A look at the Science - MMR vaccine and autism.

Amongst the scientific community there is overwhelming evidence that in immunocompetent individuals the benefits of vaccines outweigh the risk of harm and the herd immunity offered by vaccination is vital to protecting those who are immunosuppressed. There has been concern among popular media that the Measles, Mumps and Rubella (MMR) vaccine can be harmful, more specifically that it may increase the risk of developing Autism Spectrum Disorder (ASD). However there is no scientific evidence to suggest an association. ASD is a developmental disability involving how the brain functions which manifests itself in difficulties in communication, behaviour, learning and social interaction. The MMR vaccine is made of a live attenuated virus, administered to children at 12 months and 4 years of age in Australia. If unvaccinated, measles, mumps and rubella can result in fatal illness. There are claims in popular media and by anti-vaccination groups suggesting a link between MMR and autism. This misconception originates primarily from a retracted article published in The Lancet in 1998, ‘Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children’ by A J Wakefield et all which claimed a link between ASD, gastrointestinal disorders and MMR.[1] There have been numerous studies since this time disputing these claims, including widespread epidemiological studies from a number of countries. These have been published after the Wakefield article, in an attempt to subdue controversy and misunderstanding amongst the general public, which led to a larger proportion of the population avoiding vaccination. These include a 1999 article titled, ‘Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a casual association’[2], ‘Time trends in autism and in MMR immunisation coverage in California’[3], published in 2001 and a 2002 study in Denmark, ‘A population based study of measles, mumps, and rubella vaccination and autism’[4]. Furthermore, a Cochrane Review in 2012 entitled, ‘Vaccines for Measles, Mumps and Rubella in children (review)’supported these findings.[5] This report will analyse the methodologies, motives and results of the retracted Lancet article and the subsequent widely supported literature.

Wakefield’s article, ‘Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children’ investigated 12 children with chronic enterocolitis and regressive developmental disorder   This article was originally published in the Lancet, a well renowned UK medical journal but was subsequently retracted due to significant flaws, and subsequent finding of fraud against the main author, Wakefield. The study involved pathology tests to investigate intestinal abnormalities and inflammation and a survey asking parents about the interval from exposure to the MMR vaccine to them noticing the first behavioural symptom of autism spectrum disorder. The authors concluded that there was a link between gastrointestinal disease and onset of autism spectrum disorder after exposure to the MMR vaccine.

There are many issues with the methodologies used to come to this conclusion.  This was a case series with no control group and the cases were selected from a group sympathetic to the anti-vaccination message. This was not a randomised control trial or even study, rather it was based primarily on anecdotal accounts of the 12 children’s parents rather than concrete data which could be statistically compared. There was no control group, simply a collection of cases with no variables such as demographics, age, sex controlled for at all. Furthermore, the sample size was very small and Wakefield did not suggest other reasons for this slight correlation such as parents typically watching children more closely and displaying greater concern 6 months after being vaccinated. For instance, he did not declare that Child 2 who had both regressive autism and gastrointestinal issues also had 2 brothers with the same gastrointestinal issues – both of whom displayed symptoms before they were administered the MMR vaccine but did not have a diagnosis of autism spectrum disorder. Due to the failure to use proper scientific method, ten out of the twelve original authors have since retracted their support of the article.  A 2010 editorial in the British Medical Journal, ‘Wakefield’s article linking MMR vaccine and autism was fraudulent’ has revealed further issues with Wakefield’s methodologies.[6] This includes excluding evidence of existing developmental delays in some of the children before the vaccine, altering parent’s accounts and results of tests, and selecting 12 children from a larger cohort of unaffected children. This resulted in the results looking like there may be an association, when there really wasn’t. Furthermore, there was significant conflict of interest and a source of potential bias - the study was funded by an anti-vaccine group who were looking to sue the makers of the vaccine, and the 12 children in the trial were actively recruited by people who were campaigning against the MMR vaccine. This research has had huge implications for public health, leading to a distrust in public immunisation programs. In the UK there has been a significant decrease in vaccination, leading to a reemergence of measles  

The highest quality of cause and effect research is considered to be the double blinded randomized control trial. In situations like this, it would be deemed unethical to withhold MMR in the control group, and replace with placebo as MMR is known to reduce the incidence and complications, including mortality, of measles, mumps and rubella infection. Therefore, to prove an association, large epidemiological studies looking at incidence across populations compared with rates of intervention, in this case, immunisation need to be used. These following epidemiological studies consistently show no connection between autism and MMR.

