In this episode, Donovan & Matt speak with Geoffrey Hoffman, a faculty member and researcher at the University of Michigan School of Nursing. Dr. Hoffman discusses the transition to the role of caregiver for partners of persons diagnosed with dementia – and the unique emotional burden and stress that they undertake. Dr. Hoffman also discusses using the Health & Retirement Study (HRS) to better explore the emotional stress and onset of depression in caregivers.
Article Referenced in this Episode
Harris ML, Errickson J, Ha J, Hoffman GJ. Depressive Symptoms and Caregiving Intensity Before and After Onset of Dementia in Partners: A Retrospective, Observational Study. Med Care. 2022;60(11):844-851. doi:10.1097/MLR.0000000000001771 [PMID: 36038513]
Transcript
Matt Davis:
A decline of cognitive functioning can cause tremendous stress among those who experience it. There can be confusion, fear, and uncertainty about what's going on and what to expect. And for those who live with a partner, the emotional burden is often shared between the two. As one person in the dyad experiences a decline in their level of cognition and function, their partner often finds that their role in the relationship evolves as they take on new tasks to help support their person while also continuing the usual tasks they have always done before. For some, the transition into the role of caregiver can be a difficult and isolating time. Informal generally unpaid caregivers, including partners, provide an estimated 18.5 billion hours of unpaid assistance to people living with dementia. It's a huge largely hidden workforce that experiences unique emotional stress. In this episode, we'll speak with a researcher who has examined how the onset of dementia can affect depressive symptoms in their partners. I'm Matt Davis.
Donovan Maust:
And I'm Donovan Maust.
Matt Davis:
You are listening to Minding Memory, a podcast devoted to exploring research on Alzheimer's disease and other related dementias. Today we're joined by Dr. Geoffrey Hoffman. Dr. Hoffman is an assistant professor in the Department of Systems, Populations, and Leadership at the University of Michigan School of Nursing, and a faculty affiliate of the Institute for Healthcare Policy and Innovation. He's a health services researcher with a background in health policy. His research focuses on quality of care and Medicare policies affecting physical functioning and safety for older adults. He's published on a variety of topics, including risks and outcomes associated with fall injuries, family caregiving for older adults, and how Medicare reimbursements influence older adults' long-term functioning and care. Dr. Hoffman's here today to tell us about his recent study that examined depressive symptoms among partners of individuals who develop dementia. Geoff, welcome to the podcast.
Geoff Hoffman:
Thank you, Matt. It's a pleasure to be here with you and Donovan.
Matt Davis:
Dr. Hoffman was the senior author of the study titled, "Depressive Symptoms and Caregiving Intensity Before and After Onset of Dementia in Partners", that was published in the journal Medical Care. The study used data from the Health and Retirement Study, and was co-authored with his colleague, Melissa Harris, who is currently a scholar in the National Clinician Scholars Program at Duke University. We'll make sure to include a link to the study attached to this episode. So to start things off, Geoff, what got you interested in studying the health of caregivers?
Geoff Hoffman:
I've been interested in family caregiving for a long time. I think there's a personal reason and an academic reason. Personally, my grandmothers were aging. I was close with them. One just about got to the centenary mark and was full of life and vibrant and active, and engendered fall risks constantly, and so I was fascinated with how that worked with injury, but also how she cared for herself and the help that she needed. They both lived in assisted living facilities. So it got me thinking about informal caregiving. And then the other reason academically was I had lived abroad for a while in France and they had a very different system there for caring for older adults and family members living with one another.
When I came back to the states, learning a lot more about how the system that we all believe that we rely on when it comes to caring for us when we're older, Medicare doesn't actually provide a lot of the care that most adults will end up needing, which are, as you said earlier, for long-term services and supports. So the sorts of things that help with if you're going to have an injury, if you're losing your balance, your gait isn't working well, or all sorts of issues. Medicare is really not going to cover it. And the other program, Medicaid, that can cover it requires really low incomes for eligibility. So for those two reasons, I thought caregiving was a really fascinating place to sink my teeth into.
Matt Davis:
There's been other work that looked at depression among caregivers, so why did you specifically want to examine symptoms around that time of dementia onset?
