Caregiver- and Care Recipient–Related Factors Linked to Pain in Family Caregivers of Persons Living With Dementia

Abstract

Background:

Family caregivers of persons living with dementia (PLWD) often experience significant psychological, financial, and social stress, increasing their risk for both mental and physical health problems.

Objective:

This study examined factors associated with caregivers’ perceived pain in the context of dementia care.

Methods:

This was cross-sectional study using baseline data from a dyadic clinical trial evaluating a behavioral sleep intervention for PLWD and their family caregivers. A total of 148 caregivers were included in the analytic sample. Descriptive statistics, bivariate analyses, and hierarchical linear regression were used to identify caregiver- and care recipient-related factors associated with caregiver pain.

Results:

Most caregivers (84%) reported pain in the past week. In the final regression model, lower education, greater caregiver comorbidity, and poorer caregiver sleep quality independently predicted higher pain, explaining 30.6% of the variance and highlighting key caregiver health factors linked to pain.

Introduction

Caring for a person living with dementia (PLWD) is emotionally, physically, and time intensive, placing family caregivers at increased risk for adverse mental and physical health outcomes (Parker et al., 2022), such as elevated caregiver burden, psychological distress, functional limitations, and increased morbidity (Gilhooly et al., 2016). Physical health problems, including pain, are common among caregivers and are often attributed to the physical strain and chronic stress of caregiving (Blyth et al., 2008; Turner, Robinson, et al., 2024). Pain can interfere with family functioning and is associated with greater psychological distress, poorer health, and increased caregiver burden (Jones et al., 2011; Perkins et al., 2013). Caregivers of PLWD may be particularly vulnerable to pain due to greater caregiving intensity and physical demands compared with other caregiving groups (Parker et al., 2022).

Dementia caregiving-related pain is influenced by multiple factors, including caregivers’ sociodemographic characteristics such as race, age, education, and self-efficacy, health conditions, and caregiving demands (Abshire Saylor et al., 2026; Boon et al., 2023; Nahin, 2021; Turner, Robinson, et al., 2024; Turner et al., 2025). Care recipient factors such as behavioral symptoms, sleep problems, and functional dependence may further contribute to caregiver strain and pain (Rodríguez-González et al., 2021; Turner, Pillemer, et al., 2024). Yet despite growing recognition of these individual contributors, pain among dementia caregivers remains relatively understudied, and few studies have examined caregiver and care recipient factors together within a unified conceptual framework.

Pearlin’s Stress Process Model offers a theoretically grounded basis for understanding how caregiving stress translates into adverse health outcomes such as pain (Pearlin et al., 1990). Within this model, primary stressors including care recipient behavioral symptoms, sleep disturbances, and functional dependence and secondary stressors such as caregiver burden, diminished self-efficacy, and physical health conditions interact dynamically to shape caregiver wellbeing. Empirical work supports this interplay. Vernon et al. (2019) found that caregiver stress and care recipient symptoms mutually influence one another in ways that can either exacerbate or buffer the demands of dementia care. Similarly, another study (Ning et al., 2026) demonstrated that elevated caregiver burden is associated with worsening care recipient mood, underscoring the bidirectional nature of the caregiving dyad. Taken together, these findings suggest that examining caregiver pain in isolation from care recipient characteristics may yield an incomplete picture of the factors driving pain in this population.

The present study examined factors associated with perceived pain among caregivers of PLWD, accounting for both caregiver and care recipient characteristics. We hypothesized that caregivers’ pain levels would be significantly associated with both caregiver and care recipient factors.

Methods

Participants and Setting

This cross-sectional study used baseline data from an ongoing clinical trial evaluating the efficacy of Care2Sleep, a behavioral dyadic sleep intervention for PLWD and their family caregivers (NCT05452031). Participants were recruited through three healthcare systems, clinical partners, and public advertisements in California, United States.

Eligible caregivers were ≥18 years old, lived with the PLWD, provided assistance with at least one activity of daily living (ADL)(Katz et al., 1970) or instrumental ADL (IADL)(Lawton & Brody, 1969) for ≥ 6 months, had a Pittsburgh Sleep Quality Index (PSQI)(Buysse et al., 1989) score >5 indicating poor sleep quality, had a Montreal Cognitive Assessment (MoCA)(Nasreddine et al., 2005) ≥23, and were English-speaking. A MoCA cutoff of ≥23 was used, consistent with meta-analytic evidence supporting this threshold as optimal for identifying intact cognition (Carson et al., 2018; Islam et al., 2023). PLWD were eligible if they had a documented dementia diagnosis, were community-dwelling, ambulatory, had at least one sleep problem occurring more than three times per week, as indicated on the Neuropsychiatric Inventory Nighttime Behavior Scale (Cummings et al., 1994) and had an eligible caregiver who lived together. PLWD were excluded for severe untreated sleep disorders, terminal illness other than dementia (e.g., cancer with a prognosis of less than 6 months), or being bedbound. Written informed consent was obtained from caregivers. For PLWD who lacked capacity to consent, consent was obtained from their legally authorized representatives, who were often the caregivers themselves. Baseline data from 148 caregivers were analyzed, representing the sample enrolled at the time of this analysis.

