Abstract
Background
Few studies have comprehensively examined health outcomes among older caregivers. We aimed to describe older caregivers and characterize risks for mortality and hospitalization compared to noncaregivers.
Methods
Caregiving status and characteristics were determined for Atherosclerosis Risk in Communities (ARIC) Study participants via a one-time telephone assessment in 2015. All-cause mortality was identified from active surveillance, state records, and linkage to the National Death Index through December 31, 2021. Hospitalizations were identified from active cohort surveillance. Cox proportional hazard models assessed the risks of mortality and hospitalization.
Results
Among 5239 ARIC participants [mean age: 75.4 (SD 5.1) years; female: 60.0%; Black: 18.9%], 427 (8.2%) reported caregiving. Caregivers were generally female and younger as compared to noncaregivers. Most caregivers provided care for their spouse (55.0%), and 28.3% reported spending >40 hours/week on caregiving activities. Caregivers had modestly better cognitive scores but were similar to noncaregivers in the number of comorbidities and self-rated health. During a mean 5.4 (SD 1.3) years of follow-up, caregivers had a lower risk of mortality than noncaregivers (18.7% vs. 23.8%), although not statistically significant in fully adjusted time-to-event models [hazard ratio (HR) = 0.84; 95% confidence interval (CI), 0.67 to 1.06]. Caregivers and noncaregivers had similar risk of hospitalization (63.5% vs. 64.9%; HR = 1.00; 95% CI, 0.89 to 1.14).
Conclusions
Older caregivers provide substantial care while facing their own health challenges. Despite similar baseline comorbidity burdens as noncaregivers, caregivers had a lower risk of all-cause mortality over the 6 years of follow-up. Future studies should examine the potential protective factors of caregiving in older age to inform caregiver support initiatives for older adults providing care.
Keywords: Caregiving, Cognitive function, Strain, Hospitalizations, Mortality
Introduction
Nearly 15 million family and other unpaid caregivers provide financial and social support, engage in healthcare decision-making, and assist older adults with daily activities related to health and function.1 Over the last decade, there has been a shift to these caregivers being older adults themselves, with an approximately 30% increase in caregivers who are 65 years of age or older.2 Often these caregivers provide highly complex, physically demanding care critical for the care recipient’s health and well-being.1,3 Despite their significant contributions to care delivery, caregivers’ abilities and limitations are not routinely assessed as part of the clinical care, often resulting in inadequate attention or prioritization of caregiver-preventive health behaviors or overall health.
Although researchers have demonstrated worse health outcomes among caregivers than age-matched noncaregivers, including increased mortality,4 recent studies examining stress, inflammation, and event outcomes are mixed, with more evidence suggesting that individuals providing care are at lower risk of mortality.5,6 Few studies have comprehensively examined health outcomes in older caregivers specifically.
The Atherosclerosis Risk in Communities (ARIC) Study is a well-characterized community-based cohort, recruited in 1987, and consisting of mostly White and Black participants. ARIC assessed caregiving status in 2015 when participants were 70 years of age or older, which provides a unique opportunity to characterize older adults who identify as caregivers. Given the availability of expansive longitudinal surveys and biomarkers, ARIC data may provide new insights into the health of caregivers while also allowing examination of key caregiver characteristics and comorbidities within important subgroups of sex, race, and social determinants of health. This study aims to provide an in-depth understanding of the health of older caregivers as well as subsequent risk of hospitalizations and mortality using the ARIC study data.
Methods
Study population
The ARIC study recruited participants from four US locations: Forsyth County, NC, Jackson, MS, suburban Minneapolis, MN, and Washington County, MD.7 The initial sample for this study comprised 6867 ARIC participants who completed the 2015 semiannual telephonic follow-up interview. Participants with missing caregiving status were excluded (n = 21). Of the remaining participants, we excluded participants who did not attend ARIC visit 5 (2011-2013), at which time information on key participant characteristics was obtained (n = 1570). We further excluded participants whose self-identified race was neither Black nor White (n = 15), and Black participants (n = 22) from majority White locations (Minneapolis and Washington County) due to small sample cells, yielding a final analytic sample of 5239 participants. The ARIC study was approved by the institutional review boards at each participating center. Written informed consent was obtained from all participants.