Geoff Hoffman:
We think there's a lot of really good literature out there. There's a burgeoning literature on Alzheimer's disease, given it's growing prevalence in the broad caregiving literature. A lot of attention has been paid to the burden on caregivers, the stress process that creates emotional health effects, and just spillovers on work economic effects of caregiving. When it comes to the literature on dementia and caregiving, again, it's really broad, and what's typically done is, to simplify greatly here, comparisons of before and to after. So we have this term that's commonly used in the literature, which is dementia caregiver, and I think it's an important term that helps us think about these extraordinary burdens for often spouses who provide a large proportion of that family care for partners with dementia.
I think what we thought might be interesting to think about was rather than thinking about a threshold that's crossed where you suddenly become "dementia caregiver", what does the actual pathway look like as you enter into that state, if you will? And so we thought about analyzing it in a slightly different way using some modeling strategies where we could see, "Hey, what happens when the partner might have mild cognitive impairment? What happens as you cruise through those different phases to the partner, both in the intensive of care they provide and their emotional health?
Matt Davis:
That's really interesting. It's much more about a transition into caregiving, right?
Geoff Hoffman:
That's right. We thought there's probably an interesting story to tell where for sure if you're a dementia caregiver we know that the levels of burden are tremendous compared to non caregivers, and when I say the burden of care, to define my terms, the proportion of individuals with dementia receiving care, as well as amongst those receiving care, the amount of weekly care received. Both of those are much greater than in the population without dementia, including people who have heart failure and cancer and stroke. The dementia care burden is very high. But for us, for Dr. Harris and I, there was this black box of does a switch just turn on and suddenly when your partner transitions into the state where they've either been clinically diagnosed, or there's some indication through a cognitive survey that the partner has dementia, suddenly caregiving increases very steeply at that moment, or does some of that increase, is that incurred in a prior period? And if so, that would have interesting clinical and policy implications, we felt.
Donovan Maust:
So Geoff, we've talked to a couple previous guests who've used HRS, or the Health and Retirement Study, but I think you're the first person we've spoken to where your paper took advantage of how HRS uniquely can be used to study caregivers and caregiving. So can you explain to us a little bit what's unique about HRS that allowed you to do this, allowed you to look at caregivers?
Geoff Hoffman:
Sure, I'd love to. So I've been a fan of the Health and Retirement Study, the HRS, for a long time. It's hosted at Michigan, of course, and I think your listeners have heard a lot about it. What I'm drawn to in terms of studying caregiving is that it allows for identifying some of the things I mentioned. So is someone receiving care or not? If so, how much care they're receiving on average in the recent past? When I say the recent past, prior to the date of the survey when the survey is given.
So the way it works is they will ask questions of the respondent, or for individuals who may have dementia or functional limitations such that they can't do a telephone interview, they'll ask a proxy, a series of questions, and in order to get questions about care receipt, the respondent will be asked about a number of functional activities. So these will be things like, "In the past two years, have you had trouble bending down, picking a penny off a table, walking up a flight of stairs, walking one block," et cetera, and this is called the Negi checklist, which is a overly technical term for your listeners. But if someone says that they've had difficulty with one of these, they'll then be asked about some more common groupings of their daily life activities. These are called activities of daily living. So if someone has a problem, and they're flagged on that Negi checklist, the interviewer on the HRS will then ask the respondent or their proxy, "How are you doing when it comes to things like eating, feeding yourself, getting dressed every day, toileting, bathing, transferring in and out of bed?"
And they'll separately be asked a set of questions about instrumental activities of daily living, which are things like, "Hey, how's it going with things like managing money, going shopping at the grocery store, using the phone, cooking," things like that, and that's an opportunity for them to say, "Things are going great," or, "I have some difficulty with one or more of those ADLs," as they're called, or IADLs, activities of daily living or instrumental activities of daily living. And finally, if someone is flagged as having said, "Yeah, I have some difficulty some of the time with one or more of those ADLs and/or IADLs," they're then asked, "Do you get help for those? And if so, from whom?"
And that's where you start to get into the richness of the caregiving data. So they'll say, "What's the gender of the caregiver?" They'll ask if you get care for each one of those ADLs and IADLs, again, ADLs being eating, dressing, toileting, bathing, IADLs, things like managing money. "Who helps you with it? Is it a stepchild? Is it a grandparent? Is it a daughter? Is it a son? A partner? And then, on average in the past month, how much care did they give you? Was it every day? Was it just one day a month? How many hours a day on average was it, if it was every day? Was the caregiver paid?" So without belaboring it anymore, there's just a nice set of rich information that you can piece together. Did they get ADL support? Did they get IADL support? About how much of each did they get, which you can guesstimate with HRS. You can't do it perfectly because they have to protect some of the information about caregivers, so I can talk about that if people are interested. So we've just found it to be a really rich resource.
