Identifying Future Caregivers via Current Social and Economic Resources

Abstract

With smaller family sizes and increased longevity, more adults live into late old age in a context with fewer available family caregivers. Acknowledging that social services and policies take time to develop and implement, identifying members of the next generation of caregivers and anticipating their service needs become an important endeavor. Using data from the 2022 Centers for Disease Control and Prevention’s Behavioral Risk Factor Surveillance System, we applied multinomial logistic regression analyses to identify theoretically relevant variables that differentiate among current caregivers (n = 19,662), those who anticipate becoming caregivers in the near future (n = 10,258), those who are unsure of their near-future caregiving status (n = 5,214) and those who are certain they will not be providing caregiving assistance in the next few years (n = 62,780). Although age, sex, and race each emerged as a significant regressor, the pattern of those associations differed by anticipated caregiving status. For example, although those who anticipate being caregivers (OR = .99) and those who are certain they will not be caregivers (OR = 1.0) are younger than current caregivers, those who are unsure (OR = 1.01) were significantly older than current caregivers. We also note that those who were unsure regarding their future caregiving status were significantly less likely to identify as white non-Hispanic than current caregivers (OR = 1.51). Challenges related to economic Social Determinants of Health were higher for current than for anticipated (OR = .89), unsure (OR = .76), and those not anticipating becoming a caregiver (OR = .73). Our results speak to the need to plan for the needs of future caregivers, who may be more diverse than current caregivers.

Keywords

caregiving social determinants and drivers of health health disparities quantitative methods

What This Paper Adds

• Extends the caregiving literature to include an examination of correlates that differentiate among current caregivers, anticipated caregivers, those who are unsure about future caregiving, and those not expecting to provide care.

• Identifies differences in the demographic and economic social determinants of health across caregiving groups, highlighting potential disparities before caregiving begins.

• Introduces anticipated and unsure future caregivers as important and understudied groups for future research, programs and policies.

Applications of Study Findings

• Screening for economic social determinants of health among adults who are likely to become caregivers may enable communities to deploy financial and other support resources earlier.

• Findings support the need for development of caregiver preparedness programs before individuals assume the caregiving role.

• Results can inform policymakers and healthcare systems in designing equitable support that addresses the needs of current and future caregivers.

Between 2021 and 2022, approximately 37.1 million U.S. American adults provided unpaid care to a family member or friend with an illness or disability (US Census, 2023). Extensive research on informal caregiving notes that caregivers, who often provide help with activities of daily living (ADL) and instrumental activities of daily living (IADL) experience more threats to physical health, emotional well-being, social relationships, financial well-being, and household space constraints compared to non-caregivers (Berg et al., 2021; Grenard et al., 2020). The majority of caregivers tend to be female, middle-aged and non-Hispanic white (Berg et al., CDC, 2019), individual differences may influence who takes on the caregiving role.

In contrast to research comparing caregivers to non-caregivers, there is little research comparing current caregivers to those who anticipate being a caregiver in the near future or with those unsure whether they will provide care. Thus, little is known about the challenges and resources individuals bring into the caregiving role before becoming a caregiver. Specifically, it is unclear whether and how caregivers differ from those who anticipate becoming a caregiver, those who are unsure, and those who are sure they will not take on the caregiving role. Knowing these pre-existing differences would allow better planning, programs, and support for caregivers prior to the role. In order to fill this knowledge gap, the current study examined demographic indicators and social determinants of health between current caregivers, anticipated caregivers, those who are unsure if they will become caregivers in the next 2 years, and those not anticipating being caregivers.

Theoretical Framework

Social determinants of health (SDOH) are environmental conditions such as where people live, work, behave, and age that influence one’s health and quality of life (Duran & Pérez-Stable, 2019). Although inclusive of age, sex, and race, SDOH include access to health care, food insecurity, difficulty with transportation, and economic challenges (Duran & Pérez-Stable, 2019).

For instance, caregivers who are younger and not established in a career may have fewer financial resources than caregivers who are older and more established in their careers (Koumoutzis et al., 2021). The current study focuses on specific economic social determinants of health (eSDOH), including loss of employment, difficulty with finances, and difficulty with transportation. Looking at eSDOH to better understand predisposing differences in mental and physical health prior to taking on the caregiving role is essential. Additionally, there are significant gaps in the literature regarding eSDOH among current caregivers, specifically in terms of race, ethnicity, gender, age, and socioeconomic status (Young et al., 2020). While most caregivers experience strain during their role, the degree of the strain may differ across age, gender, and race (Young et al., 2020). For instance, female caregivers of all races were more likely to spend at least 20 hours per week providing care and provide ADL and IADL care than white male caregivers (Cohen et al., 2024).

The Informal Care Model (Broese van Groenou & De Boer, 2016) explains caregiving onset as a function of the presence of care needed, intention to provide care, and contextual conditions that either enable or constrain providing care. According to the informal care model, intentions of providing care are shaped by factors such as social norms, perceived obligation, available resources, and broader social determinants within the family and community (Broese van Groenou & De Boer, 2016). Once a loved one’s need for care has been established, an individual’s attitude and value of providing care drives the intention to provide care (Broese van Groenou & De Boer, 2016). Importantly, the informal care model suggests that contextual constraints, such as time, distance from the person needing care, and money can inhibit a person from providing care (Broese van Groenou & De Boer, 2016). Prior use of the informal care model has primarily focused on caregiving networks (de Klerk et al., 2021; Vos et al., 2022), and factors influencing informal care (Raiber et al., 2022) among existing or established caregivers. The current study applies this model to an earlier and more exploratory stage by examining demographics and economic social determinants of health characteristics among groups of varying probabilities of being a future caregiver. From this perspective, examining demographics and economic disparities among individuals with varying likelihood of becoming future caregivers may help identify populations who could face different levels of preparedness and support needs as caregiving demands increase.

