Abstract
Introduction:
Caregivers play an important role in Parkinson’s disease (PD) treatment, especially as the disease progresses. As the symptom profile and needs of people with PD (PwP) differ across ethnoracial groups, whether caregiving needs also differ for different ethnoracial groups should be investigated.
Methods:
Data were obtained from the Parkinson’s Foundation funded Parkinson’s Outcomes Project for PwP identifying as Hispanic (n = 495), non-Hispanic Asian (n = 170), non-Hispanic Black (n = 162), or non-Hispanic White (n = 7687). Cross-sectional and longitudinal total Multidimensional Caregiver Strain Index (MCSI) and domain-specific scores for caregiving burden were compared across the ethnoracial groups. Effect of demographics and clinical variables, interaction of these variables with ethnoracial groups for caregiver burden was assessed.
Results:
Care partners of PwP identifying as non-Hispanic Asian experienced the most burden. PwP identifying as non-Hispanic White were oldest, yet their care partners experienced the least burden. Care partners of PwP identifying as non-Hispanic Asian experienced more burden in physical and social domains, care partners of PwP identifying as Hispanic experienced more burden in financial and elder demanding/manipulative domains. Over time, burden increased similarly across the ethnoracial groups. Effect of frequency of falls, hospital admission, neuropsychiatric disorder and social support on burden over time differed across the groups.
Conclusion:
PwP from different ethnoracial groups can experience different levels of caregiving burden. Predictors for caregiving burden, such as social support and falls can have different impacts based on ethnicity and race. Caregiver needs should also be assessed and culturally competent support should be provided to benefit all affected by PD.
Keywords: Ethnicity, Race, Caregiving, Parkinson, Disparities
1. Introduction
Parkinson’s disease (PD) is the fastest growing neurological disorder, with a broad spectrum of motor and non-motor symptoms significantly impacting daily life [1]. PD not only places a burden on the affected individuals (people with PD, PwP), but also on their care partners. Ideal PD management requires a team approach with care partners playing a substantial role [1]. Providing care in the form of physical, social, and emotional tasks may come at a significant cost to the care partners of PwP [2–7]. In other neurodegenerative and chronic diseases, the burden associated with caregiving was shown to differ based on ethnicity and race [8,9]. Despite spending more time caregiving and experiencing more financial strain, people identifying as Hispanic or Black may not experience overall higher levels of caregiver burden compared to people identifying as non-Hispanic White. With budding reports examining the association between ethnicity, race and PD, the care partner aspect also needs to be considered to better understand and address ethnoracial differences.
Needs of PwP and strain on care partners can differ across ethnoracial groups. Sociocultural differences between ethnic and racial groups can lead to differences in PD symptom expression [10]. Compared to people identifying as White, people identifying as Asian, Black or Hispanic may be less likely to have spouses as their care partner and instead rely on their adult children for caregiving [11,12]. In dementia and Alzheimer’s disease, adult children have unique forms of stress and burden when caring for their older parent with dementia; adult children identifying as Black, who care for their parent, report higher physical strain compared to other caregiver relationship types [9]. Due to this higher physical strain and lower self-reported levels of health, care partners identifying as one of the minoritized ethnoracial groups may be at higher risk for increased health problems.
Previous reports on ethnoracial differences for health outcomes in PD implicates different caregiving needs of PwP [13]. As PD progresses, the role of and burden on care partners can expand. Addressing the needs of not only the PwP but also the care partners can have important implications for PD treatment and quality of life. However, ethnoracial differences for care partner strain in PD have yet to be determined in large samples. To address this gap and to initiate discussions about personalized approaches to support care partners of PwP, we investigated differences for longitudinal care partner strain across different ethnoracial groups in a large population-based cohort of PwP.
2. Methods
2.1. Participants
Data were obtained from the Parkinson’s Foundation funded Parkinson’s Outcomes Project (also known as the Parkinson’s Foundation Parkinson’s Quality Improvement Initiative [PF-QII]), an international multicenter prospective clinical study of care and outcomes at expert centers [14]. Briefly, PF-QII recruits all PwP without exclusion through a quasi-sequential process by appointment slots and without prior evaluation to ensure a representative sample of PwP seen at the clinic; information on demographics, clinical status, delivered care, health-related quality of life and caregiver strain are collected. All study visits are held approximately 10–18 months apart and PwP not returning for a scheduled subsequent visit in 24 months are automatically withdrawn. Inclusion is based on the clinical confirmation of Parkinson’s disease diagnosis by the treating movement disorders expert. Local Institutional Review Board approvals were obtained at each participating clinical site and all participants were enrolled after signing informed consent.
Data used in our analysis includes information collected from June 2009 until September 2020 restricted to sites in the United States for consistency in ethnic and racial designations. These sites include Baylor College of Medicine, Beth Israel Deaconess Medical Center, Georgetown University, Johns Hopkins University, Medical College of Georgia, Mount Sinai, Muhammad Ali Parkinson Center, New York University Langone Medical Center, Northwestern University, Oregon Health & Science University, Struthers Parkinson’s Center, The Parkinson’s Institute, University of California San Diego, University of Florida, University of Kansas, University of Miami, University of Pennsylvania, University of South Florida, and Vanderbilt Medical Center.
Race and ethnicity were self-reported by the participants. Classification of race and ethnicity were based on the categories suggested by the National Institutes of Health Office of Equity, Diversity and Inclusion. Ethnoracial groups in our analysis included people identifying as Hispanic (n = 495), non-Hispanic Asian (n = 170), non-Hispanic Black (n = 162), or non-Hispanic White (n = 7687). Based on prior reports (10), people identifying as Hispanic, non-Hispanic Asian, non-Hispanic Black were considered as underrepresented minorities for clinical research.
