Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Oct 24.
Published in final edited form as: J Appl Gerontol. 2025 Jul 30;45(6):1149–1159. doi: 10.1177/07334648251360819

Health Patterns of Dementia Caregivers With Chronic Pain: Latent Profile Analysis of the PROMIS-29 Measure

Shelbie G Turner 1, Dakota D Witzel 2, Shana Garza 1,3, Karl Pillemer 1,4, M Carrington Reid 1
PMCID: PMC12548739  NIHMSID: NIHMS2096163  PMID: 40738141

Abstract

We identified patterns of health and functioning in a national sample of dementia family caregivers experiencing chronic pain (N = 273). Utilizing latent profile analysis of the 29-item Patient-Reported Outcomes Measurement Information System (PROMIS-29), we identified three discrete groups: an average group (48%), a low health challenges group (30%), and a high health challenges group (22%). Women were more likely to be in the high (vs. low) health challenges group. There are discrete health patterns of caregivers with chronic pain, with over 20% reporting comorbid poor health and functioning that may negatively affect caregiving quality. Screening caregivers for high impact pain in clinical settings should be considered and prioritizing these caregivers for interventions may be an efficient use of resources.

Keywords: family caregivers, chronic pain, dementia, latent profile analysis

Introduction

Nearly all older adults with dementia require support with daily functioning at some point during the progression of their disease (Riffin et al., 2017). The vast majority of that care comes from family members and friends (Friedman et al., 2015). Estimates suggest there are between 8 (Freedman & Wolff, 2024) and 11 million (Alzheimer’s Association, 2024) dementia family caregivers who currently provide such care in the United States. The physical and psychological stress from dementia caregiving can be detrimental to caregivers’ own health (Sörensen & Conwell, 2011), and caregivers’ health has implications for caregiving outcomes such as care recipients’ hospitalization (Kuzuya et al., 2011) and mortality (Pristavec & Luth, 2020; Schulz et al., 2021) rates. Therefore, federal agencies (e.g., National Institute on Health, Centers for Disease Control), as well as dementia care advocacy organizations and research foundations (e.g., Alzheimer’s Association), have prioritized dementia caregivers’ health as a means to support the growing number of people across the United States with dementia.

Researchers have recently prioritized chronic pain as a prevalent health condition among caregivers that is consequential to caregiving outcomes. Significant numbers of family caregivers report experiencing pain (over 50%; Turner et al., 2024). Arthritis, a leading cause of chronic pain, is especially common among caregivers ages 65 and above (Turner et al., 2024) who represent over half of all caregivers to older adults (Wolff et al., 2018). Research further suggests that caregiving constitutes a risk factor for worsening pain intensity over time (Turner et al., 2025b). Indeed, caregivers report their pain is often exacerbated by performing specific care tasks, such as moving the care recipient in and out of a bed or chair (Turner et al., 2025a). In addition, caregivers report that pain creates diverse challenges to caregiving, including physical challenges such as not being able to perform care tasks and emotional challenges such as increased frustration from being unable to provide adequate care (Turner et al., 2025a). Understanding caregivers’ experiences of chronic pain, including how pain impacts their overall functioning and which caregivers with chronic pain are at an increased risk for other health challenges, constitutes a central component of this developing research topic within the dementia family caregiving literature.

Variability in Caregivers’ Pain Experiences

A large body of evidence conveys great between-person variability in how people experience pain, including the extent to which pain impacts various health domains (Cai & Oderda, 2012; Miller & Cano, 2009; Rouhi et al., 2023). Because pain so frequently affects and is exacerbated by other components of health (e.g., physical functioning, emotional and mental health, and sleep quality), when a caregiver has chronic pain, it is helpful to understand their pain in conjunction with their health in other domains (e.g., mental health, sleep, and physical functioning). As with the general population, caregivers with pain likely differ in their overall health. Unlike the general population, however, these differences are likely closely bound to−and dependent upon− caregiving experiences. Well-established caregiver theories reinforce the between-person variability in caregivers’ health. For example, Pearlin and colleagues’ (1990) stress process model of caregiving denotes how caregiving factors such as hours of care provided impact social relationship health (e.g., family role strain), psychological and emotional health (e.g., caregiver burden), and physical health.

Relatedly, and more specific to caregivers’ pain, Turner and colleagues’ (2025a) conceptual model of the challenges and consequences of caregivers’ pain depicts various inter-connected caregiving challenges in the physical, psychological and emotional, and familial and relational domains as influencing caregivers’ pain. Per the model, pain can create caregiving challenges across physical, psychological and emotional, and familial and relational domains. For example, pain may make performing certain care tasks more difficult (physical challenge) while also making a caregiver worry about how they will continue providing care if their pain condition worsens (psychological and emotional challenge). The model suggests that the extent of challenges in any one domain, as well as the extent to which those challenge’s impact health in other domains, differs between caregivers based on factors such as the size of the care network.

Ultimately, even among caregivers who experience chronic pain, we predict variability in their overall health profiles. For example, some caregivers with chronic pain may have poor sleep quality, high fatigue, and low physical functioning, whereas others may experience poor sleep quality along with high levels of fatigue, but preserved physical functioning. Though it is well-established that some caregivers have better health outcomes relative to other caregivers, there is limited research exploring this level of variability by simultaneously studying multiple health domains, and no research within the context of pain. Moreover, identifying specific patterns of health across a variety of health domains can capture this variability and is more informative than determining individual associations between pain and various health outcomes (e.g., via a regression), which has already been established in prior research (IsHak et al., 2018; Van Looveren et al., 2021).