Caregiving Differences in Financial and Support Resources

Difficulties with finances and working in the caregiving role have been reported in the caregiving literature. Younger caregivers (18–39 years old) have been found to experience more significant financial strain compared to older caregivers (60–80 years old; Koumoutzis et al., 2021). Similarly, prior work (Lahaie et al., 2012) showed that the number of hours devoted to caregiving differed as a function of age and employment status, with those who provided more care being older and less likely to be employed for pay. Women caregivers tend to experience poor work environments, such as a less flexible schedule and paid leave, than male caregivers (Kaufman et al., 2010). Thus, it is not surprising that women caregivers are more likely to retire from the paid work force earlier than men caregivers (Kaufman et al., 2010). However, because taking on the caregiver role is not random nor is it separated from the social and economic context in which one lives, it is important to plan for supportive services that assist current and future caregivers.

Gaps in the Research

However, little empirical research examines the differences between current caregivers, those who anticipate the caregiving role, and those who are unsure if they will become caregivers. One study compared current, anticipated, and not anticipating caregivers in a sample of emerging adults (ages 18–25) drawn from the 2015–2017 BRFSS and found that young adult Hispanic males were more likely to be anticipated caregivers than current and non-caregivers (Grenard et al., 2020). According to the Family Caregiver Alliance (2016), age of caregivers varies, with 48% of caregivers being ages 18–49 years old and 34% ages 65 and older. As such, more inclusive studies are needed to examine demographic characteristics across caregivers and non-caregivers (anticipated, unsure, and not anticipating) that could influence experiences.

While we do not know much about those who anticipate providing future care, there is some research that assesses the outcomes of the entry into the caregiving role. For instance, for some individuals, caregiving can be disruptive and overwhelming at the start of the role (Gaugler et al., 2003). However, those who experience a slower transition into the caregiving role report better experiences (Uhm et al., 2023). If individuals are aware of the possibility of needing to provide care in the future, they may be more likely to take the necessary steps to be prepared for the role resulting in better care experiences. Prior to diagnosis, about 40% of Uhm and colleagues’ (2023) sample recognized symptoms of Alzheimer’s or dementia in their loved one and experienced less overload at the onset of caregiving than those who started abruptly. Those who recognized the symptoms and had a slow entry into the role, as opposed to those who abruptly started caregiving after diagnosis, were more likely to continue to provide care in their own home but still experienced decreases in well-being across 3 years (Uhm et al., 2023). Uhm et al.’s (2023) finding echoes that of earlier studies showing that feeling prepared for the caregiving role was associated with less depression and better quality of life (Seltzer & Li, 2000). As such, further research on those who anticipate providing care may be imperative to understanding entry into the caregiving role and caregiving outcomes. By identifying those who anticipate being a caregiver, policy makers and support networks may be able to provide more targeted support to help future caregivers feel better prepared within the role.

Thus, although the extant literature shows caregivers experience worse emotional and physical health than non-caregivers (Berg et al., 2021), these effects may vary by age, race, gender, and eSDOH. Moreover, there are likely unexplored differences across the adult life span among caregivers, those who anticipate becoming a caregiver, and those who are either unsure or confident that they will not take on the role soon. Those anticipating being a caregiver may have more time to prepare for the caregiving role, but those who are unsure about the role may experience an abrupt, overloading onset of caregiving if they are not preparing themselves for caregiving, consistent with prior research that looked at abrupt onset of caregiving (Uhm et al., 2023). As such, it is important to focus on those anticipating the caregiving role and those unsure, to provide enough information about their well-being and demographics before becoming a caregiver. Examining such differences allows us to better direct supportive services to those likely to enter the caregiving role.

Current Study

The current study contributes to the caregiving literature by examining the contributions of demographic variables and economic Social Determinants of Health (eSDOH) to the empirical differentiation among these four groups of caregiving and non-caregiving adults. By identifying patterns associated with caregiving status and likelihood of anticipated caregiving, this work may inform future efforts aimed at caregiver outreach, preparedness, and equitable distribution of support resources for family caregivers. Thus, we hypothesize that current caregivers report more challenges related to eSDOH than the non-caregiving group (8 and 9). We anticipated that the four caregiving groups would be differentiated by age, sex and race (Grenard et al., 2020). Given the paucity of research in this area, the examination of differences among the four groups (i.e., current caregivers, anticipated, unsure, and not anticipating) are exploratory.

Method Data Source

The study used data from the 2022 Centers for Disease Control and Prevention’s (CDC) Behavioral Risk Factor Surveillance System (BRFSS; Centers for Disease Control and Prevention, 2023). The BRFSS is a national interview conducted annually to assess US Americans’ health, including specific risks, chronic health conditions, and use of health care services. The survey is composed of core questionnaires administered in every US state, as well as optional modules and state-specific questions. The CDC uses trained interviewers to conduct the approximately 17-minute core interview. The CDC also monitors interviews and uses callbacks to verify the data. Thus, it provides current and high-quality data regarding health. The key predictors of interest come from the Social Determinants and Health Equity module.