Measures

Caregiver pain was assessed using a single PROMIS Global Health item (Hays et al., 2009) measuring perceived pain severity as average pain intensity over the past seven days, rated on a scale from 0 (no pain) to 10 (worst imaginable pain). Pain is thus operationalized in this study as a unidimensional measure of severity rather than a multidimensional construct encompassing interference or affective dimensions. Caregiver characteristics included sociodemographic factors, caregiving context (relationship, duration, hours of care), number of comorbid conditions, and pain medication use. Depressive symptoms were measured using the Center for Epidemiological Study of Depression Scale (CES-D) (Andresen et al., 1994), caregiver self-efficacy using the Caregiver Self-Efficacy Scale (CSES) (Ritter et al., 2022), and caregiving burden using the Zarit Burden Interview (ZBI) (Zarit et al., 1980). Cognitive function was assessed with the MoCA (Nasreddine et al., 2005) and sleep quality with the PSQI (Buysse et al., 1989). Care recipient characteristics included sleep quality (PSQI), behavioral symptoms (Revised Memory and Behavior Problem Checklist [RMBPC] (Teri et al., 1992)), physical function (ADL/IADL) (Katz et al., 1970; Lawton & Brody, 1969), and quality of life (Quality of Life-Alzheimer’s Disease Scale [QOL-AD]) (Logsdon et al., 2002).

Data Analysis

Descriptive statistics were calculated for all variables. Bivariate associations between caregiver pain and continuous variables were examined using Pearson correlations, and independent samples t-tests were used for categorical variables. Variables significant at p< 0.05 were entered into a hierarchical linear regression model with pain as the dependent variable. This approach was selected to examine the incremental contribution of each block of variables to explained variance in pain, consistent with the Stress Process Model. Block 1 included caregivers’ sociodemographic characteristics (education and ethnicity), Block 2 included physical comorbidities and pain medication use, Block 3 included caregiver health variables (depressive symptoms, cognitive function, sleep quality), and Block 4 included care recipients’ characteristics (sleep quality, behavioral symptoms, and quality of life). Missing data were minimal and handled via listwise deletion. Statistical significance was set at p < 0.05. Analyses were conducted using Stata version 15.

Results

Participant Characteristics

Caregivers had a mean age of 64.1 years (SD = 13.2); most were women and non-Hispanic. Over half were caring for a spouse or partner. Caregivers had provided care for an average of 6.9 years and 58.8 hours per week. Mean CES-D score was 15.9 and mean ZBI score was 41.1, indicating moderate burden. Mean caregiver pain was 3.4 out of 10 (SD = 2.6), and 84% reported pain in the past week. Approximately one-quarter reported using pain medication. Caregivers and care recipient characteristics are summarized in Table 1.

Table 1.

Characteristics of Caregivers and Care Recipients (N = 148)

Variables	Caregivers	Care Recipients
	Mean (SD)/ n (%)	Mean (SD)/ n (%)
Age, years	64.1 (13.2) Age range (24-93)	78.8 (9.0) Age range (57-105)
Gender
Male	25 (16.9%)	88 (60.3%)
Female	123 (83.1%)	58 (39.7%)
Race
White	88 (59.5%)	93 (62.8%)
Black	14 (9.5%)	17 (11.5%)
Asian	23 (15.5%)	19 (12.8%)
Other or more than one	23 (15.5%)	19 (12.8%)
Ethnicity
Hispanic	25 (16.9%)	29 (19.9%)
Non-Hispanic	123 (83.1%)	117 (80.1%)
Education level
Less than high school	0	10 (6.9 %)
Some high school	3 (2.0%)	2 (1.4 %)
High school graduate	8 (5.4%)	22 (15.1%)
Higher than high school	136 (91.9%)	102 (68.9%)
Employed (Yes)	55 (37.2%)	-
Relationship to a care recipient		-
Spouse or significant other	87 (58.8%)	-
Adult Child	52 (35.1%)	-
Others	9 (6.1%)	-
Years of caregiving	6.9 (7.7)
Weekly caregiving duration (hours)	58.8 (45.0)
Number of comorbidities	1.3 (1.6)	1.8 (2.0)
Pain medication use (Yes/No)	36 (24.2%)	-
CES-D total score (range 0-60)	15.9 (10.0)	-
CSES total score (range 0-10)	6.1 (2.1)	-
ZBI total score (range 0-88)	41.1 (16.7)	-
MoCA total score (range 0-30)	26.8 (2.7)	-
PSQI total score (range 0-21)	8.3 (4.1)	8.4 (4.1)
ADLs score (0-6)	-	4.3 (1.9)
IADLs score (0-8)	-	1.9 (1.7)
RMBPC frequency score (1-4)	-	1.5 (0.6)
RMBPC caregiver reaction score (1-4)	1.8 (0.7)	-
QOL-AD total score (range 0-52)	-	30.3 (6.0)

