Financial Toxicity and Multidimensional Economic Strain Among Dementia Caregivers

Abstract

INTRODUCTION:

Financial toxicity is an underrecognized but significant burden among caregivers of individuals with Alzheimer’s disease and related dementias (ADRD). The interaction between economic hardship, caregiver burden, and social determinants of health (SDOH) remains poorly defined.

METHODS:

We conducted a cross-sectional survey of 79 ADRD caregivers recruited through the University of Kansas Alzheimer’s Disease Research Center. Participants completed the COST measure, Zarit Burden, Resource Utilization in dementia questionnaire, and an SDOH assessment. The primary outcome was moderate-to-severe financial toxicity (COST ≤26).

RESULTS:

Financial toxicity affected 50.6% of caregivers. After adjusting for caregiver age and household income, higher caregiver burden was associated with greater financial toxicity (OR = 1.12), while older age was associated with lower odds of financial toxicity (OR = 0.93), and low household income was associated with greater financial toxicity compared with high income households (OR = 7.63).

DISCUSSION:

Financial toxicity is common among ADRD caregivers and closely linked to burden and socioeconomic disadvantages, underscoring the need for policy and support interventions.

Background:

Financial toxicity refers to the financial burden and distress associated with illness and care, including out-of-pocket expenses, lost income, debt, and depletion of household resources. Dementia imposes not only profound emotional and physical challenges for patients and families, but also a substantial and often underrecognized economic burden. Although financial toxicity can affect both patients and caregivers, its manifestations differ across these groups. For patients, financial strain is often driven by direct medical and treatment-related costs, whereas for caregivers it more commonly reflects indirect and cumulative consequences, including reduced employment, lost wages, debt accumulation, and depletion of personal assets associated with prolonged caregiving responsibilities and, in some cases, caregivers’ own declining health¹. While financial toxicity has been most extensively studied in oncology, it is increasingly recognized as a broader construct capturing multidimensional financial strain across chronic illness contexts, including dementia caregiving².

The emotional and psychological toll of dementia caregiving has been widely documented; however, the financial consequences experienced by caregivers have received comparatively less systematic attention^3,4. Existing studies often examine financial strain as a secondary outcome or consider economic burden without explicitly situating it within broader caregiving stress processes or social contexts. Prior work shows that financial burden varies by sociodemographic characteristics, including race and income, yet many studies either combine patient and caregiver costs or omit explicit consideration of social determinants of health (SDOH)⁵. As a result, the pathways through which caregiving demands, caregiver burden, and structural disadvantages jointly contribute to financial toxicity remain insufficiently characterized⁶.

To address this gap, we draw on the Stress Process Model, which conceptualizes caregiving outcomes as the result of interactions between primary stressors (e.g., care demands), secondary stressors (e.g., role strain, financial pressure), and available coping resources, all operating within a broader social context. Within this framework, caregiving demands can lead to psychological and emotional burden, which may constrain caregivers’ ability to maintain employment, manage household finances, and recover economically over time. We further integrate insights from SDOH theory, which emphasizes that socioeconomic position, access to resources, and structural inequities shape individuals’ exposure to stressors and their capacity to cope. From this perspective, financial toxicity represents not a discrete event but an accumulated outcome arising from the interaction of caregiving stressors and pre-existing or co-occurring social and economic disadvantage^7–10.

Guided by this integrated conceptual framework, this study adopts a multidimensional approach to examining financial toxicity among Alzheimer’s caregivers by jointly considering financial strain, caregiver burden, caregiving demands, and social disadvantage. By examining these domains together, we aim to provide a more holistic understanding of caregiver financial vulnerability and the mechanisms through which burden and structural factors intersect to produce economic hardship. This approach moves beyond prior work that isolates financial strain or caregiver burden and instead situates financial toxicity within a dynamic stress process shaped by broader social conditions.

Understanding financial toxicity within the dementia caregiver population is essential for informing clinical practice, social services, and healthcare policy. We hypothesize that greater caregiver burden is associated with more severe financial toxicity, consistent with the Stress Process Model’s emphasis on secondary stress proliferation. Collectively, this work seeks to advance a more equitable framework for supporting dementia caregivers, one that addresses not only emotional and physical caregiving demands, but also the structural and financial conditions that critically shape caregiver well-being.