Clinically, identifying discrete subgroups of caregivers for whom pain is most strongly associated with poor overall health and poor caregiving outcomes would allow healthcare providers to make informed choices about which caregivers should be the focus of intervention efforts (Griffin et al., 2020; Park et al., 2023). Moreover, grouping caregivers based on patterns of overall health can help identify which types of pain management interventions might be most appropriate for a certain group. For example, caregivers with poor physical functioning but who are mentally healthy may benefit more from physical therapy directed interventions (e.g., teaching safe transfer techniques to minimize care tasks that aggravate pain), whereas caregivers whose primary challenge is poor mental health may benefit more from emotion-based pain management programs (e.g., teaching skills to better cope with the negative thoughts, emotions, and behaviors that often accompany chronic pain).

The Present Study

In this study, we aimed to categorize caregivers with chronic pain into discrete subgroups of health and functioning based on the 29-item Patient Reported Outcomes Measurement Information System (PROMIS-29), a person-centered measure of physical, mental, and social health in adults. We also compared each subgroup’s PROMIS-29 scores to the PROMIS-29 scores of the general population of adults with chronic pain, and identified how subgroups were associated with caregiver demographics (e.g., race) and caregiving characteristics (e.g., hours of care provided per week).

Method

Procedures and Participants

We partnered with Qualtrics to recruit a national sample of dementia family caregivers with chronic pain. Qualtrics recruited eligible caregivers from pre-existing survey panels. Potentially eligible participants received an electronic invitation directly from Qualtrics to complete the survey, which included three screening questions to confirm eligibility. Eligibility criteria included the following: (1) participant experienced pain, aching, burning, or throbbing sensations on most days of every month for the past 3 months; (2) participant’s pain or discomfort was not due to cancer; and (3) participant provided 20 or more hours of care per week to a relative/family member or friend with Alzheimer’s disease or a related dementia.

The survey was designed to last no more than 15 min. Qualtrics directly delivered compensation to participants, which varied based on the participants’ pre-existing agreement with Qualtrics but averaged $7.50 per survey. Weill Cornell Medicine’s Institutional Review Board deemed this study exempt from human subjects review given data were anonymous.

Measures

PROMIS-29.

The Patient-Reported Outcomes Measurement Information System (Cella et al., 2010) is a set of publicly-accessible surveys to assess general physical and emotional health and functioning. In this study, we utilized the PROMIS-29, a 29-item survey that measures seven health domains, each comprised of four questions, plus a single-item question about pain intensity. The PROMIS-29 is widely used as a standardized tool to evaluate and monitor patient health and is validated across a variety of chronic health conditions including chronic pain (Deyo et al., 2015).

Cronbach’s alphas for this study’s sample ranged from .81–.94 across each domain, suggesting high internal consistency. We computed sum scores for each of the seven domains and then converted each mean score to a t-score where the mean was 50 and the standard deviation was 1, per suggested PROMIS-29 scoring guidelines (HealthMeasures, 2024) and in line with existing studies conducting latent profile analysis of the PROMIS-29 (Stone et al., 2019). We used these t-scores and the single pain intensity item as variables included in latent profile analyses.

Caregiving Characteristics and Caregiver Demographics.

Caregivers reported the following caregiving characteristics: their relationship to the care recipient, including spouse/partner, child, or child-in-law, related in some other way (sibling, grandchild, niece, nephew, aunt, uncle, cousin, or an unrelated friend); whether they considered themselves the primary caregiver; how many other family members or friends were providing care to the person with dementia; how far away they lived from the care recipient; how long they had been providing care to the care recipient; and how many days per week and hours per day they provide care (we multiplied the number of days per week by the number of hours per day to create a single variable representing, as closely as possible with this dataset, the total number of hours per week spent providing caregiving). They also reported the following demographics: age, gender, race, ethnicity, and income.

Analytic Strategy

To identify patterns of physical and emotional functioning among caregivers with chronic pain, we ran latent profile analyses following guidelines from (Masyn, 2013) in MPLUS. We tested whether the data fit between one and five groups across four separate covariance structures (varying unrestricted, invariant unrestricted, varying diagonal, and invariant diagonal). We then compared model fit across these models, choosing the best absolute and relative fit that was parsimonious and theoretically relevant. Better absolute fit was indicated by smaller log likelihood (LL). Better relative fit was indicated smaller Akaike information criterion (AIC), Bayesian information criterion (BIC), and Akaike’s Bayesian information criterion (ABIC), larger Bayes factor, and significant Lo–Mendell–Rubin likelihood ratio test (VLMR-LRT) and the bootstrapped likelihood ratio test (BLRT). We additionally examined smallest class proportions (>.05) and entropy (closer to 1, >.8) to ensure that groups were large and separate enough for analyses.

We then built a multinomial logistic regression model using SAS PROC LOGISTIC to analyze associations between caregiver demographics and caregiving characteristics and group membership.

