select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
A study published in NPJ Vaccines looks at adjuvants in vaccines and their association with a lower risk of dementia.
Dr Julia Dudley, Head of Research at Alzheimer’s Research UK, said:
“Dementia is not an inevitable part of ageing. Identifying ways to reduce dementia risk is a priority for research, and vaccination offers an intriguing area of exploration. There have been an increasing number of studies suggesting a link between people who receive certain vaccinations, like the Shingrix vaccine, and a decreased risk of dementia. This study offers a potentially different perspective on what might be linked to this finding.
“In this latest large US-based observational study, researchers are proposing that it might be the adjuvant that is providing a protective effect, rather than the disease the vaccine is seeking to protect against. This study looked at dementia diagnoses in people who had received vaccines with the AS01 adjuvant and those who had a flu vaccine, which doesn’t contain this component.
“An adjuvant is a substance in the vaccine used to create a boosted immune response, designed to give more effective protection upon exposure to the virus.AS01 is in the shingles vaccine Shingrix, and Arexvy, the vaccine to protect against respiratory syncytial virus (RSV).
“They found people who had Shingrix, Arexvy or both of these vaccines were less likely to get a dementia diagnosis within 18 months. They found no difference between the Shingrix or Arexvy in terms of reducing dementia risk.
“One of the strengths of the study is that it adjusted for factors that could influence risk, such as underlying health conditions and some lifestyle and environmental factors. However, as the study is observational and examined past health data, the researchers cannot conclude how the Shingrix and Arexvy vaccines may protect against dementia. We also cannot rule out that the link between vaccine and dementia risk is due to other factors not captured in this study, such as social and lifestyle factors.
“One of the limitations highlighted by the authors was around people not having a dementia diagnosis when they could be living with the condition, which could skew the findings. We do not know if the adjuvant is reducing the risk of dementia or delaying its onset. The follow-up period was only 18 months, so more research is needed to determine the potential long-term effects of the vaccines.
“As we understand more about the biological mechanisms behind any protective effects seen with vaccines, we may be able to investigate new treatment approaches.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:
“This is an interesting, worthwhile and statistically competent piece of work, but a lot more research needs to be done to make good sense of its possible implications for health care. In fact it’s a good example of how scientific and medical knowledge has to be built up through a series of studies, not just a single piece of work.
“Previous research has provided pretty convincing evidence that vaccination against shingles, in older people, can reduce dementia risk. A recent study (published 2024), by the same research team responsible for the new study, found that the reduction in dementia risk is greater in people who had the shingles vaccine now in most widespread use, including in the UK (it’s called Shingrix), than with the previous vaccine (Zostavax).
“However, that study could not provide direct evidence on the reason for the risk difference between the two shingles vaccines. One possibility is that having shingles might increase dementia risk, and that the new vaccine provides better protection against shingles than the old one did, so reducing dementia risk. Another is that there’s some component in the new vaccine that reduces dementia risk. Or it could be some combination of these possibilities.
“The new Shingrix vaccine differs from the old Zostavax vaccine in several ways, but one difference is Shingrix vaccine contain a substance called AS01, while Zostavax does not. Some previous research has indicated that it’s possible that AS01 somehow provides in itself a reduction in dementia risk. AS01 is not the ingredient of the vaccine that directly incites the immune system to develop immunity against shingles. Instead it is an adjuvant – a substance that is intended to help the immune system to respond to the vaccination.
“AS01 is also used as an adjuvant in another vaccination offered to older people; the vaccination against the respiratory virus infection RSV (respiratory syncytial virus), which is now recommended for people in the UK aged 75-79. One of the vaccines in use against RSV also contains AS01. (There’s another available vaccine that does not contain it.) So the research team responsible for the new study used data from a large set of American electronic health records, to compare dementia risk over a period of 18 months after vaccination for older people who had had various different vaccinations containing AS01 (just the RSV vaccine, just the shingles vaccine, or both). These people were compared with older people who had had a flu vaccine, not containing AS01.
“The researchers found that those who had had either of the two AS01-containing vaccines (against RSV or shingles) had a lower risk of being diagnosed with dementia in the 18 months after vaccination than those who had had the flu vaccine. This pattern of lower risk showed up in people who had had just one of the AS01 vaccines, or both, though there were some relatively small (and statistically uncertain) differences in the average size of the risk reduction, compared to the flu vaccine, for different groups.
“What’s still not known from this study is exactly why these risk differences occur. The researchers mention that, in some way, they could occur because having either RSV or shingles might in itself increase dementia risk, so that having a vaccine that makes it less likely to have one or both of those diseases might reduce dementia risk. Or it could be because of some protective effect of the AS01 adjuvant, which is in these vaccines but not in the flu vaccine. (Or some combination of these possible effects.)
