Stress, Burden, and Well-Being in Dementia and Nondementia Caregivers: Insights From the Caregiving Transitions Study

Abstract

Background and Objectives

Few population-based studies have directly compared caregivers of persons with dementia to caregivers of persons with other disabilities (nondementia caregivers). We enrolled dementia and nondementia caregivers who were providing substantial and sustained care and compared these groups on measures of caregiver stressors, appraisals of burden, and well-being.

Research Design and Methods

Caregivers (N = 251) who provided continuous care for at least 1 year and at least 5 h per week were recruited from the population-based REasons for Geographic And Racial Differences in Stroke (REGARDS) study. Caregivers reported on dementia caregiving status, stressors, burden, and well-being.

Results

Forty-seven percent (n = 117) reported caring for a person with dementia. Dementia caregivers reported more stressors, providing more care for self-care and behavioral problems than nondementia caregivers. Dementia caregivers also reported higher appraisals of stress and burden, and more depressive symptoms, but did not differ from nondementia caregivers on mental and physical health quality of life. In multivariable-adjusted models, adjustment for the total number of care recipient problems attenuated differences between dementia and nondementia caregivers on burden and depression measures.

Discussion and Implications

Dementia and nondementia caregivers showed relatively few differences in indicators of overall well-being in this population-based sample, perhaps because both groups of caregivers in this study were providing substantial care. Dementia caregivers may require special assistance with dementia-specific problems such as behavioral problems. Clinical interventions and policy changes targeting highly burdened caregivers are needed to support them in allowing their care recipients to age in place at home.

Caregiving, Older adults, Stress process theory

There are an estimated 50 million people worldwide currently living with dementia and dementia prevalence is predicted to triple in the next 30 years (Ferri et al., 2005; World Health Organization [WHO], 2020). Dementia is a major cause of disability and dependency among older people with nearly 10 million new cases diagnosed each year (WHO, 2020). The physical, psychological, social, and economic impact of dementia affects not only people with dementia, but also their caregivers, families, and our whole society. The majority of persons with dementia live in the community and receive care from family and friends. In stark contrast to the predicted rise in cases of dementia (Alzheimer’s Association, 2018), the number of available caregivers per frail older adult is predicted to sharply decline from seven potential caregivers to just four within the next 10 years (American Association of Retired Persons [AARP], 2013). Lower fertility rates, smaller family size, low unemployment, lower marriage rates, and geographic mobility resulting in disparate families are just some of the factors contributing to the decline in caregiver availability.

National representative studies comparing caregivers for persons with dementia and caregivers for persons with other disabilities but not dementia have found that dementia caregivers report more physical and mental health problems, caregiver strain, financial difficulty, employment and relationship complications, and have less time for social participation (AARP, 2018; Badana et al., 2019; Ory et al., 1999; Riffin et al., 2017; Wolff et al., 2016). A report from the National Caregiver Survey first suggested that there was “something unique about caring for a demented older adult, apart from caregiver characteristics and level of caregiving involvement, which leads dementia caregivers to experience greater strain” but could not determine what made dementia caregiving so unique (Ory et al., 1999). The increased burden of dementia caregiving is often attributed to the behavioral and psychological symptoms of dementia (BPSD; Fauth et al., 2016). However, persons with dementia and BPSD are more likely to attend specialist clinics and, therefore, are more likely to be over-represented in studies that use convenience samples by recruiting dementia caregivers from these settings (Fernández et al., 2010).

Studies comparing dementia and nondementia caregivers often do not require minimum thresholds of caregiving or current caregiving and in many cases include caregivers providing relatively little care, especially in nondementia caregivers. For example, the Ory et al. (1999) study required only that caregivers report some unpaid caregiving during the previous 12 months. Dementia caregivers in this study provided more hours of care per week (17 h) than nondementia caregivers (12 h) and were more than twice as likely to provide more than 40 h of care per week or continuous care (38.1% dementia and 17% nondementia). Increased caregiver burden leads to negative consequences for the care recipient with increased rates of institutionalization and higher health care costs (Gaugler et al., 2009; Zhu et al., 2015). Identifying why dementia caregiving is more burdensome is important to target interventions to relieve the burden.

