Jump to Transcript

In this episode, Lauren and Matt talk with Leah Richmond-Rakerd, PhD who is an Assistant Professor in the Department of Psychology at the University of Michigan. Her research focuses on emotional and behavioral dysregulation across the life course. Here, Dr. Richmond-Rakerd will discuss her recent study on “The Associations of Hospital-Treated Infections with Subsequent Dementia: Nationwide 30-year Analysis” that was published in Nature Aging.  

More resources

Richmond-Rakerd LS, Iyer MT, D'Souza S, Khalifeh L, Caspi A, Moffitt TE, Milne BJ. Associations of hospital-treated infections with subsequent dementia: nationwide 30-year analysis. Nat Aging. 2024 Jun;4(6):783-790. doi: 10.1038/s43587-024-00621-3. Epub 2024 May 7. PMID: 38714911. 

Milne BJ, Atkinson J, Blakely T, Day H, Douwes J, Gibb S, Nicolson M, Shackleton N, Sporle A, Teng A. Data Resource Profile: The New Zealand Integrated Data Infrastructure (IDI). Int J Epidemiol. 2019 Jun 1;48(3):677-677e. doi: 10.1093/ije/dyz014. Erratum in: Int J Epidemiol. 2019 Jun 1;48(3):1027. doi: 10.1093/ije/dyz054. PMID: 30793742. 

NZ Integrated Data Infrastructure: https://www.stats.govt.nz/integrated-data/integrated-data-infrastructure/ 

Visit the Center to Accelerate Population Research in Alzheimer's (CAPRA) website to learn more.

Transcript

Matt Davis:

Public health and medical advancements through the widespread use of vaccinations and antibiotics have made great progress in the reduction of infectious disease. However, infectious diseases continue to be a major public health issue in some parts of the world. Further, no matter how much the science advances, there will always be new bugs for humans to deal with.

But why am I talking about infection on a dementia podcast? You see, the way the body responds to infection is with inflammation, and this whole body inflammatory response isn't just limited to the vascular and musculoskeletal systems. It can also affect the nervous system, including the brain. Some researchers believe that neuroinflammation may be a risk factor for dementia, and studies have started to uncover a link between infection and cognitive outcomes.

Today, we're going to speak with a researcher who used a national registry to examine the relationship between a hospital-treated infection and dementia. Welcome to Minding Memory, a podcast devoted to exploring research on Alzheimer's disease and other related dementias. Here, we'll discuss some of the most compelling research and talk with leaders in the field about how their work is improving the detection and treatment of dementia. I'm Matt Davis.

Lauren Gerlach:

And I'm Lauren Gerlach.

Matt Davis:

We're both researchers at the University of Michigan. I have a Ph.D. in data science.

Lauren Gerlach:

And I'm a geriatric psychiatrist who specializes in diagnosis and management of dementia.

Matt Davis:

I'll work to minimize the use of medical jargon in our discussions.

Lauren Gerlach:

And I'll make sure that the research we talk about has practical, real-world applications to people living with dementia and their care partners.

Matt Davis:

Thanks for joining us, and let's get started. Today, we're joined by Dr. Leah Richmond-Rakerd. Dr. Richmond-Rakerd is an assistant professor in the Department of Psychology at the University of Michigan.

Her research focuses is on emotional and behavioral dysregulation across the life course, which seeks to unpack the mechanisms behind self-regulation difficulties for physical health and the aging process. She's here today to speak with us about her recent study. Leah, welcome to the podcast.

Leah Richmond-Rakerd:

Thank you so much for having me on.

Matt Davis:

Dr. Richmond-Rakerd was the lead author of the study titled Associations of Hospital-Treated Infections with Subsequent Dementia: Nationwide 30-Year Analysis that was published in the Journal Nature Aging. We'll include a link to the study. Make sure to check it out. What got you interested in looking at infection and dementia?

Leah Richmond-Rakerd:

My interest in this topic stemmed from another line of research that my colleagues and I have in which we, as other teams have shown, have found that individuals who experience mental health disorders like depression and other types of mental health conditions are at elevated risk for dementia. There's a variety of mechanisms that might help to explain that association. Inflammation is one.