Methods:

Study Design and Oversight

We conducted a cross-sectional observational study to examine associations between caregiver burden, social determinants of health (SDOH), and financial toxicity among caregivers of individuals with dementia. The study was approved by the University of Kansas Medical Center Institutional Review Board (STUDY00160985). All participants provided written informed consent before participation. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Setting and Participants

Caregivers were recruited between November 2024 and November 2025 through the University of Kansas Alzheimer’s Disease Research Center (KU ADRC) and affiliated caregiver outreach efforts. The Outreach, Recruitment, and Engagement (ORE) Core at the University of Kansas Alzheimer’s Disease Research Center (KU ADRC) assisted in identifying potential participants from the KU ADRC database and through social media digital marketing campaigns. Eligible participants were adults (≥18 years) who provided unpaid care to an individual diagnosed with dementia and who resided in the same household as the care recipient. Participants were required to complete all study instruments in English. Individuals were excluded if they submitted incomplete surveys, duplicate responses (with the most recent response retained), or reported serving as their own caregiver.

Data Collection and Measures

All data were collected using standardized, self-administered electronic surveys administered via REDCap^11,12.

The primary outcome was financial toxicity (FT), operationalized using the Comprehensive Score for Financial Toxicity-Functional Assessment of Chronic Illness Therapy (COST-FACIT) questionnaire, a validated 12-item instrument with a total score ranging from 0 to 44, with higher scores indicating better financial well-being (i.e., lower financial toxicity).⁷ Financial toxicity was dichotomized as moderate-to-severe (COST-FACIT ≤26) versus not affected (>26), consistent with prior literature.¹³

The primary variable of interest (exposure) was caregiver burden, measured continuously using the Zarit Caregiver Burden Survey. Covariates considered in the analysis included caregiver age, household income, caregiving duration, caregiving hours, employment status, and the presence of at least one social determinant of health challenge.

Caregiver burden was measured using the 22-item Zarit Caregiver Burden, a widely used and validated instrument with scores ranging from 0 to 88, where higher scores reflect greater caregiver burden.⁸

Caregiving time was assessed using the Resource Utilization in Dementia (RUD) questionnaire, which captures formal and informal caregiving activities for individuals with dementia.¹⁰ Total weekly caregiving hours were calculated by summing time spent assisting with meals, grooming, and household activities. Reported minutes were rounded to the nearest hour, and implausibly high values were capped to limit undue influence on regression estimates.

Social determinants of health were assessed using the University of Kansas Health System SDOH screening tool, adapted from the Health Leads framework.⁹ A binary SDOH indicator was created to reflect the presence of at least one reported challenge related to food insecurity, housing instability, utility insecurity, or healthcare access.

Demographic variables included caregiver age, sex, race, ethnicity, marital status, employment status, household income (categorized as low, medium, or high), rural–urban residence based on Rural–Urban Commuting Area (RUCA) classification, caregiving duration, and number of additional caregivers.

Sample Size Calculation

An a priori power analysis for logistic regression was conducted using G*Power. Financial toxicity was the binary outcome, and the primary predictor was Zarit caregiver burden score. Assuming a two-sided α = 0.05, 80% power, an odds ratio of 1.10 per 1-point increase in Zarit score, a baseline outcome probability of 0.50, Zarit score approximated as normally distributed (mean = 25, SD = 15), and R² = 0.25 between Zarit and other covariates, the required total sample size was 70 participants¹⁴.

Data Processing and Bias Mitigation

Of 204 survey responses collected, 91 complete records were identified. Duplicate entries (n=11) were resolved by retaining the most recent response, and one response was excluded because the participant reported serving as their own caregiver, resulting in a final analytic sample of 79 caregivers. Standardized, validated instruments were used to reduce measurement bias, and surveys were completed anonymously to mitigate social desirability bias.

Statistical Analysis

Descriptive statistics were used to summarize caregiver characteristics and study variables. Continuous variables were reported as means with standard deviations and medians with ranges, and categorical variables were summarized as counts and percentages. Bivariate associations with financial toxicity were examined using independent-sample t tests for continuous variables and ${\chi }^2$ or Fisher’s exact tests for categorical variables, as appropriate. Pearson correlation coefficients were calculated among key continuous measures.