Both general stress process models (Lazarus, 1984) and stress process models specific to caregiving (Haley et al., 1996; Pearlin et al., 1990) distinguish between stressors (e.g., primary or objective caregiving stress), appraisals (subjective stress or burden), and outcomes (e.g., depression, quality of life, and health). We adopted this model to organize our approach to evaluating the impact of dementia and nondementia caregiving on caregiving stressors, caregiver appraisals of burden, and caregiver outcomes including depressive symptoms and quality of life. Our aim was to compare dementia and nondementia caregivers on these elements of the stress process and to examine whether differences between the groups in appraisal and outcomes were due to factors including demographics, hours of care, and total care recipient problems in a population-based prospective study of incident caregivers.

Design and Methods

Data Source

The Caregiving Transitions Study (CTS) is a nested case–control ancillary study of the REasons for Geographic And Racial Differences in Stroke (REGARDS) national cohort study. The REGARDS study is a national longitudinal investigation of 30,239 African American or White adults in the United States who were 45 or more years of age when enrolled from 2003 to 2007. Residents of the “stroke belt” (southeastern states with higher stroke incidence) and African Americans were oversampled by design. Verbal informed consent was sought from eligible persons and a baseline computer-assisted telephone interview (CATI) was used to obtain demographics, risk factors, and medical history information. Subsequent in-home physical exams were performed including blood pressure, biologic specimens, and physical measurements. Participants are followed with semiannual follow-up CATIs that inquire about possible strokes, other major health events, hospitalizations, and screen for changes in cognitive functioning. These semiannual follow-up interviews continue at the present time for more than 13,000 participants who are still actively participating in the REGARDS study. Approximately 9 years (mean 9.4 years 2013–2016) after the first in-home visit, active participants were invited to undergo a second in-home exam and extensive CATI assessment using a protocol similar to the first. Only persons who transitioned into a caregiving role between these two time points were eligible for inclusion in the ancillary study. Additional information on the design, sampling, enrollment, second assessment, and long-term follow-up procedures used in the REGARDS study has been described in detail elsewhere (Howard et al., 2005, 2017; Roth et al., 2013).

CTS Enrollment

During the REGARDS baseline CATI, each participant was asked: “Are you currently providing care on an on-going basis to a family member with a chronic illness or disability? This would include any kind of help such as watching your family member, dressing or bathing this person, arranging care, or providing transportation.” Participants who answered “yes” were categorized as caregivers, and those who answered “no” were characterized as noncaregivers (Roth et al., 2019). Approximately 12 years after enrollment into REGARDS (M = 11.7 years, range = 9.1–14.1 years) and about 9 years after the first in-home visit, a similar question was asked during a routine follow-up CATI to obtain updated information on caregiving status. For those who answered “yes,” additional screening questions were asked about the care including the age of the care recipient, their relationship with that person, how many hours of care they provided per week, when (year and month) they started providing care due to that persons’ disability, and major health problems that the person had. Caregivers were asked specifically if their care recipients had “Alzheimer’s disease, another form of dementia, or serious memory problems?” and those who answered “yes” were considered to be dementia caregivers. They were asked if they were “the main or primary family member or friend who provides care for this person.” If they responded yes they were classified as primary caregiver. All caregivers were required to provide at least 5 h of caregiving per week and to have been providing care for at least 1 year. Participants who reported that they were not caregivers at the first REGARDS baseline CATI but had become caregivers at some point between the first and second REGARDS in-home assessments were considered to have transitioned into caregiving. This ancillary study was reviewed and approved by the Institutional Review Boards of the University of Alabama at Birmingham and Johns Hopkins University. Further details on eligibility criteria and recruitment are described in detail elsewhere (Roth et al., 2019).

Measures

Dementia

The Washington University Dementia Screening Test (AD8) was administered to all caregivers (including those who reported that they were not caring for someone with dementia; Galvin et al., 2005). Scores of 2 or more have been shown to reliably differentiate nondemented from those with impaired cognitive function (Larner, 2015; Ryu et al., 2020). The AD8 was used to confirm the validity of the caregivers’ reports that they were or were not caring for a person with dementia.

Caregiving stressors

Caregivers reported the number of hours per week spent on caregiving activities. Participant reports of hours per day were capped at 12. The number of hours per day was multiplied by 7 to calculate the average hours of care per week. In addition, caregivers reported on the number out of six activities of daily living (ADL) and eight instrumental activities of daily living (IADL) impairments in the care recipient, and with how many the caregiver provided assistance (Katz et al., 1963; Lawton & Brody, 1969) The 24-item Memory and Behavior checklist allowed caregivers to report the specific behaviors exhibited by the care recipient, with subscales for cognitive and memory problems, emotional problems, and disruptive behaviors (Teri et al., 1992).