Results

Characteristics of the Study Sample

A total of 4,400 people responded to Qualtrics’ invitation to complete the survey. Of this number, the vast majority (4,099) were ineligible because they did not meet the criterion for experiencing chronic pain. Twenty-eight additional participants’ responses were removed due to data quality issues (e.g., straight lining), resulting in a final sample of 273 caregivers.

Caregivers’ demographics and caregiving characteristics were fairly similar to national statistics of dementia family caregivers (Alzheimer’s Association, 2024). Caregivers ranged in age from 19 to 83, with an average age of 45 and most (56%) were women. Seventy-five percent were White, 17% were Black, and 6% were Asian, American Indian/Alaskan Native, and 3% identified as more than one race. Thirty-one percent were of Hispanic, Latino, or Spanish origin.

Fifty-three percent were adult-child caregivers, 14% were a spouse/partner of the person with dementia, 20% were related to the person with dementia in another way, and 12% were an unrelated friend of the person with dementia. The vast majority (89%) identified as the primary caregiver to the care recipient, while 41% were the sole caregiver. Fifty-one percent lived in the same home as the care recipient, and an additional 38% lived less than 10 miles away. Seventy-seven percent of caregivers had been providing care for over 1 year but less than 6 years. Caregivers provided, on average, 66 h of care each week (range = 20–168). Additional characteristics of the sample broken down by subgroups are presented in Table 1.

Table 1.

Sample Characteristics by Subgroup

High health challenges Average group Low health challenges
group (n = 57) (n = 127) group (n = 83)
Caregiving characteristics
 Relationship to care recipient
 Spouse 21% 12% 11%
 Adult-child 58% 54% 52%
 Unrelated friend 18% 10% 8%
 Related in some other way 3% 24% 29%
 Primary caregiver 88% 90% 89%
Number of other caregivers
 None 44% 37% 46%
 1+ 56% 63% 54%
Distance lived from the care recipient
 Live in the same home 51% 54% 48%
 Less than 10 miles away 35% 37% 42%
 11+ miles away 14% 9% 10%
Duration of caregiving
 Less than a year 9% 10% 8%
 1–2 years 44% 43% 53%
 3–5 years 35% 33% 28%
 6–10 years 7% 6% 6%
 More than 10 years 5% 8% 5%
 Hours of caregiving per week 65.50 h (SD = 43.19) 73.61 h (SD = 54.37) 54.37 h (SD = 42.51)
Caregiver demographics
 Age 47.05 years (SD = 12.86) 44.73 years (SD = 12.41) 44.52 years (SD = 11.74)
 Gender 64% women 40% women 47% women
 Race
  White 81% 76% 69%
  Black or African American 14% 13% 24%
  Asian, American Indian/Alaskan Native 2% 9% 4%
  Multiple races 4% 2% 4%
Ethnicity 33% Hispanic 30% Hispanic 33% Hispanic
 Income
  Under $30,000 16% 24% 17%
  $30,000–$49,999 23% 18% 22%
  $50,000–$74,999 23% 27% 14%
  $75,000–$99,999 11% 9% 16%
  $100,000–$149,999 16% 17% 20%
  $150,000 or more 12% 5% 11%
Open in a new tab

Latent Profile Analysis

We compared models across four covariance-variance structures (Table 2). Two solutions provided adequate fit above others: a three-group solution with an invariant diagonal covariance-variance structure, and a four-group solution with an invariant unrestricted variance-covariance structure based on relative and absolute fit statistics, as well as classification diagnostics. Although the four-group solution fit better with regard to BIC, LL, Bayes Factor, and entropy, the three-group solution provided significant differences based on VLMR-LRT and larger proportion sizes, suggesting good fit and more equal group sizes. As such, we chose the more parsimonious 3-group model as the profile solution.

Table 2.

Model Fit Indices from Latent Profile Analysis

# of Classes # of Parameters Smallest Class Proportion Absolute Fit
 Relative Fit
LL Scale AIC BIC ABIC BF p (VLMR-LRT) p (BLRT) EK
1 44 - −7357.15 1.1826 14802.31 14960.15 14820.64 - - - -
Invariant diagonal
2 25 0.43 −7583.83 1.1438 15217.66 15307.35 15228.08 4.04494E − 76 <.0001 <.0001 0.87
3 34 0.22 −7043.19 1.2004 14154.37 14276.34 14,168.54 7.5991 E + 223 0.0083 <.0001 0.88
4 43 0.12 −7006.51 1.1691 14099.03 14253.28 14,116.94 101620.4425 0.1011 <.0001 0.85
5 52 0.10 −7403.64 1.2254 14911.29 15097.82 14,932.95 4.0701E − 184 0.3106 <.0001 0.85
Invariant unrestricted
2 53 0.15 −7336.61 1.1672 14779.22 14969.34 14,801.3 0.010092124 0.08 <.0001 0.85
3 62 0.15 −7316.59 1.1793 14757.18 14979.59 14,783.02 0.005934337 0.23 <.0001 0.79
4 71 0.12 −6867.87 1.1903 13877.74 14132.44 13907.33 9.0605 E + 183 0.38 <.0001 0.91
5 No convergence
Varying diagonal
2 No convergence
Varying unrestricted
2 89 0.12 −7236.61 1.2317 14651.23 14970.49 14688.31 0.005667541 0.56 <.0001 .98
3 No convergence
Open in a new tab

Figure 1 details the three groups that emerged. We created t-scores for each PROMIS domain score, where 50 was the mean and the United States average per PROMIS scoring guidelines (HealthMeasures, 2024). Given this, the largest group (n = 127) reported approximately average scores across all variables and was thus termed the “average group.” The second group was the smallest, representing approximately 22% of the sample (n = 57). This group reported higher levels of anxiety, depression, fatigue, sleep problems, pain interference, and pain intensity relative to the other two groups. In addition, this group reported poorer physical functioning and less engagement in social activities compared to the other two groups. We termed this group the “high health challenges” group. The final group, termed the “low health challenges” group, included approximately 30% of the sample (n = 83). This group reported better functioning across all measures, including lower anxiety, depression, and pain, and more engagement in social activities (Table 3).