So, for instance, depression and some other types of mental health disorders have associated with increased levels of inflammation in the body. So I started looking into the research on the role of inflammation in dementia, and I came upon this body of literature on potential associations of infections with dementia because infections, as you've noted in the intro, might be sort of a seed for neuroinflammation. And so that's really what kind of sparked my interest in this, in this topic.

Matt Davis:

How recent... I mean, when did some of this work start coming about? It's just the last five, 10 years, or long... or does it go farther back than that?

Leah Richmond-Rakerd:

Some of it goes a bit farther back, but I would say in the last five to 10 years is when we've seen a growth in the use of sort of epidemiologic methods similar to what we've done in our study. So long-term longitudinal studies with large population registries or other epidemiologic studies.

Also, an increase in the use of non-human animal studies as well, where they can employ more kind of tightly controlled designs to test the role of inflammation in brain disease. So I think that... But in particular, I think the growth of large observational studies longer longer-term prospective studies has been more recent. Yes.

Lauren Gerlach:

And can you tell us a little bit more about the theory or link between inflammation and cognitive function and how infection and inflammation may impact kind of risk for dementia?

Leah Richmond-Rakerd:

Yeah, so inflammation is a natural response from our body's immune system to harmful invaders or harmful stimuli, so things like bacteria or viruses or a sprained ankle. And so indicators of an inflammatory response would be things like a fever in response to infection or swelling at the site of that, a sprained ankle. So this is, as I mentioned, a natural part of our body's attempt to heal and protect itself.

But when inflammation becomes significant or longer lasting, then it can potentially confer harm to the body and possibly the brain. So evidence suggests that individuals who experience systemic inflammation, so inflammation throughout their bodies, may be at elevated risk for cognitive decline. And also neuroinflammation, so inflammation in the brain is implicated in dementia, in particular in Alzheimer's disease.

So there's been this sort of greater discussion in recent years around the hypothesis that infections might seed risk for dementia. I should note a couple of areas that are still, I think, certainly up for debate. One is the extent to which systemic inflammation, not necessarily just infections that involve the nervous system, is sufficient to impact the brain and lead to neurodegeneration, and also the extent to which inflammation is a cause rather than a consequence of neurodegeneration.

So is it the inflammation that leads to brain decline or rather vice versa? I think there's growing research from longer-term longitudinal studies that can measure inflammation or infection well in advance of dementia, as well as more tightly controlled kind of animal experiments that have linked this inflammation to neurodegeneration that suggests that inflammation might be, in factor, a risk factor itself. But I think that's still very much up for debate, and more research is ongoing.

Matt Davis:

I'm sure we'll get into this later. But this idea about inflammation and dementia risk, I immediately started wondering about autoimmune disorders and things that are more like chronic because, I mean, an infection is a transient thing presumably. But I mean, there's probably thinking... people thinking about this and starting to look at it, right?

Leah Richmond-Rakerd:

Absolutely. And I think one of the... There's certainly a debate. Is it more about that acute inflammatory response, or rather, is it more about the chronic inflammatory burden over the long term that might be particularly harmful? And I think many feel that it may in fact more be about the chronic inflammation or inflammatory responses that are quite significant in nature. So for instance, in our current study, we measured infections that were treated in inpatient hospitals, right, so that tended to be more severe.

Other studies have found that infections that are also treated in inpatient hospitals tend to show a stronger association with dementia risk than less severe cases. So that's one sort of indication, is the level of inflammatory response, but then also might this be, these associations be stronger for infections that simmer over the long term and confer more chronic inflammatory burden. I think there's a lot of sort of questions around that too.

Matt Davis:

So most of our listeners are based here in the United States and are familiar with some of the big US sources of data for dementia, but they're likely less familiar with international data. I was wondering if you could tell us a little bit about the data source that you use for your study.

Leah Richmond-Rakerd:

Yes. We worked with the data from the New Zealand Integrated Data Infrastructure or IDI. This is a collection of whole of population administrative data sources that have been de-identified and linked at the individual level. This data resource was established in 2011, but some of the registers have information that goes back much earlier. So, for instance, with respect to hospitalizations, which we used in the current study, criminal offending, social welfare use, deaths, and others.