Appraisals of stress and burden

Several widely used single-item indicators of subjective strain were included. These include an item described by Schulz and Beach (1999) asking how much of a mental or emotional strain it was on the caregiver to provide the care (no strain, some strain, or a lot of strain). In previous studies, this one-item caregiving strain measure has proven an effective measure of the mental and emotional strain associated with caregiving and to predict caregiver well-being and mortality (Fauth et al., 2016; Schulz & Beach, 1999; Selwood et al., 2007). Three single items from the National Study of Caregiving (NSOC) asked caregivers about the overall emotional, physical, and financial difficulties associated with care provision with responses on a Likert scale (0 = no difficulty; 5 = very difficult; Kasper et al., 2013; Wolff et al., 2016). A score of 1 or higher was characterized as experiencing difficulty.

The four-item version of Cohen’s Perceived Stress Scale (PSS) in which higher scores indicate greater perceived stress was used as an indicator of general perceived stress (not specific to caregiving; Cohen et al., 1983). Several indicators of the stressfulness appraisals of specific types of caregiving tasks were used, which address a different type of subjective burdens of caregiving (Haley et al., 1987, 1996). For each ADL, IADL impairment, or Memory and Behavior checklist problem reported, caregivers were also asked how much on a scale from 1 (not at all) to 5 (extremely) the impairment or problem bothered or upset the caregiver.

Caregiver appraisals of the stressfulness of assisting relatives with health care-related tasks, a relatively new and novel indicator termed treatment burden, were assessed using a four-item scale (Wolff & Boyd, 2015). Caregivers were asked to reflect “on the things you do to help your family member stay healthy or to treat their health problems.” A positive score was indicated by answering yes to at least one of the four items: how difficult the health care-related tasks are for the caregiver, for the family member, how often health care tasks for the family member get delayed or not done and if they feel that providers ask their family member to do too much.

Well-being and health outcomes

Well-being indicators were included that go beyond asking about caregiving strain and reflect broader psychological and physical health which can be affected by caregiving. Depressive symptoms were assessed using the 10-item Center for Epidemiologic Studies Depression scale (CES-D; Andresen et al., 1994; Melchior et al., 1993). Frequency of each symptom was rated on a 0 (rarely) to 3 (most of the time) scale with higher scores indicating greater depression. The 12-item Short Form Health Survey (SF-12) assessed health-related quality of life measured by the Physical Component Summary and Mental Component Summary scores with higher scores representing better health (Ware et al., 1996).

We also examined two items assessing caregivers’ subjective evaluation of how caregiving stress affected their own health and their ability to complete health care-related tasks for the care recipient. A commonly used indicator of caregivers’ perceptions of whether caregiving has affected their health was also included (AARP, 2015), although concerns have been raised about the likelihood of bias with such measures relying on the subjective recollection of health (Blome & Augustin, 2015). Caregivers were asked if they felt their health was better, worse, or the same than a year ago and if they felt that their caregiving responsibilities affected their health. We also analyzed a single item from the treatment burden scale noted above, which asked how often the things the caregiver does to help their family member stay healthy or treat health problems get delayed or do not get done.

Analyses

Initial analyses were conducted to compare the demographic characteristics of dementia and nondementia caregivers. The groups were then compared for differences in caregiving stressors (caregiving duration, hours of care provided, problems, and behaviors of the care recipient), caregiver appraisals of burden (strain, difficulty, and treatment burden), and well-being and health (depressive symptoms, quality of life, reports of impact on own health, and health of care recipient). Ordinary least squares regression analyses were performed on these variables to compare dementia and nondementia caregivers. Sensitivity analyses were performed recategorizing those persons who denied caring for a person with memory problems but who had an AD8 score of 2 or more and reported more than two ADL or IADL impairments. Using multivariable regression, two models were constructed to examine the differences observed between dementia and nondementia caregivers. In Model 1, age of the care recipient, age of the caregiver, and hours of care were added as covariates. In Model 2, the total number of problems of the care recipient as reported by the caregiver (which was the sum of ADL, IADL, and Revised Memory and Behavior Problems Checklist [RMBPC] problems) was added as an additional covariate or explanatory variable in analyses of caregiving health and well-being. All analyses were performed using SAS 9.4.

