The Influence of Race and Culture on Rehabilitating Veterans with Stroke through Informal Caregiving

MISSING IMAGE

Material Information

Title:
The Influence of Race and Culture on Rehabilitating Veterans with Stroke through Informal Caregiving
Physical Description:
1 online resource (83 p.)
Language:
english
Creator:
Perrin,Paul Bernard
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Counseling Psychology, Psychology
Committee Chair:
Heesacker, Martin
Committee Members:
Berg, William K
Rice, Kenneth G
Torres-Rivera, Edil

Subjects

Subjects / Keywords:
caregivers -- ethnicity -- rehabilitation -- stroke -- veterans
Psychology -- Dissertations, Academic -- UF
Genre:
Counseling Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The purpose of this study was to create a set of culturally sensitive mental-health-intervention recommendations for the caregivers of Latino/Puerto-Rican, Black, and White veterans with stroke. Accordingly, the study examined whether the mental health of stroke caregivers and functioning of veterans with stroke differed according to race/ethnicity, changed differentially over time according to race/ethnicity, and showed relationships between the two sets of constructs that differed according to race/ethnicity. Data on caregiver mental health and functioning of veterans with stroke were collected from 124 (n = 248) ethnically diverse dyads at 1, 6, and 12 months post-hospital-discharge, with approximately equal representation from White, Black, and Latino/Puerto-Rican dyads. Out of the three racial/ethnic groups, Latino/Puerto-Rican veterans with stroke consistently showed the lowest functioning, and their caregivers showed the poorest mental health, though the mental-health effects did not reach statistical significance. Consistent patterns emerged over time in the relationships between caregiver mental health and functioning of individuals with stroke, though the patterns differed as a function of race/ethnicity. Critical knowledge may be lost regarding the connections between caregiver mental health and veteran functioning when researchers and clinicians look only across caregiver race/ethnicity as opposed to also within race/ethnicity. A monolithic approach to stroke rehabilitation, in particular one that ignores the role of race/ethnicity, is likely to limit the quality of care. Race/ethnicity may affect the nature of the reciprocal influences of caregiver mental health and the functioning of veterans with stroke. Accordingly, a first step in stroke rehabilitation may and perhaps should involve culturally-sensitive mental-health interventions for stroke caregivers.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Paul Bernard Perrin.
Thesis:
Thesis (Ph.D.)--University of Florida, 2011.
Local:
Adviser: Heesacker, Martin.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0041934:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 THE INFLUENCE OF RACE AND CULTURE ON REHABILITATING VETERANS WITH STROKE THROUGH INFORMAL CAREGIVING By PAUL B. PERRIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT O F THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2010 Paul B. Perrin

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS I thank my family for their support through this process and my advisor Dr. Martin Heesacker for his guidance and mentorship. They have been instrumental throughout my entire graduate school career. Without their support, I would not be where I am today.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ...... 4 LIST OF TABLES ................................ ................................ ................................ ................ 7 LIST OF FIGURES ................................ ................................ ................................ .............. 8 ABSTRACT ................................ ................................ ................................ .......................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ........ 11 2 LITERATURE REVIEW ................................ ................................ .............................. 14 Background ................................ ................................ ................................ ................. 14 Mental Health of Stroke Caregivers ................................ ................................ ........... 14 Caregiver Mental Health and Functioning of Persons with Stroke ........................... 15 Cau sal Reciprocity of Caregiver Mental Health and Functioning of Persons with Stroke ................................ ................................ ................................ ....................... 16 Race/Ethnicity and Caregiving ................................ ................................ ................... 22 Implicatio ns for Caregiver Interventions ................................ ................................ .... 26 Hypotheses ................................ ................................ ................................ ................. 27 3 METHOD ................................ ................................ ................................ ..................... 30 Part icipants ................................ ................................ ................................ ................. 30 Procedure ................................ ................................ ................................ .................... 32 Measures: Caregiver ................................ ................................ ................................ .. 33 Geriatric Depres sion Scale (GDS) ................................ ................................ ..... 33 Sense of Competence Questionnaire (SCQ) ................................ ...................... 34 Sense of Coherence (SOC) ................................ ................................ ................. 34 Measures: Veteran with Stroke ................................ ................................ .................. 35 Geriatric Depression Scale (GDS) ................................ ................................ ..... 35 Functional Independence Measur e (FIM) ................................ .......................... 35 Frenchay Activities Index (FAI) ................................ ................................ ............ 36 Data Analysis Plan ................................ ................................ ................................ ...... 36 R epeated Measures Multivariate Analyses of Variance ................................ ..... 36 Canonical Correlations ................................ ................................ ......................... 38 4 RESULTS ................................ ................................ ................................ .................... 41 Hypotheses 1a and 2a ................................ ................................ ................................ 41 Hypotheses 1b and 2b ................................ ................................ ................................ 43 Hypothesis 3 ................................ ................................ ................................ ............... 45