Univariate logistic regression models were fitted to estimate unadjusted associations between each covariate and financial toxicity. These models were used descriptively and were not used as the basis for automated variable selection for the multivariable model. Multivariable logistic regression was used to estimate the adjusted association between caregiver burden and moderate-to-severe financial toxicity. Caregiver burden, measured by the Zarit score, was treated as the primary variable of interest (exposure) and was included in all multivariable models. Results from logistic regression models were reported as odds ratios (ORs), 95% confidence intervals (CIs), and p-values.

Covariate selection for the primary adjusted model was guided by a directed acyclic graph (DAG) representing hypothesized relationships among caregiver burden, financial toxicity, and caregiver characteristics (Supplementary Figure S1). The DAG was used as a conceptual tool to guide variable adjustment for the primary model, and findings were interpreted as associations given the cross-sectional study design. Caregiver age and household income were included in the primary adjusted model because they were considered clinically relevant and plausibly associated with both caregiver burden and financial toxicity. These variables were selected to support adjustment while reducing the risk of overfitting given the modest sample size. Additional covariates of interest, including SDOH, employment status, caregiving duration, and caregiving hours, were evaluated in sensitivity analyses. These analyses added each covariate to the primary model individually and in combination to assess the robustness of the association between caregiver burden and financial toxicity. Prespecified interaction terms between Zarit score and caregiver age and between Zarit score and household income were evaluated.

Missing Zarit scores were imputed using corresponding short-Zarit responses. Missing caregiving hours were imputed using the cohort mean. All statistical tests were two-sided with a significance threshold of α = 0.05. Analyses were conducted using SAS version 9.4 and R version 4.5.1.^15,16

Results:

The analytic cohort consisted of 79 caregivers as shown in Figure 1, approximately half of whom were classified as experiencing moderate-to-severe financial toxicity (50.6%). Participants were predominantly female (78.5%), White (89.9%), married (78.5%), and resided in urban areas (84.8%). The mean caregiver age was 63.5 years, and caregivers had provided care for a median of 3 years, reporting a mean Zarit caregiver burden score of 41.3 and an average of 9 caregiving hours per week. The number of additional caregivers per patient, as assessed using the Resource Utilization in Dementia (RUD) instrument, ranged from 0 to 4, with a median of 0, indicating that many participants were the sole caregiver, and 16.5% screened positive for at least one social determinant of health (SDOH) challenge. Fewer than 30% of participants (22, 27.8%) reported “Low” household income. Financial toxicity responses were nearly evenly split between the two categories (Moderately to Severely Affected vs. Not Affected; 40 vs. 39 participants), and the mean and median COST-FACIT scores on the continuous scale (23.44 and 25, respectively) were both below the cutoff value of 26, indicating that, on average, the cohort tended toward moderate-to-severe financial toxicity. Table 1 summarizes key demographic characteristics and study variables for the analytic cohort.

Figure 1.

summarizes participant flow from initial survey responses to the final analytic cohort.

Table 1:

Summary Statistics of Study Variables

Categorical Characteristic	Count (%)
Financial Toxicity (binary)
Moderately to Severely Affected	40 (50.6)
Not Affected	39 (49.4)
SDOH $
0	66 (83.5)
1	13 (16.5)
Sex
Male	17 (21.5)
Female	62 (78.5)
Ethnicity
Hispanic or Latino	6 (7.6)
Not Hispanic or Latino	71 (89.9)
Unknown/Not Reported	2 (2.5)
Race
Black or African American	5 (6.3)
White	71 (89.9)
American Indian/Native Alaskan	1 (1.3)
More than one race	2 (2.5)
Marital Status
Married	62 (78.5)
Divorced	6 (7.6)
Single	9 (11.4)
Widowed	2 (2.5)
Employment Status
Employed	33 (41.8)
Unemployed	44 (55.7)
Unable to work	2 (2.5)
Household Income
Low	22 (27.8)
Medium	34 (43.1)
High	23 (29.1)
RUCA Score
Rural	11 (13.9)
Urban	67 (84.8)
Unknown	1 (1.3)
Continuous Characteristic	Mean (SD); Median (Range)
Financial Toxicity (Continuous) *	23.44 (10.45); 25 (0,43)
Age +	63.53 (14.00); 67 (25,85)
Duration of care	3.62 (2.95); 3 (1,15)
Zarit	41.33 (17.07); 39 (10,78)
Caregiving hours +	9.06 (9.56); 6 (0,45)
Number of additional caregivers	0.66 (0.99); 0 (0,4)

^*Financial toxicity measured using the Comprehensive Score for Financial Toxicity–Functional Assessment of Chronic Illness Therapy (COST-FACIT), range 0–44; lower scores indicate greater financial toxicity. Moderate-to-severe financial toxicity defined as COST ≤26.