Caregiving measures
Caregiver status was ascertained from the 2015 semiannual follow-up interview, at which time participants were asked: “Are you currently providing care on an ongoing basis to a family member or friend with a chronic illness or disability?” Those who responded “yes” were classified as caregivers, while those who responded “no” were classified as noncaregivers. For caregivers, a set of questions similar to those used within the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study8 was administered to collect information on caregiver characteristics and attributes of care, including relationship to care recipient (spouse, relative other than spouse, friend, and neighbor/other), time spent on care activities per week (<5 hours, 5-13 hours, 14-40 hours, and >40 hours), whether care recipient had memory difficulty (yes/no), whether they helped with daily tasks (yes/no), perceived strain in physical, emotional, and financial domains (no strain, low amount of strain, moderate amount of strain, and a lot or extreme amount of strain), and length of time caregiving (6 months or less, 6 months to 1 year, 1-5 years, and 5 years or longer).
Covariates
Trained study personnel took physical measurements and administered questionnaires following a standardized protocol during the visit 5 clinical examination. Demographic characteristics of study participants included age (in years), self-reported sex (female/male), race-center (five categories of race and the ARIC center), and living arrangement (living alone vs. other). Indicators of socioeconomic positions included educational attainment (less than completed high school, high school or equivalent, and at least some college/graduate school), self-reported annual household income (under $25 000, $25 000-50 000, $50 000-75 000, and greater than $75 000), and self-reported Medicaid enrollment (yes/no).
A comorbidity count profile, consisting of 10 medical conditions,9 was used to assess participants’ health status. The 10 conditions were ascertained by a combination of self-report and objective measures: hypertension (systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or use of antihypertensive medications), diabetes [glycated hemoglobin (HbA1c) ≥ 6.5%], hyperlipidemia (total cholesterol in SI units ≥5.17 or use of cholesterol-lowering medications in the past 4 weeks), chronic kidney disease [estimated glomerular filtration rate using creatinine (eGFRcr) < 60 mL/min/1.73 m2 or albumin-to-creatinine (ACR) ratio > 30 mg/g], prevalent cardiovascular diseases: coronary heart disease, heart failure, and stroke (self-reported at visit 1 or ascertained by medical record review between visits 1 and 5), cancer (self-reported cancer site during a follow-up interview from June 1, 2011 to visit 5), lung disease [including emphysema or chronic obstructive pulmonary disease (COPD), asthma, or chronic lung disease; self-reported at visit 5] and dementia (ascertained by reviewer and algorithm diagnoses, hospitalization, and death records).10 Comorbidities were examined as a categorical variable (0, 1, 2, 3, and 4 or more conditions).
Other health metrics included self-reported health (excellent, good, fair, and poor), depressive symptoms measured by the Center for Epidemiological Studies Depression (CES-D) scale (CES-D score ≥ 9),11 and cognitive impairment measured with the Mini-Mental State Examination (MMSE). Physical activity was assessed by the interviewer‐administered Baecke questionnaire.12
Outcomes
The primary outcomes of interest were all-cause mortality and hospitalizations. All-cause mortality, routinely captured in this cohort, was determined through semiannual telephone calls to participants or their proxies, state records, and linkage to the National Death Index up to December 31, 2021. Hospitalization records were collected through active cohort surveillance and linkage of records from collaborating hospitals. Specifically, for cohort surveillance, trained study personnel complete hospital abstraction forms to obtain information on each hospital event reported by eligible participants during annual or semiannual follow-up interviews. For the purposes of our analysis, we examined two major aspects of inpatient service use: time-to-first hospitalization event and count of hospitalizations over the 6 years of follow-up.
Statistical analyses
We used descriptive statistics, including means, medians, and proportions to compare caregivers and noncaregivers and describe caregiver characteristics.
For the time-to-event analyses, we adopted the completion date of the 2015 semiannual follow-up interview, when caregiving status was assessed, as the time origin. Participants who did not experience the outcome of interest (death or hospitalizations) were right-censored at 6 years or for the case of hospitalizations, time of death when applicable. Unadjusted Kaplan–Meier survival curves for all-cause mortality and first hospitalization during follow-up were computed by caregiving status and log-rank tests were performed.
