select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
A systematic review published in Alzheimer’s and Dementia: Translational Research & Clinical Interventions looks at the association between prescribed drugs and dementia risk.
Rachel Richardson, Manager at the Methods Support Unit at Cochrane, said:
“Although systematic reviews are a great way of bringing together and analysing the results of multiple research studies, there are many important steps that should be taken. An important part of a review is an assessment of the overall certainty of the findings and this does not appear to have been carried out. Assessing the certainty means considering factors such as potential biases in the included studies, whether the results are likely to reflect important clinical effects, and how relevant the findings are to different groups of people.
“Although the press release acknowledges the possibility of confounding, the potential impact of this cannot be overstated. For example, the authors state that antidepressants may be associated with an increased risk of dementia. However, there are many factors that increase the risk of dementia, for example, loneliness. Loneliness is also associated with depression, for which antidepressants may be prescribed. Therefore, the link between antidepressants and dementia may not be a causal one.”
Dr Richard Oakley, Associate Director of Research and Innovation at Alzheimer’s Society, said:
“One in three people born today will develop dementia in their lifetime, making it the biggest health and social care issue of our time. If we can repurpose drugs that have already been shown to be safe and approved for use for other conditions, this could save millions of pounds and decades it takes to develop a new dementia drug from scratch, and get us closer to beating dementia.
“This research provides some initial groundwork and indicates which drugs have potential for being repurposed for dementia and should be prioritised for further investigation. Some antibiotics and antivirals, certain vaccines, anti-inflammatories, and medication to treat high blood pressure and high cholesterol were among the drugs most associated with a lower risk of dementia.
“But we can’t draw definite conclusions from this study alone. This was an observational study so more work is needed to understand the careful balance between how the drugs work, their specific doses, interactions with other drugs and the individual’s own health, and how that can impact a person’s risk of developing dementia. Clinical trials will help to answer some of these questions in the context of dementia.
“Research will beat dementia – find out how to help at alzheimers.org.uk.”
Dr Baptiste Leurent, Associate Professor in Medical Statistics, UCL, said:
“This study is of limited relevance to the public.
“It summarises findings from 14 studies that had looked at the association between data from medical records and dementia. As expected in large dataset, studies have identified various characteristics ‘associated’ with dementia.
“This does not indicate in any way that the medications mentioned are linked with increasing or decreasing the risk of dementia, which is how the public may generally interpret ‘associated’. For example, medications commonly prescribed in older people will appear ‘associated’ with dementia, but this is not very exciting news.
“The quality of the studies reviewed appear generally poor. This review is only based on observational studies, and does not include any randomised clinical trial, limiting the causal link that can be made. Unfortunately, the review excluded any study which aimed to look specifically at associations between the medications and dementia, and only selected studies that ‘explored’ the association between dementia and multiple information. The risk of spurious findings is such study is very high.”
Prof Francesco Tamagnini, Neurophysiologist at the Reading School of Pharmacy, University of Reading, said:
“This research strengthens our understanding that inflammation is important in how dementia develops. We still don’t fully understand exactly how this process works, especially since inflammation in the brain is complicated and can have different, sometimes opposite effects. One interesting possibility from this research is that common anti-inflammatory medications that older people take for heart health (like low-dose aspirin to prevent blood clots) might also help protect brain function and reduce dementia risk.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:
“This is an interesting study and some of its conclusions could be important in the future. However, the results cannot lead directly to treatments for dementia. They could point the way to more research into some existing drugs, which in turn could lead to some treatments. But that would require a lot more research, over a fairly long period of time.
“The top line of the press release and its first paragraph could be interpreted to mean that all or most of the existing drugs in the above classes (antibiotics, vaccines, and anti-inflammatories) are linked to reducing the risk of dementia. But that’s not what the study found. It found evidence of a relationship for only a few drugs in each of the categories. For example, of all the vaccines, only four are specifically mentioned in the paper as being linked to reducing the risk of dementia (hepatitis A, typhoid, a combination vaccine for hepatitis A and typhoid, and diphtheria).
“It makes sense to look for existing medications for other diseases that may affect dementia risk. That approach had some successes in treating COVID-19, for example. Much is already known about the safety and potential side effects of existing drugs, thanks to clinical trials of those drugs and monitoring of their routine use. This could shorten the time it takes to start using one of these dementia medicines, possibly reducing the time and expense of safety testing. In addition, many of the existing drugs would have already been in use long enough for their patents to expire, so they are likely to be relatively cheap to produce and acquire.
“Some studies have already been done on the use of existing dementia medications, so it makes sense to review the studies systematically. However, this new systematic review takes a particular approach that is perhaps unusual. It reviews only previous studies that looked at a very wide range of existing medications in relation to dementia risk. Those studies often took data from electronic health records, and examined which drugs had been prescribed to a broad population of people in some defined region of the world.