2.2. Study measures
Detailed methods of Parkinson Foundation-QII have been previously described elsewhere [15,16]. A specific range of information was retrieved from the study for our analyses. For care partner burden, Multidimensional Caregiver Strain Index (MCSI), a brief self-report tool including 18 items, was used [17]. The scale includes at least two items for six domains including Physical Strain, Social Constraints, Financial Strain, Time Constraints, Interpersonal Strain and Elder Demanding/Manipulative. Each item is scored on a five-point Likert scale (0 = Never, 1 = A Little, 2 = Moderate, 3 = A lot, 4 = A great deal) based on the applicability of each statement to the care partner. Higher scores indicate more burden. Baseline and longitudinal MCSI scores were included as outcome measures in our analyses.
Further variables on care partner and PwP were included to assess whether these predictor variables are associated with care partner burden differently across the ethnoracial groups. Care partner related data included (1) care partner relationship with PwP, whether they were regular care partners or not, and (2) their employment status. Demographics for PwP in our analysis included data on (1) age, (2) sex and (3) education. PD characteristics included data on (1) disease duration, (2) Hoehn & Yahr stage, (3) frequency of falls, (4) hospital admissions, (5) cognition (z-score based on combination of immediate and delayed 5-word recall, animal fluency; higher scores indicating better cognition), (6) presence of neuropsychiatric disorder (at least one of the following; report of depression or psychosis, use of antidepressant, antipsychotic and benzodiazepine prescribed for anxiety), and (7) engaging in outside activities unaccompanied. PD treatment characteristics included data regarding (1) presence of levodopa-related issues (at least one of the following; dyskinesias and wearing-off that limit patient activities), (2) deep brain stimulation, (3) social support (social worker/counseling support). Presence of at least one type of chronic comorbidity implicated report of at least one of the following; high blood pressure, heart problem, anemia or other blood disease, lung disease, cancer, diabetes, ulcer/stomach disease, liver disease, kidney disease, osteoarthritis, degenerative arthritis and rheumatoid arthritis.
2.3. Statistical analysis
We compared the baseline data of PwP demographics, care partner related variables, PD-related disease and treatment characteristics, presence of comorbidity, domain-specific and total MCSI across the four ethnoracial groups using analysis of variance (ANOVA), Chi-square or Fisher’s exact tests as appropriate. p-values were adjusted for multiple comparisons using the Bonferroni method. p < 0.0021 was considered significant.
We evaluated the interaction effects of ethnicity or race and PwP related and care partner related variables at baseline on care partner burden by using generalized linear models with adjustment for age, sex, cognition, disease duration, and Hoehn and Yahr Stage. Baseline total and domain-specific MCSI scores were used as outcomes to represent care partner burden. For significant interactions, subgroup analyses were performed. Subgroup analyses were also adjusted for age, sex, cognition score, disease duration, and Hoehn and Yahr Stage.
To assess change over time for care partner burden across ethnoracial groups, generalized linear mixed models with random intercept and random slope were applied with ethnoracial group as a predictor, the changes of total and domain-specific MCSI scores at follow-up as outcomes. Models included baseline MCSI scores, demographics (age, sex, education), disease duration, length of follow-up, longitudinal clinical variables (cognition, Hoehn and Yahr stage, hospital admission) as covariates. The interaction of ethnoracial group and predictor variables were also included. As the observation numbers of non-Hispanic Asian, Black, and Hispanic groups were small, mixed models for subgroup analyses included only the random intercept with age, sex, cognition, disease duration, Hoehn and Yahr stage and the predictor. R (version 4.2.1) was used for statistical analysis.
3. Results
3.1. Baseline comparisons for care partner and PwP related information
There were several group differences for demographics, PD disease and treatment characteristics at baseline (Table 1). PwP identifying as non-Hispanic White were oldest; PwP identifying as non-Hispanic Asian had the highest percentage of bachelor’s degree or higher; PwP identifying as Hispanic were youngest and had the lowest percentage of bachelor’s degree or higher. PwP identifying as non-Hispanic White had the highest mean z-score for cognition and the highest likelihood of engaging in outside activities unaccompanied. PwP identifying as non-Hispanic Black were the least likely to have a regular care partner, least likely to have their spouse as their care partner, and more likely to have another relative as their regular care partner.
Table 1.