 ‘Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a casual association is an epidemiological study released in 1999 in The Lancet in response to Wakefield and his colleagues work  The study involved a more robust scientific method with a much larger sample size than Wakefield’s study. The researchers identified children with Autism Spectrum Disorder from a computer based special needs register, rather than ‘hand picking’. Furthermore, the diagnosis of autism was done by 2 practitioners, to eliminate ambiguity and records of immunisation were obtained independently of the child’s clinical records. Unlike Wakefield, the researchers used a statistical analysis, rather than relying on anecdotal accounts. Another advantage of the methodologies is that the researchers identified bias and limitations in their study. For instance, they recognised that in some individuals, there were clinical notes of poor quality and in a few cases, it was difficult to classify children according to diagnostic criteria, particularly in regards to Asperger’s syndrome or atypical autism

Similarly, a 2001 study in the Journal of the American Medical Association (JAMA), ‘Time trends in autism and in MMR immunisation coverage in California’ involved a study of 10,000 kindergarten children  JAMA is a high impact peer reviewed medical journal with a focus on public health. The study published specifically investigated the increasing prevalence of autism – a 373% increase in diagnoses between 1980 and 1994 and whether this was linked with an increase in administration of the MMR vaccine, finding that a relative increase in vaccination rates of 14% was not a statistically significant explanation for an increase of autism spectrum disorders  Similar to the 1999 study, the researchers used statistical analysis of school vaccination records and autism diagnosis and used a large sample over a number of years. Furthermore, the researchers also noted possible limitations such as an inability to examine individual vaccination and autism records. This resulted in a robust method and consistent empirical results which were consistent with findings of other studies worldwide, finding no correlation between vaccination against MMR and a diagnosis of ASD.

A 2002 article published in the New England Journal of Medicine, ‘A population based study of measles, mumps, and rubella vaccination and autism’ studied 537,000 children in Denmark between 1991 and 1998  This study looked at the rate of gastrointestinal disorders, the MMR vaccine and children with and without developmental disorders, in an attempt to confirm or refute the findings of Wakefield et al. The study involved all children born, using information from a national registry of immunisation and a central psychiatric register. This study has the advantage of using data from a whole population and had complete data for vaccination records and diagnosed cases of autism spectrum disorder. No link was found between MMR and ASD, or gastrointestinal disease

A comparison of the literature strongly supports that the idea that there is no link between the MMR vaccine and autism spectrum disorder. ‘Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children’ claims to have discovered a new gastrointestinal disorder, enterocolitis linked with regressive autism  Due to flawed methodology, bias and conflict of interest, these results should be dismissed. However, the other studies investigated have been undertaken across a large sample and have shown no link. These studies use a good scientific method and explore many explanations for their results unlike Wakefield’s study. ‘Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a casual association’ researchers studied children in the UK, like Wakefield, but found very different evidence using good scientific method with no undeclared limitations or biases  This resulted in strong evidence that there was no casual association between the MMR vaccine and autism. This is consistent with the California based, ‘A population based study of measles, mumps, and rubella vaccination and autism’  This was a larger scale epidemiological study which similarly found no correlation between immunisation rates and autism diagnosis, concluding that an increase in autism was too large to be explained by the small increase in vaccination rates. Furthermore, the national study of children in Denmark supported these findings again, finding no difference in risk of autism between vaccinated and unvaccinated children, no temporal cluttering of autism cases after vaccination and no link between MMR and ASD  These findings are robust with no obvious bias from researchers and good scientific methodology. This has produced statistically significant results which showed no relationship between ASD and MMR vaccine.

This analysis of primary literature is supported by the Cochrane review of the available scientific literature, comparing their results regarding the safety of the MMR vaccine  The Cochrane is a systematic, peer reviewed and independent study into primary research related to health care and is considered to be the highest international standard.  The data used by the review involved numerous studies, investigating a total of 14,700,000 children. Specifically, in regards to autism, a review of the literature found no statistically significant link between the MMR and autism – there were no cluster intervals, no difference in the onset of autism depending on both the age of vaccination or if children were vaccinated at all.

There is overwhelming evidence that the Measles, Mumps and Rubella vaccine does not cause autism spectrum disorder. In immunocompetent individuals, the benefits of vaccines to the individual and the wider community greatly outweigh any possible harm. A lack of understanding amongst the public after Wakefield’s article significantly impacted vaccination rates and has thus led to many widespread epidemiological studies that conclusively support that there is no relationship. These studies were conducted ethically, using scientific methodologies which produced comparative statistical results unlike the original study and thus should be considered as valid.

[1] A J Wakefield, S. H.-S. (1998, February). Illeal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet, 351, 637-641.

[2] Brent Taylor, Elizabeth Miller, C Paddy Farrington, Maria-Christina Petropoulos, Isabelle Favot-Mayaud, Jun Li, Pauline A Waight, Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a causal association, The Lancet, 353, 9169, 12 June 1999, 2026-2029.

[3] Dales L, Hammer S, Smith NJ. Time Trends in Autism and in MMR Immunization Coverage in California. JAMA. 2001;285(9):1183-1185. doi:10.1001/jama.285.9.1183.

[4] Taylor B, Miller E, Farrington CP, Petropoulos MC, Favot-Mayaud I, Li J, Waight PA. Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a causal association. Lancet. 1999 Jun 12;353(9169):2026-9.

[5] Demicheli V, Rivetti A, Debalini MG, Di Pietrantonj C. Vaccines for measles, mumps and rubella in children. Cochrane Database of Systematic Reviews 2012, Issue 2.

[6] Godlee Fiona, Smith Jane, Marcovitch Harvey. Wakefield’s article linking MMR vaccine and autism was fraudulent BMJ 2011; 342.