PAGE 6

6 Comparisons of Canonical Correlations by Race/Ethnicity ................................ 46 Dyads with a White Caregiver ................................ ................................ ............. 47 Dy ads with a Latino/Puerto Rican Caregiver ................................ ...................... 49 Dyads with a Black Caregiver ................................ ................................ .............. 51 5 DISCUSSION ................................ ................................ ................................ .............. 55 Overv iew of Findings ................................ ................................ ................................ .. 55 Caregivers: Hypotheses 1a and 2a ................................ ................................ ............ 55 Veterans with Stroke: Hypotheses 1b and 2b ................................ ........................... 57 Hypothesis 3 ................................ ................................ ................................ ............... 5 9 Implications: Dyads with a White Caregiver ................................ .............................. 60 Implications: Dyads with a Lat ino/Puerto Rican Caregiver ................................ ....... 61 Implications: Dyads with a Black Caregiver ................................ ............................... 62 General Implications ................................ ................................ ................................ ... 63 Limitations and Future Directions ................................ ................................ ............... 66 Conclusions ................................ ................................ ................................ ................ 68 APPENDIX A GERIATRIC DEPRESSION SCALE ................................ ................................ .......... 70 B SENSE OF COMPETENCE QUESTIONNAIRE ................................ ....................... 71 C SENSE OF COHERENCE ................................ ................................ ......................... 72 D FUNCTIONAL INDEPENDENCE MEASUR E ................................ ........................... 73 E FRENCHAY ACTIVITIES INDEX ................................ ................................ ............... 74 LIST OF REFERENCES ................................ ................................ ................................ ... 75 BIOGRAPHICAL SKETCH ................................ ................................ ................................ 83

PAGE 7

7 LIST OF TABLES Table page 3 1 Care recipient ICD 9 classifications. ................................ ................................ ...... 30 4 1 Caregiver scale alphas and means and standard deviations by racial/ethnic group over time ................................ ................................ ................................ ...... 42 4 2 Veteran scale alphas and means and standard deviation s by racial/ethnic group over time ................................ ................................ ................................ ...... 44 4 3 Standardized canonical coefficients over time in dyads with a White caregiver .. 49 4 4 S tandardized canonical coefficients over time in dyads with a Latino/Puerto Rican caregiver ................................ ................................ ................................ ...... 51 4 5 Standardized canonical coefficients over time in dyads with a Black caregiver .. 53 4 6 Correlation coefficients among all variables ................................ .......................... 54

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Perrin et al ....... 18 2 2 ........... 19 2 3 lagged panel of canonical correlations 1 and 6 months after hospital discharge. ................................ ................................ ............ 20 2 4 lagged panel of ca nonical correlations 6 and 12 months after hospital discharge. ................................ ................................ ............ 21 2 5 lagged panel of canonical correl ations 1 and 12 months after hospital discharge. ................................ ................................ ............ 22 2 6 Conceptual framework for the current study. ................................ ........................ 28 3 1 Visual depiction of the canonical correlation model applied separately for Black, Latino/Puerto Rican, and White caregivers ................................ ................ 38 4 1 Canonical correlations between veteran and caregiver variable sets over time by caregiver race/ethnicity ................................ ................................ ............. 47

PAGE 9

9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE INFLUENCE OF RACE AND CULTURE ON REHABILITATING VETERANS WITH STROKE THROUGH INFORMAL CAREGIVING By Paul Perrin August 2011 Chair: Martin Heesacker Major: Counseling Psychology The purpose of this study was to create a set of culturally sensitive mental health intervention recommendations for the caregivers of La tino/Puerto Rican, Black, and White veterans with stroke. Accordingly, the study examined whether the mental health of stroke caregivers and functioning of veterans with stroke differed according to race/ethnicity, changed differentially over time accordin g to race/ethnicity, and showed relationships between the two sets of constructs that differed according to race/ethnicity. Data on caregiver mental health and functioning of veterans with stroke were collected from 124 (n = 248) ethnically diverse dyads a t 1, 6, and 12 months post hospital discharge, with approximately equal representation from White, Black and Latino/Puerto Rican dyads. Out of the three racial/ethnic groups, Latino/Puerto Rican veterans with stroke consistently showed the lowest function ing, and their caregivers showed the poorest mental health, though the mental health effects did not reach statistical significance. Consistent patterns emerged over time in the relationships between caregiver mental health and functioning of individuals w ith stroke, though the patterns differed as a function of race/ethnicity. Critical knowledge may be lost regarding the connections between caregiver mental health and veteran functioning

PAGE 10

10 when researchers and clinicians look only across caregiver race/ethni city as opposed to also within race/ethnicity. A monolithic approach to stroke rehabilitation in particular one that ignores the role of race/ethnicity, is likely to limit the quality of care. R ace/ethnicity may affect the nature of the reciprocal influen ces of caregiver mental health and the functioning of veterans with stroke Accordingly, a first step in stroke rehabilitation may and perhaps should involve culturally sensitive mental health interventions for stroke caregivers.