CES-D, Center for Epidemiological Studies Depression; CSES, Caregiver Self-Efficacy Scale; ZBI, Zarit Burden Index; MoCA, Montreal Cognitive Assessment; PSQI, Pittsburgh Sleep Quality Index; ADL, activities of daily living; IADL, instrumental ADLs; RMBPC, Revised Memory and Behavior Problems Checklist; QOL-AD, Quality of Life in Alzheimer’s Disease

Bivariate Associations with Caregiver Pain

Higher caregiver pain was significantly associated with greater depressive symptoms (CES-D: r = 0.20), more comorbid conditions (r = 0.30), poorer sleep quality (PSQI: r = 0.30; all p < 0.05), lower cognitive function (MoCA: r = −0.20), and fewer years of education (r = −0.20) (all p< 0.05). Hispanic caregivers (mean 4.4 [SD 2.7] vs. 3.3 out of 10 [SD 2.5], p = 0.049) and those using pain medication reported higher pain than their counterparts (mean 4.4 [SD 2.5] versus 3.2 [SD 2.5] out of 10, p = 0.016).

Care recipient factors associated with higher caregiver pain included poorer sleep quality (PSQI: r = 0.30, p = 0.004), greater behavioral symptom frequency (RMBPC: r = 0.20, p = 0.026), and lower quality of life (QOL-AD: r = −0.20, p = 0.005). Caregiver burden, self-efficacy, and care recipients’ ADL and IADL performance were not significantly associated with caregiver pain. Detailed bivariate results are presented in Supplemental Tables 1 (Pearson correlations) and 2 (t-tests).

Multivariable Regression Models

Hierarchical regression results are presented in Table 2. Sociodemographic characteristics (Block 1) explained 5.9% of the variance in caregiver pain (p = 0.012), physical health variables (Block 2) explained an additional 13.3% (p < 0.001), and other caregiver health factors (Block 3) explained an additional 7.6% (p < 0.001). The full model, including care recipient characteristics (Block 4), explained 30.6% of the variance (adjusted R² = 24.4%). In the final model (Supplemental Table 3), higher education was associated with lower pain (B = −0.469, p = 0.008), such that each additional year of education corresponded to approximately half a point reduction in pain on the 0-10 scale. Greater comorbidity burden was associated with higher pain (B = 0.460, p = 0.002), with each additional comorbid condition corresponding to nearly half a point increase in pain severity. Poorer sleep quality was also independently associated with higher pain (PSQI; B = 0.146, p = 0.011), with each one-unit increase in PSQI score corresponding to approximately a 0.15-point increase in pain. Other variables were not significant after adjustment.

Table 2.

Regression Models Predicting Caregivers’ Pain

Outcome	F(df)	P-value	R2	R2 change
Block 1 (caregiver sociodemographic)
Years of education, ethnicity	4.53 (2,144)	0.012	0.059	-
Block 2 (caregiver physical health)
Comorbidity, pain medication use	8.46 (4,142)	<0.001	0.192	0.133
Block 3 (caregiver other health)
Depressive symptoms (CES-D), cognitive function (MoCA), subjective sleep quality (PSQI)	6.00 (7, 115)	<0.001	0.268	0.076
Block 4 (care recipient characteristics)
Care recipients’ sleep (PSQI), problematic behavior (RMBPC frequency), quality of life (QOL-AD)	4.90 (10,111)	<0.001	0.306	0.038

CES-D, Center for Epidemiological Studies Depression; MoCA, Montreal Cognitive Assessment; PSQI, Pittsburgh Sleep Quality Index; RMBPC, Revised Memory and Behavior Problems Checklist; QOL-AD, Quality of Life in Alzheimer’s Disease

Discussion

This study examined factors associated with perceived pain among caregivers of PLWD. Most caregivers reported pain during the prior week (84%), with a mean severity of 3.4 out of 10. This prevalence is notably higher than estimates (~53%) reported in the general older adult population (LaRowe et al., 2024; Patel et al., 2013), highlighting the substantial physical burden associated with dementia caregiving (Turner, Mindlis, et al., 2024; Turner, Pillemer, et al., 2024). Our findings extend prior work by examining both caregiver and care recipient characteristics simultaneously.