^$SDOH indicates presence of ≥1 reported challenge related to food insecurity, housing instability, utility insecurity, or healthcare access.

⁺Age reported in years; duration of care reported in years.

Pearson correlation coefficients and corresponding p-values are presented in Supplementary Figure S2. The heatmap indicates a statistically significant positive correlation between COST-FACIT score (continuous) and age, suggesting that older caregivers tended to report lower financial toxicity (higher COST-FACIT scores). In contrast, COST-FACIT scores were significantly and negatively correlated with Zarit caregiver burden scores, indicating that greater caregiver burden was associated with higher financial toxicity. Zarit scores were positively associated with caregiving hours, and the number of additional caregivers was positively associated with caregiver age. Taken together, these findings suggest that caregiver burden and caregiving time are closely linked, and that higher burden is particularly related to greater financial toxicity.

Based on Supplementary Table S1, there were significant mean differences in age and Zarit caregiver burden scores between participants classified as not affected and those classified as experiencing moderate-to-severe financial toxicity. In contrast, duration of care, number of additional caregivers, and caregiving hours did not differ significantly between the two groups. Supplementary Table S2 summarizes the bivariate associations between categorical variables and financial toxicity status. Household income, SDOH, and employment status showed statistically significant associations with financial toxicity status. However, race, ethnicity, sex, and marital status were not significantly associated with financial toxicity status.

Based on the univariate logistic regression results in Table 2, the odds of moderate-to-severe financial toxicity decreased by approximately 6% for each additional year of caregiver age (OR = 0.94, 95% CI: 0.91–0.98, p = 0.004). Conversely, the odds of moderate-to-severe financial toxicity increased by approximately 11% for each one-point increase in the Zarit caregiver burden score (OR = 1.11, 95% CI: 1.06–1.16, p < 0.001). These findings are consistent with the Pearson correlation patterns observed using the continuous COST-FACIT scale. Caregivers who screened positive for at least one SDOH challenge had higher odds of moderate-to-severe financial toxicity compared with those reporting no SDOH challenges (OR = 7.02, 95% CI: 1.44–34.17, p = 0.016). Caregivers in the low household income group had higher odds of moderate-to-severe financial toxicity than those in the high-income group (OR = 4.42, 95% CI: 1.21–16.12, p = 0.024). Employment status showed a significant unadjusted association with financial toxicity, with unemployed caregivers having lower odds of moderate-to-severe financial toxicity compared with employed caregivers (OR = 0.22, 95% CI: 0.09–0.59, p = 0.003). Some unadjusted estimates for categories with very small cell counts, particularly several race categories and the “unable to work” employment group, had wide or unstable confidence intervals and should be interpreted cautiously.

Table 2:

Unadjusted Odds Ratios from Univariate Logistic Regression Models for Financial Toxicity

Variable		Odds Ratio (95% CI)	p-value
Zarit score #		1.11 (1.06,1.16)	<0.001
Age		0.94 (0.91,0.98)	0.004
Household* Income	Low	4.42 (1.21,16.12)	0.024
	Medium	0.80 (0.27,2.36)	0.692
Ethnicity *	Hispanic or Latino	0.97 (0.18,5.15)	0.974
	Unknown/Not Reported	0.97 (0.06,16.16)	0.984
Race *	American Indian/Native Alaskan	1.6×107 (0, Inf)	0.994
	Black or African American	0.69 (0.11,4.36)	0.689
	More than one race	1.6×107 (0, Inf)	0.992
Sex *	Female	1.63 (0.55,4.82)	0.381
Marital Status *	Married	0.25 (0.05,1.31)	0.100
	Divorced	0.29 (0.03,2.69)	0.274
	Widowed	0.29 (0.01,6.91)	0.441
Employment Status *	Unemployed	0.22 (0.09,0.59)	0.003
	Unable to work	6.8×106 (0, Inf)	0.993
Duration of care		1.17 (0.98,1.41)	0.078
Number of additional caregivers		1.09 (0.69,1.72)	0.702
Caregiving hours		1.03 (0.98,1.08)	0.313
SDOH *	Yes	7.02 (1.44,34.17)	0.016

Note: Estimates for sparse categorical subgroups should be interpreted cautiously because small cell counts may result in unstable confidence intervals.

