Logo for the Journal of Rehab R and D

Volume 46 Number 2, 2009
   Pages 223 — 232

Racial/ethnic variation in recovery of motor function in stroke survivors: Role of informal caregivers

Melanie Sberna Hinojosa, PhD;1-2* Maude Rittman, RN, PhD;3 Ramon Hinojosa, PhD;1,4 William Rodriguez, MD5

1North Florida/South Georgia Veterans Health System, Rehabilitation Outcomes Research Center, Gainesville, FL; 2Medical College of Wisconsin, Milwaukee, WI; 3College of Nursing, University of Florida, Gainesville, FL; 4Department of Social and Cultural Sciences, Marquette University, Milwaukee, WI; 5Department of Veterans Affairs Caribbean Health System, San Juan, PR

Abstract — Research documents that African American and Latinos who have experienced an acute stroke recover more slowly than Caucasians in the United States. This descriptive study examines (1) the variation in Caucasian, Puerto Rican, and African American motor function after stroke; (2) the association between caregiver attributes and motor recovery after stroke; and (3) the degree to which caregiver attributes explain the variation in motor recovery between different racial/ethnic groups. One hundred and thirty-five veterans who had been hospitalized after an acute stroke, released home, and identified an informal caregiver were enrolled in the study. Veterans and caregivers were surveyed at five time points over the course of 24 months. Results indicate that Puerto Ricans show greater impairment and African Americans show less impairment at discharge from the hospital compared with Caucasians. Caregiver characteristics mediate the racial/ethnic differences in impairment at discharge and motor recovery across time.

Key words: African American, caregivers, Caucasian, longitudinal, Puerto Rican, race/ethnicity, recovery, rehabilitation, stroke, veterans.


Abbreviations: FIM = Functional Independence Measure, IRB = Institutional Review Board, mFIM = motor subscale of the FIM, SD = standard deviation, VA = Department of Veterans Affairs, VAMC = VA medical center.
*Address all correspondence to Melanie Sberna Hinojosa, PhD; Medical College of Wisconsin, 8701 Watertown Plank Rd, PO Box 26509, Milwaukee, WI 53226; 414-456-4095; fax: 414-456-6523. Email: mhinojosa@mcw.edu
DOI:10.1682/JRRD.2007.10.0175
INTRODUCTION

An important aspect of recovery from stroke after discharge from the hospital is the presence of a caregiver. A majority of stroke survivors return home for rehabilitation, usually to an informal caregiver who is a spouse, child, or friend [1]. Little is known about how the number of caregivers and the amount of care provided is related to the recovery of motor function in stroke survivors. African Americans and Latinos generally have poorer recovery outcomes poststroke, but evidence supports that, for these groups, the number of caregivers and the amount of care positively affects these lower outcomes. This study examines the association between the number of caregivers, the amount of care provided, and impairment at discharge from the hospital and recovery of motor function over the course of 24 months.

BACKGROUND
Racial/Ethnic Differences in Stroke

Stroke is the leading cause of serious long-term disability affecting more than 4 million people in the United States [2-4]. Research indicates that racial/ethnic variations occur in the incidence and mortality of stroke, with African Americans and Latinos having a higher incidence of stroke and greater stroke mortality than Caucasians [5]. Latino and African-American stroke patients are twice as likely as Caucasians to experience a recurrent stroke within 2 years of their first stroke [6]. In addition, African American stroke survivors have greater residual physical impairment after the stroke [7]. Moreover, significant differences are apparent among Latino subgroups, indicating higher levels of stroke mortality among Puerto Ricans than Cuban or Mexican Americans [8]. One study revealed lower Functional Independence Measure (FIM) scores among Mexican Americans than Caucasians or African Americans upon admission for stroke rehabilitation, but comparable FIM scores at discharge [9]. Yet another study of Mexican Americans indicates a lower stroke burden than among Caucasians [10]. A separate study found lower levels of rehospitalization among Mexican Americans and speculated the positive effect of family care [11].

Caregiver Context

Close to 80 percent of stroke survivors are discharged home with the burden of continuing care and recovery shifting to informal caregivers, often family members with little or no preparation or training. Informal caregivers face difficulties, such as uncertainty, anxiety, depression, eroded health, and poor quality of life, due to the sudden onset of stroke and the new caregiver role [12-14]. Notably, caregivers of all races and ethnicities are at risk for depression, reduced quality of life, and increased anxiety [15-16], but these conditions also vary by race/ethnicity. Caucasian caregivers report feeling more distress and show more depressive symptoms than African-American caregivers [17]. African-American caregivers are more likely to have health problems and have higher mortality rates than Caucasian caregivers [18]. African Americans conversely are more likely to have more favorable perceptions of caregiving than Caucasians in contexts other than stroke rehabilitation (Alzheimer's caregiving) [19] and are more likely to express stronger cultural reasons for providing care than Caucasian caregivers [20].