This was a really significant feat by Statistics New Zealand to do this linkage. So in countries like Denmark or Sweden they're able to use a single national identifier to link different government databases together. But they didn't have that New Zealand, and so they needed to use probabilistic linkage procedures to complete this linkage.

And so, it was a really kind of important statistical feat, and it's a really, really valuable data resource for health services and social science researchers as well as others to evaluate questions of interest and I think public health importance really at the population level. I'd refer interested listeners to my colleague Barry Milne's nice paper in 2019 in the International Journal of Epidemiology. He wrote a nice data resource profile about the IDI, and there's also more information on the Statistics New Zealand website as well.

Matt Davis:

It always makes me feel like the US is behind some of other... what other countries are doing, which is surprising to me. You hear about these... I mean, I have some experience with Thailand data where they, like this, have this national registry, they have a national health survey, and everything's all kind of linked up and go ready for the entire population.

And here in the United States, we still are pretty compartmentalized. I mean, we can get access to 100% claims for some populations, but just I always find it interesting how sometimes, in our countries, we seem to be behind the ball, so to speak.

Leah Richmond-Rakerd:

There's a lot of reasons why. There's that... There's those differences. I think there's really interesting and exciting initiatives though happening here in the United States. So very large-scale administrative databases are available for research, some of which capture a very high portion of the population of interest.

So those would be databases like information about health through the Center for Medicare and Medicaid Services, birth records collected at the state level and aggregated up, education records, income and tax databases. So these sorts of databases, I think, are becoming... administrative databases are becoming increasingly used for research in the United States. There's also a growth in efforts to sort of build up integrated health databases to link health and social services records at the state-level and hire both using sort of state level approaches as well as through the Census Bureau.

So lots of, I think, initiatives to really facilitate this kind of large-scale data linkage at the individual level. But certainly when we don't have the same kinds of government services and things are very sort of decentralized and also sort of the challenges of working with private sector data as well, these all introduce unique kinds of issues in terms of thinking about data integration within the US.

Lauren Gerlach:

Is there anything specifically about New Zealand that listeners should be aware of regarding their healthcare system and how that might differ from the United States?

Leah Richmond-Rakerd:

Yes. So New Zealand has nationwide healthcare, so all New Zealanders have access to free or highly subsidized healthcare. So we certainly always need to think about the impact that that might have on the generalizability of our findings, sort of that unique healthcare context. I'll note that associations of... with respect to the current study, associations of infections with dementia have been found in other nations, not just those with nationwide healthcare, but also the United States.

And I think there's unique benefits that can also come from studying these kinds of questions in countries like New Zealand with universal healthcare. One is that because care is freely or free or highly subsidized to all cost barriers don't generate bias in the subset of unhealthy individuals who seek care. So that's a unique benefit to working with these kinds of healthcare records in this kind of country.

Matt Davis:

So, for your study, how did you specifically measure hospital-treated infections and dementia?

Leah Richmond-Rakerd:

So we measured infections using diagnoses of infections as well as medical conditions that can stem from infections, so things like infectious mononucleosis using diagnoses from inpatient... public inpatient hospital records.

And we also measured dementia using a previously published scheme by Katherine Walesby and colleagues. We used diagnoses of dementia made in inpatient hospitals as well as those in death records and also anti-dementia drug prescriptions from pharmaceutical records. And we collected that information across a 30-year period. So, from 1989 to 2019.

Lauren Gerlach:

Can you give us kind of a high-level overview of what your study found?

Leah Richmond-Rakerd:

We found that individuals who had been diagnosed with an infection had about three times the risk of subsequently being diagnosed with dementia relative to those without an infection. That increased risk was evident for both men and women for younger and older individuals. And it also held after we accounted for people's histories of poor physical and mental health as well as their socioeconomic backgrounds.

We've observed associations across different types of infections. So that included viral bacterial, parasitic, and other infections, as well as Alzheimer's disease and other dementias, including vascular dementias. And importantly, we found that associations persisted across long follow-up. So individuals weren't just at elevated risk for dementia in the years immediately following their infection but were also at increased risk up to 25 to 30 years later.