Participants

The Caregiving module in the 2022 BRFSS was asked of adults residing in eleven US states and one territory while the Social Determinants and Health Equity module was asked of adults in 35 US states and two territories. Those who answered both the Caregiving module and the Social Determinants and Health Equity module, resided in six US states and one territory including Georgia (n = 5,387), Mississippi (n = 2,978), New Hampshire (n = 4,388), Utah (n = 6,382), Washington (n = 17,822), Wisconsin (n = 7,212), and Puerto Rico (n = 3,952). Of the 445,132 participants who completed the BRFSS, 48,121 are included in our sample. Inclusion of participants can be seen in Figure 1. Of the sample, 9,557 (19.9%) adults answered affirmatively to currently providing “…regular care or assistance to a friend or family member who has a health problem or disability.” Of the current caregivers, 32.8% were caring for a parent or parent-in-law, 10.3% for their own child or grandchild, 22.8% for their spouse or partner, 7.8% for a sibling, 9.8% for some other relative, and 15.5% for a nonrelative/friend. The most comment chronic conditions that required assistance included dementia (14.9%), general old age/frailty (13.5%), heart disease (8.1%), cancer (8%), lung diseases (7.2%), developmental disabilities (4.9%), and mental health conditions (4.5%). Most (61.1%) current caregivers were female, middle-aged (M age = 56.7, SD = 15.9), white non-Hispanic (85.5%), married (62.5%), and well educated with 43.3% reporting a 4-year degree or more. Although 50% reported household incomes in excess of $75,000 per year, 23.6% reported annual incomes at or below $15,000. Approximately 48.2% were also employed outside the home.

Figure 1.

Inclusion Flow Chart

To those who did not respond “yes,” a follow up question was posed regarding whether they anticipated becoming a caregiver in the next 2 years. Of the 38,564 (80.1%) who were not current caregivers, 5,381 (11.2%) anticipated doing so in the next two years, 2,049 (4.3%) were unsure whether they would enter the caregiving role, and 31,134 (64.7%) replied that they did not anticipate becoming a caregiver in the next 2 years.

The responses from the two caregiving items were combined into a single caregiving item grouping participants into current caregivers (19.9%), anticipated caregivers (11.2%), those unsure in the caregiving role (4.3%), and those not anticipating the caregiving role (64.7%). Thus, based on these data, as many as 97,914 adults were available for the analyses but varied due to completion of other items in the survey (sample size for each analysis included in the results section). Only 97,594 complete cases were available to use for analyses. The missing 320 participants are due to refusal to reply, overlapping categories with different care recipients, inconsistent skip patterns, and interviewer error.

Of the participants, 47.1% were male and 52.9% identified as female. Age was measured as a continuous variable. Caregivers ranged in age from 18 to 80+ years (M = 55.30, SD = 17.09). The sample included white non-Hispanic (74.6%), African American non-Hispanic (6.5%), Asian non-Hispanic (2.4%), American Indian or Alaskan Native non-Hispanic (0.9%), Hispanic adults (13.6%), and other non-Hispanic (2.1%). Although race and ethnicity are important contributors to caregiver well-being (15), further breakdown of race was limited by small cell size. Therefore, race was treated as a dichotomous variable with non-Hispanic white (85.0%) and not white (15.0%). The mean annual household income was roughly between $50,000 and $75,000. Of note, 43.1% have an income of more than $75,000. More descriptive statistics can be found in Table 1. For analysis, marital status was used as a dichotomous variable with married and part of an unmarried couple belonging in the married status (59.8%) and all others belonging to the non-married status (40.2%). Specific marital statuses are shown in Table 1.

Table 1.

Sample Demographic Statistics (N = 48,121)

Variable	Mean% (sd)		Caregiving status
Variable		Current	Anticipating	Unsure	Not anticipating
Age	55.30 (17.1)	56.55	54.50	58.61	54.84
Sex/female (%)	0.53 (.50)	60.1	55.6	52.9	50.3
Marital status
Married	55.6	61.0	60.5	55.5	53.0
Partnered	4.4	4.2	5.2	4.8	4.3
Divorced	13.1	13.2	11.1	12.9	13.5
Widowed	9.8	6.8	6.5	11.0	11.2
Separated	1.7	1.6	1.8	2.2	1.7
Never married	15.4	13.2	14.9	13.7	16.3
Race
White	85.0	85.6	85.7	80.0	85.1
Black/African Am.	10.2	10.2	10.0	15.3	9.8
American Indian	1.4	1.7	1.4	1.3	1.3
Asian	2.7	1.6	2.2	2.8	3.1
Other	0.6	0.8	0.7	0.5	0.6
Education
0 to 8th grade	1.3	0.7	1.3	2.3	1.4
Some high school	3.0	2.8	2.8	4.7	3.0
High school/GED	21.6	20.1	19.7	22.3	22.4
Some college	28.5	30.8	28.0	29.6	27.7
College degree	45.6	45.6	48.2	40.9	45.4
Income
Less than $20,001	10.9	9.1	9.6	14.2	11.4
$20,001–$35,000	16.7	17.5	15.3	20.3	16.4
$35,001–$50,000	12.7	13.9	12.5	14.1	12.4
$50,001–$75,000	16.6	18.0	16.8	17.0	16.0
$75,001 or more	43.1	45.8	34.5	43.8	43.1
Employment
Employed	43.6	41.6	47.6	38.2	43.9
Self-employed	8.8	8.3	10.1	10.2	8.6
Unemployed <1 yr	1.6	1.9	1.8	1.7	1.5
Unemployed 1+ yr	1.7	2.4	1.8	1.6	1.5
Homemaker	4.7	5.3	4.8	4.3	4.6
Student	1.9	1.4	1.9	1.4	2.1
Retired	32.3	32.8	27.7	37.1	32.6
Unable to work	5.3	6.2	4.4	5.5	5.2
eSDOH count	.42	.50	.43	.40	.35
Lost employment	8.8	2.2	1.2	0.4	5.0
Food stamps	11.0	11.6	10.2	13.2	10.8
Difficulty with bills	8.5	11.9	9.2	8.0	7.3
Utilities disconnect	5.1	7.6	5.7	4.9	4.2
Transportation difficulty	5.8	8.2	6.3	5.5	5.1

