COVID-19 mRNA vaccination in young and middle-aged adults was associated with a 25% reduced risk for all-cause mortality between 2021 and 2025, according to a population-based cohort study drawing from France’s National Health Data System (JAMA Netw Open 2025;8[12]:e2546822).
The study—among the largest to date examining long-term safety of COVID-19 mRNA vaccination—found that this association persisted even after adjusting for sociodemographic characteristics, severe COVID-19 death, and 41 comorbidities.
‘Extraordinarily Robust’ Study
To address persistent public skepticism about the safety of COVID-19 vaccines, as well as the lack of data on their long-term risk–benefit profile, the authors examined records from more than 28 million adults—almost half of France’s population. Among this cohort of adults aged 18 to 59, more than 22 million received at least one dose of a COVID-19 mRNA vaccine during the country’s mass vaccination campaign in 2021; roughly six million remained unvaccinated.
Describing the data as “extraordinarily robust,” Juan Gea-Banacloche, MD, MS, an infectious disease physician at MedStar Georgetown University Hospital and a member of the FDA advisory committee for Pfizer’s COVID-19 vaccine, noted that “[the] authors went through great lengths to minimize bias and ‘stress-test’ their results via a number of sub-group analyses and adjustments.”
Still, he acknowledged the study’s observational nature and lack of randomization meant a portion of the reduction in four-year all-cause mortality observed likely reflected underlying differences among the vaccinated and unvaccinated groups. Dr. Gea-Banacloche was not involved with the new research.
Related Content
“The effect of this bias is in full display when the vaccine appears to significantly reduce absolute mortality risk from accidents by 9%,” Dr. Gea-Banacloche told Infectious Disease Special Edition. “As there is no biological plausibility for a vaccine reducing deaths caused by accidents, one has to assume this reflects an intrinsic difference, likely much of it behavioral, between vaccinated and unvaccinated individuals. However, that bias would explain only 9% to 10% of the difference. The relative reduction in death was 25%, so there is 15% unaccounted for. Most of that is likely truly related to the vaccine.”
Still Positive Even With Caveats
This study also reinforced earlier findings on the efficacy and short-term safety of COVID-19 mRNA vaccinations: The vaccinated group was associated with a 74% lower risk for death from severe COVID-19 and a 29% reduction in all-cause mortality within six months of vaccination.
As for the generalizability of these findings, Dr. Gea-Banacloche said potential differences in diet, comorbidities, and demographics are unlikely to change the overall conclusion.
“While these differences might affect the magnitude of the effect, they shouldn’t change the direction of the effect,” Dr. Gea-Banacloche said. “If you give the vaccine to millions of people and see no signal of harm, there is little biological plausibility that vaccinating another million people in a similar country, like the U.S., would reveal that it causes harm. With such a large sample, this study—if people don’t overpromise with it—should help support the concept that COVID mRNA vaccines are safe globally.”
Dr. Gea-Banacloche reported no relevant financial disclosures.
This article is from the February 2026 print issue.