NHS Evidence Health Information Resources formerly National Library for Health

'New study shows no link between MMR and autism' reported The Guardian on 6 July 2006. The report, based on a study which found no link between rates of autism and vaccinations including MMR, was generally accurate. The research had some limitations but supports the findings of other reliable studies.

The Guardian (1) reported that a Canadian study (2) assessing 28,000 children provided further evidence that there was no link between the MMR vaccine and autism. The article also reported that there was no evidence for a link between autism and the use of the mercury compound thimerosal in vaccines, as the autism rate in the study rose from 52 per 10,000 to 70 per 10,000 after thimerosal had been discontinued. The report states that the study provides confirmation of the results of a previous Japanese study (3).
The newspaper report was based on a cohort study of almost 28,000 children attending school in Quebec, Canada who were born between 1987 and 1998 (2). The study assessed the prevalence rates of pervasive developmental disorder (PDD), which includes autism, and examined the relationship between the prevalence of PDD and children's exposure to thimerosal and to either one or two dose MMR vaccinations.
The newspaper (1) accurately reported the findings of the research study, although the prevalence rate cited is slightly lower than the actual figure and refers to pervasive developmental disorder in a subset of the children, and not specifically to autism. The research relied on the accuracy of a number of assumptions, although the authors attempted to address these in their analysis. It is not clear whether the conclusions of the study are reliable, although the findings support those of more reliable research (3, 4).

Evaluation of the evidence base for no link between the MMR vaccine or the use of a mercury compound in vaccines and autism and autistic spectrum disorders

Where does the evidence come from?

The research was conducted by Dr Eric Fombonne and colleagues at the Department of Psychiatry at McGill University and the Montreal Children's Hospital, and at the Lester B Pearson School Board in Montreal, Canada.

What were the authors' objectives?

To assess the prevalence of pervasive developmental disorder (including autism, Asperger syndrome and pervasive developmental disorder not otherwise specified) in Montreal and to evaluate its relationship with
(1) the children's exposure to ethylmercury (thimerosal) in vaccines
(2) the rates of measles-mumps-rubella (MMR) vaccination
(3) the introduction of a second MMR vaccination into the standard vaccination schedule.

What was the nature of the evidence?

The evidence comes from an uncontrolled retrospective (cohort) study of 27,749 children who were born between 1987 and 1998 and attended the 55 schools of the largest Anglophone school board in Montreal, Canada. The children's current school grade was used to extrapolate their year of birth. This was used to determine the children's exposure to aspects of the vaccination programme implemented in Quebec. Cases of PDD were identified from data held by the school boards' special support team, set up to monitor the progress of children with PDD in its schools.

What interventions were examined in the research?

The particular factors of interest were the cumulative level of thimerosal in the vaccinations administered to children, the uptake of measles-mumps-rubella vaccination at the time children were eligible for this vaccination, and the use of an additional second measles-mumps-rubella vaccination at the time the children were eligible for vaccination.

What were the findings?

The overall prevalence of PDD was 64.9 per 10,000, with the prevalence for autism being 21.6 per 10,000 and 10.1 per 10,000 for Asperger syndrome. 82.8% of children with PDD and 83.6% with autism were boys.

Ninety-three percent of children born in Quebec over the study period received MMR vaccination. There was a significant decrease in the uptake of MMR from 96.1% of children born in 1988-1989 to 92.4% of children born between 1996 and 1998. The prevalence of PDD increased significantly by an average of 10% per year over this period, from 27.5 for children aged 16, to 107.6 per 10,000 for children born in 1998. Thus the prevalence of PDD including autism increased as MMR uptake decreased.

The prevalence of PDD in age groups not exposed to thimerosal was significantly higher than in those cohorts that were exposed (82.7 per 10,000 versus 59.5 per 10,000). Adjusting the analysis to account for the increasing PDD prevalence showed no significant effect of thimerosal exposure. Prevalence of PDD began to rise before thimerosal exposure increased from moderate to high (1992), and continued to rise after thimerosal use was discontinued in 1996.

What were the authors' conclusions?

The authors conclude that the findings of their study rule out an association between pervasive developmental disorder and either thimerosal exposure as a consequence of vaccinations or 1- or 2-dose MMR vaccinations.

How reliable are the conclusions?

A cohort study cannot demonstrate a causal link, or absence of a causal link between two events such as vaccination and diagnosis with PDD, although it can demonstrate an association between them.

This was a cohort study with a clearly defined population. However, the study had a number of methodological limitations. As the authors acknowledge, the overall prevalence of PDD may have been over-estimated, as it was based on a cohort of children in schools known to actively support children with PDD. In addition, diagnoses were made by school staff, which may have led to misdiagnosis. However, the majority of children were diagnosed at a hospital, and an analysis of children with a diagnosis of autism, which is more easily diagnosed than atypical PDD, revealed similar results to the main analysis. A population selected by attendance at school in a particular area may also be subject to selection bias which makes it more or less likely that children will have been vaccinated than is the case for the general population.

Data for the use of both thimerosal and MMR were based on vaccination programmes used in Quebec, and may not be an accurate reflection of exposure by children not born in Quebec. Additionally, individual data on vaccination were not available, so it is possible that the prevalence of autism was assessed against factors which did not apply to this population. However, almost 88% of children were born in Quebec, and the results of a subgroup analysis of only those children born in Quebec did not differ from results for the whole cohort.

It is unclear if the results of the study, which may have had a higher than average proportion of individuals with PDD, can be generalised to a wider population. The validity of the authors' conclusions is to some extent limited by these methodological concerns; however, the results support the findings of previous, more robust studies (3,4).

Systematic reviews

Information staff at CRD searched for systematic reviews relevant to this topic. Systematic reviews are valuable sources of evidence as they locate, appraise and synthesize all available evidence on a particular topic.

There was one related systematic review identified on the Cochrane Database of Systematic Reviews (CDSR) (5) and two on the Database of Abstracts of Reviews of Effects (DARE) (6, 7).

References and resources

1. New study shows no MMR link to autism. The Guardian, 6 July 2006, p14.

2. Fombonne E, Zakarian R, Bennett A, Meng L, McLean-Heywood D. Pervasive developmental disorders in Montreal, Quebec, Canada: prevalence and links with immunizations. Pediatrics 2006;118:e139-e150. DOI:10.1542/peds.2005-2993.

3. Honda H, Shimizu Y, Rutter M. No effect of MMR withdrawal on the incidence of autism: a total population study. Journal of Child Psychology and Psychiatry 2005;46:572-9.

4. Madsen KM, Hvid A, Vestergaard M, Schendel D, Wohlfahrt J, Thorsen P, Olsen J, Melbye M. A population-based study of measles, mumps, and rubella vaccination and autism. New England Journal of Medicine 2002;347(19):1477-82.

5. Demicheli V, Jefferson T, Rivetti A, Price D. Vaccines for measles, mumps and rubella in children. The Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD004407. DOI: 10.1002/14651858.CD004407.pub2.

6. Jefferson T, Price D, Demicheli V, Bianco E, European Research Program for Improved Vaccine Safety. Unintended events following immunization with MMR: a systematic review. Vaccine 2003;21(25-26):3954-3960. [DARE Abstract]

7. Wilson K, Mills E, Ross C, McGowan J, Jadad A. Association of autistic spectrum disorder and the measles, mumps, and rubella vaccine: a systematic review of current epidemiological evidence. Archives of Pediatrics & Adolescent Medicine 2003;157(7):628-634. [DARE Abstract]