I know there are other ones out there, like NHATS, National Health and Aging Trends Survey, which are arguably a little more in depth. If I may just keep going without the risk of boring listeners, the thing that's exceptionally powerful about the HRS is that they can interview the spouse or partner in the same household. And so, if the spouse or partner is actually the caregiver, you can link the spouse or partner's information to the care recipient's information, and that was the gold we were mining in this particular study.
Matt Davis:
And I must say we are all fans of the Health and Retirement Study. So that was a great just overview of activities of daily living and instrumental activities of daily living. Just to be clear, these are good measures of function that have been validated in high quality, right?
Geoff Hoffman:
Absolutely. These have been used historically and are commonly used in the literature. Sometimes people use a set of five ADLs, or a smaller set, but they've been used quite commonly and validated.
Matt Davis:
So I'm curious, one of the unique features of HRS, I mean we want to get into your study here in a second, but is its ability to link the survey data that they acquire to claims data, billing records. Can you do that for partners of enrollees as well?
Geoff Hoffman:
You absolutely can. So a lot of these data become restricted and private for obvious reasons, because the claims data contains sensitive information. But if you go through a process with the HRS, you can gain access to linked Medicare claims for individuals who are respondents to the HRS. You don't get everyone. You get about 80% or so, and there are differences we've discovered between those 80% who allow the HRS to use their Medicare beneficiary ID number and the 20% who don't, based on socioeconomic factors. But generally speaking, you can really get into a rich set of data about healthcare utilization, in addition to the extraordinarily rich and robust set of survey data, including data about cognition and caregiving that are in the HRS. That's a powerful thing. And I think you also asked can you get data on the spouse? So the nice thing is as long as two people are living in the same household and they're both interviewed during the same survey wave, you can both get a linked rich set of survey data, including cognition for both, emotional health for both, socioeconomic factors in addition to the utilization data and the claims.
Matt Davis:
That's a really unique aspect of this data. I'm just curious, was there anything else out there that you considered using, or is pretty much HRS the thing that you had to do for this type of study?
Geoff Hoffman:
Well, I believe, if I'm not mistaken, Donovan may have some insights. He may have used the NHATS data that is also linkable to Medicare claims. I have not used that data set. My understanding is there's some really rich stuff to mine there. They define caregiving a little bit differently. They actually go interview the caregivers about their care experience. Stepping back for half a second to how I was walking through how the HRS defines a caregiver, primarily through the care recipient, they define who provided care to them, how much care was received. I think that's powerful. I think it's important to ask people what do they believe they're getting? How much are they getting? Who's helping them? I think that's, for me, a conceptually powerful thing that's grounded in some of the gerontology conceptual frameworks of how we should think about aging and what people need. But I think that NHATS has this very powerful lens, which is what are the caregivers experience? What's their burden like, that's directly tied to the caregiving experience? We're maybe giving too much away here, but I think we're inferring that using the HRS data.
When we link the data to the spouse and partner, we're trying to do our best to infer that the emotional health that the partner or spouse is talking about has to do with the caregiving experience, whereas in the NHATS, I think much more direct questions are asked about that. Correct me if I'm wrong, Donovan, but I've spent a lot of time with the HRS just because I like to do some healthy econometrics, and they have so much good data. They also have geographic data about caregivers and care recipients that my understanding is you can't get always with the NHATS.
Donovan Maust:
Yeah. So I think it's always about the trade-offs of different data sources. I think specifically for your paper, Geoff, HRS was probably better just because of the scale you used for depression that I think will get into a little bit later, has a little bit more data or nuance to it than in NHATS, where I believe they ask what's called the PHQ-2, which is just really a two item screener about depression. And so I think that wouldn't have quite been what you needed for your paper. Just moving on, so for this paper, how did you all go about identifying dementia?
Geoff Hoffman:
The acknowledgement that I'm in the presence of a geriatric psychiatrist, yeah, so my knowledge base is much-
Donovan Maust:
It's plenty.