Some literature suggests that Latino caregivers are less likely to institutionalize those in their care, or delay institutionalization the longest compared with other racial/ethnic groups [21]. Latino caregivers are also more likely to be family members, and culturally, Latinos report a greater sense of duty toward the elderly in their care than do other racial/ethnic groups [22-23]. Latino caregivers also spend more hours on informal care than Caucasian and African-American caregivers [24].

The role of social support, culturally embedded family systems, and informal family care often has been used to explain unexpectedly better health outcomes among Latinos [25]. Postacute stroke disability and survival among Latinos are beginning to draw similar attention. For example, residence in a high-density Mexican- American neighborhood appears to positively affect survival after a stroke [26]. Further suggestive evidence is indicated in the observation that receipt of informal care by disabled older adults is highest among Latinos and lowest among Caucasians [24].

In this study, we identified a population of veterans hospitalized for acute stroke in several Department of Veterans Affairs (VA) hospitals who returned home for care by an informal caregiver. We add to the literature by constructing a descriptive picture of the relationship between caregiver attributes, the amount of care provided to these racially/ethnically diverse veterans, and their functional recovery outcomes over 24 months.

METHODOLOGY
Research Questions
1. How do Caucasian, Puerto Rican, and African-American stroke survivors differ in their motor function after stroke across 24 months following discharge home?
2. Are caregiver attributes and motor recovery associated after acute stroke?
3. Can caregiver attributes explain the variation in motor recovery between different racial/ethnic groups?
Subjects

The data for this analysis are drawn from a longitudinal study of culturally sensitive models of stroke recovery and caregiving among veterans [27]. Caregivers of stroke survivors were identified in several VA medical centers (VAMCs) in Florida and Puerto Rico from 2000 to 2001. The target sample was veterans who experienced acute stroke who identified a primary informal caregiver upon his or her hospital discharge directly to home. Survivor demographic and baseline health data and caregiver demographic information were collected at discharge from 125 survivor and caregiver pairs. The panel was stratified by race/ethnicity, yielding 49 Puerto Ricans, 33 African Americans, and 43 non-Latino Caucasians.

Data Collection

This study was approved by the University of Florida Health Science Center Institutional Review Board (IRB) and the individual VAMC IRBs in San Juan, Puerto Rico; Tampa, Florida; and Miami, Florida. Informed consent was obtained prior to enrollment. Subjects were given a copy of the consent form, and a copy was placed in their medical record. Demographic and baseline health data were collected at discharge from patients and measures of recovery were performed on the stroke survivor at 1, 6, 12, 18, and 24 months following the stroke. In addition, the caregivers were given a survey to complete during the home visits. One hundred thirty-five caregiver-stroke survivor dyads were originally enrolled and initial data were collected while the stroke survivor was still in the hospital. At the 1-month follow-up visit, 11 dyads withdrew, reducing our sample to 124 dyads (n = 248). Of the 124 caregivers, 45 were Caucasian non-Hispanic, 28 were African American, 2 were Asian American, and 49 were Puerto Rican (8 residing in the United States and 41 residing in Puerto Rico). Of the 124 stroke survivors, 45 were Caucasian non-Hispanic, 30 were African American, and 49 were Puerto Rican (8 residing in the United States and 41 residing in Puerto Rico). Of the stroke survivors, 124 were men and 2 were women, and of the caregivers, 16 were men, and 110 were women. The average age was 66.6 ± 10.61 standard deviation (SD) for care recipients and 59.0 ± 14.08 for caregivers. At the 6-month visit, 12 dyads withdrew; at the 12-month visit, 13 withdrew. Subsequently, 6 more dyads withdrew at the 18-month interview, and 4 more withdrew at 24 months. Of the 46 dyads that did not complete the study, 3 survivors went into a nursing home, 8 survivors died, 16 survivors withdrew consent, 9 survivors moved or were lost to follow-up, 4 survivors were too sick to continue, 1 caregiver died and was not replaced, 4 caregivers ceased to provide care, and 1 caregiver became ineligible. Analyses indicated that survivors and caregivers who withdrew from the study did not statistically differ from those who did not withdraw from the study across multiple domains, with the exception of African-American participants, who were more likely to refuse consent at the 18- and 24-month visits.

Measures
Motor Functioning

The FIM is a scale commonly used to measure physical and cognitive recovery after stroke (Appendix, available online only). This analysis uses only the motor subscale of the FIM (mFIM), which measures motor functioning in activities of daily living. The mFIM measures basic motor functioning and consists of 15 questions about toileting, bathing, dressing, eating, and mobility [28]. Trained interviewers administered the mFIM to the stroke survivors at discharge from the hospital and in their homes at 1, 6, 12, and 24 months poststroke. The mFIM scale ranges from 13 to 91, with 13 representing total assistance with motor activities and 91 representing complete independence in motor functioning.

Time

Time is included in the model to represent each interview time point. Time is coded as a continuous variable that ranges from 0 to 5. On this scale, 0 represents the baseline interview taken at discharge from the hospital, 1 represents the 1-month interviews, 2 represents the 6-month interviews and so on.