Figure 1.

Figure 1.

Open in a new tab

T-Scores for PROMIS-29 Domains by Subgroup

Table 3.

T-Scores by Group

T-scores
Mean (standard deviation)
High health challenges group Average group Low health challenges group United States’ adults with chronic paina
Physical functioning 41.84 (10.13) 48.39 (8.25) 58.07 (5.80) 36.23
Anxiety 60.35 (7.93) 51.07 (6.99) 41.26 (7.27) 53.94
Depression 59.78 (8.89) 51.46 (7.85) 41.05 (4.98) 54.45
Fatigue 61.88 (5.38) 51.74 (5.80) 39.18 (5.62) 58.50
Sleep disturbance 56.55 (8.72) 52.41 (7.57) 41.81 (8.82) 57.19
Social participation 37.59 (5.59) 49.08 (5.63) 59.93 (6.80) 40.06
Pain interference 62.12 (5.27) 50.47 (6.40) 40.95 (7.57) 64.61
Pain intensity 7.67 (1.41) 6.15 (1.51) 4.43 (2.00) 6.38
Open in a new tab

Note. Pairwise comparisons revealed domain means each group are significantly different from the other two groups at p < .0001, except for the sleep disturbance comparison between the average group and the high health challenges group, which was significant at 0.002.

a

based on results from Pope et al., 2021.

Pairwise comparisons revealed the domain means for each group were significantly different from the other two groups (all p values <.0001), except for the sleep disturbance comparison between the average group and the high health challenges group, which was significant at the p = .002 level.

Comparisons to Reference Populations

The PROMIS-29 t-score for the average adult in the United States is 50 across all domains (except the pain intensity item, which is not t-scored). For adults in the United States with chronic pain conditions, t-scores are as follows: pain interference = 64.61, sleep disturbance = 57.19, fatigue = 58.50, anxiety = 53.94, depression = 54.45, social participation = 40.06, and physical functioning = 36.23 (Pope et al., 2021). The average pain intensity of United States adults with chronic pain is 6.38 on an 11-point scale (0–10; Pope et al., 2021).

Using these metrics for comparison, our results suggest that over one-fifth of dementia family caregivers with chronic pain are not only performing poorly but are performing worse than the general adult population and the population of adults with chronic pain. In particular, the high health challenges caregivers had higher levels of anxiety, depression, fatigue, sleep disturbance, and social participation than the reference population of United States adults with chronic pain. They also reported higher mean pain intensity (7.67). The remainder of the sample was performing on par with (48% of the sample), or even healthier than (30% of the sample) adults with chronic pain as well as the general adult population (Table 3).

Associations with Caregiver Demographics and Caregiving Characteristics

Compared to spousal caregivers, caregivers who reported being related to the care recipient in another way (sibling, grandchild, niece, nephew, aunt, uncle, and cousin) were less likely to be in the high health challenges compared to the healthy group (Estimate = −3.26, p = .001, Odds Ratio (OR) = 0.04; 95%; Confidence Interval (CI): [0.01, 0.28]). Compared to caregivers who were the sole caregiver in the care network, caregivers with at least one other caregiver in the care network were more likely to be in the average group compared to the low health challenges group (Estimate = 0.71, p = .04, OR = 2.03; 95% [CI = 1.02, 4.04]. Women in the sample were more likely to be in the high (vs. low) health challenges group (Estimate = 0.90, p = .03, OR = 2.47; 95% CI = [1.07, 5.68]). Results for all caregiver demographics and caregiving characteristics are presented in Table 4.

Table 4.

Associations Between Subgroup Membership and Caregiver Demographics and Caregiving Characteristics