Geoff Hoffman:
Yeah. So we relied on the giants that come before us. And so someone who I know is probably a fan and who's been on this podcast, Ken Langa, has done a lot of work of building the cognitive measures and assessing them and validating them in the HRS. So we use what's called the TICS, the Telephone Interview Cognitive Survey, which was a shortened version of a cognitive survey that was given in person. They tried to make it shorter for use on a telephone survey. Initially it was validated and implemented on what was called the ADAMS Survey, which was an early version of the HRS. The HRS consists of different waves with different cohorts, and the ADAMS cohort was an older, I believe 75 years and older, panel, or cohort rather, that had a higher prevalence of dementia. And they, I believe, and correct me if I'm wrong, but had access to clinician information and they were able to have a pretty good gold standard to identify whether this TIC Survey that was for shortened for the telephone was doing a good job of identifying, in an accurate way, whether individuals actually had dementia.
And then building off that, since we knew that worked pretty well, we then used the ADAMS to look at different cohorts where there's lower prevalence, and found that it does a reasonably good job of identifying individuals with no cognitive impairment, those with mild impairment and those with dementia. And so this has been studied by the giants, and validated over time and used frequently, so we sat atop those giants using it. Specifically, we used Ken Langa's and Donald Weir's algorithm, which is a way to stratify individuals starting with a cognitive score that's zero to 35 based on memory tests and cognitive tests where they stratify it into those three groups I mentioned: no dementia, cognitive impairment no dementia, and dementia. So we used that as our approach.
Donovan Maust:
So you had TICS assessments for a given person for multiple waves, and then essentially your exposure of interest was when someone crossed a cognitive threshold from normal to impaired, as defined by the TICS?
Geoff Hoffman:
That is exactly correct. So we tried to collect as much information as we could about that individual using that threshold point, and lining everyone up on that threshold point, and then looking as far back as we could and as far forward as we could for the partner's information. And so, I'll try to be as clear as I can here, because this got really tricky even in the paper, to be clear, are you talking about the individual who crossed that threshold into having dementia or the partner? And we're using both of those pieces of information, so I'll try to be as clear as I can as we talk about it.
Matt Davis:
But big picture, it was determined based on the cognitive functioning of the individual, not on any kind of diagnosis or anything like that, right?
Geoff Hoffman:
That's correct.
Matt Davis:
Okay. Yeah, which is a unique thing with HRS, which is why it's special it has these ongoing cognitive assessments. We did an episode on TICS before, so hopefully listeners will check that out as well. One thing that we haven't talked much about, though, is measures around depression. Could you tell us how these depressive symptoms, we might call them, are assessed in HRS, and how that scale works that you used?
Geoff Hoffman:
Yeah, and again, I will not try to be overly humble, plead I'm not a particular expert in depressive symptomatology, but we use something that is in the HRS called the Eight-Item Center for Epidemiologic Studies Depression. It's a mouthful. We use the abbreviation CESD, which is a scale that asks individuals to basically self-report depressive symptoms based on eight questions. This also has been studied extensively. So just a brief background on it, it's a range of zero to eight, where a higher score means you have more depressive symptoms. And just to be clear, we're talking about something different than depression itself. So this is not a scale that was intended to measure depression. People try to use it to say, I've seen literature, "If you have a score of three or more, or two or more, than you're depressed." But if you look back at the original development of this scale, it wasn't intended to identify prevalence in the population or at an individual level.
The questions themselves, zero to eight, these eight questions ask things like, "Over the past week," and again, the past week being the week prior to the administration of this survey, "Did the respondent feel any of the following much of the time in that week? Were you depressed? Was everything an effort? Was your sleep restless? Did you feel alone? Did you feel sad? Did you feel it was hard to get going in the morning? Did you feel happy and did you enjoy life?" So two of them are framed in positive ways, "Were you happy? Did you enjoy life," and six in negative ways. And getting really technical, the two that are framed positively, you do what's called reverse code. So if someone was happy, they get a score of zero. If they weren't happy, they get a score of one. And you add up all these things. So if you felt depressed, if your sleep was restless, if you didn't feel happy, that's three points right there and you just add them up like that.
And I should say this was also a shortened version of the initial original CES Survey scale. That was also done with awareness that the HRS can be a really long survey for respondents, especially those who are frail and might have difficulties sitting through a really long interview.
Matt Davis:
Does HRS have other measures on anxiety and other things as well?