Stroke Survivor Variables

To test for the association between caregivers and stroke recovery, one must control for factors that may simultaneously influence recovery. In this analysis, we controlled for patient age, income, educational level, and comorbid conditions at baseline (discharge from the hospital). Age is measured as a continuous variable in years, income is measured on a scale from 1 to 5 with higher numbers indicating greater income (1 =  ≤$14,999, 2 = $15,000-$24,999, 3 = $25,000-$34,999, 4 = $35,000-$44,999, 5 = ≥$45,000); educational level is measured on a 7-point ordinal scale with larger numbers indicating higher levels of education (1 =  <7 years, 2 = 7-9 years, 3 = 10-11 years, 4 = high school graduate, 5 = some college or technical school, 6 = college graduate, 7 = graduate school); and patients' comorbid conditions are measured with the Charlson Comorbidity Index. This index ranges from 1 to 6, depending on the severity and number of comorbid conditions.

Caregiver Demographics

Baseline caregiver demographics used in this analysis are age, educational level, sex, and health status. Age is measured as a continuous variable in years, and educational level is measured on an ordinal scale that ranges from 1 to 7, with higher numbers indicating greater levels of education. Health status is a caregiver self-report variable with higher numbers indicating poorer health and is coded as follows: 1 = excellent, 2 = very good, 3 = good, 4 = fair, and 5 = poor.

Caregiver Network Characteristics

We used caregiver relationship to stroke survivor, average number of people in caregiver network over 24 months, time spent providing care, and average caregiver need for outside help to measure characteristics of caregiver networks in this study. Stroke survivors identified their relationship to their primary caregiver (spouse/partner, children, friend, other); this variable was coded so that spouse/partners = 1 and all others = 0. At each interview, caregivers were asked to name up to two others who provided caregiving assistance to them. We averaged their responses over 24 months to obtain the average number of caregivers for each stroke survivor. Caregivers also reported the actual number of hours per day they spend providing care. We averaged their responses over the 24 months poststroke to obtain the average time spent providing care. Caregivers were asked whether they could use outside help with caregiving activities at each time point. We averaged their responses across time to arrive at a variable ranging from 0 to 1 measuring need for help.

Analytic Strategy

The data analyzed in this article represent a longitudinal design with repeated measurements of functional recovery for individuals at successive points in time. The most appropriate statistical method is a multilevel model (also called hierarchical linear models, mixed models, or random effects models). We tested two components in our multilevel models: within-individual components, which measure one person's functional change over time; and between-individual components, which measure variation between individuals in functional change over time. We selected an autoregressive correlation structure for the repeated measurements and used hierarchical linear modeling (specifically, HLM6 software) [29] for the data analyses.

results
Bivariate

The first analysis examines demographic and caregiver network characteristics and compares them for Caucasian, African American, and Puerto Rican stroke survivors. Table 1 displays the results of this analysis. The average age of stroke survivors in our sample is 66.6 years and the average income corresponds roughly to between $15,000 and $25,000 per year. Educational level of stroke survivors corresponds to some college or technical school. average comorbidity index is 3.1 on a scale from 1 to 6. Caregivers have an average age of 59 years and have an average educational level that corresponds to high school graduates. Eighty-eight percent of caregivers are women and their self-rated health is 3.0 on a scale of 1 to 5, indicating average health. Stroke survivors and caregivers do not differ by race/ethnicity in their demographic characteristics.


Table 1.
One-way analysis of variance test for racial/ethnic differences in patient and caregiver characteristics (data shown as mean ± standard deviation unless otherwise indicated).

Characteristic
Total
(N = 121)
Caucasian
(n = 42)
Puerto Rican
(n = 49)
African American
(n = 30)
F-Statistic
p-Value

mFIM
74.82 ± 15.77
78.29 ± 10.14*
67.66 ± 20.28†‡
80.94 ± 8.95*
9.96
0.000
Patient Demographics
 
 
 
 
 
 
Age
66.58 ± 10.29
66.71 ± 9.71
69.31 ± 9.90
61.97 ± 10.39*
5.06
0.008
Income
1.88 ± 1.09
2.10 ± 1.14
1.61 ± 0.98
2.03 ± 1.13
2.67
0.070
Educational Level
4.45 ± 1.25
4.48 ± 1.04
4.59 ± 1.24
4.17 ± 1.51
1.09
0.34
Comorbidity Index
3.12 ± 1.53
2.93 ± 1.33
3.10 ± 1.57
3.43 ± 1.70
0.97
0.38
Caregiver Demographics
 
 
 
 
 
 
Age
59.33 ± 16.70
62.38 ± 11.13
58.77 ± 15.09
56.00 ± 14.12
2.00
0.14
Educational Level
4.08 ± 1.59
4.13 ± 1.03
4.09 ± 1.99
4.00 ± 1.55
0.06
0.94
Female
0.88 ± 0.32
0.92 ± 0.27
0.86 ± 0.35
0.87 ± 0.35
0.42
0.66
Health
3.04 ± 0.91
3.12 ± 0.91
3.02 ± 0.98
2.95 ± 0.95
0.31
0.74
Caregiver Network
 