To investigate the associations of caregiver status with all-cause mortality and first hospitalization during the study timeframe, we used Cox proportional hazard models. The association between caregiver status and number of hospitalizations over a 6-year follow-up was estimated using a negative binomial model. All models were adjusted for participants’ sociodemographics (age, female sex, race, living arrangement, education attainment, and household income) and health characteristics [self-reported health, number of comorbidities, depression, physical activity (min/week of moderate-vigorous activity), and MMSE score].
Missing covariates were imputed using multiple imputation with chained equations (MICE). Our strategy included all covariates from the fully adjusted model (in addition to age squared, self-reported Medicaid enrollment, mortality, and number of hospitalizations) with 50 sets of imputations and 100 iterations for the burn-in period.
In our secondary analyses, we further examined the moderation effects of key sociodemographic characteristics on the associations between caregiver status and study outcomes and investigated the heterogeneity among caregivers. Specifically, we performed subgroup analyses by age group, (older, age ≥ 76 years; younger, age ≤ 75 years; sample mean age = 75.4 years), sex (female vs. male), race (Black vs. White), and living arrangement (living alone vs. other). In addition, we tested the multiplicative interaction between caregiver status and these subgroup variables separately in fully adjusted models.
Furthermore, we restricted our study sample to caregivers only (n = 427) to examine caregivers’ relationship to care recipients (spousal vs. nonspousal caregiver), history of caregiving (<5 years vs. ≥5 years), and any presence of physical, emotional, or financial strains (no strain vs. any amount of strain) as additional predictors in our primary models. Missing information about the caregiver was imputed using MICE as previously described, while also including whether the care recipient had memory difficulty and whether the caregiver helped with daily tasks in the imputation model.
In sensitivity analyses, all regression models were fit among participants with a full set of covariates. A two-sided P-value <.05 was determined as the cutoff for statistical significance. Analyses were conducted using Stata version 18.0 (StataCorp LLC, College Station, TX).
Results
Among 5239 participants [mean age: 75.4 (SD 5.1) years; female: 60.0%; Black: 18.9%], 427 self-identified as caregivers. At baseline, those who identified as caregivers were generally younger (74.6 years vs. 75.4 years), female (66.0% vs. 59.5%), and less likely to live alone (15.9% vs. 29.5%) than noncaregivers (Table 1). Levels of education were similar between caregivers and noncaregivers (43.8% and 44.8% with more than a high school education, respectively) as were levels of self-rated health (26.5% and 27.1% with excellent health).
Table 1.
Baseline characteristics of study participants, stratified by caregiver status.
| Characteristics of participants | Total | Self-reported caregiver status |
||
|---|---|---|---|---|
| Noncaregivers | Caregivers | P-values | ||
| Participant counts | N = 5239 | N = 4812 | N = 427 | |
| Age at visit 5 (years), mean (SD) | 75.4 (5.1) | 75.4 (5.1) | 74.6 (4.9) | <.001 |
| Female sex, N (%) | 3143 (60.0%) | 2861 (59.5%) | 282 (66.0%) | .008 |
| Race and study site, N (%) | <.001 | |||
| NC White | 1153 (22.0%) | 1083 (22.5%) | 70 (16.4%) | |
| NC Black | 80 (1.5%) | 76 (1.6%) | 4 (0.9%) | |
| MS Black | 910 (17.4%) | 845 (17.6%) | 65 (15.2%) | |
| MD White | 1613 (30.8%) | 1494 (31.0%) | 119 (27.9%) | |
| MN White | 1483 (28.3%) | 1314 (27.3%) | 169 (39.6%) | |
| Living arrangements, N (%) | <.001 | |||
| Not living alone | 3746 (71.6%) | 3387 (70.5%) | 359 (84.1%) | |
| Living alone | 1485 (28.4%) | 1417 (29.5%) | 68 (15.9%) | |
| Education attainment, N (%) | .140 | |||
| Less than high school | 679 (13.0%) | 634 (13.2%) | 45 (10.5%) | |
| High school diploma or equivalent | 2216 (42.4%) | 2019 (42.0%) | 197 (46.1%) | |
| More than high school | 2336 (44.7%) | 2151 (44.8%) | 185 (43.3%) | |
| Annual household income, N (%) | <.001 | |||