Baseline demographics and clinical characteristics of people with Parkinson’s disease based on ethnoracial groups.
| Variable | Hispanic (N = 495) | Non-Hispanic Asian (N = 170) | Non-Hispanic Black (N = 162) | Non-Hispanic White (N = 7687) | p value |
|---|---|---|---|---|---|
| Age | 65.08 ± 10.69 | 65.22 ± 9.94 | 66.07 ± 10.32 | 67.24 ± 9.45 | <0.001 # |
| Sex, female, N (%) | 183 (37) | 59 (34.7) | 79 (48.8) | 2820 (36.7) | 0.016 |
| Level of education, N (%) | |||||
| Less than high school | 23 (7.9) | 3 (3.7) | 1 (1.6) | 47 (1.3) | <0.001 # |
| Highschool | 57 (19.7) | 7 (8.5) | 16 (25.8) | 475 (13.5) | |
| Some post-high school | 76 (26.2) | 16 (19.5) | 16 (25.8) | 762 (21.6) | |
| Bachelor’s degree | 79 (27.2) | 18 (22) | 18 (29) | 1072 (30.5) | |
| Graduate degree | 55 (19) | 38 (46.3) | 11 (17.7) | 1164 (33.1) | |
| Living situation, N (%) | |||||
| At home | 147 (96.7) | 43 (100) | 34 (100) | 1611 (97.2) | 0.245 |
| Skilled care | 1 (0.7) | 0 (0) | 0 (0) | 35 (2.1) | |
| Other | 4 (2.6) | 0 (0) | 0 (0) | 11 (0.7) | |
| Regular care partner, N (%) | |||||
| No | 82 (16.6) | 25 (14.7) | 42 (25.9) | 1158 (15.1) | <0.001 # |
| Spouse/partner | 373 (75.5) | 138 (81.2) | 99 (61.1) | 6050 (78.8) | |
| Other relative | 32 (6.5) | 5 (2.9) | 17 (10.5) | 301 (3.9) | |
| Paid caregiver | 4 (0.8) | 0 (0) | 3 (1.9) | 127 (1.7) | |
| Other | 3 (0.6) | 2 (1.2) | 1 (0.6) | 44 (0.6) | |
| Care partner employed outside of home, N (%) | 67 (55.8) | 10 (32.3) | 7 (36.8) | 471 (38.9) | 0.325 |
| Disease duration, years | 6.65 ± 5.65 | 7.06 ± 5.99 | 6.73 ± 5.37 | 6.70 ± 5.73 | 0.867 |
| Hoehn and Yahr stage, N (%) | |||||
| 1 | 55 (11.5) | 12 (7.9) | 14 (9.1) | 872 (12.1) | 0.018 |
| 2 | 297 (62) | 87 (57.6) | 80 (51.9) | 4141 (57.5) | |
| 3 | 103 (21.5) | 41 (27.2) | 40 (26) | 1755 (24.4) | |
| 4 | 20 (4.2) | 10 (6.6) | 20 (13) | 368 (5.1) | |
| 5 | 4 (0.8) | 1 (0.7) | 0 (0) | 65 (0.9) | |
| Frequency of falls, N (%) | |||||
| No | 322 (73.2) | 100 (70.4) | 92 (63) | 4430 (69.5) | 0.190 |
| Rare | 92 (20.9) | 36 (25.4) | 47 (32.2) | 1501 (23.6) | |
| Monthly | 14 (3.2) | 3 (2.1) | 5 (3.4) | 288 (4.5) | |
| Weekly | 12 (2.7) | 3 (2.1) | 2 (1.4) | 153 (2.4) | |
| Hospital admission, N (%) | 89 (24.9) | 28 (19.6) | 26 (18.1) | 1416 (22) | 0.329 |
| Cognition (Z-score) | −0.07 ± 0.73 | −0.11 ± 0.77 | −0.2 ± 0.74 | 0.08 ± 0.74 | 0.001 # |
| Neuropsychiatric disorder, N (%) | 173 (34.9) | 44 (25.9) | 46 (28.4) | 2780 (36.2) | 0.008 |
| Engaging in outside activities unaccompanied, N (%) | 129 (84.9) | 27 (62.8) | 25 (73.5) | 1488 (89.9) | <0.001 # |
| At least one type of chronic comorbidity, N (%) | 226 (45.7) | 84 (49.4) | 84 (51.9) | 3762 (48.9) | 0.449 |
| Levodopa-related issue | 92 (60.9) | 28 (66.7) | 16 (50) | 821 (50.2) | 0.015 |
| Deep brain stimulation | 4 (10.8) | 7 (35) | 0 (0) | 45 (9.2) | 0.008 |
| Social support, N (%) | 21 (4.3) | 4 (2.4) | 12 (7.4) | 565 (7.4) | 0.025 |
| Total MCSI | 14.97 ± 12.07 | 15.26 ± 10.64 | 13.93 ± 11.68 | 12.39 ± 10.96 | 0.0003 # |
| MCSI-Physical strain | 2.48 ± 2.95 | 2.62 ± 2.68 | 2.43 ± 2.78 | 2 ± 2.55 | 0.002 # |
| MCSI-Social constraints | 4.52 ± 3.71 | 4.78 ± 4.01 | 4.51 ± 3.93 | 3.84 ± 3.63 | 0.002 # |
| MCSI-Financial strain | 1.19 ± 1.8 | 1.02 ± 1.41 | 1.08 ± 1.57 | 0.68 ± 1.23 | <0.001 # |
| MCSI-Time constraints | 2.84 ± 2.38 | 2.76 ± 2.35 | 2.9 ± 2.12 | 2.46 ± 2.28 | 0.015 |
| MCSI-Interpersonal strain | 2.71 ± 3.25 | 3.11 ± 3.16 | 2.29 ± 3.05 | 2.53 ± 2.97 | 0.204 |
| MCSI-Elder demanding/manipulative | 1.27 ± 1.75 | 1.06 ± 1.59 | 1.11 ± 1.76 | 0.89 ± 1.49 | 0.001 # |
Variables are reported as mean ± SD or N (%) as appropriate. MCSI: Multidimensional Caregiver Strain Index.
Statistically significant group difference is bolded and marked with # for Bonferroni-corrected p < 0.0021 (0.05/24 = 0.0021).
Care partners for PwP identifying as non-Hispanic White had the lowest; care partners for PwP identifying as non-Hispanic Asian had the highest total MSCI score (Fig. 1). Care partners of PwP identifying as non-Hispanic Asian had the highest scores on Physical Strain and Social Constraints; care partners of PwP identifying as Hispanic had the highest scores on Financial Strain and Elder Demanting/Manipulative domains.
Fig. 1.
Violin plots of total and domain-specific MCSI scores across the ethnoracial groups.
3.2. Cross-sectional associations of care partner burden across ethnoracial groups
Effects of demographic, disease and treatment characteristics on the total MCSI scores were similar across the groups (Table 2). Effects of levodopa-related issue on Physical Strain; age on Financial Strain; chronic comorbidity on Time Constraints differed across the groups. Levodopa-related issue was associated with more Physical Strain in the Hispanic group (β±standard error (SE) = 2.955 ± 0.737, p = 0.0002). Younger age was associated with more Financial Strain in the non-Hispanic White (β±SE = 0.02 ± 0.002, p < 0.0001), Black (β±SE = 0.064 ± 0.02, p = 0.002), and Hispanic groups (β±SE = 0.039 ± 0.011, p = 0.0003). Chronic comorbidity was associated with less Time Constraints in the non-Hispanic Asian group (β±SE = 1.77 ± 0.486, p = 0.0005).