Thus, data from four groups were available for analyses. Small group differences were observed among the non-caregivers on demographic variables. More than half (56.2%) of those who anticipated becoming a caregiver in the next 2 years were female. They had an average age of 54.3 years (SD = 16.4). Most were white non-Hispanic (85.6%), married (62.8%), and well educated with 45.7% reporting a 4-year degree or more. More than half (54%) reported household incomes in excess of $75,000 per year, but 21.4% reported annual incomes at or below $15,000. More than half (55.3%) were currently employed outside the home. About 54.6% of those who were unsure of their future caregiving status were female and this group was slightly older, with a mean age of 59.1 years (SD = 16.4). Approximately 80.5% were white non-Hispanic, 57.1% were married, and 38.5% had earned a 4-year college degree or higher. About half (52%) reported annual incomes above $75,000 but 25.5% reported incomes below $15,000 per year. About 42.7% were currently employed outside the home. Of those who were not expecting to become a caregiver in the near future, 51.7% were female, with a group mean age of 55.6 years (SD = 17.5). Most (85.1%) were white non-Hispanic, and 53.7% were married or partnered. About 42.5% reported having a 4+ year college degree and 53% reported annual incomes in excess of $75,000. However, 23.3% reported annual incomes at or below $15,000 and about half (49.3%) were employed outside the home. Of course, large cell sizes make these small mean differences emerge as statistically significant, but we note the similarity in means and the significant overlap across caregiver groups.

Measures

Social Determinants of Health

In 2022, the BRFSS included an optional module with several SDOH items addressing economic support. Additionally, five items were used to assess difficulties with economic SDOH (eSDOH): employment, use of food stamps, difficulty paying rent or utility bills, discontinuation notices from utility companies, and transportation difficulties. Response choices for the five items were yes, no, and don’t know/not sure. The five items were re-coded such that “yes” responses were coded “1” and all other responses were coded as zero. The five items were then summed into a total eSDOH score ranging from 0 to 5, with 5 indicating the participant was at the highest economic disadvantage (M = .392, SD = .834). By using the eSDOH variables as a summed score, we obtain a crude index of the number of economic challenges experienced by these participants. In the sample, 76.0% reported having zero eSDOH disadvantages. As one might expect with a count measure, the reliability for the eSDOH measure was weak, α = .598.

Analyses

All statistical analyses were conducted using IBM SPSS Version 29.0 software. Bivariate associations were examined using Pearson correlation coefficients for continuous variables and Spearman correlation coefficients for categorical variables. To test whether demographics and economic SDOH would differentiate among current caregivers and other caregiving statuses, a multinomial logistic regression was used. Using current caregivers as the reference group for those who anticipate, unsure, and do not anticipate the caregiving role, the equation tested the influence of female sex, age, white race, marital status, education, income, current employment, and the sum of five economic SDOH. To assess whether the model could differentiate among the different groups of non-caregivers, a second multinomial logistic regression was conducted, using anticipated caregivers as the reference group.

Results

When examining bivariate associations, due, in part, to the large sample size, the majority of the variables were significantly correlated. Thus, only coefficients greater than .20 are highlighted in text. When examining marital status, being married was significantly associated with more income (ρ = .42) and having fewer eSDOH (ρ = −.21). Being employed was significantly correlated with being older (ρ = .52) and having a higher income (ρ = .35). Having a higher education was correlated with higher income (ρ = .41) and fewer eSDOH (ρ = −.23). A higher income was significantly correlated with fewer SDOH (ρ = −.42). In addition, inspection of VIFs and other indices suggested that multicollinearity was not present. To address the hypotheses, multinomial logistic regressions were performed using the four mutually exclusive caregiver groups: current caregivers, anticipated, unsure, and not anticipating being a caregiver.