PAGE 11

11 CHAPTER 1 INTRODUCTION African American and Latino individuals face tremendous disparities in the prevalence and course of stroke as compared to White individuals (Ayala et al., 2001; McGruder, Malarcher, Antoine, Greenlund, & Croft, 2004; Scheinart et al., 1998), a trend that is especially pronounced in the southern United States (Borhani, 1965; Cushman et al., 2008; Howard, Labarthe, Hu, Yoon, & Howard, 2007). For example, Sheinart et al. (1998) found that African Americans who had sustained a stroke were twice as likely as Wh ites to sustain a second stroke, whereas Latinos were 2.6 times as likely. Though many cultural, economic, and social factors have been implicated in the causes for these disparities, one major factor may be the differential types of care that informal car egivers provide to individuals with stroke within each racial/ethnic group. Indeed, individuals with stroke from differing racial/ethnic backgrounds can often have very different caregiving experiences because of family structure, cultural norms, and livi ng arrangements (Hinojosa, Zsembik, & Rittman, 2009). For example, as more people take on informal caregiving in a family, as has been reported in some ethnic minority families, the burden on the primary caregiver may decrease (Hinojosa, Zsembik, & Rittman 2009). A consistent link has been established between the mental health of informal caregivers and the functioning of individuals with stroke (e.g. Perrin, Heesacker, Stidham, Rittman, & Gonzalez Rothi, 2008), but this link has been established broadly a cross racial/ethnic groups. One vital area of research for culturally sensitive stroke rehabilitation, then, involves examining the links between caregiver mental health and the functioning of veterans with stroke within and between racial/ethnic groups. B etter understanding the relationships among race/ethnicity and

PAGE 12

12 caregiving is an important step in creating culturally sensitive mental health interventions for stroke caregivers, a process that may improve caregiver mental health and ultimately improve inf ormal care for veterans with stroke. Based on American Heart Association (2007) ratios of stroke incidence and prevalence, as many as 80,000 veterans receiving health care in the Veterans Health Administration (VHA) have sustained a stroke, and 9,000 11,00 0 veterans are hospitalized annually with a new stroke (Veterans Health Administration, 2003). Informal caregivers are so critical in assisting individuals with stroke who need long term care to live independently that caregivers save over $200 billion ann ually in health care costs in the United States (Arno, Levine, & Memmott, 1999). When one considers the disparities in the prevalence and course of stroke among African Americans and Latinos compared to Whites, informal caregiving may be a central contribu tor. Therefore, identifying the culturally specific psychosocial factors that optimize the quality of care that informal quality of life of veterans receiving long te rm care. The central purpose of this project is to create a set of culturally sensitive mental health intervention recommendations for the caregivers of Latino/Puerto Rican, Black, and White veterans with stroke. To achieve that goal, this project will ex amine whether the mental health of stroke caregivers and functioning of veterans with stroke differ according to race/ethnicity, whether the mental health of stroke caregivers and functioning of veterans with stroke change differentially over time accordin g to race/ethnicity, and whether there are racial/ethnic differences over time in size of the relationship between the mental health of stroke caregivers and the functioning of

PAGE 13

13 veterans with stroke. Because published research points to caregiver mental hea lth and veteran with stroke functioning having a reciprocal causal relationship (e.g., Perrin, Heesacker, Sberna Hinojosa, Uthe, & Rittman, 2009), a better understanding of the role of race/ethnicity in the connections between these two sets of variables i s a critical step in creating caregiver mental health interventions that could improve the informal care they provide to veterans with stroke.

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW Background Each year, approximately 700,000 Americans suffer a new or recurrent st roke, and currently over 5.6 million Americans are living with the effects of stroke. These statistics position stroke as the leading cause of disability in the United States (American Heart Association, 2007). Although the death rate from stroke has decre ased in recent years as a result of innovative medical technology, the prevalence of stroke related disability (National Heart, Lung, and Blood Institute, 2004). The informal caregivers, usually family members or friends, to assume primary care of persons with stroke. A National Alliance for Caregiving/American Association of Retired Persons (NAC/AARP 1997) survey revealed that within a span of 8 years, the number of informal caregivers in the United States had tripled, and a significant upward trend in the number of stroke caregivers is expected to continue (Hirst, 2001). The efforts by these informa l caregivers have a substantial impact on health care costs in the United States, saving over $200 billion a year (Arno, Levine, & Memmott, 1999). Research on the functioning of stroke caregivers is vital in order to help caregivers optimally perform their challenging work. Mental Health of Stroke Caregivers The difficult job of caregiving for people with stroke often takes a psychological and emotional toll on caregivers through stress, anxiety, depression, and decreased quality of life (Dorsey & Vaca, 199