| <$25 000 | 1276 (26.5%) | 1185 (26.9%) | 91 (22.6%) | |
| $25 000-49 999 | 1617 (33.6%) | 1457 (33.0%) | 160 (39.7%) | |
| $50 000-74 999 | 1030 (21.4%) | 931 (21.1%) | 99 (24.6%) | |
| >$75 000 | 890 (18.5%) | 837 (19.0%) | 53 (13.2%) | |
| Number of comorbidities, N (%) | .760 | |||
| 0 | 142 (2.9%) | 133 (2.9%) | 9 (2.2%) | |
| 1 | 759 (15.3%) | 699 (15.3%) | 60 (15.0%) | |
| 2 | 1359 (27.4%) | 1244 (27.3%) | 115 (28.7%) | |
| 3 | 1327 (26.7%) | 1228 (26.9%) | 99 (24.8%) | |
| ≥4 | 1377 (27.7%) | 1260 (27.6%) | 117 (29.2%) | |
| Self-rated health, N (%) | .290 | |||
| Excellent | 1412 (27.0%) | 1299 (27.1%) | 113 (26.5%) | |
| Good | 2946 (56.4%) | 2691 (56.1%) | 255 (59.7%) | |
| Fair | 752 (14.4%) | 699 (14.6%) | 53 (12.4%) | |
| Poor | 118 (2.3%) | 112 (2.3%) | 6 (1.4%) | |
| Depressive symptoms (CESD ≥ 4), N (%) | 323 (6.2%) | 292 (6.1%) | 31 (7.3%) | .350 |
| MMSE score, mean (SD) | 27.8 (2.2) | 27.7 (2.2) | 28.2 (1.9) | <.001 |
Summary statistics for available data, without restricting to individuals with complete covariate data. Other analyses include imputed covariate data: living arrangements (N = 8); education level (N = 8); income (N = 426); number of comorbidities (N = 275); self-rated health (N = 11); depression (N = 62); and MMSE (N = 34). Frequencies and percentages were presented for categorical variables, and means and standard deviations (SD) were calculated for continuous variables. Number of comorbidities among hypertension, diabetes, hyperlipidemia, chronic kidney disease, coronary heart disease, heart failure, stroke, cancer, lung disease, and dementia. P-values were calculated using Pearson’s chi-square test and independent samples t-test. Abbreviations: ADI, area deprivation index; MMSE, Mini-Mental State Examination.
Among identified caregivers, most provided care for their spouses (55.0%), and 28.3% reported spending >40 hours/week on care activities (Table S1). Moreover, 27.8% of caregivers reported having any caregiving-related financial strain, and most caregivers reported some level of mental or emotional (67.7%) and physical (48.7%) caregiving strain. Caregivers were most likely to have been providing care for 1-5 years (43.8%), but almost a third (32.2%) reported having provided care for 5 years or longer.
Over a median of 6 years of follow-up, there were 1226 deaths, with caregivers having lower mortality compared to noncaregivers. A total of 1146 (23.8%) deaths occurred among noncaregivers and 80 (18.7%) occurred among caregivers (Table S2 and Figure 1). Cox regression models adjusted for baseline sociodemographic and health variables suggested that caregivers had a lower risk of all-cause mortality (HR = 0.84; 95% CI, 0.67 to 1.06) compared to noncaregivers (Figure 2), although this effect estimate was not statistically significant. The risk of hospitalization (HR = 1.00; 95% CI, 0.89 to 1.14) and the total number of hospitalizations over a 6-year period [incidence rate ratio (IRR) = 0.99; 95% CI, 0.88 to 1.11] were not significantly different between caregivers and noncaregivers [3122 (64.9%) noncaregivers and 271 (63.5%) caregivers experienced at least one hospitalization]. There were no significant differences in the association between caregiving status, mortality, and hospitalizations by age group, sex, race, or living arrangements. Among caregivers, there were no significant differences in the association with outcomes by caregiving relationship, time spent on care activities, or levels of physical, emotional, or financial strain (Table S3 and Figure S1, see online supplementary material for a color version of this figure).
Figure 1.
Kaplan–Meier survival curves for all-cause mortality and hospitalization event by caregiver status (N = 5239). (A) Kaplan–Meier survival curves for all-cause mortality by caregiver status (N = 5239). (B) Kaplan–Meier survival curves for hospitalization event by caregiver status (N = 5239). Survival curves are unadjusted. Follow-up time is determined by years from study baseline (completion of the 2015 semiannual telephone interview) to the date when outcome event was developed; participants who did not develop the outcomes of interest were censored after 6 years of follow-up.