Results

Of the 251 incident caregivers enrolled in CTS, 117 reported providing care for a person with dementia and 134 reported other reasons for their caregiving including chronic illnesses, physical disability, and frailty. All participants who self-reported that the person they cared for had dementia had an AD8 score of 2 or more, suggesting that these caregivers were caring for individuals with cognitive impairment. No differences in race, sex, marital status, caregiving relationship, caregiving duration, or education were observed between our dementia and nondementia caregivers. Dementia caregivers were older than nondementia caregivers (73.1 vs. 70.7 years) as were the persons for whom they provided care (81.0 vs. 76.9 years). Primary caregivers made up 87% of all caregivers with men less likely than women to be primary caregivers (81% men vs. 91% women, p = .02). Spousal caregiving was the most common caregiving relationship for both dementia and nondementia caregivers (Table 1).

Table 1.

Demographic and Caregiving Characteristics of Dementia and Nondementia Caregivers

Variable	Dementia caregivers	Nondementia caregivers	p Value
N	117	134
Gender, female N (%)	73 (63)	90 (67)	.43
Race, African American N (%)	45 (38)	45 (34)	.42
Education, N (%)			.64
College graduate	49 (42)	57 (43)
Some college	33 (28)	45 (34)
High school graduate	31 (27)	27 (20)
Less than high school graduate	4 (3)	5 (4)
Marital status, N (%)			.88
Married/cohabiting	89 (76)	101 (75)
Widowed	9 (8)	10 (7)
Divorced/separated	13 (11)	13 (10)
Single/never married	6 (5)	10 (7)
Age of caregiver, M (SD)*	73.1 (8.3)	70.7 (7.7)	.02
Age of care recipient, M (SD)**	81.0 (11.3)	76.9 (11.0)	.004
Caregiving duration, years, M (SD)	5.8 (2.6)	5.7 (2.4)	.77
Hours of care per week, M (SD)**	49.7 (29.4)	37.7 (27.9)	.001
Primary caregiver, N (%)	99 (85)	121 (90)	.24
Caregiving relationship, N (%)			.26
Caring for a spouse	56 (48)	72 (54)
Caring for a parent or parent-in-law	35 (30)	28 (21)
Other	26 (22)	34 (25)

Note: M = mean; SD = standard deviation.

*p = 0.02; **p < .001.

Caregiving Stressors

Dementia caregivers reported providing more hours of care per week than nondementia caregivers (49.7 h compared to 37.7 h). Dementia caregivers reported higher numbers of impairments in ADLs among the persons they cared for (2.99 vs. 1.84 problems, p < .001). A similar pattern was seen with IADLs. Dementia caregivers reported that their care recipients exhibited more than twice the number of cognitive and memory problems and disruptive behaviors than persons without dementia. Dementia caregivers also reported a higher total problem load (19 problems) compared to nondementia caregivers (12 problems; Table 2).

Table 2.

Unadjusted and Adjusted Effects of Burden of Dementia and Nondementia Caregiving on the Caregiver

Variable	Dementia caregivers (N = 117)	Nondementia caregivers (N = 134)	Unadjusted	Model 1	Model 2
Stressors
Number of ADL problems	2.99 (2.46)	1.84 (2.01)	<.0001	.017	—
Number of IADL problems	5.82 (1.16)	4.63 (1.51)	<.0001	<.0001	—
Number of cognitive and memory problems	4.91 (1.87)	1.94 (1.94)	<.0001	<.0001	—
Number of emotional problems	2.76 (2.04)	2.27 (2.11)	.062	.048	—
Number of disruptive behaviors	2.09 (1.55)	1.06 (1.22)	<.0001	<.0001	—
Total number of problems	18.57 (4.97)	11.75 (5.40)	<.0001	<.0001	—
Appraisals of burden
Caregiving strain, M (SD)	2.28 (0.61)	1.96 (0.60)	<.0001	<.0001	.456
Emotional difficulty, N (%)	88 (75.2)	89 (66.4)	.128	.128	.521
Physical difficulty, N (%)	45 (38.5)	51 (38.1)	.948	.730	.081
Financial difficulty, N (%)	30 (25.6)	26 (19.4)	.238	.094	.596
Perceived stress	4.98 (2.96)	4.16 (2.93)	.029	.058	.377
Difficulty rating per ADL problem	0.78 (0.90)	0.50 (0.79)	.008	.036	.412
Difficulty rating per IADL problem	0.89 (0.84)	0.73 (0.77)	.124	.028	.616
Difficulty rating total number problems	1.29 (0.81)	0.96 (0.71)	.0007	<.0001	.786
Difficulty rating per cognitive and memory problem	1.30 (1.08)	0.61 (0.77)	<.0001	<.0001	.020
Difficulty rating per emotional problem	1.51 (1.17)	0.99 (1.02)	.0003	<.0001	.873
Difficulty rating per disruptive behavior	1.54 (1.18)	0.94 (1.21)	<.0001	<.0001	.551
Treatment burden, M (SD)	5.11 (2.27)	4.01 (2.14)	.0001	<.0001	.709
Well-being and health outcomes
CES-D	7.59 (6.39)	6.17 (6.08)	.072	.044	.975
SF-12 MCS	48.47 (10.46)	50.54 (10.76)	.125	.093	.460
SF-12 PCS	44.24 (11.58)	44.15 (10.54)	.840	.587	.638
Caregiving worsened health	38 (33.9)	20 (15.4)	.003	.048	.982
Delay or care recipient health tasks not done	27 (23.1)	11 (8.2)	<.0001	.0001	.184