Matt Davis:

I think it's important to point out that, I mean, you used hazard ratios, which is relative risk, I think, for most listeners who understand. Often a question comes up when I'm working with students, like what's a big association?

And typically, if you're over two, that's big. I mean, it's field-specific, right. I mean, some environmental epidemiologists I work with are looking at just 5% changes, and they consider that big, but two is really big. I mean, were you surprised about how strong the association was?

Leah Richmond-Rakerd:

I was somewhat surprised. As you note, this is a large association, and so the first thing that we did was to look to other papers in particular studies that have used hospital records, so looked at hospital-treated infections like our study. The associations that we identified were not too far off from some in the prior literature that have looked at more sort of severe infection cases.

So I think that's one thing to keep in mind, as we've mentioned, is that we're looking at infections that are severe enough to need inpatient treatment, so that may be a part of the equation. We're also relying on medical records to diagnose dementia. So we're likely... we may be capturing some dementia cases of higher severity or missing cases that are diagnosed in the community.

So I think that severity component is important to keep in mind, but I think that's a piece of the equation here. And it's true that we were able to account for important potential confounding factors of the association, but we didn't... we weren't able to account for everything that might explain this association. So were we able to do that, we might see an additional reduction in the risk that we observed.

Matt Davis:

Did you at all think about trying to capture infection from ambulatory or outpatient things? I mean, I guess I think the answer to this is that you were interested in serious infections, and that's what led you to the hospital ones. But I guess did you think about that at all or consider that?

Leah Richmond-Rakerd:

We did think about that. One of the reasons we didn't is simply data availability. So information on community treatment is available for fewer years in New Zealand, and we wanted to be able to preserve this kind of long-term follow-up. Another is, as you note, looking at hospital sort of treated infections allowed us to zero in on infections that might confer a higher inflammatory burden.

So we might sort of think of this as an important starting point to demonstrate an association and then be able to extend out to test whether this extends to, as you've identified, infections treated in ambulatory care settings or those sorts of settings, those that may be less severe in nature. This is something that we're planning to do.

So we have been fortunate to establish a collaboration with researchers in Norway, where they have information on... they have primary care records for the entire nation. And so, I'm planning replicate the current work in those nationwide general practitioner records to evaluate this question of whether these associations might extend to cases of lower severity.

Matt Davis:

So you touched on this a little bit already, but I'm wondering if we can kind of talk a little bit more. As researchers, sometimes it makes sense to lump all types of dementia together when you talk about population-level things and associations, and what have you when you're talking about disease burden to a population.

But when you start thinking about mechanisms, which I feel like your study is starting to dip into a little bit, to understanding what causes or the etiology of different types of dementias. Sometimes, it makes more sense potentially to look at different subtypes. I know you did some of these analyses, and you mentioned it, but could you talk a little bit more about how the association potentially varied across different types or perhaps didn't?

Leah Richmond-Rakerd:

Yes. So we looked at as different types of dementia, including Alzheimer's disease, as well as other dementias, including vascular conditions. We observed associations across all types of dementia that we examined. Associations were, however, somewhat larger for vascular conditions that is consistent with some prior research.

So that's interesting because it suggests perhaps that vascular mechanisms or dysregulation, so things perhaps like blood-brain barrier dysfunction or other kinds of vascular problems, might help to explain the sort of infection dementia association. We obviously can't unpack that in that level of detail in the context of our current study design, but that is something that is consistent with some prior work and sort of raises that hypothesis.

Lauren Gerlach:

Did the type of infection or the number of infections make any difference?

Leah Richmond-Rakerd:

We observed associations across all types of infections that we looked at, so that included parasitic, viral, bacterial, and other infections. This is also consistent with prior work, including work, for instance in Finnish and UK register data. And we chose to focus on these categories of infections. One, to provide some opportunity for comparison with prior studies and also because there's sort of different hypotheses about the mechanisms that might link infections with dementia.