Current Caregivers vs. Non-Caregivers

To assess whether demographics and economic SDOH would differentiate among current caregivers and other caregiving statuses, the multinomial logistic regression to adequate model fit. The overall fit of the model was evaluated using the likelihood ratio chi-square test, Pearson’s goodness-of-fit tests, and McFadden’s pseudo-R-square. The likelihood ratio chi-square test was statistically significant, χ² (24) = 1639.28, p < .001, indicating our model fit the data significantly better than a null model. However, due to the large sample size, the Pearson chi-square was statistically significant, suggesting the model may not be a good fit to the data, χ² (41718) = 45532.79, p < .001. The McFadden pseudo R² equaled .014. Next, the effect of each of the predictors in each model was tested to determine unique contributions to the overall model fit. Each regressor, age, education, income, female sex, marital status, white non-Hispanic race, current employment, and eSDOH, emerged as a unique contributor to the overall model. Based on the fit indices above, the overall fit of the model was adequate. However, inspection of the classification table suggested that the variable set, based on the caregiving and health disparities literature, were not adequate. For example, although 64.9% were correctly classified overall by this model, fewer than 1% (.5%) of the current caregivers were correctly classified. Moreover, none of those anticipating or who were unsure were correctly classified. Yet, virtually all (99.9%) of those who were not anticipating becoming a caregiver in the near future were classified correctly.

As shown in Table 2 and Figure 2, relative to current caregivers, those who anticipated becoming a caregiver were less likely to be female (b = −.145***, OR = .87), somewhat younger (b = −.006***; OR = .99), more likely to be employed (b = .176***; OR = 1.2), and reported fewer eSDOH (b = −.106***; OR = .90). Of note, race (b = −.019, OR = .98), marital status (b = −.065, OR = .94), education (b = −.017, OR = .98), and income (b = −.014; OR = .99) did not uniquely differentiate between current caregivers and anticipated caregivers.

Table 2.

Regression coefficients and odds ratios: Current Caregivers vs. Non-Caregivers

Caregiver Status	95% Confidence intervals for OR
Caregiver Status	_b_	Sig	SE (b)	OR	Lower bound	Upper bound	% Correctly classified
Current caregiver							0.5
Anticipated							0
Intercept	.025	.826	.116
Sex (F)	−.145	***	.032	.865	.812	.921
Age	−.006	***	.001	.994	.992	.996
Race (NHW)	−.019	.678	.046	.981	.897	1.07
Married	−.065	.062	.035	.937	.875	1.00
Education	−.017	.406	.020	.983	.945	1.02
Income	.014	.536	.022	1.01	.970	1.06
Employed	.176	***	.038	1.19	1.11	1.28
eSDOH	−.106	***	.020	.899	.865	.935
Unsure							0
Intercept	−.639	***	.157
Sex (F)	−.257	***	.043	.774	.711	.842
Age	.009	***	.002	1.01	1.01	1.01
Race (NHW)	−.490	***	.057	.613	.548	.685
Married	−.244	***	.046	.784	.716	.858
Education	−.161	***	.027	.851	.808	.898
Income	.040	.176	.029	1.04	.982	1.10
Employed	−.011	.832	.052	.989	.893	1.10
eSDOH	−.168	***	.028	.846	.800	.894
Not anticipating							99.8
Intercept	2.858	***	.080
Sex (F)	−.379	***	.022	.685	.656	.715
Age	−.008	***	.001	.992	.990	.993
Race (NHW)	−.086	**	.032	.918	.862	.977
Married	−.479	***	.024	.619	.591	.649
Education	−.074	***	.014	.929	.904	.955
Income	.041	**	.015	1.04	1.01	1.07
Employed	−.095	***	.026	.909	.864	.957
eSDOH	−.284	***	.014	.753	.733	.774

Note. Current caregiver is the referent category; sample size is 48,121; overall correctly classified = 64.9%. **p < .01; ***p < .001

Figure 2.

Odds ratio with 95% confidence intervals: current caregivers vs non-caregivers. Note. Current caregivers served as the reference category. Bars depict odds ratios and error bars indicate 95% confidence intervals. The dashed horizontal line represents OR = 1.00. NHW = Non-Hispanic White; eSDOH = Economic Social Determinants of Health

Compared to current caregivers, those who were unsure of taking on the caregiving role were less likely to be female (b = −.257***, OR = .77), were somewhat older (b = .009***; OR = 1.01), were less likely to be white (b = −.490***; OR = .61), were less likely to be married or partnered (b = −.244***, OR = .78), reported less education (b = −.161***; OR = .85), reported fewer eSDOH (b = −.168***, OR = .85). Variables that did not differentiate between current caregivers and those who were unsure include income (b = .040, OR = 1.04) and current employment (b = −.011; OR = .99).

Relative to current caregivers, those who were not anticipating to take on the caregiving role were less likely to be female (b = −.379***, OR = .69), were somewhat younger (b = −.008***; OR = .99), were less likely to be white non-Hispanic (b = −.086***, OR = .92), less likely to be married (b = −.479***, OR = .62), reported less education (b = −.074***; OR = .93), reported somewhat higher annual incomes (b = .041***, OR = 1.04), were less likely to be employed (b = −.095***, OR = .91), and reported fewer eSDOH (b = −.284***, OR = .75).