Geoff Hoffman:
Yeah. Yes, they do. The HRS, again, in light of having a really long survey... So keep in mind it was designed by economists and so it has maybe the best warehouse of information about household resources and assets that just go into, "Did you have a second home and a third home and stocks and mutual funds?" And there's a lot of good stuff. It's called the Health and Retirement Study that help you think about the aging process outcomes in the peri-retirement period, if you will. And so there's a host of other types of information that the HRS will try to go out and collect not every wave, not all the time. So they trust interviews individual respondents every two years, but what they will sometimes do are put out modules that ask about specific questions. So recently they put out one on COVID. They've done things on multiple different topics.
There is a survey that they put out called the Psychosocial Survey that they do effectively every other wave for individuals. They actually administer it every wave, but it only goes to a random half of the original sample. And so if you're person A, you'll get it, let's say, in year 2010, and then 2014, if you're still taking the survey. A long-winded way to say yes, there's a lot of rich psychosocial stuff, so measures of anxiety, other measures of depressive symptomatology, not just the CESD, loneliness, did you have chronic life stressors, the list goes on. It's a really rich set. The problem as a researcher is if you only have half the sample in the HRS when you're trying to do big analyses already, sometimes the sample sizes can get a little small when you start going down to those who have dementia and then link to their partner's data. So it's a little hard to dig into those extra amazing data sometimes.
Donovan Maust:
Sometimes in papers, say you're interested in looking at what happens before and after a diagnosis of dementia, you might look at one group before there's a diagnosis of dementia, and then another group after there's a diagnosis of dementia. In this particular case, you have the same people, and you're looking at the same people before the diagnosis, and then the same people afterwards. So can you take a step back and say what are the advantages of being able to use the same group of people to ask your question before and after?
Geoff Hoffman:
The first way that you described where we effectively take a cross-section of the population at a given point in time and we say, "Okay, this is a group that's dementia caregivers, and this is a group of spouses or partners whose spouse or partner does not have dementia," however you want to slice and dice the data, those say something interesting about the world. They describe the world as it looks, if you will. So when you look at studies like that, and we ourselves did that. We showed that the amount of care, the care intensity, the hours per week of care provided for dementia caregiving partners versus not was much, much, much greater. And that basically tells you, wow, there's a group out there that's doing a whole bunch more.
The issue is if you're trying to make inferences from that cross-sectional analysis where you simply compare this group to that group, what you want to do, what everything inside you wants to do after reading that data is to say, "Oh my gosh. If my partner suddenly develops severe cognitive impairment, my caregiving is going to skyrocket from here to here, and my depressive symptoms are going to skyrocket from here to here." That's the inference we want to make. And the type of inference that we want to make when we read those data is called a causal inference, that A causes B. My partner develops this, A. This is what will happen to me, B. And there's nothing wrong with assuming that there's most likely some sort of causal mechanism there. That's absolutely correct. My partner develops dementia and I have a lot more responsibilities to care for behavioral psychological symptoms, for self-management of other diseases my partner may have, of going to appointments, diagnostic tests, managing medications, for sure. And maybe as the partner, I also have less time for myself. All those things can be at play.
The issue is we don't actually know if there are other things happening at the same time that are different across those two groups for dementia caregivers and non dementia caregivers. There may be wildly different things going on in those two groups that we don't learn about, more or less, mostly, if we do the type of analysis that you mentioned, Donovan. And so when you look at it the way we did, what we're trying to do is to say if you look over time at the same individual, what's going to be so different about that individual over time? If we're going, let's say, five years before their partner develops dementia and five years after, and you look every two years, what else would explain these sharp breaks that we're seeing in trends of their care intensity and their emotional health, if it's not the partner developing cognitive impairment? Because, using technical fancy terms, we're risk adjusting for other things that might explain that.
And so when you exploit the fact that you're looking at someone over time, you're basically using them as to address an account for those unaddressed unaccounted for factors when you simply compare two groups, dementia caregivers and not. I'm not sure how good a job I did describing that, but it's a powerful way to control for other explanations, other mechanisms, so you can just focus in intensely on what we hope is the instrument, the mechanism, that describes these changes in the partner, which is the increasing cognitive impairment of the partner.
Matt Davis:
So overall, what did you find?