Table 2.
p values for the interaction between ethnoracial group and predictor variables for baseline MCSI scores.
| Variable | Total MCSI | Physical strain | Social constraints | Financial strain | Time constraints | Interpersonal strain | Elder demanding/manipulative |
|---|---|---|---|---|---|---|---|
| Age | 0.047 | 0.040 | 0.047 | 0.003 # | 0.340 | 0.206 | 0.140 |
| Sex | 0.009 | 0.057 | 0.006 | 0.246 | 0.164 | 0.215 | 0.029 |
| Level of education | 0.812 | 0.222 | 0.764 | 0.092 | 0.469 | 0.546 | 0.656 |
| Regular care partner | 0.690 | 0.675 | 0.497 | 0.561 | 0.520 | 0.836 | 0.484 |
| Care partner employed outside of home | 0.629 | 0.080 | 0.381 | 0.354 | 0.700 | 0.826 | 0.470 |
| Disease duration, years | 0.998 | 0.830 | 0.873 | 0.945 | 0.779 | 0.988 | 0.035 |
| Hoehn and Yahr stage | 0.076 | 0.007 | 0.295 | 0.024 | 0.304 | 0.157 | 0.012 |
| Frequency of falls | 0.350 | 0.220 | 0.117 | 0.520 | 0.051 | 0.457 | 0.901 |
| Hospital admission | 0.514 | 0.176 | 0.981 | 0.979 | 0.410 | 0.642 | 0.242 |
| Cognition (Z-score) | 0.278 | 0.308 | 0.039 | 0.541 | 0.329 | 0.719 | 0.524 |
| Neuropsychiatric disorder | 0.601 | 0.494 | 0.378 | 0.490 | 0.767 | 0.646 | 0.136 |
| Engaging in outside activities unaccompanied | 0.592 | 0.335 | 0.620 | 0.499 | 0.794 | 0.750 | 0.799 |
| At least one type of chronic comorbidity | 0.416 | 0.697 | 0.727 | 0.984 | 0.002 # | 0.455 | 0.949 |
| Levodopa-related issue | 0.236 | <0.0001 # | 0.281 | 0.431 | 0.056 | 0.339 | 0.055 |
| Deep brain stimulation | 0.121 | 0.426 | 0.010 | 0.271 | 0.332 | 0.268 | 0.165 |
| Social support | 0.250 | 0.258 | 0.132 | 0.009 | 0.867 | 0.763 | 0.053 |
Statistically significant interaction is bolded and marked with # for Bonferroni-corrected p < 0.0031 (0.05/16 = 0.0031). General linear model for interaction between ethnoracial group and predictor variables are adjusted for main effect of age, sex, cognition, disease duration, Hoehn and Yahr stage at baseline, ethnoracial group and corresponding predictor.
3.3. Change over time from baseline care partner burden across ethnoracial groups
Of the 8514 PwP, 2689 PwP had follow-up data available after excluding PwP with missing data on sex, age, cognition, disease duration, Hoehn and Yahr stage and MCSI score (Supplementary Tables 1, 2, 3). Total and domain-specific MCSI scores changed similarly over time across the groups (Fig. 2). Male sex of PwP, worsening in cognition and Hoehn and Yahr stage, longer follow-up time, social support were significantly associated with increase in burden over time from baseline (Table 3).
Fig. 2.
The change in total MCSI scores over the follow-up period across the ethnoracial groups.
Table 3.
Longitudinal associations of the follow-up MCSI score with baseline demographic and time-dependent clinical predictors.
| Variable | β | Standard error | t-value | p value |
|---|---|---|---|---|
| Intercepta | 2.154 | 1.231 | 1.750 | 0.080 |
| Ethnoracial group | 0.449 | |||
| Hispanic vs non-Hispanic White | −0.286 | 0.706 | −0.404 | 0.686 |
| Non-Hispanic Asian vs non-Hispanic White | 1.493 | 1.175 | 1.270 | 0.204 |
| Non-Hispanic Black vs non-Hispanic White | −1.198 | 1.315 | −0.911 | 0.362 |
| Age at baseline | 0.007 | 0.016 | 0.411 | 0.681 |
| Sex (Male vs female) | 0.921 | 0.301 | 3.062 | <0.001 # |
| Disease duration at baseline | −0.014 | 0.025 | −0.547 | 0.584 |
| MCSI at baseline | −0.292 | 0.014 | −20.756 | <0.001 # |
| Length of follow-up | 0.770 | 0.060 | 12.747 | <0.001 # |
| Cognitiona | −1.353 | 0.149 | −9.072 | <0.001 # |
| Hoehn and Yahr stagea | <0.001 # | |||
| Stage 2 vs 1 | 0.748 | 0.471 | 1.589 | 0.112 |
| Stage 3 vs 1 | 2.729 | 0.502 | 5.415 | <0.001 # |
| Stage 4 vs 1 | 4.261 | 0.573 | 7.430 | <0.001 # |
| Stage 5 vs 1 | 4.917 | 0.801 | 6.142 | <0.001 # |
| Other predictors b | ||||
| Education | 0.287 | |||
| Post-high school or associate degree vs high school graduate | −0.431 | 0.606 | −0.710 | 0.478 |
| Bachelor degree vs high school graduate | 0.355 | 0.586 | 0.605 | 0.545 |
| Graduate degree vs high school graduate | −0.019 | 0.585 | −0.032 | 0.974 |
| Regular care partner (Yes vs. No)+ | 2.797 | 0.940 | 2.977 | 0.003 |
| Frequency of falls | 0.394 | 0.361 | 1.091 | 0.276 |
| Hospital admission+ | 0.072 | 0.384 | 0.188 | 0.851 |
| Neuropsychiatric disorder+ | 0.223 | 0.294 | 0.758 | 0.448 |
| At least one type of chronic comorbidity+ | 0.425 | 0.279 | 1.523 | 0.128 |
| Deep brain stimulation+ | 0.123 | 0.318 | 0.388 | 0.698 |
| Social support (Yes vs No)+ | 2.641 | 0.282 | 9.363 | <0.001 # |
Time-dependent predictors are marked with +. Statistically significant interaction is bolded and marked with #.