So some propose that specific types of microbes invade the brain and lead to neurodegeneration. So that would be things like herpes virus or influenza virus. But as an alternative, an alternative hypothesis is that it's more about systemic inflammation throughout the body, generally, that can lead to neuroinflammation and dementia. And so our finding that these associations were evident across different kinds of infections is consistent with that. Also, consistent with that is our finding that as you... You asked about number of infections.

Individuals who accrued more hospital admissions for infections were at a elevated risk for dementia relative to those with fewer. So people who had two or more hospitalizations for infection had about two and a half times the risk for dementia relative to those without, and individuals who had only one had about one and a half times the risk relative to those without. So that's also potentially consistent with this idea that seeing that dose-response association is potentially consistent with this idea that it's about a level of inflammatory burden.

Lauren Gerlach:

That makes sense. And I'm not sure if, in this study, you were able to delve into it, but did you look at all by specific bodily symptom... or system? So, for instance, urinary tract infections, respiratory infections, CNS, or central nervous system infections? Was there any opportunity to do that here?

Leah Richmond-Rakerd:

We did not make those categorizations, but prior research has. So, for instance, as one example, a 2021 study in UK Primary Care Data looked at a breakout by exactly these kinds of different sites that you've mentioned. For instance, urinary tract infections or skin and soft tissue infections and pneumonia. They found associations across infections at different sites.

Another paper from the same year in Finnish and UK cohorts looked at CNS versus extra CNS infections, and they also observed associations across both types, which speaks to that question of whether systemic inflammation is sufficient to affect the brain and suggests that it may not necessarily... infections may not necessarily need to incur central nervous system involvement to do that.

Yeah, so I think interesting findings from that work and would be useful, I think, in our future work to try and sort of break down into those kinds of comparisons to be able to provide some parallel... parallelism there.

Matt Davis:

So there are a lot of different risk factors for dementia that may be associated with infection. I was wondering to what degree did your analyses account for some of these other factors.

Leah Richmond-Rakerd:

Yeah, so we accounted for a few key risk factors in our work. So we controlled for people's histories of poor, physical, and mental health, so specifically their histories of chronic disease, things like cancer, diabetes, cardiovascular disease, as well as previous diagnoses of mental health disorders. And then we also accounted for their socioeconomic backgrounds. So, in New Zealand, there's a measure of the level of deprivation in the neighborhoods in which people live.

And so we accounted for that as well. So we controlled for those three key alternative explanations and continued to observe associations, but we couldn't account for all potential risk factors. Associations could arise from, for instance, a more general vulnerability to poor brain or bodily health, including things like age-related kinds of vulnerabilities like frailty or low education, other factors such as these.

And so those kinds of factors, and other studies have aimed to include those as well, could be a part of the equation here. I do think it's important to note that even if infections aren't necessarily themselves a causal risk factor for dementia, our study and others sort of suggest that they might be an important early warning sign. At least infections of significant severity might be an important kind of early warning sign of risk for a later neurodegeneration.

Lauren Gerlach:

Interesting. This is in the weeds a bit, but considering there's improvements in reducing and treating infections that changes over time, for instance, people born in the 1930s versus 1960s, how did your analysis deal with this or try to handle it?

Leah Richmond-Rakerd:

I think this is a really important question. So different birth cohorts in New Zealand, including those in our study, lived through different diagnosis and treatment regimens because, as you note, there's been historical changes in medical practices to diagnose and treat infections and dementia. A key example is penicillin, the first antibiotic.

So that came available in the 1940s and really markedly changed infectious disease treatment. So the earliest born kind of oldest cohort in our study would not have had access to penicillin as children, but more recently born younger cohorts would have. So that's a difference in kind of early life exposure. The medical records that we used for this study followed a 30-year period much later.

So that was from 1989 to 2019. So all of the participants in our study essentially were experienced the same conditions of diagnosis and treatment in that three decade period. But it's certainly possible that those early life experiences, such as the one that I've described regarding penicillin, might have sort of affected their medical status and infected medical status differently across different birth cohorts in ways that we couldn't fully account for that were sort of unknown to us. So I think it's a really important question.

Lauren Gerlach:

And this is also in the weeds a bit, but I wonder if we might have a brief discussion of ascertment bias or reverse causation, which I think are issues that really come up a lot with these studies that look at associations and dementia risk.