Figure 2.
Summary of measures of associations between caregiver status and study outcomes, Atherosclerosis Risk in Communities (ARIC) Study, 2015-2021 (N = 5239). Cox-proportional hazard regression was used to model all-cause mortality and first-time hospitalization even, while negative binomial regression was applied to the modeling of count of hospitalizations. All models were adjusted for caregivers’ age, female sex, site-race, living arrangement, education, income, number of comorbidities, self-reported health, MMSE scores, and depression. Analyses were repeated on subgroups of study participants by age group, sex, race, and living arrangements. All analyses include imputed covariate data: living arrangements (N = 8); education level (N = 8); income (N = 426); number of comorbidities (N = 275); self-rated health (N = 11); depression (N = 62); physical activity (N = 270); and MMSE (N = 34). P-values for the interaction correspond to the coefficient of the interaction term between caregiver status and subgroup variable. Abbreviations: HR, hazard ratio; IRR, incident rate ratio; MMSE, Mini-Mental State Examination.
In sensitivity analyses, complete case analyses produced similar results as with multiple imputation (Table S4 and Figure S2, see online supplementary material for a color version of this figure).
Discussion
In this study of older adults (mean age 75.4 years), those identifying as caregivers had similar comorbidity profiles as those not identifying as caregivers and they experienced hospitalizations at a comparable rate. Despite these similarities, caregivers showed a trend of lower mortality risk over 6 years than noncaregivers.
Our findings are consistent with previous literature showing that caregivers are at a lower risk of mortality.5 A recent study of older women showed a similar lower risk of mortality among caregivers.13 In our study, more than half of caregivers were providing care for a spouse, which may suggest they were the “healthier” spouse. However, our results do not support the “healthy caregiver hypothesis” which suggests that individuals who take on caregiving roles are generally healthier and that performing caregiving tasks maintains health.14,15 Performing the tasks of caregiving may increase physical activity and provide caregivers with satisfaction and a sense of purpose, both of which are well-established factors that can decrease mortality risk.16,17 The caregivers in this study had similar baseline comorbidities and subsequent rates of hospitalization to noncaregivers. Previous studies have associated better survival for caregivers experiencing less strain15; however, in our study the nonsignificant trend toward lower mortality risk among caregivers did not differ by the level of reported strain, similar to other reports.13
We did not obtain information about the caregiving network within this study, so these caregivers may not be the primary caregiver or the only individual providing care. For more than a quarter of participants, they were spending more than 40 hours/week providing care. This very high level of caregiving may lead to feelings of isolation and disconnection that can be detrimental to health as caregiving time replaces other valued activities.18,19 However, positive support from family and formal programming can increase feelings of confidence, resilience, and connection for the caregiver, as well as satisfaction that the care recipient is being well cared for.20 Effective interventions to improve social connection for older adults in general are often rigorous and involve being outside of the home, which may not be possible for older adults who are caregivers. There have been small studies on early planning interventions within disease-specific caregiving21,22; however, this is an area of need for study.
This study examined a well-characterized and diverse sample of older adults, assessing caregiving status with long-term health and healthcare outcomes. However, this study had a number of limitations. We measured caregiving status at only one timepoint, which does not allow for an assessment of change in role status over time or accurate estimate of the amount of time an individual has been in a caregiving role. Prior studies demonstrate that the benefits of caregiving on mortality are most pronounced for individuals actively caregiving and diminishes after caregiving ends.14 Also, participants labeled as noncaregivers in this study may have transitioned into caregiving roles within the follow-up time. Our sample of caregivers is a smaller proportion of the sample than expected, and small sample size may have impacted our ability to detect differences between caregivers and noncaregivers. This may be due to the question assessing caregiving status focusing on ongoing care for chronic illness or disability, which may have excluded individuals providing more intermittent care or care less focused on physical and medical needs. However, if that is the case, then our selected caregiving sample likely included those individuals providing more intense care. Caregiving status was captured via telephone interviews at a different time from the in-person visit, so we cannot account for any changes in health status or covariates that may have occurred in the intervening period. The study does not capture information on any potential positive impacts from caregiving, which may explain reduced mortality among caregivers.