Notes: ADL = activity of daily living; CES-D = Center for Epidemiologic Studies Depression scale; IADL = instrumental activities of daily living; MCS = mental component score; PCS = physical component score; SD = standard deviation. ADL range 0–8; IADL range 0–7; cognitive and memory problems range 0–7; emotional problems range 0–9; disruptive behaviors range 0–8; the total number of problems range 0–39. Difficulty rating refers to how much a problem bothers or upsets the caregiver and is scored from 0 (not at all) to 4 (extremely). Model 1: Adjusted for age of care recipient, age of care provider, and hours of caregiving. Model 2: Adjusted for age of care recipient, age of care provider, hours of caregiving, and the total number of problems reported by the caregiver. Stressors were not adjusted for the total number of problems.

Appraisals of Burden

For the single-item indicators of caregiving strain, dementia caregivers reported higher overall mental or emotional strain to provide care, but did not differ on any of the indicators of emotional, physical, or financial difficulty with providing care from the NSOC. General perceived stress was also higher in dementia caregivers. In terms of appraisals of the stressfulness of specific caregiving tasks, dementia caregivers reported greater strain from ADL impairments, but did not differ from nondementia caregivers in ratings of how bothersome IADL problems were. Dementia caregivers reported more subjective stress on the RMBPC for overall problems and for cognitive and memory problems, emotional problems, disruptive behaviors than nondementia caregivers.

Well-Being and Health Outcomes

No differences in overall well-being, including depressive symptoms and physical or mental health quality of life, were observed between the two groups with similar scores on SF-12 and CES-D. However, on the subjective retrospective report of changes in health and interference with health care for the care recipient, dementia caregivers were more likely than nondementia caregivers to report these effects.

A third of dementia caregivers reported that their caregiving role made their health worse compared to only 15.4% of nondementia caregivers (p = .003; Figure 1). When caregivers were asked how often the things that the care recipient needed to do for the care recipient to stay healthy or treat health problems were delayed or did not get done, dementia caregivers were far more likely to report problems (23.1% vs. 8.2%, p = .007).

Figure 1.

Health effects of caregiving on the care recipient and caregiver as reported by the caregiver.

Multivariable Regression

Two models are also described in Table 2: the first adjusted for age of care recipient, age of the care provider, and hours of care per week and the second adjusted for age of care recipient, age of the care provider, hours of care per week, and the total number of problems reported by the caregiver. Adjusting for the ages of the caregiver and care recipient and hours of care per week did not eliminate any of the previously significant differences between dementia and nondementia caregivers on indicators of appraisal and burden, except that the PSS no longer differed between groups but dementia caregivers reported significantly higher depressive symptoms than nondementia caregivers. The addition of the total number of problems reported by the caregiver to the model eliminated significant differences between dementia and nondementia caregivers on overall caregiving strain, reported difficulty caused by ADL impairments, total RMBPC problems and the three subscales, and treatment burden. Depressive symptoms, reports that caregiving had worsened health, and reports of delay or care recipient tasks not being done were also no longer significantly different between dementia and nondementia caregivers after adjustment.

Sensitivity Analysis

Among the 134 caregivers who reported caring for a person without dementia, 83 (62.2%) had a score of 2 or more on the AD8 dementia assessment and reported at least two ADL or IADL impairments for their care recipients. Sensitivity analyses were conducted after these 83 persons were reclassified as dementia caregivers. Differences observed between this expanded group of dementia caregivers and the remaining nondementia caregivers were largely unchanged (Supplementary Table 1).