PAGE 15

15 Howse, Bagg, & Bartfay, 2007). As many as 40 50% of stroke caregivers experience depression (Anderson, Linto, & Stewart Wynne, 1995; Silliman, Fletcher, Earp, & Wagner, 1986; Stein, Gordon, Hibbard, & Sliwi nski, 1992), and many experience high Fleissner, 1995; Thommessen et al., 2002). Stroke typically has a sudden and unexpected onset, which forces people who are experien cing stroke and their caregivers to make major life changes very quickly (Van Puymbroeck & Rittman, 2005). As a result, there is little time for stroke caregivers to adjust to the role they must caregivers (Bugge, Alexander, & Hagen, 1999). The first month following hospital discharge is one of the most stressful periods for stroke caregivers, as the limitations of the person with stroke become salient at home (Grant, Glandon, Elliott, Giger, Weaver, 2004). This stress can exacerbate the already high rates of physical health problems to which stroke caregivers are vulnerable (Baumgarten et al., 1992; Schulz et al., 1995). Caregiver Mental Health and Functioning of Perso ns with Stroke Specific links exist among reduced functioning of persons with stroke and poor mental health in stroke caregivers. Research has shown that as the mental health and cognitive ability of persons with stroke decrease, caregiver burden increases (Tooth, McKenna, Barnett, Prescott, & Murphy, 2005). Perhaps accordingly, caregiver burden is more prevalent with left hemisphere strokes than with right hemisphere strokes (Blonder, Langer, Pettigrew, & Garrity, 2007). White, Mayo, Hanley, & Wood Dauphin ee (2003) have suggested that because the left hemisphere is associated with language, left hemisphere stroke often creates poor verbal communication between caregivers and persons with stroke, resulting in caregiver fear, hopelessness, depression, and

PAGE 16

16 bur den. Grant, Weaver, Elliot, Bartolucci, and Giger (2004) found that the best predictors of stroke caregiver depression were impaired coherence of the person with stroke, coupled with a caregiver negative problem orientation and inadequate preparedness. Oth er psychosocial factors linked with caregiver depression include a disruption to family and marital life (Evans, Bishop, & Haselkorn, 1991), time constraints (Kinney, Stephens, Franks, Ann, & Norris, 1995; Periard & Ames, 1993), reduced social support (Kot ila, Numminen, Waltimo, & Kaste, 1998), and caregiver physical health (Bugge, Alexander, & Hagen, 1999). Similarly, research has identified connections between caregiver mental health and the physical functioning of individuals with stroke. For example, po sitive associations have emerged between stroke severity and caregiver depression Caregivers with physical and basic care responsibilities experience more burden than do caregiver s with less physical and fewer basic care responsibilities (Chumbler, Rittman, Van Puymbroeck, Vogel, & Qin, 2004; Perrin et al., 2008). Causal Reciprocity of Caregiver Mental Health and Functioning of Persons with Stroke A pervasive assumption in the rese arch literature is that lower functioning of persons with stroke causes a decline in caregiver mental health. Although this assumption has a strong theoretical basis, study based evidence has also suggested the opposite causal direction. Poor caregiver men tal health has been shown to lead to lower functioning of people with stroke. For example, caregiver depression is positively associated with less than optimal care of people with stroke (Evans, Bishop, & Haselkorn, 1991) and inversely related to successfu l, long term stroke recovery (Evans,

PAGE 17

17 Hendricks, Haselkorn, Bishop, & Baldwin, 1992). Accordingly, overwhelmed caregivers who experience burnout are more likely to institutionalize dependent family members (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1 995). High caregiver depression, high burden, and low social support may lead to higher depression in people with stroke (Suh et al., 2005). Investigating causal reciprocity, Perrin et al. (2008) uncovered a medium sized negative correlation between caregi ver mental health and the physical and emotional impact of stroke. See Figure 2 equation model representing this relationship.

PAGE 18

18 Figure 2 1. onfirmatory f acto r a nalysis with a dequate f it i ndices (Note. CR = Care recipient. CG = Caregiver ) To explain this broad correlation, Perrin et al. (2008) proposed a causal feedback loop whereby the poor mental health of caregivers adversely impacts the quality of care th ey provide, thus slowing the recovery of people with stroke. This poor recovery then negatively affects the mental health of stroke caregivers, and the loop continues. See Figure 2 CR : Stroke Severity CG: Mental Health u5 u6 u7 u3 u2 u1 .00 CR : Stroke Impact .10 CR : Depression .16 CR : Language .51 CG: Coherence .34 CG: Burden .56 CG: Depression .23 CR : Recovery u4 .71 .75 .31 .31 .48 .40 1.00 .58

PAGE 19

19 Figure 2 2. eedback l oop i ncorporating q uality of i nformal c aregiving Perrin et al. (2009) then tested this theory using longitudinal data from a sample of stroke caregiver/care recipient dyads. They employed a cross lagged panel design with canonical correlations to examine whether caregiver mental health or care recipient functioning were more causally dominant in the relationships between the two variable cross lagged panels examining these two variable sets across 1 to 6 months (Figure 2 3), 6 to 12 months (Figure 2 4), and 1 to 12 months (Figure 2 5) post hospital discharge appear below. In these figures, larger correlations across time between earlier c aregiver mental health and later care recipient functioning indicate a causal predominance of caregiver mental health in the relationship between the two sets of variables. Conversely, larger correlations across time between earlier care recipient function ing and later caregiver mental health indicate a causal predominance of care recipient functioning in the same relationship. Caregiver Mental Health Quality of Informal Caregiving Stroke Severity/Recovery