 
 
 
 
 
Caregiver is Spouse
0.65 ± 0.48
0.71 ± 0.46
0.64 ± 0.48
0.57 ± 0.50
0.84
0.4
Number of Caregivers
1.19 ± 0.34
1.13 ± 0.31*
1.31 ± 0.40
1.11 ± 0.27*
4.06
0.0
Caregiver Needs Help
0.22 ± 0.32
0.17 ± 0.29
0.27 ± 0.35
0.24 ± 0.31
1.13
0.33
Hours Per Day Providing Care
8.53 ± 6.14
7.78 ± 7.02
10.21 ± 5.18
6.82 ± 5.76*
3.35
0.04

*Distinct from Puerto Ricans.
Distinct from Caucasians.
Distinct from African Americans.
mFIM = motor subscale of the Functional Independence Measure.
   


Sixty-five percent of caregivers were the spouse of the stroke survivor, and the average number of caregivers in our sample was 1.19 over the course of 24 months. Twenty-two percent of caregivers report the need for help with caregiving activities, and caregivers report spending an average of 8.5 hours per day providing care. The number of caregivers providing care for stroke survivors is higher for the Puerto Rican group. Puerto Ricans have 1.31 caregivers compared with 1.13 for Caucasians and 1.11 for African Americans. In addition, Puerto Rican caregivers provide more hours per day of care for stroke survivors compared with African Americans. Puerto Rican caregivers average 10.2 hours per day compared with 6.8 for African Americans; Caucasians average 7.8 hours per day, which is not statistically distinct from African Americans.

Multivariate

The second analysis uses hierarchical linear modeling to estimate the average growth trajectory for stroke survivors over the five time points of data collection (24 months) poststroke. Our first model is an unconditional model with only motor function. This unconditional model indicates that the average mFIM score for all veterans across all time points in our sample is 80.24. Model 2 adds time and the quadratic of time to the model to estimate the average mFIM score of stroke survivors in our sample. This model allows us to examine the average trajectory while controlling for time and the quadratic of time. In this model we see that adjusting for time and time 2, the average trajectory is 74.95 with an average change of 5.58 + 2(-0.90) (0-1) from time 0 to time 1. Figure 1 displays the actual trajectories of 8individual stroke survivors in this study as an example of how these trajectories can differ between individuals.


Figure 1. Sample of motor recovery trajectories for 8 stroke survivors. mFIM = motor subscale of the Functional Independence Measure.

Model 3 adds race/ethnicity to examine the average trajectories for Caucasian, African-American, and Puerto Rican stroke survivors. This model demonstrates that, controlling for time, African Americans have an average mFIM score that is 4.66 points higher than Caucasians (p < 0.01) and Puerto Ricans have an average mFIM score that is 5.64 points lower than Caucasians (p < 0.05). Table 2 and Figure 2 display the average recovery trajectories over time for the average stroke survivor, as well as the average Caucasian, Puerto Rican, and African-American stroke survivor.


Table 2.
Average recovery trajectories by race/ethnicity.

Recovery
Time
Trajectories
Average
Caucasian
Puerto
Rican
African
American

Discharge
75.0
76.1
70.5
80.8
1 month
78.7
80.8
73.5
84.0
6 months
80.7
81.5
76.3
85.9
12 months
80.9
81.1
76.7
86.2
18 months
79.3
79.0
74.6
85.1
24 months
75.9
75.4
70.0
82.4

Figure 2. Stroke recovery trajectories: motor subscale of the Functional Independence Measure scores by race/ethnicity.

Higher numbers represent greater functional independence, and Figure 2 demonstrates that the trajectories of stroke survivors are curvilinear. Recovery starts low for all racial/ethnic groups, peaking at 12 months for African Americans and Puerto Ricans and peaking at 6 months for Caucasians before a subsequent decline.

The fourth, fifth, and sixth models add patient demographics, caregiver demographics, and caregiver network characteristics separately in order to examine their effects independently. The seventh and final model adds all variables into the model.

Table 3 displays the results of these models. These models demonstrate that several patient and caregiver attributes predict mFIM scores at baseline. Racial/ethnic differences in mFIM scores are evident in model 3, but Puerto Ricans become similar to Caucasian stroke survivors when patient demographics (model 4) and caregiver network characteristics (model 6) are added. This indicates that these two sets of covariates can explain the lower mFIM score for Puerto Rican stroke survivors. Conversely, African-American stroke survivors have a higher mFIM score than Caucasians even when holding all covariates constant in the model (06 = 3.72, p < 0.05). In the full model (model 7), educational level of the stroke survivor, number of caregivers, and number of hours spent providing care are predictors of mFIM. Patients who have lower educational levels have higher mFIM scores (010 = -1.66, p < 0.05). Patients who have a greater number of caregivers have lower mFIM scores (01 = -7.74, p < 0.05), and patients with greater hours devoted to providing care have lower mFIM scores (014 = -0.49, p < 0.01).


