Researchers at the University of Washington reviewed health records for 2,560,035 American children and followed them up to age eight.
Children who got their first MMR shot before age two were no more likely to develop autism than children who had not yet been vaccinated.
The data is from 2015 to 2024. In earlier, mainly European studies, autism was only a third as common as it is in the US today.
Children who get their first shot against measles, mumps and rubella before age two are no more likely to develop autism than other children. That is the finding of a study of 2,560,035 children in the US, published in The Pediatric Infectious Disease Journal.
The study was carried out by physician Todd Burstain at UW Medicine in Seattle, together with Dale Jacques and Jennifer Burstain.
Records from 2.5 million children
The researchers used Cosmos, a database of electronic health records from hospitals and clinics across the US. They included children who had a routine checkup between 12 and 24 months of age during the years 2015 to 2024. The children's records also had to be linked to their mother's. Each child was followed until age eight, or until the last health care visit before December 2024.
Of the children, 91.4 percent had received their first MMR dose before age two. In total, 2.55 percent of the children were diagnosed with autism in childhood. Children diagnosed at 11 months of age or earlier were excluded.
Same risk with and without the vaccine
The researchers divided the children into age windows between 11.5 and 24 months of age. The first windows were one month long, the later ones longer. In each window, children vaccinated during that specific period were compared with children who had not yet been vaccinated. Only diagnoses made after the window began were counted. The method is meant to keep the time before vaccination from being counted incorrectly, which can skew results in this type of study.
The researchers also adjusted for the mother's age, preterm birth, diabetes during pregnancy, the region where the family lived and how disadvantaged the neighborhood was.
The result was an adjusted hazard ratio of 0.97. A hazard ratio of 1 means the same risk in both groups. The upper limit of the confidence interval was 1.03, which rules out an increase in risk of more than about 3 percent. No age window showed any increased risk.
The method is also designed to reduce the effect of children being followed for different lengths of time and seeking care at different rates. The vaccinated children were followed for a median of 853 days, compared with 366 days for the unvaccinated children.
Another vaccine gave the same result
To test whether hidden factors affected the result, the researchers ran the same analysis for a booster dose of the pneumococcal vaccine. It is given at 12 to 15 months of age, and there is no theoretical link between it and autism. That analysis gave a hazard ratio of 1.01. According to the researchers, the fact that both vaccines gave nearly the same result makes it unlikely that hidden factors explain the findings.
The result also held when the researchers added more factors related to birth, removed the neighborhood measure and excluded children with only a single recorded autism diagnosis.
The study is based on health record data and can show associations, but cannot prove causation. Todd Burstain told PsyPost that the size of the study makes it easier to detect even small differences in risk. The researchers write that the findings strengthen the existing evidence that the MMR vaccine is safe.
Newer data than earlier studies
Vaccines prevented at least 154 million deaths over the 50 years from 1974 to 2024, as we have written before. Still, concerns about the MMR vaccine have persisted. They go back to a 1998 study of 12 children, which was later retracted due to scientific misconduct.
Several large studies, mainly in Europe, have since found no link. A 2015 US study of children with older siblings with autism reached the same conclusion. The European studies were based on data from 2000 to 2012. According to Todd Burstain, autism was then about a third as common as it is in the US today, where about 3.2 percent of children are estimated to have autism. The new study is based on data from a period when the diagnosis is much more common.
WALL-Y
WALL-Y is an AI bot created in Claude. A human selects the news, WALL-Y writes the text and a human checks the results before publishing.