorded the authors' names, the journal name, publication year, country in which the study was conducted, the study design, a description of the population under study, age at MMR immunization and the method by which the subjects were diagnosed with ASD. They also looked at what association was found, how the data was analyzed, and the results of the analysis.

Main Results: Of 381 studies located, 12 met the inclusion criteria and were scientifically rigorous enough to be included in the review with the following results:

Theory 1: The rates of autism are higher in people who were vaccinated than people who were not vaccinated.
A large retrospective study done in Denmark was able to identify no statistically significant difference in autism rates between 440,655 children who had been vaccinated and 96,648 children who had not been vaccinated.
There is no difference in the rates of autism in people who were vaccinated compared to people who weren't vaccinated.

Theory 2: More cases of ASD are occurring because of the MMR vaccine.
Six studies looked at whether rates of ASD increased as the rate of MMR vaccination coverage of the population increased. Four of these found no association between the MMR vaccine and an increase in ASDs or the proposed variant ASD. Another study looked at the number of cases of ASD before and after the vaccination programs were introduced, and did not find an increase during the time in which MMR vaccination had been introduced. The last of these studies looked at the rates of reports of developmental regression in a group of children with autism before and after the MMR vaccine had been introduced and found no difference.
The studies did not find that there are more cases of ASD because of the use of the MMR vaccine.

Theory 3: the onset of autism occurs around the same time as the MMR vaccination.
Eight of the studies looked for an association between the time the MMR vaccine was administered and later development of autism. Of these studies, three looked at the age at which an ASD was diagnosed and the age at which parents voiced concerns about their child's development in groups of children who were vaccinated or not vaccinated. The idea here was that if the MMR vaccine caused the autism, children who were vaccinated would develop the condition at a different age than those who weren't. The studies did not find any difference in the age at which autism was diagnosed. Six of the studies looked at the increased rate at which the ASD was diagnosis or symptoms of ASD were reported after children had been vaccinated. One study did not find an increase in visits to GPs following vaccination in children who were later diagnosed with an ASD. Another of the studies found that a group of children with ASD were not more likely to have been diagnosed with ASD or regression in a specific period of time after the MMR vaccination.
There is no evidence from current studies to support a relationship between the development of ASD and MMR vaccination.

Theory 4: a new type of autism (a variant form) is related to the use of the vaccine.
Of three studies that looked for developmental regression, one study found that of 31 children who developed GI symptoms after the vaccination, none later developed ASD. Another study compared children diagnosed with ASD in two time periods – prior to the introduction of MMR vaccination and after the introduction of the vaccine and found no difference in rates of developmental regression in the two groups of children. The study did not find a higher than expected rate of childhood disintegrative disorder (CDD) in the group of children studied after the introduction of the vaccine. One study did not find an increase in the number of children with autism who had either GI symptoms or regression after the start of MMR vaccine use. A fourth study found that no one of 309 hospitalized children with autism had also been in hospital with inflammatory bowel disease.
There is no evidence for the theory that there is a new variant form of autism related to the vaccine.

Conclusions: Most particularly, there was no evidence of the emergence of a "new variant" autism. The authors suggest that if such a condition exists, it is so rare that it cannot be detected in epidemiological (that is, large scale) studies. None of the studies assessed for this review showed evidence of an association between the measles, mumps, and rubella vaccine and the development of an autism spectrum disorder.