In conclusion, this study found that older caregivers provide substantial care to others while experiencing financial, emotional, and physical strain, as well as their own health challenges. This work further supports a growing body of evidence that the role of caregiving itself does not increase the risk of hospitalization or mortality, even among older adults. Although we were unable to assess social connection and shared caregiving experience as potential positive and protective factors, these qualities should be further examined. As the population of caregivers who are themselves older adults has been on the rise,2 future studies examining the potential protective factors of caregiving in older age may inform caregiver support initiatives.
Supplementary Material
Funding and Acknowledgments
The Atherosclerosis Risk in Communities study has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under Contract nos. (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, and 75N92022D00005). The ARIC Neurocognitive Study is supported by U01HL096812, U01HL096814, U01HL096899, U01HL096902, and U01HL096917 from the NIH (NHLBI, NINDS, NIA, and NIDCD). The authors thank the staff and participants of the ARIC study for their important contributions.
Contributor Information
Shoshana H Ballew, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States.
Emmanuel E Garcia Morales, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States; Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States.
Wuyang Zhang, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States.
Martha Abshire Saylor, School of Nursing, Johns Hopkins University, Baltimore, Maryland, United States.
Danielle S Powell, Department of Speech and Hearing Sciences, University of Maryland College Park, College Park, Maryland, United States.
James Russell Pike, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States.
Anna Kucharska-Newton, Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States.
Nancy L Schoenborn, Division of Geriatric Medicine and Gerontology, School of Medicine, John Hopkins University, Baltimore, Maryland, United States.
Silvia Koton, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States; Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States; Stanley Steyer School of Health Professions, Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel.
Erin E Kent, Department of Health Policy and Management, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States.
David L Roth, Division of Geriatric Medicine and Gerontology, School of Medicine, John Hopkins University, Baltimore, Maryland, United States; Center on Aging and Health, Johns Hopkins University, Baltimore, Maryland, United States.
Josef Coresh, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States.
Jennifer L Wolff, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States.
Nicholas Reed, Optimal Aging Institute, New York University Grossman School of Medicine, New York, New York, United States.
Katherine A Ornstein, School of Nursing, Johns Hopkins University, Baltimore, Maryland, United States.
Supplementary material
Supplementary material is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online.
Conflict of interest
None declared.
Data availability
Investigators may approach the Atherosclerosis Risk in Communities Study for data sharing at https://aric.cscc.unc.edu/aric9/researchers/Obtain_Submit_Data.
References
- 1. Wolff JL, Spillman BC, Freedman VA, Kasper JD. A national profile of family and unpaid caregivers who assist older adults with health care activities. JAMA Intern Med. 2016;176:372-379. 10.1001/jamainternmed.2015.7664 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Wolff JL, Cornman JC, Freedman VA. The number of family caregivers helping older us adults increased from 18 million to 24 million, 2011-22. Health Aff (Millwood). 2025;44:187-195. 10.1377/hlthaff.2024.00978 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Wolff JL, Mulcahy J, Huang J, Roth DL, Covinsky K, Kasper JD. Family caregivers of older adults, 1999-2015: trends in characteristics, circumstances, and role-related appraisal. Gerontologist. 2018;58:1021-1032. 10.1093/geront/gnx093 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Schulz R, Beach SR. Caregiving as a risk factor for mortality: the caregiver health effects study. JAMA. 1999;282:2215-2219. 10.1001/jama.282.23.2215 [DOI] [PubMed] [Google Scholar]