^*Reference categories: Financial Toxicity (FT) = Not Affected (0); Ethnicity = Not Hispanic or Latino; Race = White; Sex = Male; Marital Status = Single; Employment Status = Employed; SDOH = 0; Household Income = High.

^#Zarit score was treated as the primary exposure variable.

In the primary adjusted model (Table 3), caregiver burden remained significantly associated with financial toxicity after adjustment for caregiver age and household income. Each one-point increase in Zarit score was associated with approximately 12% higher odds of moderate-to-severe financial toxicity (adjusted OR = 1.12, 95% CI: 1.06–1.19, p < 0.001). Caregiver age remained associated with financial toxicity, with each additional year associated with approximately 7% lower odds of moderate-to-severe financial toxicity (adjusted OR = 0.93, 95% CI: 0.88–0.97, p = 0.005). Compared with caregivers in the high household income group, caregivers in the low household income group had substantially higher odds of moderate-to-severe financial toxicity (adjusted OR = 7.63, 95% CI: 1.38–53.46, p = 0.027), whereas the medium-income group did not differ significantly from the high-income group (adjusted OR = 0.71, 95% CI: 0.15–3.27, p = 0.653).

Table 3:

Adjusted Odds Ratios from the Primary Multivariable Logistic Regression Model for Financial Toxicity

Variable		Odds Ratio (95% CI)	p-value
Zarit score #		1.12 (1.06, 1.19)	<0.001
Age		0.93 (0.91, 0.97)	0.005
Household* Income	Low	7.63 (1.38, 53.46)	0.027
	Medium	0.71 (0.15, 3.27)	0.653

Note: The primary adjusted model included Zarit score, age, and household income based on DAG-guided confounder selection (Supplementary Figure S2). No statistically significant interaction effects were observed in interaction analyses (Supplementary Table S3).

^*Reference categories: Financial Toxicity (FT) = Not Affected (0); Household Income = High.

^#Zarit score was treated as the primary exposure variable.

Sensitivity analyses are summarized in Table 4. The association between Zarit caregiver burden and financial toxicity remained stable across alternative adjustment sets. The Zarit OR ranged from approximately 1.11 to 1.12 across models additionally adjusting for SDOH, employment status, caregiving duration, caregiving hours, and the expanded covariate set. These findings suggest that the association between caregiver burden and financial toxicity was robust to additional covariate adjustment.

Table 4:

Sensitivity Analyses Assessing the Association Between Zarit Caregiver Burden Score and Financial Toxicity Across Alternative Adjustment Sets

Model	Zarit score Odds Ratio (95% CI)	p-value
Unadjusted (univariate): FT~ Zarit score	1.11 (1.06, 1.17)	<0.001
Primary: FT~ Zarit score+ Age+ Household Income	1.12 (1.06, 1.19)	<0.001
Primary + SDOH	1.11 (1.06, 1.19)	<0.001
Primary + Employment Status	1.12 (1.06, 1.19)	<0.001
Primary + Caregiving Duration of care	1.12 (1.06, 1.19)	<0.001
Primary + Caregiving Hours	1.12 (1.06, 1.19)	<0.001
Expanded: Primary + SDOH + Employment Status + Duration of care + Caregiving Hours	1.12 (1.06, 1.20)	<0.001

Note: Reference categories: Financial Toxicity (FT) = Not Affected (0); Employment Status= Employed; SDOH = 0; Household Income = High. The primary adjusted model included Zarit score, age, and household income based on DAG-guided confounder selection (Supplementary Figure S2). Primary + ”x” indicates that the listed x-covariate was added to the primary adjusted model.

Prespecified interaction terms among variables in the primary model, including Zarit score, caregiver age, and household income, were evaluated (Supplementary Table S3). No statistically significant interactions were observed.