PAGE 20

20 However, neither caregiver mental health nor care recipient functioning were found to causally dominate the relationship between th e caregiver and care recipient variables, indirectly supporting the theory that the constructs operate in a feedback loop. Figure 2 3. c ross lagged p anel of c anonical c orrelations 1 and 6 m onths after h ospital d ischarge ( Note. FIM 1(6) = Functional Independence Measure at one (six) month(s) after discharge. FAI 1(6) = Frenchay Activities Index at one (six) month(s) after discharge. CR(CG): GDS 1(6) = care recipient (caregiver) Geriatric Depression Scale at one (six) month(s) after discharge. SOC 1(6) = Sense of Coherence at one (six) month(s) after discharge. SCQ 1(6) = Sense of Competence Questionnaire at one (six) month(s) after discharge )

PAGE 21

21 Figure 2 4. ross lagged p anel of c anonical c orrelations 6 and 1 2 m onths after h ospital d ischarge. ( Note FIM 6(12) = Functional Independence Measure at six (twelve) months after discharge. FAI 6(12) = Frenchay Activities Index at six (twelve) months after discharge. CR(CG): GDS 6(12) = care recipient (caregiver) Geria tric Depression Scale at six (twelve) months after discharge. SOC 6(12) = Sense of Coherence at six (twelve) months after discharge. SCQ 6(12) = Sense of Competence Questionnaire at six (twelve) months after discharge )

PAGE 22

22 Figure 2 5. 9) c ross lagged p anel of c anonical c orrelations 1 and 12 m onths after h ospital d ischarge. ( Note. FIM 1(12) = Functional Independence Measure at one (twelve) month(s) after discharge. FAI 1(12) = Frenchay Activities Index at one (twelve) month(s) after disc harge. CR(CG): GDS 1(12) = care recipient (caregiver) Geriatric Depression Scale at one (twelve) month(s) after discharge. SOC 1(12) = Sense of Coherence at one (twelve) month(s) after discharge. SCQ 1(12) = Sense of Competence Questionnaire at one (twelve ) month(s) after discharge ) Race/Ethnicity and Caregiving Research to date suggests that the effects of stroke are especially pronounced in African American and Latino individuals, who have been reported to experience more strokes and to recover more slow ly from stroke than White individuals ( Ayala et al., 2001; McGruder, Malarcher, Antoine, Greenlund, & Croft, 2004; Scheinart et al., 1998). For example, Sheinart et al. (1998) reported that African Americans who sustained a stroke were twice as likely as W hites to sustain a second stroke, whereas Latinos were 2.6 times as likely. Further, evidence has been reported that these differences have been even more pronounced in the Southern states (Cushman et al., 2008; Howard, et

PAGE 23

23 al., 2007), and research suggests that there has long been a 50% greater mortality rate (Borhani, 1965). One of the major contributors to these racial/ethnic disparities in stroke could be the types o f informal care that individuals with stroke receive. The socio cultural contexts in which African American, Latino, and White stroke caregivers provide care have some key differences. White culture is heavily influenced by individualism, a social outlook that stresses independence and self reliance. On the other hand, Latino culture is generally collectivist, as Latinos often place significant value on the well being of the group rather than on the individual (Marn & Marn, 1991; Triandis, 1995). Conseque ntly, the care structures of each ethnic group tend to be individualist and collectivist, respectively (Dilworth Anderson, Williams, & Gibson, 2002). African American care structures exist somewhat in between the extremes of White and Latino care structure s, in that compared to that of Whites, African American care structures are larger and more likely to include friends and neighbors than relatives (Cox, 1993; Lawton, Rejagopal, Brody, & Kleban, 1992). In Puerto Rican culture, individuals often display a s trong adherence to the cultural value of familism, a feeling of loyalty and obligation to provide support when family members are sick or hurt, whether the interpersonal relationships are among members of the nuclear or extended family (Delgado & Tennstedt 1997; Snchez Ayendz, 1998; Villarreal, Blozis, & Widamen, 2005; Zea, Quezada, & Belgrave, 1994; Zsembik & Bonilla, 2000). In contrast, Whites, who have fewer cultural ties, are less likely to feel a sense of obligation toward all extended family member s and more likely to have smaller,

PAGE 24

24 less diverse networks, and African Americans tend to help people who are most important to them, rather than specifically helping those who are members of their family (Dilworth Anderson, Williams, & Gibson, 2002). Though research has examined differences in the structure of caregiving between ethnic groups, very little research has examined the connections between race/ethnicity and the mental health of stroke caregivers, and the little research that has been performed th us far appears inconsistent. For example, Wright, Hickey, Buckwalter, over time, whereas African other hand, Grant (1999) found the opposite, that African American stroke caregivers had a higher prevalence of depression than did White caregivers. The reliability and generalizability of these two studies would be difficult to determine if it were not for a study that i (Thommessen et al., 2002). This study found a similar type and degree of poor mental health among caregivers independent of disease, thus cautiously extending the discussion of stroke care giver mental health to dementia caregiver mental health, which has attracted a considerably greater amount of research. A number of dementia caregiver studies have triangulated on a general consensus that African American caregiver stress, burden, and dep ression tend to be lower than that of Whites. (Connell & Gibson, 1997; Covinski et al., 2003; Farran, Miller, Kaufman, & Davis 1997; Janevic & Connell, 2001). Connell and Gibson (1997) attributed the disparities to the higher likelihood in African America ns of the caregiver being an adult child, friend, or other family member and the tendency to use prayer or faith as a coping