High health challenges groupversus low health challenges group* Average group versus low health challenges group*
Odds ratio (95% Confidence Interval) Odds ratio (95% Confidence Interval)
Caregiving characteristics
Relationship to care recipient
 Spouse ref ref
 Adult-child 0.45 (0.13–1.53) 0.98 (0.32–2.96)
 Un-related friend 1.05 (0.21–5.26) 1.78 (0.41–7.81)
 Related in some other way 0.04 (0.01–0.28) 0.74 (0.20–2.71)
Number of other caregivers
 None ref ref
 1+ 1.68 (0.74–3.83) 2.03 (1.02–4.04)
Distance lived from the care recipient
 Live in the same home ref ref
 Less than 10 miles away 0.88 (0.32–2.43) 0.86 (0.40–1.87)
 11+ miles away 2.12 (0.56–7.97) 1.03 (0.33–3.19)
Duration of caregiving
 Less than a year ref ref
 1–2 years 1.93 (0.43–8.56) 1.16 (0.38–3.56)
 3–5 years 3.55 (0.74–17.12) 1.95 (0.58–6.50)
 6+ years 1.72 (0.30–9.75) 1.46 (0.38–5.37)
Hours of caregiving per week 1.01 (1.00–1.02) 1.01 (1.00–1.02)
Caregiver demographics
 Age 1.00 (0.96–1.02) 0.99 (0.95–1.02)
 Gender
  Men ref ref
  Women 2.47 (1.07–5.68) 1.53 (0.80–2.90)
Race
 White ref ref
 Black or African American 0.56 (0.15–1.17) 0.38 (0.17–0.86)
 Asian, American Indian/Alaskan Native 0.56 (0.05–6.24) 2.02 (0.48–8.51)
 Multiple races 1.05 (0.13–8.73) 0.44 (0.07–2.60)
 Ethnicity
  Non-Hispanic ref ref
  Hispanic 1.33 (0.56–3.19) 0.85 (0.49–1.71)
 Income
  Under $30,000 ref ref
  $30,000–$49,999 1.26 (0.37–4.30) 0.56 (0.21–1.48)
  $50,000–$74,999 2.90 (0.81–10.35) 1.60 (0.59–4.32)
  $75,000–$99,999 1.30 (0.30–5.65) 0.53 (0.17–1.61)
  $100,000–$149,999 0.88 (0.23–3.39) 0.53 (0.19–1.48)
  $150,000 or more 1.48 (0.34–6.38) 0.27 (0.07–1.04)
Open in a new tab

Note. Bold text indicates significant difference from the reference category.

*

= significant difference from the reference group.

Discussion

We sought to determine whether discrete subgroups existed based on health and functioning among dementia family caregivers living with chronic pain. Within this sample, latent profile analysis revealed three distinct subgroups of caregivers based on health and functioning across pain intensity and seven additional health domains. These results are indicative of between-person variability in chronic pain experiences and are an endorsement of latent profile analysis as a helpful analytic tool to identify which dementia family caregivers with pain have especially poor health outcomes.

Key Findings

Comparisons to Reference Populations.

The substantial prevalence of high health challenges caregivers in our sample (more than 1 in 5) reinforces caregivers with chronic pain as population at a particularly high risk for negative health outcomes, especially as it pertains to mental health outcomes. Notably, however, the high health challenges caregivers in our sample had less pain interference and better physical functioning than the reference population of United States adults with chronic pain (Figure 2). These caregivers may be overcoming their pain in order to perform care tasks, and/or performing caregiving tasks may help caregivers maintain their physical functioning (i.e., “use it or lose it”). Future research that probes physical functioning specifically as it pertains to completing care tasks may help elucidate the specific strengths and limitations of physical functioning and pain interference in caregivers with chronic pain.

Figure 2.

Figure 2.

Open in a new tab

T-Scores for PROMIS-29 Domains, Comparison Between Poor Performing Subgroup and U.S. Adults With Chronic Pain

The remainder of the sample was performing on par with (48% of the sample), or even healthier than (30% of the sample) adults with chronic pain as well as the general adult population. These caregivers could be experiencing the healthy caregiver effect (Li et al., 2023), whereby those who enter the caregiving role do so because they are healthier and/or whereby the act of caregiving helps a person maintain or even improve their health. Indeed, there are positive benefits to caregiving for some caregivers, such as increased physical activity, social interaction, and purpose and meaning (Lloyd et al., 2016; Quinn & Toms, 2019). Given all caregivers in our sample endorsed chronic pain, and thus our sample was a less healthy sample, the high prevalence of healthy caregivers in this sample raises new questions about the healthy caregiver effect, such as how caregivers cope with chronic conditions (e.g., chronic pain) in order to continue providing care. We encourage future research in this area.

Ultimately, finding that a portion of the caregivers in this sample endorsed chronic pain but had low pain intensity and were otherwise healthy across other domains, reinforces between-person variability with respect to the pain experience. It further indicates that a subset caregivers exist who are in greater need of support related to managing their pain and associated conditions.

Health and Functioning Domain Correlations Across Subgroups.

Rather than identifying multiple subgroups with variability across health domains (i.e., one group having high anxiety and high sleep disturbance and another having high anxiety but low sleep disturbance), the profile analysis identified three groups with consistent ratings across domains. That the health and functioning domains correlated together so clearly reinforces the extent to which pain is associated with major comorbid health challenges in this population, particularly for the 20% of the sample that had considerably poorer health and functioning across all seven domains compared to the remainder of the sample.

An avenue for future research is to determine the extent to which pain, versus some other component of caregivers’ health (e.g., poor sleep and fatigue), contributes to caregiving difficulty. It may be challenging methodologically to isolate or partition, but determining strategies to do so can help advance targeted pain management solutions for these caregivers and thus may be a worthwhile endeavor.

Demographics and Caregiving Characteristics.