Table 3.
Hierarchical linear modeling: Patient and caregiver characteristics regressed on baseline motor recovery (data shown as unstandardized beta coefficient/standardized beta coefficient).

Characteristic
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7

Intercept
80.25/1.13*
74.95/1.28*
74.94/1.29*
74.96/1.26*
74.96/1.45*
74.94/1.22*
74.96/1.45*
Time
-
5.58/0.72*
5.08/0.77*
5.00/0.76*
4.43/1.09*
5.28/0.82*
4.82/0.81*
Time 2
-
-0.90/0.14*
-0.91/0.14*
-0.93/0.14*
-0.90/0.14*
-0.93/0.14*
-0.93/0.14*
Race/Ethnicity
 
 
 
 
 
 
 
African American
-
-
0.87/0.55
0.77/0.56
0.79/0.64
0.70/0.54
0.44/0.50
Puerto Rican
-
-
0.71/0.47
1.01/0.48
0.72/0.71
1.48/0.48*
1.56/0.64
Patient Demographics
 
 
 
 
 
 
 
Age
-
-
-
-0.06/0.02
-
-
-0.05/0.03
Income
-
-
-
-0.02/0.19
-
-
-0.13/0.24
Educational Level
-
-
-
-0.32/0.17
-
-
-0.44/0.18
Comorbidity Index
-
-
-
-0.53/0.14*
-
-
-0.34/0.15
Caregiver Demographics
 
 
 
 
 
 
 
Age
-
-
-
-
-0.03/0.02
-
-0.02/0.02
Educational Level
-
-
-
-
-0.08/0.19
-
-0.04/0.16
Female
-
-
-
-
0.71/0.84
-
0.99/0.78
Health
-
-
-
-
0.34/0.36
-
0.12/0.28
Caregiver Network
 
 
 
 
 
 
 
Caregiver is Spouse
-
-
-
-
-
-0.67/0.45
-0.55/0.53
Number of Caregivers
-
-
-
-
-
-2.20/0.66*
-1.93/1.31
Caregiver Needs Help
-
-
-
-
-
-0.68/0.74
-0.24/0.92
Hours Per Day Providing Care
-
-
-
-
-
-0.12/0.03*
-0.12/0.04*

Within Variation
-
73.65
73.23
71.45
72.65
71.24
69.82
Between Variation
-
132.57
134.60
126.47
135.96
117.27
116.27
Deviance
-
4,548.50
4,544.70
4,533.10
4,546.50
4,515.40
4,515.60

*p < 0.01.
p < 0.05.
             


The next set of analyses mirrors the first with the exception of the focus of the coefficients. Table 4 displays the results of these analyses. In these models, the dependent level-2 variable becomes time, and each of the person-level covariates are regressed on time, which represents recovery (mFIM scores) over time.


Table 4.
Hierarchical linear modeling: Patient and caregiver characteristics regressed on motor recovery growth (data shown as unstandardized beta coefficient/standardized beta coefficient).

Characteristic
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7

Intercept
80.25/1.13*
74.95/1.28*
76.14/1.55*
75.70/1.49*
77.41/2.53*
76.54/1.90*
78.00/2.50*
Time
-
5.58/0.72*
5.59/0.77*
5.57/0.77*
5.58/0.78*
5.59/0.78*
5.57/0.78*
Time 2
-
-0.90/0.14*
-0.90/0.14*
-0.90/0.14*
-0.90/0.14*
-0.90/0.14*
-0.90/0.14*
Race/Ethnicity
 
 
 
 
 
 
 
African American
-
-
4.66/1.66*
3.98/1.75
5.08/1.85*
4.13/1.51*
3.72/1.84
Puerto Rican
-
-
-5.64/2.62
2-4.18/2.41
-5.47/2.74
-2.36/2.24
-1.86/2.10
Patient Demographics
 
 
 
 
 
 
 
Age
-
-
-
-0.17/0.11
-
-
-0.08/0.10
Income
-
-
-
1.20/0.74
-
-
0.60/0.80
Educational Level
-
-
-
-1.43/0.85
-
-
-1.66/0.83
Comorbidity Index
-
-
-
-1.71/0.62*
-
-
-0.99/0.60
Caregiver Demographics
 
 
 
 
 
 
 
Age
-
-
-
-
0.23/0.11
-
-0.01/0.10
Educational Level
-
-
-
-
1.11/0.95
-
1.28/0.93
Female
-
-
-
-
-1.65/2.35
-
-2.99/2.83
Health
-
-
-
-
1.19/1.59
-
-0.27/1.50
Caregiver Network
 
 
 
 
 
 
 
Caregiver is Spouse
-
-
-
-
-
-2.49/1.80
-0.84/1.82
Number of Caregivers
-
-
-
-
-
-7.13/3.29
-7.74/3.48
Caregiver Needs Help
-
-
-
-
-
-6.61/3.88
-5.68/3.83
Hours Per Day Providing Care
-
-
-
-
-
-0.61/0.13*
-0.49/0.14*

Within Variation
-
73.65
73.65
73.67
73.61
73.68
73.64
Between Variation
-
132.57
117.99
109.82
117.63
94.39
93.50
Deviance
-
4,548.50
4,527.10
4,514.00
4,518.10
4,419.20
4,475.80

*p < 0.01.
p < 0.05.
             