Linear mixed model with random intercept, random slope, age, sex, cognition, disease duration, Hoehn and Yahr stage, baseline MCSI score, length of follow-up and ethnoracial group.
Linear mixed model with random intercept, random slope, age, sex, cognition, disease duration, Hoehn and Yahr stage, ethnoracial group, and the corresponding predictor.
Subgroup analyses with only random intercept showed that neuropsychiatric disorder was associated with decrease; social support was associated with increase in burden for the non-Hispanic Asian group (Supplementary Table 4). Frequency of falls and social support were associated with increase in burden for the non-Hispanic White group.
For domain-specific score changes, subgroup analyses with only random intercept are summarized in Supplementary Table 5. For the Hispanic group, having a bachelor’s degree compared to a high school degree was associated with decrease in Financial Strain. For the non-Hispanic Asian group, social support was associated with increase in all domains except for Interpersonal Strain; neuropsychiatric disorder was associated with decrease in Interpersonal Strain and Elder Demanding/Manipulative. For the non-Hispanic Black group, having a bachelor’s degree compared to a high school degree was associated with increase in Physical Strain, hospital admission was associated with decrease in Social Constraints over time. For the non-Hispanic White group, social support was associated with increase in all domain scores; frequency of falls was associated with increase in all domains except for Financial Strain; hospital admission was associated with increase in Interpersonal Strain; having a graduate degree compared to a high school degree was associated with decrease in Elder Demanding/Manipulative scores.
4. Discussion
In this study, we investigated the ethnoracial differences for care partner burden in PD and specific factors contributing to care partner burden for ethnoracial groups using a large longitudinal data from the Parkinson’s Foundation QII. To our knowledge, this is one of the largest studies evaluating ethnoracial differences for caregiver burden in PD with a longitudinal design. Although PwP identifying as non-Hispanic White were older, care partner burden was significantly lower in this group compared to PwP identifying as Hispanic, non-Hispanic Asian and non-Hispanic Black. On the other hand, PwP identifying as non-Hispanic Asian had the highest level of education and care partner burden, with particularly more burden related to physical and social domains. PwP identifying as Hispanic were youngest with lower levels of education, accompanied by highest burden in Financial and Elder Demanding/Manipulative domains. Longitudinal analyses showed that burden increased similarly across the ethnoracial groups over time, suggesting overcoming disparities at any given time can benefit PwP across different ethnoracial groups.
Overall, most of the care partners in our cohort were spouses of PwP. In community-dwelling adults over the age of 45, spouses are often the caregivers for all ethnoracial groups, except for women identifying as Black who reported their children as caregivers [18]. This is reflected in our cohort of PwP; PwP identifying as Black were more likely than other groups to have another relative than their spouse/partner as their care partner. Although similar proportions of care partners were employed outside of home across the groups in our cohort, the type of employment and the level of income gained from the employment can differ and be affected from providing care. Income has been recognized as an important predictor of burden [19]. Care partners identifying as Black or Hispanic have been shown to spend more time caregiving, which may lead to early retirement or unemployment and consequently lower income [8,20]. Although disease progression and later onset PD is associated with more care partner burden [21], we found younger age to be associated with more burden at baseline for PwP identifying as non-Hispanic Black or White. Younger age can be particularly important for PwP identifying as Black, as they are more likely to have their children as their care partners compared to other ethnoracial groups [18]. They may be at a higher risk for social and financial burden due to longer hours spent caregiving. For people identifying as Hispanic, high levels of financial strain can be due to the relatives of PwP taking on caregiving roles and spending more time for caregiving duties, with the strain decreasing over time as the major shift in jobs occur in the beginning of the caregiving [8]. Highest level of education in PwP identifying as non-Hispanic Asian, accompanied by lowest financial strain can be due to type of jobs held by care partners. Spending more time caregiving can lead to more burden both socially and financially.
Cross-cultural differences can play an important role in the domain-specific scores across ethnoracial groups. We found higher levels of physical strain and social constraints experienced by the care partners of PwP identifying as non-Hispanic Asian. In a large study from Malaysia with a multi-ethnic cohort of PwP, the majority of the care partners indicated they would not consider their loved one residing in a nursing home, in line with the traditional Asian value of caring for the older family members at home [22]. Stigma associated with PD symptoms was also prevalent with a particular lack of awareness on non-motor symptoms, which are more burdensome and bothersome for PwP and caregivers [22]. Having a chronic comorbidity along with PD could indicate a previously initiated caregiving role. Having PD without chronic comorbidities may have led to more recent adjustments at the household with the care partner increasing the time spent on caregiving tasks and contributing to the higher levels of physical and social burden on the care partner. The sense of obligation to family is also common in the Hispanic culture. Care partners of PwP identifying as Hispanic experienced more strain in the Elder Demanding/Manipulative domain than other ethnoracial groups in our cohort. People identifying as Hispanic are less likely to seek formal support services and more likely to provide caregiving themselves in their homes [8]. Compared to PwP identifying as White seeking care at tertiary centers, PwP identifying as Hispanic and their care partners may be less aware about PD symptoms and care [23]. As levodopa-related issues pose a higher physical strain on care partners of PwP identifying as Hispanic, providing education about what PD entails, what type of resources and support are available might be a potential solution to overcome these disparities.
Female sex was associated with more burden only for PwP identifying as non-Hispanic White at baseline. Many studies in PD do not show an association between care partner gender and level of burden [19,24]. However, sex and gender of PwP have been associated with different range and severity of symptoms, rates of decline, treatment preferences and effects in PD [1]. This implicates different caregiving needs based on sex and gender of PwP, which is supported by significant sex association with burden in our analysis. A prior study in the same cohort showed that females with PD were less likely to receive informal caregiving at baseline [25]. In community-dwelling older adults, female sex and identifying as White have been previously associated with lower caregiving availability [18]. Older adults identifying as non-Hispanic White received less informal care than older adults identifying as Hispanic [26]. This lower availability to care for female PwP can increase the burden particularly for people identifying as non-Hispanic White, as shown in our analysis. During follow-up, females with PD received formal caregiving at a higher rate than males with PD in this cohort [25]. Thus, the increased burden likely shifts to informal care partners of males with PD as suggested by our finding of the significant association between male sex of PwP and higher levels of care partner burden over time.