Discussion:

This study was designed to examine how caregiver burden, socioeconomic disadvantage, and social determinants of health relate to financial toxicity among caregivers of individuals with Alzheimer’s disease and related dementias. Our findings suggest that financial toxicity in dementia caregiving is not simply a function of household finances alone but rather emerges at the intersection of caregiving strain and structural vulnerability. In particular, the strong association between caregiver burden and financial toxicity supports the view that the demands of caregiving can contribute to economic hardship through disrupted employment, reduced earnings, and diminished financial resilience over time. This interpretation is consistent with the Stress Process Model and with social determinants of health frameworks, both of which emphasize that distress is shaped not only by direct care demands but also by the social and economic conditions in which caregiving occurs.^2,5–10

These findings align with prior literature in dementia caregiving showing that financial hardship is patterned by sociodemographic and caregiving factors rather than distributed uniformly across caregivers. Recent work in nationally representative dementia caregiver samples has shown that financial toxicity is more common among socially and economically vulnerable groups, while other studies have linked caregiver stress, lower quality of life, and caregiving burden to worse caregiver outcomes.^5,6,13,17 Our results extend this literature by showing that caregiver burden remains strongly associated with financial toxicity even when socioeconomic factors are considered simultaneously, suggesting that burden is not merely a parallel correlate but may be one pathway through which financial strain accumulates. This pattern is also consistent with evidence from other chronic illness contexts, particularly oncology, where caregiver and family financial toxicity has been recognized as a distinct and consequential dimension of illness burden.^1,18

At the same time, caregiver burden and financial toxicity should not be interpreted as interchangeable constructs. Caregiver burden reflects the emotional, physical, time-related, and role-related strain associated with providing care, whereas financial toxicity captures the material and psychological consequences of economic hardship, including reduced income, out-of-pocket costs, debt, and constrained financial recovery. Their overlap likely reflects a reinforcing cycle: more intensive caregiving can limit workforce participation and destabilize household finances, while financial strain can heighten stress and worsen the perceived burden of caregiving. Clarifying this distinction is important because it suggests that financial toxicity warrants attention as a separate outcome and intervention target, rather than being treated solely as a byproduct of general caregiver distress.^2,6,17,18,19

In our cohort, caregiver burden and lower household income showed the strongest and most consistent associations with financial toxicity, while older caregiver age appeared to be protective. This age pattern may reflect greater financial stability, Medicare eligibility, retirement-related insulation from wage loss, or fewer competing work and caregiving responsibilities among older caregivers. By contrast, younger caregivers may be particularly exposed to financial toxicity because they are more likely to be balancing active employment, household expenses, and caregiving simultaneously.^17,20,21 Although SDOH challenges were associated with financial toxicity in univariate analyses, adjustment for SDOH in sensitivity analyses did not materially alter the association between caregiver burden and financial toxicity. Some of the observed SDOH effects may overlap conceptually with household income and caregiver burden. Even so, the direction of association supports the broader conclusion that structural disadvantage contributes meaningfully to caregiver financial vulnerability.^5,9,17

These findings have important implications for policy and practice. If financial toxicity is shaped jointly by caregiver burden and socioeconomic disadvantage, then interventions focused only on coping skills or psychosocial support are unlikely to be sufficient. More effective approaches will likely require upstream strategies that reduce the economic consequences of caregiving itself, such as paid family leave, flexible workplace protections, caregiver tax credits or stipends, and stronger access to respite and community-based support services.^21–24 Such strategies may be especially important for lower-income and younger caregivers, who may have fewer financial reserves and less flexibility to absorb caregiving-related disruptions. Without more comprehensive structural support, financial toxicity is likely to remain concentrated among caregivers facing the greatest burden and the fewest resources.

Several limitations should be considered. The cross-sectional design precludes causal inference, and the sample size may have limited power to detect more modest associations or fully disentangle overlapping socioeconomic indicators. In addition, patient-level clinical characteristics were not collected, which may be relevant to understanding how disease severity and caregiving intensity influence financial hardship. These factors may influence caregiving intensity and caregiver time demands, thereby shaping financial strain. The absence of these data limits our ability to fully characterize the relationships between caregiving burden and financial toxicity. Longitudinal studies that integrate caregiver-reported outcomes with patient-level data will be important for clarifying how financial toxicity develops over time and for identifying the most actionable points for intervention across the caregiving trajectory.

Future studies that integrate caregiver-reported outcomes with patient-level clinical and functional data will be essential to disentangle how disease severity and care needs translate into economic burden for caregivers over time.

Availability of data and materials:

The data supporting the findings of this study are not publicly available because they contain sensitive participant-level information; however, they are available from the corresponding author upon reasonable request and with approval from the University of Kansas Medical Center Institutional Review Board.