Non-Caregiving Groups

A multinomial logistic regression to assess whether demographics and economic SDOH would differentiate among non-caregiving group statuses had poor model fit. With the exception of income [χ² (2) = 1.70, p = .43], the other variables contributed to the omnibus model. However, poor model fit was suggested by both the overall model’s likelihood ratio chi-square test, χ² (16) = 693.74, p < .001 and the Pearson chi-square, χ² (24260) = 26689.28, p = .001. McFadden pseudo R² was equaled .012. Moreover, the classification table showed that although overall correct classification was 80.5%, no anticipated caregivers or unsure caregivers were correctly classified. Thus, the model was a poor fit to the data.

We provide the data in Table 3 to inform future attempts to identify variables that differentiate among potential future caregivers. We briefly highlight the results in text. Compared to anticipated caregivers, those who are unsure of whether they will be a caregiver were less likely to be female (OR = .89), were somewhat older (OR = 1.02), were less likely to be white non-Hispanic (OR = .63), were less likely to be married/partnered (OR = .85), reported less education (OR = .87), were less likely to currently be employed (OR = .83), and reported fewer eSDOH (OR = .88). Income did not differentiate significantly between those who anticipated becoming a caregiver and those who were unsure (OR = .96).

Table 3.

Regression coefficients and odds ratios: Anticipated Caregivers vs. other Non-Caregivers

Caregiving status	95% confidence intervals for OR						Percent correctly classified
Caregiving status	_b_	Sig	SE(b)	OR	Lower bound	Upper bound	Percent correctly classified
Anticipated (reference)							0
Unsure							0
Intercept	−.621	***	.168
Sex (F)	−.114	*	.047	.892	.815	.978
Age	.015	***	.002	1.02	1.01	1.02
Race (NHW)	−.470	***	.062	.625	.554	.705
Married	−.161	***	.050	.851	.772	.939
Education	−.140	***	.029	.869	.821	.920
Income	.023	.475	.032	1.02	.961	1.09
Employed	−.192	***	.056	.825	.739	.922
eSDOH	−.070	*	.031	.933	.878	.990
Not anticipating							100
Intercept	1.964	.116	<.001
Sex (F)	−.005	.001	<.001	1.00	.99	1.0
Age	−.049	.018	.006	.95	.92	.98
Race (NHW)	−.007	.008	.434	.99	.98	1.0
Married	−.254	.030	<.001	.78	.73	.82
Education	−.383	.034	<.001	.68	.64	.73
Income	−.056	.044	.202	.95	.87	1.0
Employed	−.270	.038	<.001	.76	.71	.82
eSDOH	−.196	.020	<.001	.82	.79	.86

Note. Anticipated caregiver is the referent category, sample size is 48,121; percent correctly classified overall = 80.5%.

Similarly, as shown in Table 3, relative to those who anticipated becoming caregivers, those who were certain they would not take on the role in the near future were less likely to be female (OR = .79), somewhat younger (OR = 1.0), less likely to be married or partnered (OR = .66), reported less education (OR = .95), were less likely to be employed (OR = .76) and reported fewer eSDOH (OR = .84). Of note, neither race (OR = .93) nor income (OR = .99) differentiated between the two groups.

Discussion

With more than 37 million American adults serving as family caregivers, at a significant cost to themselves and significant cost savings to the long-term care and medical systems (US Census, 2023), identifying economic disparities in current and future caregivers and their needs may represent an important area for future public health research and intervention development. Earlier identification of populations more likely to have caregiving responsibilities could help inform interventions aimed at improving caregiver preparedness, resource awareness, and support before substantial negative caregiver outcomes. To date, however, most caregiving research has addressed predictors of caregiving burden, with an eye towards alleviating those stresses among current caregivers (Koumoutzis et al., 2021). We argue that additional avenues of research should examine the next generation of family caregivers in order to anticipate their needs. To that end, we examined four groups of adults: those who were currently providing care, those who anticipated being called into the role in the near future, those who were unsure of whether they would assume the caregiving role in the next 2 years, and those who were certain that they would not become caregivers in the near future. We used predictors from both the family caregiving literature (Broese van Groenou & De Boer, 2016; Young et al., 2020) and the health disparities literature (Duran & Pérez-Stable, 2019) to test whether these variables could discriminate among these four groups.

Using a large public data set, the results were clear: demographics and eSDOH do differentiate among current caregivers and other non-caregiving groups. Although current caregivers have several resources, such as more education and a partner, current caregivers face several economic challenges, including less annual income and more eSDOH. Current caregivers may be spending more resources than they have due to the costs of being a caregiver, such as paying for medical bills, needing to take time off work, and other costs needed to maintain the care receiver’s activities of daily living (Silva-Smith, 2007).

Compared to current caregivers, those who anticipate caregiving are a bit younger, more likely to be male, more likely to be employed, and report fewer eSDOH than current caregivers. Thus, an adult who anticipates that they will become a caregiver may be the person who has the resources to do so. Of course, these anticipated caregivers may have already begun to prepare for the caregiving role, including allocating resources to provide care (Revenson et al., 2016). As a society, we must monitor their needs as their caregiving situation changes.

In many ways, those who are unsure whether they will take on a caregiving role in the near future are similar to those who anticipate doing so. The two differences that stand out in the current analyses are that unlike those who anticipate being a caregiver, those who are unsure are a bit older than current caregivers and are less likely to be employed. Although the BRFSS does not provide data regarding to whom one might provide care, unsure caregivers were less likely to be married than current caregivers. Thus, it is unclear whether they would provide care to a partner, parent, sibling, friend or other relative. But the fact that there are many adults who do not have a clear indication that they may become a caregiver suggests the need to broaden our safety nets to help prepare these adults. It is noteworthy that those who are unsure of future caregiving roles are more likely to identify as racially or ethnically diverse.