PAGE 25

2 5 strategy. This line of thinking traces back to the general traits of the two racial/ethnic groups found in the scientific literatur e; Whites lean toward individualism, whereas African Americans find a greater balance between individualism and collectivism. Only one study attributed more depressive symptoms to African American caregivers than e may have resulted in statistical significance of a small effect (Williams, 2005). Knight and McCallum (1998), in their study of heart rate reactivity and depression in African American and White dementia caregivers, developed two related explanations for the high levels of African American caregivers may report less depression and burden than White caregivers, when in reality that may not be the case. However, because most measures of depression and burden have been self report, this hypothesis is difficult to test. The second explanation is the positive reappraisal hypothesis, which holds that, compared to Whites, African American caregivers tend to fo cus more on positive aspects of the stress experience. In principle, this posited reappraisal reduces the negative emotional outcomes in response to stressors. A follow up study supported both hypotheses, as African American dementia caregivers, in compari son to Whites, reported lower levels of burden, but equal levels of depression and anxiety (Knight, Silverstein, McCallum, & Fox, 2000). Dilworth due to their collectivist roots that is, their strong identification with a cultural sense of duty, expectation, and obligation to family members. Perhaps accordingly, Covinsky et al. (2003) compared African

PAGE 26

26 American, White, and Latino dementia caregivers finding that Latino caregivers showed the highest rates of depression and personal role strain. Further evidence for the implications of Dilworth found that Latino dementia caregivers delay inst itutionalization significantly longer than White caregivers do (Mausbach et al., 2004). This study also found that Latinos who were less American acculturated, less individualistic, and more in touch with their collectivist Latino culture viewed caregiving more positively than did those who were more heavily influenced by American culture. Implications for Caregiver Interventions As the above review highlights, the majority of research examining the connections between race/ethnicity and caregiver mental h ealth has been performed on dementia caregivers. With the strong racial/ethnic disparities in the prevalence and course of stroke, a vital next step is to bring a racial/ethnic lens to studying the relationships between the mental health of stroke caregive rs and the rehabilitation of veterans with stroke. Better understanding of the role of race/ethnicity on this relationship could help the VA healthcare system create culturally sensitive mental health interventions for caregivers targeting particular ethni c groups, a process that may improve caregiver mental health and ultimately improve informal care for veterans. Visser Meily, van Heugten, Post, Schepers, and Lindeman stroke caregiver interventions suggested that of the predominant typ es of intervention, counseling programs had the most positive outcomes, especially those involving active problem solving strategies. Therefore, based on the particular needs of caregivers in each ethnic group, one potential culturally sensitive interventi on could involve cognitive behavioral counseling for caregivers in the ethnic groups exhibiting the highest

PAGE 27

27 depression and burden and the poorest coping. Accordingly, improved caregiver mental health in turn could lead to better informal caregiving and ult imately to more effective stroke rehabilitation, as based on the causal feedback loop theory articulated by Perrin et al. (2008). Such an assertion, if supported by findings from the proposed study, could prove invaluable in follow up experiments examining whether counseling interventions for caregivers improve the quality of informal care provided to veterans with stroke in ethnic groups facing health disparities. Hypotheses The conceptual framework for this project encompasses two components diagrammed in Figure 2 6 : (a) The top three boxes represent a feedback loop whereby poor functioning of veterans with stroke leads to poorer caregiver mental health, which lowers the quality of informal caregiving provided to veterans. This lowered quality of informal care results in reduced functioning of veterans with stroke, and ultimately in limited stroke rehabilitation. (b) The bottom box highlights the impact of race/ethnicity/cultural background on each stage in the feedback loop.

PAGE 28

28 Figure 2 6 C onceptual frame work for the current s tudy. Understanding the ways in which racial/ethnic background affects caregiver mental health and the functioning of veterans with stroke over time is a key step in creating a set of culturally sensitive mental health intervention r ecommendations for caregivers. With this theoretical model in mind, the current study has three overarching research questions and resulting sets of hypotheses: (a) Do the mental health of stroke caregivers and functioning of veterans with stroke differ a ccording to race/ethnicity? Hypothesis 1a: Black caregivers will have better mental health than White caregivers, who will in turn have better mental health than Latino/Puerto Rican caregivers, across measures of burden, depression, and coping. Hypothesis 1b: Black and Latino/Puerto Rican veterans with stroke will have poorer functioning than White veterans across measures of depression, activities of daily living, and quality of life.

PAGE 29

29 (b) Does the mental health of stroke caregivers and the functioning of veterans with stroke change differentially over time as a function of race/ethnicity? Hypothesis 2a: The mental health of Black caregivers will improve at a more rapid rate than that of White caregivers, which will in turn improve at a more rapid rate tha n that of Latino/Puerto Rican caregivers. Hypothesis 2b: The functioning of Black and Latino/Puerto Rican veterans with stroke will improve more slowly than that of White veterans. (c) Are there racial/ethnic differences over time in the size of the rela tionship between the mental health of stroke caregivers and the functioning of veterans with stroke? Hypothesis 3: The broad correlations between these two sets of variables at 1 6 and 12 months post discharge will be larger for dyads with a White car egiver than for dyads with a Black or Latino/Puerto Rican caregiver. (See the Data Analysis section below for a visual model outlining Hypothesis 3.)