- 5. Roth DL, Fredman L, Haley WE. Informal caregiving and its impact on health: a reappraisal from population-based studies. Gerontologist. 2015;55:309-319. 10.1093/geront/gnu177 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. O’Reilly D, Rosato M, Maguire A. Caregiving reduces mortality risk for most caregivers: a census-based record linkage study. Int J Epidemiol. 2015;44:1959-1969. 10.1093/ije/dyv172 [DOI] [PubMed] [Google Scholar]
- 7. Wright JD, Folsom AR, Coresh J, et al. The ARIC (atherosclerosis risk in communities) study: JACC focus seminar 3/8. J Am Coll Cardiol. 2021;77:2939-2959. 10.1016/j.jacc.2021.04.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Roth DL, Haley WE, Hovater M, Perkins M, Wadley VG, Judd S. Family caregiving and all-cause mortality: findings from a population-based propensity-matched analysis. Am J Epidemiol. 2013;178:1571-1578. 10.1093/aje/kwt225 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jones SM, Amtmann D, Gell NM. A psychometric examination of multimorbidity and mental health in older adults. Aging Ment Health. 2016;20:309-317. 10.1080/13607863.2015.1008988 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Gottesman RF, Albert MS, Alonso A, et al. Associations between midlife vascular risk factors and 25-year incident dementia in the atherosclerosis risk in communities (ARIC) cohort. JAMA Neurol. 2017;74:1246-1254. 10.1001/jamaneurol.2017.1658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385-401. 10.1177/014662167700100306 [DOI] [Google Scholar]
- 12. Richardson MT, Ainsworth BE, Wu HC, Jacobs DR Jr., Leon AS. Ability of the atherosclerosis risk in communities (ARIC)/Baecke questionnaire to assess leisure-time physical activity. Int J Epidemiol. 1995;24:685-693. 10.1093/ije/24.4.685 [DOI] [PubMed] [Google Scholar]
- 13. Chavan PP, Weitlauf JC, LaMonte MJ, et al. Caregiving and all-cause mortality in postmenopausal women: findings from the women’s health initiative. J Am Geriatr Soc. 2024;72:24-36. 10.1111/jgs.18620 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Fredman L, Lyons JG, Cauley JA, Hochberg M, Applebaum KM. The relationship between caregiving and mortality after accounting for time-varying caregiver status and addressing the healthy caregiver hypothesis. J Gerontol A Biol Sci Med Sci. 2015;70:1163-1168. 10.1093/gerona/glv009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Fredman L, Cauley JA, Hochberg M, Ensrud KE, Doros G; Study of Osteoporotic F. Mortality associated with caregiving, general stress, and caregiving-related stress in elderly women: results of caregiver-study of osteoporotic fractures. J Am Geriatr Soc. 2010;58:937-943. 10.1111/j.1532-5415.2010.02808.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Boyle PA, Barnes LL, Buchman AS, Bennett DA. Purpose in life is associated with mortality among community-dwelling older persons. Psychosom Med. 2009;71:574-579. 10.1097/PSY.0b013e3181a5a7c0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Watts EL, Matthews CE, Freeman JR, et al. Association of leisure time physical activity types and risks of all-cause, cardiovascular, and cancer mortality among older adults. JAMA Netw Open. 2022;5:e2228510. 10.1001/jamanetworkopen.2022.28510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Greenwood N, Mezey G, Smith R. Social exclusion in adult informal carers: a systematic narrative review of the experiences of informal carers of people with dementia and mental illness. Maturitas. 2018;112:39-45. 10.1016/j.maturitas.2018.03.011 [DOI] [PubMed] [Google Scholar]
- 19. Robison J, Fortinsky R, Kleppinger A, Shugrue N, Porter M. A broader view of family caregiving: effects of caregiving and caregiver conditions on depressive symptoms, health, work, and social isolation. J Gerontol B Psychol Sci Soc Sci. 2009;64:788-798. 10.1093/geronb/gbp015 [DOI] [PubMed] [Google Scholar]
- 20. Leggett AN, Meyer OL, Bugajski BC, Polenick CA. Accentuate the positive: the association between informal and formal supports and caregiving gains. J Appl Gerontol. 2021;40:763-771. 10.1177/0733464820914481 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Orsulic-Jeras S, Whitlatch CJ, Szabo SM, Shelton EG, Johnson J. The SHARE program for dementia: implementation of an early-stage dyadic care-planning intervention. Dementia (London). 2019;18:360-379. 10.1177/1471301216673455 [DOI] [PubMed] [Google Scholar]
- 22. Abshire Saylor M, Pavlovic N, DeGroot L, et al. Feasibility of a multi-component strengths-building intervention for caregivers of persons with heart failure. J Appl Gerontol. 2023;42:2371-2382. 10.1177/07334648231191595 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Investigators may approach the Atherosclerosis Risk in Communities Study for data sharing at https://aric.cscc.unc.edu/aric9/researchers/Obtain_Submit_Data.