“The researchers give some arguments why they feel AS01 itself is likely to play a protective role against dementia. I don’t have expertise in virology so can’t comment directly on those arguments. But it’s at least a possibility, from all the existing evidence, that AS01 could have a protective effect. This study also doesn’t provide direct evidence on how AS01 might work to reduce dementia risk, but the researchers give some suggestions based on other studies as to what could be happening. Again I can’t comment on those.
“It’s because of this inevitable lack of knowledge about exactly how AS01 might be involved in reducing dementia risk that the researchers are asking for more studies, some of them using other research methods, to find out more. I agree with this recommendation, because in my view the results of this study provide a clear justification for looking further. But we’re not yet anywhere near the stage of using the results of the new study to change clinical practice. Also, the new study can’t make the timescale of risk reduction very clear, because the follow-up period to look for dementia diagnoses was relatively short at 18 months.
“This was an observational study – the people weren’t assigned at random to receive a particular pattern of vaccinations, but just did what they would have done anyway in consultation with health professionals. In any observational study, there can be issues about what is causing what. The basic problem is that people who receive different vaccinations will also differ in terms of many other factors – age, sex, what diseases they have previously had or still have, and many more. Some of these factors may be potential confounders, as they are called – that is, there’s a possibility that they are the cause of differences in dementia risk, and not the actual vaccinations at all.
“The researchers did a very thorough job of allowing for potential confounders, by doing something called propensity score matching. This involves setting up a statistical model that predicts people’s chances of having a dementia outcome, regardless of what vaccines they had had, and then matching people who (for example) had had the RSV vaccine but not the shingles vaccine with people who had had the flu vaccine. In this research the statistical model for the matching involves a very wide range of potential confounders. Then direct comparisons are based on these matched pairs of people. That means one can get a lot closer to comparing like with like groups, who don’t differ (on average) in terms of potential confounding factors.
“The process can’t entirely avoid the possibility that there are confounding factors that couldn’t be dealt with in this way, and that’s why the research paper says clearly that unmeasured confounding can’t be entirely ruled out. So there has to remain doubt about whether the risk differences are caused by the different vaccines. This is in addition to the inevitable doubts about which aspects of the vaccines (AS01 or something else as well) might be causes of the risk differences – if indeed it’s the vaccines that do turn out to cause the differences. These are yet more reasons why this research is nowhere near being the last word.”
Prof Sir Andrew Pollard FMedSci, Ashall Professor of Paediatric Infection and Immunity and Director of the Oxford Vaccine Group, University of Oxford, said:
“There are now a number of studies which have shown an association between shingles vaccination in older adults and a reduced rate of dementia in the vaccinated population. The fact that two different vaccine platforms (both live attenuated shingles vaccines and the adjuvanted shingles subcomponent vaccine) saw similar associations supported the idea that the mechanism was as a result of vaccine-prevention of reactivation of the usually dormant shingles virus in the brain. Another virus from the same family, herpes simplex virus (the cold sore virus) has also been associated with dementia raising the possibility that both of these viruses (shingles and herpes simplex) could cause infection, possibly silently and recurrently, in the brain that led eventually to dementia. Unfortunately, there is no licensed vaccine for herpes simplex at this time. However, this latest study published in npj vaccines shows that another vaccine, against the completely unrelated respiratory virus, RSV, is also associated with a reduced rate of dementia. The authors argue that this is because of a non-specific effect of these vaccines on the immune system which generates an environment in our bodies which is somehow protective against dementia, though further studies are needed to confirm this. Such a mechanism could account for the effects driven by both shingles and RSV vaccines. The various studies of the impact of vaccination on dementia are all observational studies which could have a risk of bias, as it can be challenging to adequately control for differences between those who seek vaccination and those who don’t, but the consistent finding across multiple studies makes the observation more convincing. It is premature to be too certain about the mechanism by which vaccines might reduce dementia risk, but these observations provide further incentive for those eligible to turn up for their scheduled vaccination visits to prevent the unpleasant and potentially serious and life-threatening infections for which they were designed, but with the added possible benefit of a longer dementia-free life-span. What’s not to like?”
‘Lower risk of dementia with AS01- adjuvanted vaccination against shingles and respiratory syncytial virus infections’ by Maxime Taquet et al. was published in npj vaccines at 10:00 UK time Wednesday June 25th 2025.
DOI: 10.1038/s41541-025-01172-3
Declared interests
Prof Kevin McConway: No conflicts.
For all other experts, no reply to our request for DOIs was received.