For example, patients with dementia may or may not have less healthcare utilization, as well as changes in hygiene due to kind of cognitive and functional changes that happen with the disease that can put them at greater risk for infection.

And these are always challenges in these types of studies. I think there's some things you did with your design to try to address some of those things, and I was hoping you could tell our listeners about that.

Leah Richmond-Rakerd:

Yeah, I appreciate you asking about this because it is a really... both reverse causation and ascertainment bias are really important considerations in dementia research in part because, as you note, dementia has a very long preclinical phase. So the cognitive decline that accompanies dementia can onset many years in advance of the actual diagnosis.

And in addition, in medical registers in particular, we're navigating the challenge of diagnostic delay for dementia. So what all of that means is that we ideally can use very long follow-up periods to try and address that challenge. So we followed individuals for up to 30 years, and we found that individuals who had an infection diagnosis were not just at elevated risk for dementia in the years immediately following their infection but were also at elevated risk at subsequent follow-up periods all the way up to 25 to 30 years later.

So because those dementia diagnoses were not occurring only very close in time to the infection diagnoses, that reduces the likelihood that they're stemming sort of solely from people presenting to hospital and for infections and then subsequently being assessed for their mental status and cognition, and also reduces the likelihood of that reverse causation kind of process that you've given an example of. Yes.

Matt Davis:

So this next question is probably a bit of a stretch, but I'm curious if you think about sort of at the population level and you see variations in the rates, the incidence of infectious diseases, and we also see differences in dementia prevalence as well, do you think that there's to any degree a correlation between differences in infectious disease rates and any association with dementia prevalence?

Leah Richmond-Rakerd:

If infection is in fact a causal risk factor for dementia, then it could potentially be sort of a piece of the puzzle in helping to explain country-level differences in dementia prevalence. But it's not entirely straightforward as I think sort of the nature of your question. It's interesting if we look at the most recent age-standardized dementia rates from the Global Burden of Disease Study, and age standardization accounts for differences in the age structures of population, so allows for a fair comparison. They estimated the population prevalence of dementia in high-income countries at about 0.7% and slightly lower in low and middle-income nations.

Now, I should note that that's population-wide prevalence. So if we were to estimate the prevalence in older adults, right, we would find much higher estimates, but that's the prevalence in the overall population estimated at about 0.7% and slightly lower in low and middle-income nations. But communicable diseases, including infectious diseases, account for a much higher burden of disease in low and middle-income nations relative to high-income countries. So those two things aren't necessarily operating in parallel.

I think there's a few things we need to think about here. One is we need to think about infections as a potential risk factor alongside many others for dementia. So that would include things like non-communicable diseases like cancers or cardiovascular disease, which also confer meaningful disease burden in particular and including in high-income nations. And also, I think there's sort of the interesting question of whether, say, the impact of infection on dementia risk might vary as a function of some of these other kinds of comorbidities.

So, for instance, might be elevated among people with diabetes or cardiovascular disease. So we sort of need to consider infections alongside other sets of risk factors that may also vary in prevalence across countries. I also think it's important to note that most of the research on infections as a potential risk for dementia, including the data from our study, that comes from high-income countries, but we need more information from lower and middle-income nations where there is this greater burden of infectious disease I think to speak more to this question.

Lauren Gerlach:

I think we've already touched on this a little bit before, but thinking about that, the underlying mechanism relates to inflammation, and there's a lot of other things that cause inflammation, have other people started to look at conditions like autoimmune diseases like lupus and dementia risk or how that might be related?

Leah Richmond-Rakerd:

Yeah, I think this is a really interesting area of research. The answer is yes. So there have been studies of different kinds of autoimmune conditions and risk for dementia, including lupus, as you've mentioned. Also, things like arthritis or inflammatory bowel disease, certain types of thyroid conditions. I would say the data are relatively mixed. So some find associations, other studies either don't or find associations that are much more modest or only for particular kinds of conditions.

So that includes a recent study in Danish administrative registers. I think that those... there's been some mixed findings with respect to autoimmune conditions. I mentioned at the start that something that we've been involved in, and other teams as well, is testing associations between mental health disorders and risk for dementia. So depression has been identified as a dementia risk factor, and our team and others has also shown associations with other kinds of mental health disorders as well.