Geoff Hoffman:
And I want to be sure to couch this very carefully so that the results don't sound in any way like this is minimizing anything about this exceptionally challenging development for families that happen all the time. So what we found was that absolutely, if you simply compare, as we said, partners who are caring for someone with dementia versus not, their depressive symptoms are higher. The amount of care or care intensity they provide is higher, but when you look over time, as we said, we saw that in the period prior to the partner developing dementia, there were increases in depressive symptoms and in care intensity, and those were 3%, 9% increases every two years, every survey wave, so this steady increase over time in emotional health burden and the care burdens for the partner. And then once you crossed that threshold where the partner did develop dementia, things seemed to stabilize a little bit. By stabilizing, I mean care intensity stayed roughly constant and emotional health more or less stayed constant, if ticked down a little bit.
And so we can talk about what this means, and I'm curious if there are other thoughts, but what this suggests is that if we stick to this historical way of interpretation of there's a dementia caregiver and non-dementia caregiver, we're missing some important developments that occur during the process of becoming a dementia caregiver, which is increasing burdens and emotional health burdens in what presumably is the mild cognitive impairment period.
Donovan Maust:
Were you surprised by what you found?
Geoff Hoffman:
I think we were surprised by what we found. That was probably not what we were expecting. I think what we expected to see was, when that threshold has crashed, there's a sudden increase in demand for care and emotional health burdens. Thinking about it some more after the fact, and talking to some other individuals who've had these experiences, I think we started to understand that this might make sense. For instance, someone actually wrote to us, a dementia caregiver who read the paper, and said this aligned with her experience.
And I hope it's okay to share the story. I will not, of course, give anyone's name, but I'll just generalize what I heard was that what can happen before there's a clear signal to the family that something is really amiss with cognition, where trust is lost and broken over a period of time, and the partner doesn't understand why their partner is having challenges, forgetful, maybe not meeting expectations that have been built into a relationship over a long period of time, and that when this particular individual sat during a cognitive screening test with their partner, and that partner couldn't answer those questions well on the memory aspects of the TICS, a light bulb went off for her, and this outwelling of empathy occurred where it was like, "Okay, the pieces come together for me now." I think that made a lot of sense to us, whereas I think initially when we saw the results we thought, well, maybe what happens is that over time as individuals have more experience, we know that you can become accustomed to greater burdens, and that emotional health also is what economists call sticky.
We know from the retirement literature, after people retire, they can become depressed for a little time, but then their emotional health rises back to where it was initially, that we have equilibrium as we do with our weight and other aspects of life. So I thought that was our driving explanation previously, but I think there are others available as well. And I just have to say there are also other methodological explanations. It could have been something about the way that we've designed this that the sicker people drop out of the survey, for instance. But we tried our best to address as many of those things as we could with some very extremely boring additional analysis that I won't bother you with.
Donovan Maust:
That's a good scientist, to work the limitations of your paper into the podcast, so we appreciate it.
Matt Davis:
What I found interesting was, I mean, the rise in depressive symptoms before onset makes sense. But the decline afterwards I found interesting, but then I started thinking maybe there was some kind of a value, and at least in some of the people that I've talked to that have experienced this in real life, the value of knowing a diagnosis. Because there's debates around, "Should we diagnose? What's the value of diagnosis," and all that kind of stuff. But it made me start to wonder maybe there is indirect values to having an answer. And then you shift gears from, "What's going on," the confusion, to, "Let's handle this, and let's get those systems in place." But I found it interesting. But you didn't look at diagnosis specifically. It was actual onset, so that's a different thing. But I do wonder how many people in the cohort that you looked at actually got that diagnosis, got that answer from the medical community.
Geoff Hoffman:
Yeah, these are such great points, Matt. I agree with everything you say. I think what's particularly interesting in this disease is that unfortunately there's not a cure for it, so this is different than if you're diagnosed with cancer and you can proceed to a series of known steps that hopefully can mitigate or get rid of the disease. But what's particularly unique about this set of diseases is that while there's no cure, it does require a substantial amount of these things that we led off the podcast talking about: long-term care needs, declines in functioning in those ADLs and IADLs that precipitate potential crises in care availability. You need maybe to move in with a child as your ability to get dressed every day, or go to the doctor, or manage, if you have heart disease, or, well, other diseases that often are comorbid with dementia. Crises can be precipitated. You need to set up care plans. So it's different than you develop cancer and the spouse needs to jump into gear and contact all the specialists and whatnot.