Male sex association with more burden as the disease is progressing can also be in part related to cognitive changes. Male sex is a risk factor for cognitive decline in PD [27]. Cognitive decline for PwP increases the time spent to provide care by care partners and burden on the care partners [28,29]. Ethnoracial differences for quality of life of PwP is partially mediated by cognitive performance [13]. Accordingly, we found cognitive decline was associated with increasing caregiving burden over time similarly across the ethnoracial groups. At baseline, PwP identifying as non-Hispanic White had the best cognitive performance and the lowest level of care partner burden. This highlights the importance of supporting cognitive function of PwP for all ethnoracial groups. Although there are currently no disease-modifying treatments for cognitive decline in PD, personalized treatment approaches need to consider social determinants associated with ethnoracial differences in PD. For instance, living alone, lower level of education, and cardiovascular comorbidities can be more prevalent for people identifying as Hispanic or Black, compared to people identifying as White with Lewy body disease related cognitive decline [30]. In our cohort, PwP identifying as non-Hispanic Black were the least likely to have a regular care partner. Addressing these social factors associated with cognitive decline can help prevent or delay cognitive decline in PD. Such efforts can be particularly beneficial for minoritized groups with a higher risk for these life and health conditions.
Social support was associated with more care partner burden in the physical, financial, social and time domains, especially for the non-Hispanic Asian and White groups. This finding may be due to the definition of social support in our analysis, as informal support from family, friends and the community are not included. Social support in our study included social worker/counseling. However, efforts can expand beyond these. It is likely that the PwP with more caregiving burden are more likely to be referred to a social worker or for counseling, or more likely to accept referral to such services. We were unable to test the association between social support and caregiving burden further to better understand this finding. Social support may be less accessible for minoritized groups, despite its importance for caregiver wellbeing, burden and even PD symptoms [31]. Subramanian and colleagues noted social support is lacking for many PwP [32]. They outlined approaches, such as education, to increase awareness about both motor and non-motor symptoms in PD, wellness strategies with counseling and education on modifiable lifestyle choices, realigned healthcare teams in community with peer-to-peer mentoring, health coaches for wellness and exercise, community health workers, psychotherapists, proactive outreach in a global scale to engage, train and provide access to resources, social prescribing, supporting self-agency can be impactful for PwP [32]. Developing and utilizing strategies for social support should be prioritized, as such efforts can benefit PwP and their care partners across all ethnoracial groups and help overcome disparities. Nonetheless, in our analyses, having social support was associated with more caregiving burden during the disease course. In a prior study, caregivers of PwP have reported that limited availability and insurance coverage for services are important barriers for social support [33]. In addition, people identifying as non-Hispanic Asian tend to underuse resources for mental health unless the resources are culturally and linguistically compatible [34]. Therefore, it is likely that social support itself is not a burden on the caregiver, but the type of social support and barriers to access supportive resources can pose a significant burden and should be addressed. The need for social support can also implicate more severe non-motor symptom burden for PwP, and their caregivers can consequently experience higher levels of burden.
There are several limitations in our study. Our findings cannot be generalized to all PwP given the higher proportion of PwP identifying as White included in our cohort and restriction of the analyses to PwP from only sites in the United States. The small number of participants in Hispanic, non-Hispanic Asian and non-Hispanic Black groups limited the analysis and findings should be interpreted cautiously. Improving the engagement of minoritized groups in research and in healthcare needs to be prioritized to better understand and address the needs of PwP from different backgrounds. Data included in our analyses stems from people treated at Parkinson’s Foundation Centers of Excellence, which likely represents PwP with more awareness and access to care and support. Disparities are likely even more striking in communities lacking access to neurologists and movement disorders specialists. There is only limited data available for care partner characteristics, socioeconomic and sociocultural factors. For instance, immigrant generation status can play a role in the cultural impact on the outcomes. Cross-cultural validity of the MCSI has not been previously determined. Cultural adaptation of research measures is crucial in the reliable interpretation of findings, as the applicability and meaning of the outcomes from the scales can differ across the cultures. A more detailed and focused data collection, including qualitative data and culturally appropriate forms, can provide a better insight into disparities across ethnoracial groups.
In conclusion, we have demonstrated that minoritized ethnoracial groups with PD can be at a higher risk for more caregiver strain. Ethnicity and race are not only important social constructs in PD diagnosis and treatment, but also in the caregiving domain in PD. The impact of demographics, PD features, PD treatment and comorbidities on caregiving burden can differ across ethnoracial groups. To support care partners of PwP, their individual needs should be assessed, and resources should be provided as possible.
Supplementary Material
Acknowledgements
Data collection for the Parkinson’s Outcomes Project was funded by the Parkinson’s Foundation. The study acknowledges the contributions of the Parkinson’s Outcomes Project participants, site principal investigators and the steering committee (www.parkinson.org/advancing-research/our-research/parkinsons-outcomes-project/steering-committee). Dr. Bayram receives research support from the National Institute on Aging (K99AG073453). Dr. Litvan’s research is supported by the National Institutes of Health grants: 2R01AG038791-06A, U01NS100610, U01NS80818, R25NS098999; U19 AG063911-1 and 1R21NS114764-01A1; the Michael J Fox Foundation, Parkinson Foundation, Lewy Body Association, CurePSP, Roche, AbbVie, Biogen, Centogene. EIP-Pharma, Biohaven Pharmaceuticals, Novartis, and United Biopharma SRL, UCB. She is a member of the Scientific Advisory Board for Amydis, but does not receive funds and from the Rossy PSP Program at the University of Toronto. She receives her salary from the University of California San Diego and as Chief Editor of Frontiers in Neurology.