PAGE 30

30 CHAPTER 3 METHOD Participants The sample was initially comprised of 135 ( n = 270) stroke caregiver/vet eran with stroke dyads. Veterans with stroke had all been discharged directly to home following hospital care for their strokes. The dyads were selected from five geographically and ethnically diverse VAMCs in Florida and Puerto Rico. These VAMCs are part of the Veterans Integrated Service Network (VISN 8, including South Georgia, Florida, Puerto Rico, and the U.S. Virgin Islands) and were selected specifically to gain a better understanding of how ethnically diverse veterans with stroke and their caregiver s manage the stroke recovery process. All veterans met classification for the International Classification of Diseases and Related Health Problems (ICD 9) codes 430 438 (See Table 3 1 for a breakdown of ICD classifications). Table 3 1 Care recipient ICD 9 c lassifications. ICD 9 Code (Title) n 43 0 (subarachnoid hemorrhage) 1 431 (intracerebral hemorrhage) 5 432 (other and unspecified intracrania l hemorrhage) 0 433 (occlusion and ste nosis of precerebral arteries) 18 434 (occ lusion of cerebral arteries) 49 435 ( transient cerebral ischemia) 3 436 (acute, but ill def ined, cerebrovascular disease) 49 437 (other and ill defined cerebrovascular disease) 3 438 (late effect s of cerebrovascular disease) 7 Total 135 (Note. ICD 9 code 435 is a transient ischemic attac k (TIA) and technically not a stroke, but these three participants were included in the current study because TIA effects often closely resemble those of strokes). The stroke caregivers either identified themselves or were identified by veterans with stro ke as the primary, informal caregiver. Veterans were recruited, gave consent, and were surveyed while they were in the hospital for the acute stroke event. In order to

PAGE 31

31 be included in the 1 month follow up, they had to be discharged directly home (i.e. be m edically stable), needed to have a score of 18 or higher on the Mini Mental State Exam (Folstein, Folstein, & McHugh, 1975), and must have been able to communicate orally at discharge. Of the veterans with stroke, 61 reported weakness on the right side of their bodies, 59 on the left side, 4 reported bilateral weakness, and 11 reported that they had no body weakness. No data were collected on whether the veterans had sustained prior strokes. There was no exclusion of participants based on comorbidities, but their diagnosis was coded throughout the study by reviewing patient records at each of the three time points. One hundred thirty five caregiver/veteran dyads were originally enrolled, and initial data were collected while the veterans were still in the ho spital. At the 1 month follow up visit, 11 dyads withdrew, reducing the final sample to 124 dyads ( n = 248). Of the 124 caregivers, 45 (36.3%) identified as White, 28 (22.6%) as Black, 2 (1.6%) as Asian, and 49 (39.5%) as Latino/Puerto Rican (8 residing in the U.S. and 41 residing in Puerto Rico). The two Asian caregivers were excluded from data analyses because the focus of this study was on Black, White, and Latino/Puerto Rican caregivers. However, the care recipients of these two Asian caregivers, who we re Latino/Puerto Rican, were included in the analyses that examined veteran functioning by racial/ethnic group. Of the 124 veterans, 45 (36.3%) identified as White, 30 (24.2%) as Black, and 49 (39.5%) as Latino/Puerto Rican (8 residing in the U.S. and 41 r esiding in Puerto Rico). These racial/ethnic categories were used in the current study because they are the

PAGE 32

32 ethnic category, so it is possible to identify as both Black or White and Latino/Puerto Rican. However, in the current study, participants were asked to choose only one category that best described their racial/ethnic identifica tion. Of the veterans, 122 (98.4%) were men, and 2 (1.6%) were women, and of the caregivers, 16 (12.9%) were men, and 108 (87.1%) were women. The average age was 66.13 ( SD = 10.61) for veterans and 59.0 ( SD = 14.08) for caregivers. In terms of caregiving n etwork size, dyads with a Latino/Puerto Rican caregiver reported the largest caregiving network, with an average of 1.31 caregivers, whereas dyads with a White or Black caregiver reported an average caregiving network size of 1.13 and 1.11 caregivers, resp ectively. At the 6 month visit, 12 dyads withdrew, and 13 more withdrew at the 12 month visit. At the end of the 12 month time period, 99 dyads remained, producing a retention rate of 74.1%. Of the dyads who withdrew, 2 veterans went into a nursing home, 3 veterans were too ill to continue, 7 veterans died, 12 dyads withdrew consent for study participation, 9 dyads moved away or were unable to be re contacted at follow up, 1 caregiver became ineligible for study, and 1 caregiver died and was not replaced. Dyads received incentives of $25 for participating at the 1 month follow up, $25 for the 6 month follow up, and $50 for the 12 month follow up. Procedure This project is part of a larger study that examined quantitative and qualitative aspects of post stro ke recovery and stroke caregiving. Data for the larger study were collected in the homes of persons with stroke at 1, 6 and 12 months after hospital discharge through four methods: in depth face to face interviews, observations, questionnaires, and medical record abstraction. Interviews of persons with stroke and