Existing scholarship on the prevalence of pain among caregivers suggests that women and men are equally likely to report pain (Turner et al., 2024). Results from this study suggest that even though women may not be more likely to have pain compared to men, when they do have pain, they have worse health outcomes (more comorbidities and worse functioning) than men. Indeed, we found that women were more likely to be in the high health challenges group (when compared to the low health challenges group). The inter-section between caregiving, pain, and gender is an important one to explore given women are disproportionally impacted by both pain (Osborne & Davis, 2022) and caregiving (Xiong et al., 2020). For example, women often provide more intense caregiving by performing care tasks that are more physically and emotionally taxing. They also spend more time caregiving than men; in this sample, gender was significantly associated with hours of care provided per week, with women providing 72 h per week compared to men who were providing 58 h per week. These more intense and time-consuming caregiving roles likely limit women’s ability to manage health challenges and may also contribute to the development of new health challenges, rendering results like what we found in this study wherein women have poorer overall health than men.

Additionally, caregivers with at least one other care-giver in the care network were more likely to be in the average group compared to the low health challenges group. This finding is in line with existing research, which suggests that the presence of other members of the care network can actually create more caregiving challenges for the caregiver with pain because it creates more opportunity for family relationship strain (Turner et al., 2025a). It is also possible, however, that the reason care networks are larger is out of necessity because of the poor health and associated limitations of the caregiver with pain. Ultimately, how care networks contribute to health and well-being of caregivers with pain constitutes yet another avenue for further inquiry.

Clinical Implications

Because of the extent to which the high health challenges caregivers reported worse health outcomes across all domains, clinical assessment and interventions focused solely on pain reduction may not adequately support these caregivers who are also likely experiencing impaired sleep, poor mental health, reduced physical functioning, and disrupted social involvement. Rather, once a care provider identifies pain, it is important to assess for mental and physical health, sleep, and social challenges as well. Likewise, multi-component interventions that have some specific focus on pain but also target other health domains will likely be necessary, although admittedly challenging to deliver given time demands from caregiving. To this end, it may be particularly useful to address the social and emotional aspects of both caregiving and pain, especially by harnessing the sense of purpose and meaning known to frequently exist among caregivers. Early efforts to address this possibility are exemplified through Turner’s (2024–2029) work to adapt and test a positive psychology intervention that helps promote positive emotions among caregivers with pain.

Relatedly, as the research on caregivers’ pain expands, researchers may experience difficulty recruiting caregivers with pain into pain management programs and clinical studies. Based on the results of our study, we posit that this may be because many caregivers with pain are doing well. Caregivers who endorse chronic pain but who do not have poor health may not feel that studies on caregivers’ pain are relevant to them, nor that enrolling would be helpful. Thus, researchers may require more targeted recruitment strategies to identify, recruit, and retain these caregivers.

Limitations

There are several limitations of this study that warrant discussion. First, we utilized Qualtrics survey sampling to identify caregivers with chronic pain. It is possible that Qualtrics’ recruitment strategies are biased towards healthy respondents, and we did not ask them to oversample for caregivers with poor health. Thus, it is possible that this sample was healthier than the general caregiving population which could explain why 30% of our sample had better health than the general U.S. population. Second, given the cross-sectional nature of the survey, we are unable to measure how long a caregiver resides in each group. Perhaps, for example, some caregivers momentarily “dip” into the high health challenges group during period of acute distress, but do not remain in that group long.

Additionally, once we divided the sample into three subgroups, the cell sizes for logistic regression analysis became small. Thus, for some of the between-group comparisons, we were underpowered to detect between-group differences, which could be a rationale for the wide confidence intervals seen in our models. For example, even though we oversampled for Black and/or Hispanic caregivers, there were only 8 Black caregivers in the high health challenges group. The smaller sample sizes should be taken into consideration when interpreting results from the logistic regression model.

Finally, the three profiles we identified may reflect different levels along a continuum rather than wholly distinct subgroups. Indeed, the presence of the three groups might be indicative of the “salsa effect” (Sinha et al., 2021) whereby a model forces an underlying continuum into ranked classes (e.g., low, average, and high). Nonetheless, there is value in identifying groups using non-arbitrary, and potentially clinically useful, cut-offs (Hickendorff et al., 2018), though we do encourage future research that attempts to identify these groups in other samples.

Conclusion

In this study, for the first time, we documented the overall health and functioning patterns of dementia family caregivers with chronic pain, which is a substantial subset of a growing population. Though many dementia caregivers experience their own chronic pain, results from this study reinforce that there is great individual variability in pain intensity and overall health and functioning among this group of caregivers. Notably, over one-fifth of the caregivers had pain intensity and overall health and functioning that was worse than the general population with chronic pain, identifying them as a severely unhealthy group in urgent need of solutions to address the health challenges that likely impact caregiving quality. Because these low functioning caregivers are more likely to have multiple comorbid health challenges, addressing caregivers’ pain likely requires multi-component intervention strategies that address multiple domains of health. Ultimately, this study extends the body of research on caregivers’ pain by revealing significant comorbid health consequences, including which caregivers are more prone to these comorbidities, that are likely responsible for challenges providing care to the growing number of older adults with dementia.

What this paper adds

  • Over one-fifth of the caregivers in the current study reported levels of pain intensity and overall health and functioning that were worse than the general United States population with chronic pain.

  • Efforts to identify and intervene on these affected caregivers, whose pain and other health challenges likely impact caregiving quality, should be prioritized.

Application of study findings

  • Because the caregivers in the high health challenges group are more likely to have multiple comorbid health conditions, addressing caregivers’ pain will likely require multi-component interventions that address multiple domains of health.