Role of the funding source
Data collection for the Parkinson’s Outcomes Project was funded by the Parkinson’s Foundation. EB receives research support from the National Institute on Aging (K99AG073453). Funding sources did not play a role in the design, analysis, interpretation of data, writing or submission of the article.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
All authors report data were provided by National Parkinson Foundation. Ece Bayram reports financial support was provided by National Institute on Aging. IL’s research is supported by the National Institutes of Health grants: 2R01AG038791-06A, U01NS100610, U01NS80818, R25NS098999; U19 AG063911-1 and 1R21NS114764-01A1; the Michael J Fox Foundation, Parkinson Foundation, Lewy Body Dementia Association, CurePSP, Roche, AbbVie, Biogen, Centogene. EIP-Pharma, Biohaven Pharmaceuticals, Novartis, and United Biopharma SRL, UCB. She is a member of the Scientific Advisory Board for Amydis, but does not receive funds and from the Rossy PSP Program at the University of Toronto. She receives her salary from the University of California San Diego and as Chief Editor of Frontiers in Neurology.
Footnotes
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.parkreldis.2023.105927.
References
- [1].Bloem BR, Okun MS, Klein C, Parkinson’s disease, Lancet 397 (2021) 2284–2303, 10.1016/S0140-6736(21)00218-X. [DOI] [PubMed] [Google Scholar]
- [2].Macchi ZA, Koljack CE, Miyasaki JM, Katz M, Galifianakis N, Prizer LP, Sillau SH, Kluger BM, Patient and caregiver characteristics associated with caregiver burden in Parkinson’s disease: a palliative care approach, Ann. Palliat. Med 9 (2020) S24–S33, 10.21037/apm.2019.10.01. [DOI] [PubMed] [Google Scholar]
- [3].Kalampokini S, Hommel ALAJ, Lorenzl S, Ferreira JJ, Meissner WG, Odin P, Bloem BR, Dodel R, Schrag A-E, Caregiver burden in late-stage Parkinsonism and its associations, J. Geriatr. Psychiatr. Neurol 35 (2022) 110–120, 10.1177/0891988720968263. [DOI] [PubMed] [Google Scholar]
- [4].Heine J, von Eichel H, Staege S, Höglinger GU, Wegner F, Klietz M, Relationship satisfaction in people with Parkinson’s disease and their caregivers: a cross-sectional observational study, Brain Sci 11 (2021) 822, 10.3390/brainsci11060822. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Jensen I, Lescher E, Stiel S, Wegner F, Höglinger G, Klietz M, Analysis of transition of patients with Parkinson’s disease into institutional care: a retrospective pilot study, Brain Sci 11 (2021) 1470, 10.3390/brainsci11111470. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].von Eichel H, Heine J, Wegner F, Rogozinski S, Stiel S, Groh A, Krey L, Höglinger GU, Klietz M, Neuropsychiatric symptoms in Parkinson’s disease patients are associated with reduced health-related quality of life and increased caregiver burden, Brain Sci 12 (2022) 89, 10.3390/brainsci12010089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Schmotz C, Richinger C, Lorenzl S,High burden and depression among late-stage idiopathic Parkinson disease and progressive supranuclear palsy caregivers, J. Geriatr. Psychiatr. Neurol 30 (2017) 267–272, 10.1177/0891988717720300. [DOI] [PubMed] [Google Scholar]
- [8].Smith ER, Perrin PB, Tyler CM, Lageman SK, Villaseñor T, Cross-cultural differences in Parkinson’s disease caregiving and burden between the United States and Mexico, Brain Behav 10 (2020), 10.1002/brb3.1753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Badana ANS, Marino V, Haley WE, Racial differences in caregiving: variation by relationship type and dementia care status, J. Aging Health 31 (2019) 925–946, 10.1177/0898264317743611. [DOI] [PubMed] [Google Scholar]
- [10].Sauerbier A, Lenka A, Aris A, Pal PK, Nonmotor symptoms in Parkinson’s disease: gender and ethnic differences, Int. Rev. Neurobiol 133 (2017) 417–446, 10.1016/bs.irn.2017.05.032. [DOI] [PubMed] [Google Scholar]
- [11].Pinquart M, Sörensen S, Spouses, adult children, and children-in-law as caregivers of older adults: a meta-analytic comparison, Psychol. Aging 26 (2011) 1–14, 10.1037/a0021863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Pereira C, LaRoche A, Arredondo B, Pugh E, Disbrow E, Reekes TH, Brickell E, Boettcher A, Sawyer RJ, Evaluating racial disparities in healthcare system utilization and caregiver burden among older adults with dementia, Clin. Neuropsychol 36 (2022) 353–366, 10.1080/13854046.2021.1951844. [DOI] [PubMed] [Google Scholar]
- [13].Di Luca DG, Luo S, Liu H, Cohn M, Davis TL, Ramirez-Zamora A, Rafferty M, Dahodwala N, Naito A, Neault M, Beck J, Marras C, Racial and ethnic differences in health-related quality of life for individuals with Parkinson disease across centers of excellence, Neurology 100 (2023) e2170–e2181, 10.1212/WNL.0000000000207247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Okun MS, Siderowf A, Nutt JG, O’Conner GT, Bloem BR, Olmstead EM, Guttman M, Simuni T, Cheng E, Cohen EV, Parashos S, Marsh L, Malaty IA, Giladi N, Schmidt P, Oberdorf J, Piloting the NPF data-driven quality improvement initiative, Parkinsonism Relat. Disorders 16 (2010) 517–521, 10.1016/j.parkreldis.2010.06.005. [DOI] [PubMed] [Google Scholar]
- [15].Zeldenrust F, Lidstone S, Wu S, Okun MS, Cubillos F, Beck J, Davis T, Lyons K, Nelson E, Rafferty M, Schmidt P, Dai Y, Marras C, Variations in hospitalization rates across Parkinson’s Foundation centers of excellence, Parkinsonism Relat. Disorders 81 (2020) 123–128, 10.1016/j.parkreldis.2020.09.006. [DOI] [PubMed] [Google Scholar]