Results from model 3 indicate that no racial/ethnic differences exist in the recovery of motor function over time as measured by mFIM. Puerto Rican stroke survivors show better growth over time when controlling for patient demographics and caregiver network characteristics, as well as in the full model, model 7. In this final model, controlling for all covariates, we see that patient comorbidity, patient educational level, and the number of hours per day care is provided for patient are predictors of improvement in mFIM. Patients with lower levels of education have greater mFIM growth (19 = -0.44, p < 0.05). Patients who score higher on the comorbidity index have slower growth in mFIM scores over time (110 = -0.34, p < 0.05). Patients whose caregivers spend more hours per day providing care have lower mFIM growth (113 = -0.12, p < 0.01).

DISCUSSION

Several patterns of recovery among our participants across the 2-year period were of interest. FIM motor scores steadily increase over the course of the first year and then begin to drop again. For African Americans and Puerto Ricans, unadjusted 24-month mFIM scores are lower than baseline scores. The curvilinear pattern across time in functional independence is similar in all three racial/ethnic groups.

Baseline mFIM scores were significantly higher for African Americans than Caucasians in our sample when we adjusted for caregiver factors. This finding is counter to recent studies that show African-American stroke survivors with a lower FIM score than Caucasians at discharge from the hospital [30]. African-American stroke survivors in our sample were younger, had fewer caregivers, and had fewer hours spent on informal care than Puerto Rican stroke survivors. The higher baseline mFIM scores can be due to two factors that have to do with excess stroke burden and family caregiving. African Americans, and African-American men specifically, are most at risk for experiencing a fatal stroke and experience worse poststroke rehabilitation [5,7]. A recent study indicates that African Americans are more likely to be discharged to extended care facilities or nursing homes than other racial/ethnic groups [31]. Being discharged to a nursing home or other institutional facility was an exclusion criterion for our study; therefore the possibility exists that we have an overrepresentation of healthier African Americans in our sample. Additionally, African Americans as a group are less likely to be married; this lack of spousal and child relationships may make being released to the community more difficult, even if they have a higher relative functional status.

Puerto Rican stroke survivors had lower baseline mFIM scores and similar but lower curvilinear recovery trajectories than Caucasians and African Americans in the unadjusted models. Once the models were adjusted for caregiver network characteristics, we found that Puerto Rican stroke survivors had an average increase in mFIM scores between time points that was significantly greater than for Caucasians. In other words, when caregiver characteristics are taken into account, recovery outcomes across time become better for Puerto Rican stroke survivors. This finding is consistent with a study that speculates caregivers provide a positive effect on recovery for Latinos [11].

Caregiver demographics and network characteristics are associated with baseline motor function for stroke survivors. Controlling for patient demographics, we see that two caregiver network characteristics predict motor impairment at baseline: number of caregivers and number of hours spent per day providing care. Stroke survivors who have more caregivers also have a lower baseline motor functioning. Stroke survivors who have caregivers who spend more time per day providing care also have a lower baseline motor functioning.

Caregiver demographics and network characteristics are associated with recovery of motor function across time. Our analyses demonstrate that independent of patient and caregiver demographics, the number of hours spent per day providing care for stroke survivors is associated with slower motor recovery across time. The more hours spent per day providing care is associated with slower motor recovery across time. In other words, stroke survivors who receive more informal care have smaller gains in recovery across time.

This study has several limitations that center on the study population. First, males make up our VA population of stroke survivors and only two women participated in the study. Participants were mostly World War II and Korean War veterans experiencing chronic conditions that are typical of older populations. Women did not begin joining the military in any significant numbers until the 1970s. We expect to see women with these types of health conditions in veteran populations in the future. Second, the majority of caregivers in our sample are spouses and female family members of these male stroke survivors. Thus, the relationship between caregiver characteristics and stroke survivors may not hold true for male spouses of female stroke survivors. The patterns may differ in ways that are related to traditional gender roles and caretaking activities that partners take on in marital relationships. A third limitation is that our sample consisted of stroke survivors living in Florida and Puerto Rico. Specific interest in multiple components of stroke recovery and informal caregiving among veterans in our integrated service network resulted in our choice of sample. Caregivers and stroke survivors may be unique and dissimilar to caregiver-stroke survivor dyads elsewhere on the U.S. Mainland or other portions of the Caribbean.