“Studies of this type often use very large data sets, as they involve a large number of people and medicines. That’s why they often use quite sophisticated statistical methods, such as machine learning, to analyze the data. And different studies use different statistical methods. This can make it very difficult to compare the results of different studies.
“Often, in a systematic review, researchers use a specific statistical approach, called a meta-analysis, to get a joint estimate of (for example) the strength of the association between taking a particular drug and some health outcome of interest. But (despite the mention of study pooling in the press release) the researchers made the decision not to use any formal pooling methods. This is because the studies they reviewed were very diverse in their statistical methods and, indeed, in their overall approach.
“I think it is the right decision not to group formally. But it means that it is completely impossible for researchers to give a joint measure of the strength of any of the associations they discuss.
“So they can say (for example) that there seems to be an association between some antibiotics and a lower risk of dementia, but they can’t give any estimate of how much less risk of dementia there seems to be in people who take a certain amount of a certain antibiotic, compared to people who don’t take it. That’s one reason why this research isn’t close to proposing any treatment for dementia.
“Because there is no formal pooling, researchers use (in some cases) what is known as a “vote count” approach, of looking at and reporting how many studies found evidence of a particular link between a drug or class of drugs and dementia risk.
“But vote-counting approaches are known to be potentially very misleading. This is because they generally do not take into account the different strength of the evidence from the different studies being reviewed.
“In this review, the studies analysed not only used different statistical methods, but had very different objectives. Half of them (7 out of 14) were specifically looking for associations between existing medications and dementia risk.
“But the other half did not. Instead, they looked for combinations of many factors, including drug prescriptions, but also many other things, such as patient characteristics and histories, that could produce a good predictor of dementia risk.
“In these types of predictive studies, a factor strongly related to dementia could be left out of the statistical model, simply because of the way it is related to other factors in the model.
“Therefore, these predictive models would have a different basis for identifying links between drugs and dementia risk than the other studies that directly looked at the links between drugs and dementia risk. So simply counting the classes of drugs that appeared in each of the 14 studies could give a misleading impression (the researchers don’t seem to have quite fallen into that trap, but it adds one more complication).
“One more reason why these findings are not close to leading to treatments for dementia is that the associations being described between drugs and dementia risk are not necessarily describing cause and effect. The studies being reviewed did not necessarily look for cause and effect.
“The researchers in this new review place a lot of emphasis on this. A medication may be associated with a reduced risk of dementia, but that does not mean that the medication causes the risk reduction.
“For example, the disease for which the medication was prescribed could actually be the cause of the change in risk.
“Or, as the researchers mention, there’s the possibility of reverse causation, in which the early stages of dementia or cognitive decline, or some condition that increases the risk of dementia, could lead to a particular treatment being prescribed.
“Some of the individual studies that were reviewed would have been careful to try to take these questions of cause and effect into account, but in very large observational studies of the kind that were reviewed it is impossible to do so perfectly, and in any case it cannot be considered in detail in a review of this nature.
“So, in summary, we have a review that suggests some classes of drugs that might be worth investigating further, to learn more about whether and how they are associated with dementia risk. In fact, it cannot go any further.
“And it reports mostly on broad classes of drugs rather than on specific drugs, because the researchers found very few specific drugs that were associated with dementia risk in more than one of the studies they reviewed. They found only nine specific drugs, out of the thousands included in the studies they reviewed, that had evidence of an association with dementia risk in two studies (and none with evidence in more than two studies).”
Dr Julia Dudley, Head of Research Strategy at Alzheimer’s Research UK, said:
“It is encouraging to see large studies exploring whether medications already licensed could be repurposed as a dementia treatment. As these medications have already been shown to be safe for use in people, this could potentially speed up the process of testing in clinical trials, as well as being significantly cheaper.
“This study looked at a range of medications including vaccines, antibiotics and anti-inflammatories, and showed that these could be linked to a lower risk of dementia.
“Although these findings are interesting, it is too early to say if these medications can reduce the risk of dementia. Researchers will need to confirm these findings in clinical trials and investigate how these medications could potentially protect the brain against dementia.
“Future research would also need to address if factors such as education level, socioeconomic status, biomarker information, or having Alzheimer’s risk genes could be influencing these results.
“At Alzheimer’s Research UK, we know that if we are to cure dementia, we must develop a toolkit of treatments which tackle different areas of the disease and can be used alongside each other. Research into repurposing medication may give us the answer to do just that.”
‘Data-driven discovery of associations between prescribed drugs and dementia risk: A systematic review’ by Benjamin R. Underwood et al. was published in Alzheimer’s and Dementia: Translational Research & Clinical Interventions at 12 noon UK time on Tuesday 21 January.
Declared interests
Rachel Richardson: I am employed by Cochrane, who produce and publish systematic reviews.
Dr Baptiste Leurent: No conflict of interest
Prof Francesco Tamagnini: None
Prof Kevin McConway: Previously a Trustee of the SMC and a member of its Advisory Committee.
Dr Julia Dudley: No conflicts of interest