With the best of intentions, American parents immunize an average of four million children each year with the measles-mumps-rubella (MMR) vaccine, a practice so common that no one thought twice about it until a British researcher in 1998 raised the controversial possibility that the vaccine could produce autism in small children.

The study created questions in the US and a furor in England, where MMR vaccination rates have since dropped. The Institute of Medicine formed an Immunization Safety Review Committee to look at the evidence behind a possible MMR-autism link and found no such danger. Strict guidelines ensured there were no conflicts of interests between the committee members and vaccine manufacturers. Marie McCormick, Sumner and Esther Feldberg Professor of Maternal and Child Health and chair of the Department of Maternal and Child Health at HSPH, headed the 15-member committee, which issued a report and testified before Congress on the issue last month.

McCormick recently discussed the report with Harvard Public Health NOW:

"The committee could find no evidence to support an association between the MMR vaccine and autism. In fact, the epidemiological evidence is quite consistent in the other direction. Having said that, the committee did recommend continued research into the possible association between the vaccine and the development of autism. This vaccination is something that we require of children and that means that the government should meet the highest standard in terms of assuring safety.

"Getting sick with measles is not insignificant. People forget what it’s like because they haven’t seen the disease in a generation, but measles epidemics have a case fatality rate of 1 to 2 per thousand, so that for every 1,000 cases, two children will die and between one and six percent will get pneumonia. For every 100,000 cases, 100 to 200 will get encephalitis. And one to 18 kids per 100,000 cases will develop a very bad form of encephalitis called subacute sclerosing panencephalitis that is an ongoing, progressive form of dementia.

"So the consequences of real measles are severe. Mumps and rubella are different. They are not quite as severe, but they can still cause a variety of side effects. Mumps in adults can cause sterility in males, and rubella, or German measles, can cause congenital malformations in children born to women who contract rubella early in their pregnancies.

"The trade-off between a potential and poorly documented outcome of autism and the real consequences of disease I think leads to one conclusion–until we know of a definitive link between MMR and autism, we should keep inoculating with the vaccine.

"There are really two parts to the story about why this vaccination became an issue. First, detecting autism in infancy can be difficult because the signs are very subtle. Usually, a child is diagnosed during the second year of life, which is exactly the same year in which children receive the MMR vaccine. So the diagnosis of autism may occur coincidentally at the same time as immunization.

"The second part of the story starts in 1998 when a British gastroenterologist by the name of Andrew Wakefield published a paper in the Lancet about 12 individuals, 10 of whom had autism and the other two being characterized as having post-vaccine encephalitis. They also had severe gut problems. He endoscoped them and saw that they had collections of lymph cells in the walls of their guts. He felt this picture was consistent with a subacute, ongoing viral infection. The parents of the children claimed that the symptoms had started after the children had been inoculated with MMR, but Wakefield himself said in his paper that this did not in any way establish a relationship between MMR and this constellation of gut problems, or enteropathy, and autism.

"Wakefield has continued to investigate, and he presented to our committee some preliminary results that he claims show persistence of measles virus in the gut after MMR inoculation. The immunologists on the committee were not particularly convinced of his suggestion, but again this is preliminary information.

"Even if he does show that measles survive in the guts of some children after MMR vaccination, that is only one step in the process. The mechanisms by which this persistent measles and enteropathy contribute to autism are totally unknown, and the mechanisms that he has postulated in some of his papers were actually argued against as not being possible by another person who presented to our committee.

"And so the first step of documenting the persistence of virus has not even occurred, but clearly the issue has raised a lot of anxiety, even though the research right now is fragmentary.

"Wakefield never said to stop vaccinating children for measles, mumps, and rubella. What he has said was to have separate vaccinations for each. He feels there is something in having the three viruses in the one vaccine that sets up some sort of abnormal reaction.

"Although it is possible to vaccinate children separately, the risk you run is that you prolong the time that you leave a child vulnerable to the disease. Also, you may increase the risk that the immunizations will not be completed, and you will certainly increase the cost of vaccination by three separate visits, which may present a financial hardship to some parents.

"There is another aspect that is concerning everybody. The number of autism cases appears to be increasing, although the epidemiology is so incomplete that the increase is not entirely clear. The higher numbers may reflect changing definitions that include more children or better diagnoses. California, in particular, has reported a dramatic rise in the number of cases they are managing in their educational system.

"Even the onset of the problem is still controversial. Most cases appear to be the result of genetic predisposition or early fetal insult. However, some observers have described a regressive form of autism in children who appear to be progressing normally. More recently, reviews of family videotapes of children thought to have regressive autism reveal subtle indications of developmental problems earlier than the second year.

"Clearly, this question is of great importance in the possible MMR-autism link, since it is regressive autism that is thought to result from MMR. Even if this form of autism exists, it’s the minority and not the majority number of cases.

"That’s not to say the issue should be dropped. The burden of disease plus the burden of autism is so significant that continued attention should be given to be absolutely sure that an association between the MMR vaccine and autism doesn’t exist, even if only in rare cases."