Discussion

In this cross-sectional analysis from our study of people who transitioned into a caregiving role during their participation in a longitudinal epidemiological study, it proved useful to distinguish between different aspects of the stress process, including stressors, appraisals and subjective burden, and well-being and outcomes. We found that dementia caregivers reported higher levels of nearly all caregiving stressors than nondementia caregivers. On unadjusted analyses, dementia caregivers reported greater levels of burden or negative stress appraisal than nondementia caregivers across nearly all of the many indicators we examined, but did not report higher levels of depression, or poorer quality of life, than nondementia caregivers. However, on these unadjusted analyses, dementia caregivers were more likely to report that caregiving had negatively affected their health and had led to delays or not completing care recipient health tasks.

Adjusting for age of the caregiver and care recipient and hours of care provided resulted in few changes to the finding that dementia caregivers reported significantly higher appraisals of stress or burden than nondementia caregivers, although after this adjustment dementia caregivers reported significantly more depressive symptoms than nondementia caregivers. However, additional adjustment for the total number of problems (ADL, IADL, disruptive behaviors, memory and emotional problems) in care recipients virtually eliminated all significant differences in burden, depression, or negative stress appraisal between dementia and nondementia caregivers, suggesting that high levels of these problem behaviors explain the higher burden and subjective stress reported by dementia caregivers. This supports other work which found that controlling for resistance to care explained differences in caregiver outcomes (Fauth et al., 2016) and adds weight to the need for targeted interventions to relieve caregiver burden that focus on specific issues such as the BPSD as well as on more general problems confronting an individual caregiver (Selwood et al., 2007). This, combined with efforts to reduce the overall volume of problems the caregiver is juggling, for example, providing a care assistant to assist with bathing, has the potential to address the increased strain of dementia caregiving and prevent the negative effects of dementia caregiving on the caregiver and care recipient.

A novel finding from our work is that while dementia caregivers report more stressors, and more subjective strain and burden than nondementia caregivers, they did not report poorer quality of life. Depression differed only after adjustment for caregiver and care recipient age and hours of care. The lack of change in the quality of life in our study may be due to dementia and nondementia caregivers in our study who provided substantial (at least 5 h of care per week) and sustained caregiving (for over a full year). In our study, dementia caregivers reported 49.7 h of care, and nondementia caregivers provided 37.7 h of care per week, nearly 3 times as many hours as reported by a widely cited previous paper in this area (Ory et al., 1999). To the best of our knowledge, the only other population-based study to provide a threshold for caregiving is the NSOC, which does not set a threshold for hours of care but requires caregivers to provide assistance with care tasks (Riffin et al., 2017; Wolff et al., 2016).

On unadjusted analyses, dementia caregivers were more than twice as likely to report a worsening of their health because of their caregiving. These findings are consistent with the AARP 2011 report which found that dementia caregivers were more likely than other caregivers to put off medical care because of their caregiving responsibilities than nondementia caregivers (Feinberg et al., 2011). This is in contrast to the lack of differences between dementia and nondementia caregivers in depression and quality noted above, using well-validated indicators. It is unclear whether this type of retrospective report of worsening health represents an accurate indicator of change in health, or whether it might be subject to retrospective bias (Blome & Augustin, 2015), an issue that merits further research. Of note, this difference was no longer significant after adjustment for the total number of problems.

Interestingly, despite the subjective reports of worsening of health, measures of physical and mental health-related quality of life from the SF-12 did not differ between the two groups, despite the consistent differences in stressors and appraisals, which are higher in dementia caregivers. This study is unique in examining persons who have transitioned into a caregiving role and provide direct comparisons between those who became dementia caregivers and caregivers of persons with other disabilities. It is also one of few studies that report national data comparing large samples of both dementia and nondementia caregivers (Ory et al., 1999; Riffin et al., 2017; Wolff et al., 2016). Our findings of increased strain and burden in a representative group of dementia caregivers support previous findings and highlight the true magnitude of the burden on dementia caregivers. Few differences were observed between dementia and nondementia caregivers on demographic or caregiving variables. It is possible, however, that other variables not measured in our study could be relevant in understanding differences between dementia and nondementia caregivers, such as the use of respite services, adult day centers, and paid caregivers. Although strain and burden were seen more frequently in dementia caregivers, it is worth noting that 16% of nondementia caregivers reported being under a lot of strain and would, therefore, likely also benefit from targeted interventions to reduce caregiver burden. Another major strength of this study is the number of different assessments of caregiver well-being. When caregivers were asked about how their caregiving role affected them, the language used in questions differed and included words such as difficulty, strain, stress, and burden. The different responses to seemingly similar questions highlight how words can have different meanings to different individuals. Dementia caregivers reported higher rates of caregiving strain and treatment burden but similar rates of financial, emotional, and physical difficulty compared to nondementia caregivers. Asking questions in different ways is critical to identify the subset of caregivers who are feeling overburdened and address the differences in interpretation of language by different individuals.