This article provides a broad portrait of the association between caregiver characteristics and motor function recovery for veterans over 24 months. These findings are of interest to clinicians and researchers who may develop strategies to aid in the recovery process and ease the burden of the caregiving role. Future research should continue to focus on the differential outcomes of stroke recovery and the role of caregivers in this process.

conclusions

This article provides a broad portrait of the association between caregiver characteristics and motor function recovery for veterans over a 24-month period. These findings are of interest to clinicians and researchers who may develop strategies to aid in the recovery process and ease the burden of the caregiver role. Future research should continue to focus on the differential outcomes of stroke recovery and the role of caregivers in this process.

ACKNOWLEDGMENTS
Author Contributions:
Study concept and design: M. Rittman.
Acquisition of data: M. Rittman.
Analysis and interpretation of data: M. Sberna Hinojosa,
R. Hinojosa.
Drafting of manuscript: M. Sberna Hinojosa, R. Hinojosa.
Critical revision of manuscript for important intellectual content: M. Sberna Hinojosa, M. Rittman, R. Hinojosa, W. Rodriguez.
Statistical analysis: M. Sberna Hinojosa.
Obtained funding: M. Rittman.
Study supervision: M. Rittman.
Financial Disclosures: The authors have declared that no competing interests exist.
Funding/Support: This material was based on work supported by the VA Rehabilitation Outcomes Research Center of Excellence and a VA Nursing Research Initiative Award from Health Services Research and Development to Maude Rittman, RN, PhD (grant 98-183).
Additional Contributions: We thank Avron Spiro III, PhD, for his thoughtful review of the manuscript.
REFERENCES
1. Anderson CS, Linto J, Stewart-Wynne EG. A population-based assessment of the impact and burden of caregiving for long-term stroke survivors. Stroke. 1995;26(5):843-49. [PMID: 7740578]
2. American Heart Association National Center. Heart disease and stroke statistics-2005 update. Dallas (TX): American Heart Association; 2005.
3. Post-Stroke Rehabilitation Guideline Panel. Post-stroke rehabilitation. Clinical guideline 16. AHCPR Publication 95-0662. Rockville (MD): U.S. Department of Health and Human Services; 1995.
4. Schwamm LH, Pancioli A, Acker JE 3rd, Goldstein LB, Zorowitz RD, Shephard TJ, Moyer P, Gorman M, Johnston SC, Duncan PW, Gorelick P, Frank J, Stranne SK, Smith R, Federspiel W, Horton KB, Magnis E, Adams RJ. Recommendations for the establishment of stroke systems of care: recommendations from the American Stroke Association's Task Force on the development of stroke systems. Stroke. 2005;36(3):690-703. [PMID: 15689577]
DOI:10.1161/01.STR.0000158165.42884.4F
5. McGruder HF, Malarcher AM, Antoine TL, Greenlund KJ, Croft JB. Racial and ethnic disparities in cardiovascular risk factors among stroke survivors: United States 1999 to 2001. Stroke. 2004;35(7):1557-61. [PMID: 15192252]
DOI:10.1161/01.STR.0000130427.84114.50
6. Scheinart KF, Tuhrim S, Horowitz DR, Weinberger J, Goldman M, Godbold JH. Stroke recurrence is more frequent in blacks and hispanics. Neuroepidemiology. 1998;17(4):188-98. [PMID: 9701833]
DOI:10.1159/000026172
7. Horner RD, Swanson JW, Bosworth HB, Matchar DB. Effects of race and poverty on the process and outcome of inpatient rehabilitation services among stroke patients. Stroke. 2003;34(4):1027-31. [PMID: 12624220]
DOI:10.1161/01.STR.0000060028.60365.5D
8. Gillum RF. Epidemiology of stroke in Hispanic Americans. Stroke. 1995;26(9):1707-12. [PMID: 7660419]
9. Chiou-Tan FY, Keng MJ Jr, Graves DE, Chan KT, Rintala DH. Racial/ethnic differences in FIM scores and length of stay for underinsured patients undergoing stroke inpatient rehabilitation. Am J Phys Med Rehabil. 2006;85(5):415-23. [PMID: 16628148]
DOI:10.1097/01.phm.0000214320.99729.f3
10. Lisabeth LD, Risser JM, Brown DL, Al-Senani F, Uchino K, Smith MA, Garcia N, Longwell PJ, McFarling DA, Al-Wabil A, Akuwumi O, Moye LA, Morgenstern LB. Stroke burden in Mexican Americans: The impact of mortality following stroke. Ann Epidemiol. 2006;16(1):33-40. [PMID: 16087349]
DOI:10.1016/j.annepidem.2005.04.009