- [16].Rafferty MR, Schmidt PN, Luo ST, Li K, Marras C, Davis TL, Guttman M, Cubillos F, Simuni T, Regular exercise, quality of life, and mobility in Parkinson’s disease: a longitudinal analysis of National Parkinson Foundation quality improvement initiative data, J. Parkinsons Dis 7 (2017) 193–202, 10.3233/JPD-160912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Stull DE, The multidimensional caregiver strain Index (MCSI): its measurement and structure, J. Clin. Geropsychol 2 (1996) 175–196. [Google Scholar]
- [18].Roth DL, Haley WE, Wadley VG, Clay OJ, Howard G, Race and gender differences in perceived caregiver availability for community-dwelling middle-aged and older adults, Gerontol 47 (2007) 721–729, 10.1093/geront/47.6.721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Mosley PE, Moodie R, Dissanayaka N, Caregiver burden in Parkinson disease: a critical review of recent literature, J. Geriatr. Psychiatr. Neurol 30 (2017) 235–252, 10.1177/0891988717720302. [DOI] [PubMed] [Google Scholar]
- [20].Cohen SA, Sabik NJ, Cook SK, Azzoli AB, Mendez-Luck CA, Differences within differences: gender inequalities in caregiving intensity vary by race and ethnicity in informal caregivers, J. Cross Cult. Gerontol 34 (2019) 245–263, 10.1007/s10823-019-09381-9. [DOI] [PubMed] [Google Scholar]
- [21].Martinez-Martin P, Skorvanek M, Henriksen T, Lindvall S, Domingos J, Alobaidi A, Kandukuri PL, Chaudhari VS, Patel AB, Parra JC, Pike J, Antonini A, Impact of advanced Parkinson’s disease on caregivers: an international real-world study, J. Neurol 270 (2023) 2162–2173, 10.1007/s00415-022-11546-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Choo XY, Lim S-Y, Chinna K, Tan YJ, Yong VW, Lim JL, Lau KF, Chung JY, Em JM, Tan HT, Lim JH, Tan SB, Tan CT, Tan AH, Understanding patients’ and caregivers’ perspectives and educational needs in Parkinson’s disease: a multi-ethnic Asian study, Neurol. Sci 41 (2020) 2831–2842, 10.1007/s10072-020-04396-4. [DOI] [PubMed] [Google Scholar]
- [23].Damron L, Litvan I, Bayram E, Berk S, Siddiqi B, Shill H, Hispanic perspectives on Parkinson’s disease care and research participation, J. Alzheim. Dis 81 (2021) 809–819, 10.3233/JAD-210231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Klietz M, von Eichel H, Schnur T, Staege S, Höglinger GU, Wegner F, Stiel S, One year trajectory of caregiver burden in Parkinson’s disease and analysis of gender-specific aspects, Brain Sci 11 (2021) 295, 10.3390/brainsci11030295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Dahodwala N, Shah K, He Y, Wu SS, Schmidt P, Cubillos F, Willis AW, Sex disparities in access to caregiving in Parkinson disease, Neurology 90 (2018) e48–e54, 10.1212/WNL.0000000000004764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Willert B, Minnotte KL, Informal caregiving and strains: exploring the impacts of gender, race, and income, Appl. Res. Qual. Life 16 (2021) 943–964, 10.1007/s11482-019-09786-1. [DOI] [Google Scholar]
- [27].Gonzalez-Latapi P, Bayram E, Litvan I, Marras C, Cognitive impairment in Parkinson’s disease: epidemiology, clinical profile, protective and risk factors, Behav. Sci 11 (2021) 74, 10.3390/bs11050074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Jones AJ, Kuijer RG, Livingston L, Myall D, Horne K, MacAskill M, Pitcher T, Barrett PT, Anderson TJ, Dalrymple-Alford JC, Caregiver burden is increased in Parkinson’s disease with mild cognitive impairment (PD-MCI), Transl. Neurodegener 6 (2017) 17, 10.1186/s40035-017-0085-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Klietz M, Schnur T, Drexel S, Lange F, Tulke A, Rippena L, Paracka L, Dressler D, Höglinger GU, Wegner F, Association of motor and cognitive symptoms with health-related quality of life and caregiver burden in a German cohort of advanced Parkinson’s disease patients, Parkinsons Dis 2020 (2020) 1–8, 10.1155/2020/5184084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Kurasz AM, Smith GE, McFarland MG, Armstrong MJ, O’Bryant S, Ethnoracial differences in Lewy body diseases with cognitive impairment, J. Alzheim. Dis 77 (2020) 165–174, 10.3233/JAD-200395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Yang Z, Tian Y, Fan Y, Liu L, Luo Y, Zhou L, Yu H, The mediating roles of caregiver social support and self-efficacy on caregiver burden in Parkinson’s disease, J. Affect. Disord 256 (2019) 302–308, 10.1016/j.jad.2019.05.064. [DOI] [PubMed] [Google Scholar]
- [32].Subramanian I, Hinkle JT, Chaudhuri KR, Mari Z, Fernandez HH, Pontone GM, Mind the gap: inequalities in mental health care and lack of social support in Parkinson disease, Parkinsonism Relat. Disorders 93 (2021) 97–102, 10.1016/j.parkreldis.2021.11.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Lageman SK, Mickens MN, Cash TV, Caregiver-identified needs and barriers to care in Parkinson’s disease, Geriatr. Nurs 36 (2015) 197–201, 10.1016/j.gerinurse.2015.01.002. [DOI] [PubMed] [Google Scholar]
- [34].American Psychiatric Association, “Treating Asian Americans.” Stress & Trauma Toolkit for Treating Historically Marginalized Populations in a Changing Political and Social Environment, 2019. https://www.psychiatry.org/psychiatrists/diversity/education/stress-and-trauma/asian-americans.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.