In the present study, caregiver education, comorbidity burden, and sleep quality were the strongest predictors of pain. Higher education was associated with lower pain, possibly reflecting differences in health literacy, access to resources, or occupational exposures. Greater comorbidity burden was associated with higher pain, consistent with prior evidence linking multimorbidity with increased physical strain among caregivers (Turner et al., 2025), underscoring the importance of chronic disease management as part of caregiver health support. Poor sleep quality was also independently associated with higher pain, consistent with the well-established bidirectional relationship between sleep disruption and pain (Finan et al., 2013). Given that this sample was recruited specifically for poor sleep, this finding is particularly notable and suggests that sleep-focused interventions may have downstream benefits for caregiver pain. Because eligibility required poor sleep quality, the range of sleep scores in this sample was restricted, which can attenuate observed correlations relative to an unselected population; that the association remained significant despite this restriction is notable, though whether it generalizes across the full spectrum of sleep quality, including caregivers with normative sleep, remains to be determined.

Average pain severity in the present study was modest compared to some earlier studies of caregivers of PLWD (Anderson et al., 2025; Moss et al., 2021), which may reflect differences in pain assessment methods or sample characteristics. Only one-quarter of participants reported using pain medications, substantially lower than rates reported elsewhere (Wojtyna & Popiołek, 2015). This suggest that pain may be substantially undertreated in this population and warranting greater clinical attention to pain management in dementia caregiving contexts.

Depressive symptoms were positively associated with caregiver pain in bivariate analyses but did not remain significant after adjustment. This finding is consistent with evidence that pain and depression often co-occur in caregivers (Finan et al., 2013; Jones et al., 2011) and suggests that their shared variance may be captured by comorbidity and sleep quality in multivariable models.

At the bivariate level, poor care recipient sleep quality, greater behavioral symptom frequency, and lower quality of life were each associated with higher caregiver pain; however, these associations did not remain statistically significant in the final multivariable model after accounting for caregiver-level factors. This pattern suggests that the influence of care recipient characteristics on caregiver pain may be partially explained by caregiver health factors such as comorbidity and sleep quality. Nevertheless, these findings are consistent with the Stress Process Model framework, which posits that care recipient primary stressors contribute to caregiver strain, and future studies with larger samples may have greater power to detect these effects after adjustment.

This study has several limitations. The cross-sectional design precludes causal inference, and the direction of observed associations cannot be established. Caregiver pain was assessed using a single-item measure capturing average pain severity only, which does not account for other dimensions such as pain interference or affective components. Reliance on self-reported data may have introduced bias, potentially affecting the accuracy of reported pain levels and caregiving characteristics. The sample was predominantly older, female, and non-Hispanic, and was recruited in California, which may limit generalizability to more diverse caregiving populations and other geographic regions. As noted above, the restricted range of sleep scores resulting from eligibility criteria may limit generalizability of the sleep-pain association to caregivers with normative sleep. Future studies using real-time or ecological momentary assessment approaches may better capture fluctuations in pain and analgesic use. Finally, unmeasured factors such as socioeconomic status, healthcare access, and caregiving-specific physical demands may have influenced results.

In summary, pain among family caregivers of PLWD is associated with comorbidity burden, sleep quality, and education. Addressing caregiver health and care recipient symptoms may help improve caregiver well-being and sustain dementia care.

Supplementary Material

Supplementary tables

Supplemental Table 1. Bivariate Pearson Correlations of Caregiver- and Care Recipient–Related Factors with Caregiver Perceived Pain

Supplemental Table 2. Bivariate Analyses of Pain Differences by Caregiver-Related Characteristics

Supplemental Table 3. Final Regression Model Predicting Caregivers’ Pain

What This Paper Adds

• This study provides new evidence that perceived pain is highly prevalent among family caregivers of persons living with dementia, with the majority reporting pain in the past week.

• By simultaneously examining caregiver and care recipient factors, the study identified caregiver education, comorbidity burden, and sleep quality as key predictors of caregiver pain.

• Care recipient sleep disturbances and behavioral symptoms showed preliminary associations with caregiver pain, underscoring the need for future research in this area.

Application of Study Finding

• Pain should be routinely assessed as part of caregiver health evaluations, particularly among dementia caregivers with multiple chronic conditions or poor sleep.

• Interventions targeting sleep improvement and chronic disease management may help reduce pain and improve overall well-being among caregivers.