  • As clinical providers are increasingly considering caregivers’ health as a part of patient treatment plans, screening for caregivers’ physical pain may be helpful.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institutes of Health/National Institute on Aging [T32 AG049666 to SGT, K99 AG086523 to SGT, R03AG088758 to DDW, T32 AG049676 to DDW, K24 AG053462 to MCR, and P30 AG022845 to MCR and KAP].

Footnotes

Ethical Approval

This study was deemed exempt from review from the Weill Cornell Medicine Institutional Review Board.

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

The research team will consider reasonable requests to use anonymous data. Please send a request to the corresponding author.

References

  1. Alzheimer’s Disease Facts and Figures. (2024). Alzheimer’s & Dementia, 20(5), 3708–3821. 10.1002/alz.13809 [DOI] [Google Scholar]
  2. Cai B, & Oderda GM (2012). The association between pain and depression and some determinants of depression for the general population of the United States. Journal of Pain & Palliative Care Pharmacotherapy, 26(3), 257–265. 10.3109/15360288.2012.703292 [DOI] [PubMed] [Google Scholar]
  3. Cella D, Riley W, Stone AA, Rothrock N, Reeve BB, Yount SE, Amtmann DA, Buysse D, Choi S, Cook K, Devellis R, DeWalt D, Fries JF, Gershon R, Hahn EA, Lai JS, Pilkonis P, Revicki D, & PROMIS Cooperative Group. (2010). The patient-reported outcomes measurement information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005–2008. Journal of Clinical Epidemiology, 63(11), 1179–1194. 10.1016/j.jclinepi.2010.04.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Deyo RA, Ramsey K, Buckley DI, Michaels L, Kobus A, Eckstrom E, Forro V, & Morris C (2015). Performance of a patient-reported outcomes measurement information System (PROMIS) short form in older adults with chronic musculoskeletal pain. Pain Medicine, 17(2), 314–324. 10.1093/pm/pnv046 [DOI] [Google Scholar]
  5. Freedman VA, & Wolff JL (2024). Dementia and caregiving trends and the emerging policy environment. In Presented at the 2024 NHATS/NSOC research conference on caregiving to older adults with dementia, June 20–21, 2024. Ann Arbor, Michigan. [Google Scholar]
  6. Friedman EM, Shih RA, Langa KM, & Hurd MD (2015). U.S. prevalence and predictors of informal caregiving for dementia. Health Affairs, 34(10), 1637–1641. 10.1377/hlthaff.2015.0510 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Griffin JM, Riffin C, Havyer RD, Biggar VS, Comer M, Frangiosa TL, & Bangerter LR (2020). Integrating family caregivers of people with alzheimer’s disease and dementias into clinical appointments: Identifying potential best practices. Journal of Applied Gerontology: The Official Journal of the Southern Gerontological Society, 39(11), 1184–1194. 10.1177/0733464819880449 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. HealthMeasures. (2024). PROMIS adult profile instruments scoring manual. Retrieved from. https://www.healthmeasures.net/images/PROMIS/manuals/Scoring_Manual_Only/PROMIS_Adult_Profile_Scoring_Manual_16Sept2024.pdf
  9. Hickendorff M, Edelsbrunner PA, McMullen J, Schneider M, & Trezise K (2018). Informative tools for characterizing individual differences in learning: Latent class, latent profile, and latent transition analysis. Learning and Individual Differences, 66(2018), 4–15. 10.1016/j.lindif.2017.11.001 [DOI] [Google Scholar]
  10. IsHak WW, Wen RY, Naghdechi L, Vanle B, Dang J, Knosp M, Marcia L, Gohar Y, Eskander L, Yadegar J, Hanna S, Sadek A, Aguilar-Hernandez L, Danovitch I, & Louy C (2018). Pain and depression: A systematic review. Harvard Review of Psychiatry, 26(6). 352, 363. 10.1097/HRP.0000000000000198 [DOI] [PubMed] [Google Scholar]
  11. Kuzuya M, Enoki H, Hasegawa J, Izawa S, Hirakawa Y, Shimokata H, & Akihisa I (2011). Impact of caregiver burden on adverse health outcomes in community-dwelling dependent older care recipients. The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry, 19(4), 382–391. 10.1097/JGP.0b013e3181e9b98d [DOI] [PubMed] [Google Scholar]
  12. Li L, Wister AV, & Mitchell BA (2023). Examination of the healthy caregiver effect among older adults: Findings from the Canadian Longitudinal Study on Aging. Gerontology, 69(3), 289–300. 10.1159/000526251 [DOI] [PubMed] [Google Scholar]
  13. Lloyd J, Patterson T, & Muers J (2016). The positive aspects of caregiving in dementia: A critical review of the qualitative literature. Dementia, 15(6), 1534–1561. 10.1177/1471301214564792 [DOI] [PubMed] [Google Scholar]
  14. Masyn KE (2013). Latent class analysis and finite mixture modeling. In Little TD (Ed.), The Oxford handbook of quantitative methods: Statistical analysis (pp. 551–611). Oxford University Press. [Google Scholar]
  15. Miller LR, & Cano A (2009). Comorbid chronic pain and depression: Who is at risk? The Journal of Pain, 10(6), 619–627. 10.1016/j.jpain.2008.12.007 [DOI] [PubMed] [Google Scholar]
  16. Osborne NR, & Davis KD (2022). Chapter eight - sex and gender differences in pain. In Moro E, Arabia G, Tartaglia MC, & Ferretti MT (Eds.), International review of neurobiology (Vol. 164, pp. 277–307). Academic Press. 10.1016/bs.irn.2022.06.013 [DOI] [PubMed] [Google Scholar]
