A natural experiment
When the United States switched from a live shingles vaccine to a recombinant one, it accidentally created a clean comparison. Researchers writing in Nature Medicine used that shift to ask whether the type of vaccine matters for heart health years later.
They drew on electronic health records from the TriNetX US Collaborative Network, which covers more than 100 million people at roughly 60 healthcare organisations. The primary analysis compared two groups of 36,460 adults aged 60 and older, matched on demographics, diagnoses and medications. One group was vaccinated between April and September 2017, when 98.6% received the live attenuated vaccine. The other was vaccinated in the same months of 2018, when 93.5% received the recombinant vaccine. None of the participants had immune deficiencies or blood cancers, and none had received immunosuppressants, steroids, chemotherapy or radiotherapy in the year before vaccination.
What the records showed
Over follow-up of up to seven years, the group that predominantly received the recombinant vaccine had a 9.0% lower overall cardiovascular disease burden. Heart failure was 12% lower and ischemic heart disease 10% lower across both sexes. Atrial fibrillation was 7.0% lower. Ischemic stroke was not clearly lower overall, though men showed a 12% lower burden.
No clear differences appeared for myocarditis, peripheral arterial disease, haemorrhagic stroke or transient ischaemic attack. The associations were strongest in the first years after vaccination and weakened later, and they held up when the researchers excluded the COVID-19 pandemic period and tested for secular trends using recipients of other vaccines.
Biological hints, and limits
The authors suggest two possible explanations. Vaccination may change endothelial and immune function, and the AS01 adjuvant used in the recombinant vaccine may reprogramme monocytes in a lasting way, including weaker interleukin-6 responses that could reduce cardiovascular risk. If those immune effects fade, it would explain why the benefit shrinks over time.
The design reduces the bias that flaws studies comparing vaccinated and unvaccinated people, but it still cannot prove cause and effect. Unmeasured factors such as smoking, diet, income and education could not be excluded, and outcomes came from recorded diagnoses. The authors say clinical trials would be needed to confirm the pattern.