In this case, it's a process that involves making sure that care supports are set up, and that's very hard to do in the US. And so it does fall on the backs of caregivers, who tend to be spouses in these particular cases. And then I'll just do a tip of the hat. Maybe a conflict of interest, but the three of us have worked on a study as well where... And Matt, you and I have worked on a paper where we know that not everyone who is flagged on this routine cognitive screening every two years, the TICS, with the HRS survey, not all of them, in fact, only about a third of them get a clinical diagnosis within three years. And so it's particularly fascinating that this isn't a diagnosis. Just suddenly a light is beamed down and people know. It's something that the families themselves must know that's going on that they see every day.
Matt Davis:
So you also looked at some of these relationships among different subgroups. What did those results show?
Geoff Hoffman:
Yeah, so we were curious. We did the same analyses by gender and by race and ethnicity. We know, again, from the broad caregiving literature, that there's a disproportionate amount of care in the US by some tens of millions, 30, 40 million caregivers, that are done by women, and this has important tolls on the workforce of people having to drop out or go down part-time and forego retirement contributions, all sorts of important things that are part of this weird long-term care system that we have in the US that's on the backs of informal care family and friend caregivers. So we know since there's a disproportionate burden, we were curious... And also different prevalence of depressive symptomatology, for instance, by gender, women versus men. If we would see different patterns by sex and race and ethnicity, we also know in the caregiving literature that there are importantly different considerations that go into deciding when a family member might say, "Mom, Dad, you have to come live with me." Or for a spouse, their willingness to accept help from the medical system or to put a family member or a spouse into a nursing home, for instance.
And because of these important different views on aging across cultures, we thought it would be important to look by race and ethnicity. However, methodologically, when we threw this into the blender, we didn't see substantially different results that seemed to generalize across. So in other words, we saw different levels of care intensity and depressive symptomatology across these groups, as we know exist in the population from millions of other studies, but in terms of this process that we are looking at over time, it didn't look like there were notable differences worth talking about by sex or by race and ethnicity. And I just want to say, I tried to be careful about this before but I should really hit this one more time is, what we're not saying is does this in any way our results diminish that there are important cultural considerations, or by gender, extraordinarily important things to think about about our caregivers in the US, or, importantly, in no way do our findings suggest that being a dementia caregiver is not as bad as we thought or anything of that nature. That is not what this means.
We know, the literature is clear, the burdens on dementia caregivers, the need for clinical and community supports and services, which are underfunded in the US and only available to really low income people. They are not sufficient or satisfactory in any possible way. I think what we're just simply adding to that important correct literature is that we might want to move the lens over to not just incorporate that period of dementia caregiving, but a little before that to recognize these supports and services are already limited as is, but we might want to open them up more, have clinicians start checking in to family members who have partners who have mild cognitive impairment.
Matt Davis:
So any ideas for next steps in terms of where your science might go following this?
Geoff Hoffman:
I absolutely must do a tip of the hat to the first author of this paper, Melissa Harris, who is a wonderful scholar who graduated from the University of Michigan in our program at the School of Nursing, is now a National Clinician Scholars Program fellow at Duke University. And I know she's been doing a lot of work on trying to qualitatively understand the experiences of dementia caregivers, caregivers of individuals with dementia, and non-pharmacological interventions. So she's been doing some wonderful work on that, that maybe at some point might be worth hearing about on this podcast or elsewhere. For me, I'm trying to think about the health system aspects of this, because so much, as we've been discussing, of the burden on dementia caregiving falls on families and friends.
What other services out there, whether in the Medicare program, the Medicaid program, community-based supports through area agencies on aging, differences in intensity of treatment across the country, how do these things influence whether someone is diagnosed as having dementia and "gets into the system" so that we can at least start providing referrals to these supports and services? Are treatment patterns different depending on if you have a caregiver or not, or live in a part of the country that has high intensity care that is more likely to treat patients? Is it good care? Is it bad care? So I'm trying to look at the health system part now that we've looked at what I like to think of as the black market part, which is this giant family and friend caregiving unit in the US, and see how those interact.
Matt Davis:
This is all really important work. Geoff, thanks so much for joining us, and thanks to all of you who listened in.
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 our seminar series and the data products we've created for dementia research. Music and engineering for this podcast was provided by Dan Langa. More information available at www.danlanga.com. Minding Memory is part of the Michigan Medicine Podcast Network. Find more shows at uofmhealth.org/podcast. 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.
Listen to more Minding Memory podcasts - a part of the Michigan Medicine Podcast Network.
