Intended for healthcare professionals


MMR vaccination and autism 1998

BMJ 1998; 316 doi: (Published 07 March 1998) Cite this as: BMJ 1998;316:715

This article has a correction. Please see:

Déjà vu—pertussis and brain damage 1974?

  1. Angus Nicoll, consultant epidemiologist,
  2. David Elliman, Consultant community paediatrician and district immunisation coordinator,
  3. Euan Ross, professor of community paediatrics
  1. Public Health Laboratory Service, London NW9 5EQ
  2. St George's Health Care, London SW17 0QT
  3. Kings College School of Medicine and Dentistry, Mary Sheridan Centre, London SE11 4QW

    News p 724

    The media excitement and public concern after a Lancet report linking measles, mumps, and rubella (MMR) vaccine with autism1 kindles a sense of déjà vu. It is highly reminiscent of similar scares over pertussis in the 1970s,2 which resulted in much suffering and many deaths from pertussis both in Britain and internationally. 2 3

    Britain's vaccination programme has hugely reduced the incidence of diphtheria, haemophilus meningitis, measles, polio, pertussis, congenital rubella, and tetanus.4 As the incidence of these diseases has fallen vaccine safety has assumed greater importance, especially in parents' minds. Any safety issue requires cool scientific consideration.3 Here the hypothesis is that MMR leads to a non-specific gut condition permitting the absorption of non-permeable peptides, which in turn cause serious developmental disorders.1 Supportive evidence consists of cases referred to a gastroenterology group. The data published comprises 11 boys and one girl, each with bowel abnormalities and serious developmental regression (nine had autism). In eight children parents reported regression starting shortly after the children received MMR.1

    An editorial accompanying the article and a recent review by the World Health Organisation list the considerable evidence against this and previous related theories from the same group. 3 5 Since each year over 600 000 British children receive MMR in their second year, an age when autism can typically manifest itself, chance alone dictates that some cases will appear shortly after vaccination.3 Cases will be selectively referred to a group known for its interest in MMR, inflammatory bowel disease, and autism, so the hypothesis rests on clinical anecdote rather than an epidemiologically sound base.

    Proved serious vaccine reactions are characterised by specific clinical or laboratory findings, but the non-specific nature of the developmental and gut abnormalities in these cases is striking, and no precise case definition is offered.1 No vaccine viruses were reported in the children's biological specimens, though the researchers have previously reported viruses in bowel tissues of children with inflammatory bowel disease, findings which others have been unable to confirm.3

    Epidemiological evidence is unsupportive: the WHO found no links between measles, MMR, and inflammatory bowel disease5; and a survey of conditions associated with autism did not mention inflammatory bowel disease.6 National data seem to indicate a rise in the incidence of autism, but it started over a decade before MMR's introduction in 1988 and showed no change at that time (M Bax, D Lawton, Family Fund Trust, unpublished data). This evidence suggests either no causative association or one that is exceedingly rare. These and many other data relating to MMR safety have been reviewed by the Joint Committee on Vaccination and Immunisation, which found no case for changing vaccination policy. Unproved theories are no basis for dropping a vaccine of proved global safety and effectiveness. 3 5

    Despite the lack of evidence of a causal relation, and the experience of other hypotheses from the same group (linking first wild measles, then measles vaccine, and latterly MMR with bowel disease) not standing up to independent scrutiny 5 7 much parental anxiety has resulted. MMR immunisation rates have begun to decline and those at the “sharp end” of immunisation— general practitioners health visitors and community paediatricians—are experiencing parental inquiries.8 Any decline in immunisation, or the giving of MMR as three injections at annual intervals (as suggested by one of the report's authors), will undo the recent near elimination of measles and rubella in the UK.8

    The experience with pertussis in the 1970s was also based on anecdotal case reports linking pertussis vaccination with infant brain damage.9 Again a temporal link between a vaccine and a devastating childhood condition whose natural peak onset was at the very time when most children received that vaccine was misinterpreted as a causal relation. A national study eventually showed that, while there was a temporal association with encephalopathy, any risk of lasting damage was so rare as to be unquantifiable.10 But the initial report, then as now, attracted media attention; parental and professional anxiety soared; and national immunisation rates fell from 80% to 30%. The number of susceptible children rose, and in the 12 years after 1976 three major pertussis epidemics accounted nationally for over 300 000 notifications and at least 70 deaths. The suffering of families experiencing long miserable illnesses was considerable, and in some cases long term damage ensued. Some parents came to believe that an immunisation they had approved had damaged their child.

    There are differences between then and now. The connection of encephalopathy with pertussis vaccine was biologically more plausible than the link proposed for MMR and autism. The original national study10 has already shown no link between measles vaccine and long term developmental disorders.11 Detection of vaccine reactions is more efficient, with international data sharing and a careful eye on safety by independent scientific experts on the Joint Committee on Vaccination and Immunisation and committees of the Medicines Control Agency. Surveillance results in product withdrawal when there is clear evidence of a safety issue.

    In the 1970s immunisation had a low priority, and evidence based information for those doing the immunising was minimal. District immunisation coordinators did not exist, and vaccination rates slumped partially because it was unclear whose responsibility it was to do anything about them.12 The pertussis experience must not be repeated with MMR vaccine. While vaccine can be guaranteed to be without any risk, this has to be weighed against the huge advantages of protection against disease. Seeds of concern have been sown among parents and no doubt will continue to be spread. Those advising families must make sure parents can base their decisions on hard science and evidence.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    View Abstract