A cohort of 2.5 million children shows no association between first MMR dose timing and autism risk, using a study design built specifically to rule out the biases that have complicated past research
Few vaccine safety questions have been studied as extensively, or remain as persistently contested in public debate, as whether the MMR vaccine is linked to autism. A new study adds one of the largest and most methodologically careful entries yet to that evidence base.
The research, published in the Pediatric Infectious Disease Journal, set out to address a specific, recurring weakness in prior observational research on this question: studies examining vaccine timing are vulnerable to bias from how families use the healthcare system, since parents who already have developmental concerns about a child may delay vaccination, which can create the appearance of an association where none causally exists.
To guard against that problem, the team used a retrospective cohort design drawing on the Cosmos electronic health record database, identifying 2,560,035 US children with linked birth-parent data and following them through age 8. Rather than simply comparing vaccinated and unvaccinated children, the researchers used what’s known as an age-based landmark analysis: they anchored their comparisons to specific ages between 11.5 and 24 months and asked whether receiving a first MMR dose at that point was associated with a later autism diagnosis, a design intended to strip out the timing-related bias that has complicated earlier work.
The results showed no meaningful association. In the primary cohort, children vaccinated between 11.5 and 24 months showed a near-identical autism risk compared with those vaccinated later. The same pattern held across every other age-specific cohort the researchers examined.
The researchers also ran a check on hidden bias. They looked at a completely different vaccine, the pneumococcal booster typically given around the same age, one with no plausible biological reason to affect autism risk at all. If something in how families use the healthcare system was skewing the results without the researchers realizing it, that unrelated vaccine should have shown a similar artificial link to autism, even though no real connection exists. It didn’t: children who got the pneumococcal booster showed essentially the same autism risk as those who didn’t. That’s reassuring on two fronts. It suggests the study wasn’t being tricked by the kind of bias that has muddied earlier vaccine research, and it makes the MMR finding, no link to autism, more likely to reflect the real picture rather than a statistical fluke.
The authors frame their contribution as methodological as much as substantive, arguing that age-anchored analytic approaches of this kind should become standard practice for evaluating vaccine safety using observational health record data going forward, given how much past controversy in this area has stemmed from studies vulnerable to exactly the biases this design was built to exclude.
This study follows two decades of research reaching similar conclusions, including a 2004 CDC-linked study in Pediatrics that found no meaningful difference in vaccination timing between children with and without autism, and a large UK database study published in The Lancet the same year that found no increased risk of developmental disorders following MMR vaccination. The current study’s scale, 2.5 million children, and its bias-resistant landmark design place it among the more rigorous additions to that existing body of evidence.
Sources and further reading:
Burstain TL, Jacques D, Burstain JM. Association Between First MMR Vaccination Before Age 2 Years and Childhood Autism in a U.S. EHR Cohort of 2.5 Million Children. Pediatric Infectious Disease Journal, July 28, 2026.
This article was researched and sourced by Global Biodefense editors and reported with Claude AI assistance for drafting and editing.