Implications

In a national cohort of incident caregivers, we compared the effects of dementia and nondementia caregiving, using the framework of stress process models of caregiving to examine differences in stressors, stress appraisals and burden, and caregiver’s well-being. Both dementia and nondementia caregivers reported similar levels of overall well-being, but differed substantially in stressors and subjective burdens, and higher levels of problems largely explained differences in burden and appraisal between the two groups. This suggests that caregiving strain and burden are not driven by a specific diagnosis, for example, dementia, or by qualitative aspects of the dementia caregiving experience, but are a result of the number of different problems for which caregivers provide assistance and are greatly influenced by a person’s functional capacity, behavioral problems, and need. Many patient and caregiver interventions target specific diseases or symptoms, for example, weight monitoring in heart failure or addressing challenging behaviors associated with dementia. Broader interventions to increase mobility, functionality, and capacity that are not disease-specific, such as CAPABLE (Szanton et al., 2014), or interventions that address social needs, may have greater, more sustainable impacts on caregiver burden and warrant more widespread use.

Routine screening of caregivers to assess well-being is rare despite substantial evidence that unavailability of a caregiver and caregiver distress are important predictors of nursing home placement (Stall et al., 2019; Wolff et al., 2018). Health care systems, providers, and community health workers all need education in the importance of assessing burden and how best to screen for and evaluate caregiver strain. Previous reports have documented the limited training of health care professionals in best practices for dementia care and provided specific recommendations to improve professional education about dementia issues (Warshaw & Bragg, 2014). Family caregivers receive little or no training despite evidence that caregiver training may improve health and reduce health care utilization of those for whom they provide care (Burgdorf et al., 2019; Nuckols et al., 2017; Wolff et al., 2016). The Centers for Medicare and Medicaid Services (CMS) recently mandated that home health clinicians assess caregivers’ ability to provide assistance and offer training as needed to caregivers providing care to persons receiving skilled home health care (CMS, 2017). Expansion of this benefit to other care settings and adoption of widespread screening for caregiving responsibilities and burden by adding, for example, the one-item caregiving strain measure (Schulz & Beach, 1999) to the annual wellness visit have the potential to allow better targeting of resources and facilitate early intervention for strained caregivers. Limited evidence exists of the benefits of caregiving interventions for many demographic subgroups including men, minority populations, those living in rural areas, and multiple caregiver situations (Gitlin et al., 2015), so further work is needed to understand their unique needs and the effects of the care recipient and caregiver interventions on these populations (Dilworth-Anderson et al., 2020; Gitlin et al., 2015).

As health systems move away from traditional fee for service and embrace more value-based models of care, there is a new opportunity to provide wraparound care which includes attention to caregiver well-being with the ultimate aim of keeping persons with high-care needs including those with dementia living in the community supported by their caregivers. Policy changes are needed to support this model of care through provider and caregiver education and training, expanded provision of care assistant and respite services, and expanded access to proven interventions. As the supply of informal caregivers falls and demand continues to rise, society and our health care system need to acknowledge caregivers, especially dementia caregivers, as a valuable resource and target supportive interventions appropriately or face rising rates of institutionalization and soaring health care expenditure.

Acknowledgments

The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at https://www.uab.edu/soph/regardsstudy/

Funding

This research project is supported by cooperative agreement U01 NS041588 co-funded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), National Institutes of Health, Department of Health and Human Service. The Caregiving Transitions Study is further supported by an investigator-initiated grant [RF1 AG050609] from the NIA. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or the NIA. Representatives of the NINDS were involved in the review of the manuscript but were not directly involved in the collection, management, analysis, or interpretation of the data.

Conflict of Interest

None declared.