11. Ottenbacher KJ, Smith PM, Illig SB, Fiedler RC, Gonzales V, Granger CV. Characteristics of persons rehospitalized after stroke rehabilitation. Arch Phys Med Rehabil. 2001; 82(10):1367-74. [PMID: 11588739]
DOI:10.1053/apmr.2001.26088
12. Dorsey M, Vaca KJ. The stroke patient and assessment of caregiver needs. J Vasc Nurs. 1998;16(3):62-67. [PMID: 9883149]
DOI:10.1016/S1062-0303(98)90003-6
13. Doty P. Family care of the elderly: The role of public policy. Milbank Q. 1986;64(1):34-75. [PMID: 3084932]
DOI:10.2307/3350005
14. O'Connell BO, Baker L. Managing as carers of stroke survivors: Strategies from the field. Int J Nurs Pract. 2004; 10(3):121-26. [PMID: 15149459]
DOI:10.1111/j.1440-172X.2004.00469.x
15. Glozman JM. Quality of life of caregivers. Neuropsychol Rev. 2004;14(4):183-96. [PMID: 15796115]
DOI:10.1007/s11065-004-8158-5
16. Low JT, Payne S, Roderick P. The impact of stroke on informal carers: A literature review. Soc Sci Med. 1999; 49(6):711-25. [PMID: 10459884]
DOI:10.1016/S0277-9536(99)00194-X
17. Janevic MR, Connell CM. Racial, ethnic, and cultural differences in the dementia caregiving experience: Recent findings. Gerontologist. 2001;41(3):334-47. [PMID: 11405431]
18. Dilworth-Anderson P, Goodwin PY, Williams SW. Can culture help explain the physical health effects of caregiving over time among African American caregivers? J Gerontol B Psychol Sci Soc Sci. 2004;59(3):s138-45. [PMID: 15118019]
19. Roff LL, Burgio LD, Gitlin L, Nichols L, Chaplin W, Hardin JM. Positive aspects of alzheimer's caregiving: The role of race. J Gerontol B Psychol Sci Soc Sci. 2004;59(4): P185-90. [PMID: 15294922]
20. Dilworth-Anderson P, Brummett BH, Goodwin P, Williams SW, Williams RB, Siegler IC. Effect of race on cultural justifications for caregiving. J Gerontol B Psychol Sci Soc Sci. 2005;60(5):S257-62. [PMID: 16131626]
21. Mausbach BT, Coon DW, Depp C, Rabinowitz YG, Wilson-Arias E, Kraemer HC, Thompson LW, Lane G , Gallagher-Thompson D. Ethnicity and time to institutionalization of dementia patients: A comparison of Latina and Caucasian female family caregivers. J Am Geriatr Soc. 2004;52(7): 1077-84. [PMID: 15209644]
DOI:10.1111/j.1532-5415.2004.52306.x
22. Clark M, Huttlinger K. Elder care among Mexican American families. Clin Nurs Res. 1998;7(1):64. [PMID: 9526315]
DOI:10.1177/105477389800700106
23. Phillips LR, De Ardon ET, Komnenich P, Killeen M, Rusinak R. The Mexican American caregiving experience. Hisp J Behav Sci. 2000;22(Pt 3):296-313. DOI:10.1177/0739986300223002
24. Weiss CO, Gonzalez HM, Kabeto MU, Langa KM. Differences in amount of informal care received by non-Hispanic whites and Latinos in a nationally representative sample of older Americans. J Am Geriatr Soc. 2005;53(1):146-51. [PMID: 15667392]
DOI:10.1111/j.1532-5415.2005.53027.x
25. Zsembik BA. Latinos, families and health. In: Crane DR, Marshall ES, editors. Handbook of families and health: Interdisciplinary perspectives. Thousand Oaks (CA): Sage Publications; 2006. p. 4061.
26. Eschbach K, Ostin GV, Patel KV, Markides KS, Goodwin JS. Neighborhood context and mortality among older Mexican-Americans: Is there a barrio advantage? Am J Public Health. 2004;94(10):1806-12. [PMID: 15451754]
DOI:10.2105/AJPH.94.10.1807
27. Rittman MR. Culturally sensitive models of stroke recovery and caregiving after discharge home. Washington (DC): Health Services Research & Development Service; 2000. Nursing Research Initiative: 98-183.
28. Granger CV, Hamilton BB, Fiedler RC. Discharge outcome after stroke rehabilitation. Stroke. 1992;23(7):978-82. [PMID: 1615548]
29. Raudenbush S, Bryk A, Congdon R. HLM6: Hierarchical linear and nonlinear modeling. Chicago (IL): Scientific Software International. 2004.
30. Ottenbacher KJ, Campbell J, Kuo YF, Deutsch A, Ostir GV, Granger CV. Racial and ethnic differences in postacute rehabilitation outcomes after stroke in the United States. Stroke. 2008;39(5):1514-19. [PMID: 18340094]
DOI:10.1161/STROKEAHA.107.501254
31. Onukwugha E, Mullins CD. Racial differences in hospital discharge dispositions among stroke patients in Maryland. Med Decis Making. 2007;27(3):233-42. [PMID: 17502447]
DOI:10.1177/0272989X07302130
Submitted for publication November 13, 2007. Accepted in revised form September 22, 2008.

Go to TOP
Go to the Table of Contents of Vol. 46 No. 2

Last Reviewed or Updated  Friday, August 28, 2009 8:04 AM

Valid XHTML 1.0 Transitional