And now, a variety of mechanisms, including psychosocial factors, could link mental health disorders with dementia, but inflammation is one potential mechanism. So I think those associations are useful to consider here as well. But I do think to the extent that we can sort of broaden the search space and be looking for or evaluating associations with other potential risk factors that in which... or excuse me, in which inflammation is implicated, would be really useful to speak to this hypothesis.

Matt Davis:

So you mentioned a little bit some work that you're getting started in Norway, but we're curious regarding this line of work, what's next for your team?

Leah Richmond-Rakerd:

So yeah, so I am excited about that replication sort of plan. So, as I mentioned, we are planning to replicate the current findings in nationwide general practitioner data from Norway to evaluate how these associations might vary depending on kind of infection or dementia severity. Also, we are making use of linked family data. So we have been engaged in digitization and linkage of sibling birth records in the IDI. Some of that with support from CAPRA, which we're really appreciative for, and that is something that will allow us... will be a useful causal inference tool.

So what we're planning to do is to conduct sibling difference analyses, both with respect to the current question around infection and dementia risk, but also potential other dementia risk factors to better account for some of those potential unmeasured risk confounders that we weren't able to in the context of our current study. So that would be things like genetic background or other family-level risk factors. So that kind of linked sibling data resource will be made available to other users of the IDI as well.

So that's sort of a next step with respect to this specific question. I'm also, I think, following some of the interesting work happening with other teams around things like vaccination studies and how that might reduce risk for dementia and other sort of tests of these questions using other kinds of experimental methods that can help to zero in on whether this association might be causal and also if so trying to parse the mechanisms in ways that we can't do in this sort of population-level observational design.

Matt Davis:

Do you think the field is getting... in terms of the amount of work being done in this space, is it getting closer towards being able to do systematic reviews or some kind of meta-analyses?

Leah Richmond-Rakerd:

Yes. I think the amount of research that is, or amount of empirical studies that are now coming out, have facilitated some systematic reviews and will continue to do that. So, for instance, there have been some syntheses with respect to vaccination and other topics. I think as more research comes out, then that we'll facilitate more robust kinds of qualitative and quantitative syntheses as well.

Lauren Gerlach:

Great. Is there anything we haven't covered that you want listeners to know?

Leah Richmond-Rakerd:

There's a couple important caveats to our study that I think are important to mention. One, as we mentioned, is that this... we focused on infection cases of higher severity. So this question of the extent to which associations might extend not just to infection cases of lower severity but also potentially earlier stage dementia, I think, is an open question, especially acknowledging that medical registers will miss some... many cases of dementia that are diagnosed in community settings.

It's also, I think, important to note that dementia was still a rare event, right. So of the individuals in our study population who had an infection hospitalization, only 5% were also diagnosed with dementia in the 30-year period in the registers that we used. So that means the vast majority were not. So individuals who have an infection should get their infections treated, and that may include more severe infections that require a hospitalization.

Matt Davis:

Regarding your work in Norway, if you're looking for collaborators to go visit Norway, please let us know. We are standing by. Leah, thanks so much for joining us today.

Leah Richmond-Rakerd:

Thank you so much for having me. I really enjoyed talking with you.

Matt Davis:

If you enjoyed our discussion today, please consider subscribing to our podcast. Other episodes can be found on Apple Podcasts, Spotify, and SoundCloud, as well as directly from us at capra.med.umich.edu, where a full transcript of this episode is also available on our website. You'll also find links to other resources we've created specifically for dementia research. The music and engineering for this podcast was provided by Dan Langa.

More information is available at www.danlanga.com. Minding Memory is part of the Michigan Medicine Podcast Network. Find more shows at michiganmedicine.org/podcasts. Support for this podcast comes from the National Institute on Aging at the National Institutes of Health, as well as the Institute for Healthcare Policy and Innovation at the University of Michigan. The views expressed in this podcast do not necessarily represent the views of the NIH or the University of Michigan. Thanks for joining us, and we'll be back soon.