A clearer profile emerges for those who do not expect to become a caregiver in the near future. Relative to current caregivers, these adults are more likely to be male, are a bit younger, may be more racially or ethnically diverse than current caregivers, and are unpartnered. Those who do not expect to become caregivers in the next 2 years present a mixed economic picture. They report less education and less current employment, but they also report more annual income and fewer eSDOH challenges.

Our analyses raise interesting questions. For example, although women continue to be more likely than men to be caregivers (Berg et al., 2021), the field has not asked questions related to anticipating becoming a caregiver. It is unclear whether the mostly-middle-aged men in the current sample are preparing to become caregivers. Their needs remain largely unexamined. As social and employment roles change, society must interrogate views that caregiving is “women’s work” and start preparing adults, especially men, to participate in the caregiving role.

Our analyses examining differences among the non-caregivers yielded less stable results. The model was a poor fit to the data and no variable, despite significance, discriminated strongly between any two groups. Although marital status and racial identity may suggest some groups Who may become caregivers, a broader selection of variables is necessary to ascertain potential differences among these non-caregiving adults.

Limitations and Future Directions

National health surveillance data sets provide critical insights for public good (CDC, 2019).However, these data sets also have limitations. For example, the questions used to create the count variable of eSDOH were dichotomous in terms of whether the participant experienced the challenge, but lacked the severity of the disparity and its effect on the participant. Another limitation of the public-access data is the lack of key identifiers for the anticipated caregivers and other demographics. We do not have information about who they are anticipating providing care, the conditions that warrant the person receiving care, or whether caregiving would be shared or solo. As such, future studies should examine differences in those key demographics and follow through to see if those who anticipate being a caregiver actually become one. While we do not know if those who anticipated the role became caregivers, the current study was able to examine whether certain demographics and economic factors provide insight into future caregiving. Other personal and contextual factors may be needed, including sexual orientations, gender identities, kinship position and geographic influences. Future studies should explore how these shapes the likelihood of becoming a caregiver, offering insight into which individuals may be more likely to assume caregiving responsibilities at different life stages. Additionally, due to the large sample size, many of the significant results had odds ratios close to 1.0, indicating a small effect. We caution readers against over interpretation of the results but note that small effects are still relevant in caregiving research when considering public health implications and the influence of multiple risk factors. Finally, because states may choose not to administer the Caregiving or the Social Determinants of Health and Health Equity Modules, we were able to analyze data from only eleven states and one territory. Thus, generalizability of results is limited. However, the large sample size and diversity of respondents provide valuable insights that are likely relevant across the nation.

Conclusion

The current study sought to understand the differences among caregiving groups to better understand those who may be likely to be caregivers in the future. By doing so, we open the possibility for more research on those who are likely to become caregivers to help implement policies, interventions, and information to better help the caregiving experience before the role. While we recognize that current caregivers need continued support, the purpose of this study was to increase awareness of the economic burden prior to when caregiving starts and during the role. By doing so, this research may help inform earlier identification of populations who could benefit from more resource awareness or interventions targeting being prepared for the role. Additionally, as the results of this study are exploratory and descriptive in nature, this study provides a preliminary step towards future research focusing on anticipated caregiver roles, preparedness, and caregiver outcomes. The current study also opens research on those unsure if they will be a caregiver in the next 2 years. For instance, we do not know why they are unsure about being a caregiver, and future research may be able to determine if it is due to their own health, their current resources, or other factors, such as not knowing if another family member will take on the caregiving role. With some evidence of trends toward caregivers being younger, minoritized, and an increase in male caregivers (2), this study provided more evidence of those trends. Additionally, resource differences (i.e., food security and financial security) across caregiving and non-caregiving groups inform policy and organizations where more resources should be allocated, specifically to those anticipating the caregiving role who may be unrecognized or unsupported. By screening eSDOH in primary care for individuals anticipating providing care, micro-grants or vouchers for food and transportation can be allocated more effectively for those in need. As suggested by Beach and colleagues (2022), taking a more person-centered approach to focus on individual caregiver needs may vary depending on their caregiving status.

Footnotes

Ethical Considerations

This study was approved by WVU protocol # 2107363352.

Credit Author Statement

Julie Hicks Patrick: Conceptualization, Formal Analysis, Supervision, Writing – review & editing. S. Alison Bolling: Conceptualization, Formal Analysis, Methodology, Resources, Visualization, Writing – original draft, Writing – review & editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Publicly available datasets were analyzed in this study. These data can be found at:

Generative AI disclosure

No Generative AI was used in the preparation of this manuscript.

