Do vaccines kill babies?

A recently released preprint written by Brian Hooker claims that babies who get their recommended vaccines at 2 months old are more likely to die in the next month, especially girls and Black infants, and that vaccines may be linked to higher infant death rates.

This preprint says it found that babies who got their 2-month shots were more likely to die in the next month, but its study design has big problems that can fool people.

The biggest issue is that it only looked at children who had already died and then compared groups inside that “only deaths” set, so it cannot tell whether vaccines raise the chance of death in the real world, where most babies live; this kind of picking your study group based on the outcome can create a false link. This is called selection bias.

Another major problem is that their “unvaccinated” group is not truly unvaccinated: the paper says a child had to have an immunization record match and “minimally have been immunized at least once,” and “unvaccinated” just means “not vaccinated in days 60–90,” so the comparison is not “vaccinated vs. unvaccinated,” but “vaccinated in this window vs. vaccinated at some other time or not recorded in this window.”

Timing also matters a lot: babies who are sick, premature, or have health problems may get vaccines later, may miss visits, or may have different care, and those same health problems also change the risk of death, so the groups can be different from the start. This is called a confounding variable.

The study also does not clearly fix these differences with strong, detailed adjustments (like prematurity, birth weight, NICU stay, major diagnoses, and access to care), so it can mix up “who got vaccinated on time” with “who was healthy enough to come in on time.”

On top of that, vaccine safety is normally judged with much stronger methods: careful trials before approval and large safety monitoring systems after approval that follow vaccinated and unvaccinated people over time using proper denominators, instead of studying only people who died.