  17. Park T, Pillemer K, Loeckenhoff C, Suitor JJ, & Riffin C (2023). What motivates physicians to address caregiver needs? The role of experiential similarity. Journal of Applied Gerontology: The Official Journal of the Southern Gerontological Society, 42(5), 1003–1012. 10.1177/07334648231151937 [DOI] [Google Scholar]
  18. Pearlin LI, Mullan JT, Semple SJ, & Skaff MM (1990). Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist, 30(5), 583–594. 10.1093/geront/30.5.583 [DOI] [PubMed] [Google Scholar]
  19. Pope JE, Fishman M, Chakravarthy K, Hanes M, Gerling M, Heros R, Shah J, Orhurhu V, Urits I, & Deer TR (2021). A retrospective, multicenter, quantitative analysis of patients’ baseline pain quality (PROMIS-29) entering into pain and spine practices in the United States (ALIGN). Pain and Therapy, 10(1), 539–550. 10.1007/s40122-021-00238-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Pristavec T, & Luth EA (2020). Informal caregiver burden, benefits, and older adult mortality: A survival analysis. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 75(10), 2193–2206. 10.1093/geronb/gbaa001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Quinn C, & Toms G (2019). Influence of positive aspects of dementia caregiving on caregivers’ well-being: A systematic review. The Gerontologist, 59(5), e584–e596. 10.1093/geront/gny168 [DOI] [PubMed] [Google Scholar]
  22. Riffin C, Van Ness PH, Wolff JL, & Fried T (2017). Family and other unpaid caregivers and older adults with and without dementia and disability. Journal of the American Geriatrics Society, 65(8), 1821–1828. 10.1111/jgs.14910 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Rouhi S, Topcu J, Egorova-Brumley N, & Jordan AS (2023). The impact of sleep disturbance on pain perception: A systematic review examining the moderating effect of sex and age. Sleep Medicine Reviews, 71, Article 101835. 10.1016/j.smrv.2023.101835 [DOI] [Google Scholar]
  24. Schulz R, Beach SR, & Friedman EM (2021). Caregiving factors as predictors of care recipient mortality. The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry, 29(3), 295–303. 10.1016/j.jagp.2020.06.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sinha P, Calfee CS, & Delucchi KL (2021). Practitioner’s guide to latent class analysis: Methodological considerations and common pitfalls. Critical Care Medicine, 49(1), e63–e79. 10.1097/CCM.0000000000004710 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Sörensen S, & Conwell Y (2011). Issues in dementia caregiving: Effects on mental and physical health, intervention strategies, and research needs. The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry, 19(6), 491–496. 10.1097/JGP.0b013e31821c0e6e [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Stone AL, Holley AL, Dieckmann NF, & Wilson AC (2019). Use of the PROMIS-29® to identify subgroups of mothers with chronic pain. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association, 38(5), 422–430. 10.1037/hea0000663 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Turner SG, Mindlis I, Reid MC, & Pillemer KA (2025a). Caregiving challenges from persistent pain among family caregivers to people with dementia. The Gerontologist, 65(2), Article gnae164. 10.1093/geront/gnae164 [DOI] [Google Scholar]
  29. Turner SG (2024. –2029). Adaptation of path-pain for dementia family caregivers living with persistent pain and negative emotions, 1K99AG086523–01 [Grant]. National Institute on Aging. https://reporter.nih.gov/project-details/10865920 [Google Scholar]
  30. Turner SG, Reid MC, & Pillemer KA (2025b). Pain prevalence and intensity among family caregivers versus non-caregivers in the United States. Journal of Aging and Health, Article 89826432. [advanced online publication]. 10.1177/08982643251331247 [DOI] [Google Scholar]
  31. Turner SG, Robinson JRM, Pillemer KA, & Reid MC (2024). Prevalence estimates of arthritis and activity-limiting pain among family caregivers to older adults. The Gerontologist, 64(5), Article gnad124. 10.1093/geront/gnad124 [DOI] [Google Scholar]
  32. Van Looveren E, Bilterys T, Munneke W, Cagnie B, Ickmans K, Mairesse O, Malfliet A, De Baets L, Nijs J, Goubert D, Danneels L, Moens M, & Meeus M (2021). The association between sleep and chronic spinal pain: A systematic review from the last decade. Journal of Clinical Medicine, 10(17), 3836. 10.3390/jcm10173836 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wolff JL, Mulcahy J, Huang J, Roth DL, Covinsky K, & Kasper JD (2018). Family caregivers of older adults, 1999– 2015: Trends in characteristics, circumstances, and role-related appraisal. The Gerontologist, 58(6), 1021–1032. 10.1093/geront/gnx093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Xiong C, Biscardi M, Astell A, Nalder E, Cameron JI, Mihailidis A, & Colantonio A (2020). Sex and gender differences in caregiving burden experienced by family caregivers of persons with dementia: A systematic review. PLoS One, 15(4), e0231848. 10.1371/journal.pone.0231848 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The research team will consider reasonable requests to use anonymous data. Please send a request to the corresponding author.

RESOURCES