References

Beach

Bélanger-Hardy

Harding

Rodrigues Perracini

Garcia

Tripathi

Gillis

Dow

(2022). Caring for the caregiver: Why policy must shift from addressing needs to enabling caregivers to flourish. Frontiers in Public Health, 10, 997981. https://doi.org/10.3389/fpubh.2022.997981

Berg

C. A.

Kelly

C. S.

Utz

R. L.

(2021). Family care exchanges across the lifespan. In Warner Schaie

Willis

S. L.

(Eds.), Handbook of the Psychology of Aging (9th ed., pp. 263–274). Academic Press. https://doi.org/10.1016/b978-0-12-816094-7.00002-7

Broese van Groenou

M. I.

De Boer

(2016). Providing informal care in a changing society. European Journal of Ageing, 13(3), 271–279. https://doi.org/10.1007/s10433-016-0370-7

Centers for Disease Control and Prevention . (2023). CDC - 2022 BRFSS survey data and documentation. Centers for Disease Control and Prevention. https://www.cdc.gov/brfss/annual_data/annual_2022.html. (Accessed June 12, 2025).

Cohen

S. A.

Nash

C. C.

Greaney

M. L.

(2024). Place-based, intersectional variation in caregiving patterns and health outcomes among informal caregivers in the United States. Frontiers in Public Health, 12, 1423457. https://doi.org/10.3389/fpubh.2024.1423457

de Klerk

de Boer

Plaisier

(2021). Determinants of informal care‐giving in various social relationships in the Netherlands. Health & Social Care in the Community, 29(6), 1779–1788. https://doi.org/10.1111/hsc.13286

Duran

D. G.

Pérez-Stable

E. J.

(2019). Novel approaches to advance minority health and health disparities research. American Journal of Public Health, 109, S1–S10. https://doi.org/10.2105/ajph.2018.304931

Family Caregiver Alliance . (2016). Caregiver statistics: Demographics. https://www.caregiver.org/resource/caregiver-statistics-demographics/. (Accessed October 9, 2024).

Gaugler

J. E.

Zarit

S. H.

Pearlin

L. I.

(2003). The onset of dementia caregiving and its longituinal implications. Psychology and Aging, 18(2), 171–180. https://doi.org/10.1037/0882-7974.18.2.171

10.

Grenard

D. L.

Valencia

E. J.

Brown

J. A.

Winer

R. L.

Littman

A. J.

(2020). Impact of caregiving during emerging adulthood on frequent mental distress, smoking, and drinking behaviors: United States, 2015–2017. American Journal of Public Health, 110(12), 1853–1860. https://doi.org/10.2105/ajph.2020.305894

11.

Kaufman

A. V.

Kosberg

J. I.

Leeper

J. D.

Tang

(2010). Social support, caregiver burden, and life satisfaction in a sample of rural African American and white caregivers of older persons with dementia. Journal of Gerontological Social Work, 53(3), 251–269. https://doi.org/10.1080/01634370903478989

12.

Koumoutzis

Cichy

K. E.

Dellmann-Jenkins

Blankemeyer

(2021). Age differences and similarities in associated stressors and outcomes among young, midlife, and older adult family caregivers. The International Journal of Aging and Human Development, 92(4), 431–449. https://doi.org/10.1177/0091415020905265

13.

Lahaie

Earle

Heymann

(2012). An uneven burden: Social disparities in adult caregiving responsibilities, working conditions, and caregiver outcomes. Research on Aging, 35(3), 243–274. https://doi.org/10.1177/0164027512446028

14.

Raiber

Verbakel

de Boer

(2022). Testing the informal care model: Intrapersonal change in care provision intensity during the first lockdown of the COVID-19 pandemic. European Journal of Ageing, 19(4), 1287–1300. https://doi.org/10.1007/s10433-022-00713-2

15.

Revenson

T. A.

Griva

Luszczynska

Morrison

Panagopoulou

Vilchinsky

Hagedoorn

(2016). Gender and caregiving: The costs of caregiving for women. Caregiving in the illness context (pp. 48–63). https://doi.org/10.1057/9781137558985_5

16.

Seltzer

M. M.

L. W.

(2000). The dynamics of caregiving: Transitions during a three-year prospective study. The Gerontologist, 40(2), 165–177. https://doi.org/10.1093/geront/40.2.165

17.

Silva-Smith

A. L.

(2007). Restructuring life. Journal of Family Nursing, 13(1), 99–116. https://doi.org/10.1177/1074840706297425

18.

Uhm

K. E.

Jung

Woo

M. W.

Kwon

H. E.

Oh-Park

Lee

B. R.

Kim

E. J.

Kim

J. H.

Lee

S. A.

Lee

(2023). Influence of preparedness on caregiver burden, depression, and quality of life in caregivers of people with disabilities. Frontiers in Public Health, 11, 1153588. https://doi.org/10.3389/fpubh.2023.1153588

19.

U.S. Bureau of Labor Statistics . (2023). Celebrating national family caregivers month with BLS data. U.S. Bureau of Labor Statistics. https://www.bls.gov/blog/2023/celebrating-national-family-caregivers-month-with-bls-data.htm. (Accessed May 10, 2025).

20.

Vos

E. E.

Hilderink

H. B.

de Bruin

S. R.

van der Beek

A. J.

Proper

K. I.

(2022). The working informal caregiver model: A mixed methods approach to explore future informal caregiving by working caregivers. Sustainability, 14(6), 3519. https://doi.org/10.3390/su14063519

21.

Young

H. M.

Bell

J. F.

Whitney

R. L.

Ridberg

R. A.

Reed

S. C.

Vitaliano

P. P.

(2020). Social determinants of health: Underreported heterogeneity in systematic reviews of caregiver interventions. The Gerontologist, 60(Supplement_1), S14–S28. https://doi.org/10.1093/geront/gnz148