PAGE 33

33 caregivers occurred during three home visits and included qualitative and quantitative assessments. Interviewers representing the three racial/ethnic groups were trained for data collection. Spanis h speaking data collectors were used for Puerto Rican participants as needed. Data collectors of the same racial/ethnic origin as participants have been found to be more successful in obtaining authentic data (National Institutes of Health, 1995), and thus all the interviewers were matched to participants by race/ethnicity. Data collectors received training in the administration of the Functional Independence Measure and the Frenchay Activities Index. Caregivers were present during the interviews of the vet erans with stroke and concurrently filled out a survey focusing on 2.5 hours. Measures: Caregiver Geriatric Depression Scale (GDS) The GDS (Yesavage & Brink, 1983; Appendix A) contains 30 dichotomous choi ce items and assesses the presence and degree of depression in older adults. Example lect the presence of significant depressive symptoms. The GDS has been shown to successfully differentiate depressed from non depressed older adults (Sheikh & Yesavage, 1986), and Yesavage and Brink (1983) reported the scale to have internal consistency ( = .94) and adequate reliability across one week ( r = .85). Numerous Izal, 1998).

PAGE 34

34 Sense of Competence Questionnaire (SCQ) The SCQ (Vernooij Dassen, Persoon, & Felling, 1996; Appendix B) contains 27 4 Higher scores indicate higher burden. The SCQ has been repeatedly used to assess caregivers of individuals with stroke (Scholte op Reimer, de Haan, Pijnenborg, Limburg, & van den Bos, 1998), as well as caregivers of persons with dementia (Vernooij Dassen, Persoon, & Felling, 1996). The scale has been shown to have adequate reliability ( = .83) and to be stable across one week (Intraclass Correlation = .93; Scholte op Reimer et al., 1998). Jansen et al. (2007) have also found the SCQ to have satisfactory Cronbach's alpha and item total correlations. Sense of Coherence (SOC) The SOC (Antonovsky, 1987; Appendix C) is comprised of 13 7 point items and assesses the extent to which a person views the world as comprehensible, meanin gful, and manageable. Sense of coherence is defined as: A global orientation that expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence that (1) the stimuli deriving in the course of living are structured, predictable, and explicable; (2) the resources are available to meet the demands posed by these stimuli; and (3) these demands are challenges, worthy of investment and engagement (Antonovsky, 1987). In the stroke car (Schnyder, Buchi, Sensky, & Klaghofer, 2000, p. 296).

PAGE 35

35 feeling that you are being surprised by the behavior of people whom you thought you indicate higher coherence, or in other words better coping ability. The scale has been shown to have high internal consistency ( s in the .90s; Frankenhoff, 1998) and adequate stability across time ( r s > .70 for 6 12 months; Schny der et al., 2000). Measures: Veteran with Stroke Geriatric Depression Scale (GDS) A description of the GDS is omitted here. P section above. A full description of the scale appears in the above section. Functional Indep endence Measure (FIM) The FIM (Granger & Hamilton, 1990; Appendix D) is the most widely used method of assessing activities of daily living in persons with a disability. The FIM consists of 18 7 point items. The FIM has six subscales: self care, sphincter control, transfer capability, locomotion, communication, and social cognition. The validity and reliability of the FIM are well established (Hamilton, Laughlin, Fiedler, Granger, 1994; Corrigan, Smith Knapp, & Granger, 1997; Stineman et al. 1996; Dodds, Ma rtin, Stolov, & Deyo, 1993). In a meta analysis study by Ottenbacher, Hsu, Grander, & Fiedler (1996), the median interrater reliability for the total FIM was .95, and the median test retest and equivalence reliability values were .95 and .92, respectively. All data collectors in the current study were trained in completing the FIM assessment and were tested by professionals with appropriate training to certify inter rater reliability.

PAGE 36

36 Frenchay Activities Index (FAI) The FAI (Holbrook & Skilbeck, 1983; Appen dix E) contains 15 4 point items and was developed for measuring disability and handicap in persons with stroke relating to the instrumental activities of daily living. It assesses quality of life beyond the necessary skills for basic activities of daily l iving. Subscales include domestic, leisure/work, and outdoor activities. Schuling, de Haan, Limburg, and Groenier, (1993) found support for Barthel, 1965) and Sickness Impac t Profile (Bergner, Bobbitt, Carter, & Gilson, 1981). Data Analysis Plan RMANOVAs The first set of data analyses will be two omnibus repeated measures multivariate an alyses of variance (RMANOVAs) to determine whether the mental health of stroke caregivers and functioning of veterans with stroke differ according to caregiver and veteran race/ethnicity, change over time, or change differentially over time according to ca regiver and veteran race/ethnicity. In the first RMANOVA, the independent variables will be caregiver race/ethnicity, time, and the caregiver race/ethnicity*time interaction. The dependent variables will be caregiver Geriatric Depression Scale (depression) caregiver Sense of Competence Questionnaire (burden), and caregiver Sense of Coherence (coping). In the second RMANOVA, the independent variables will be veteran race/ethnicity, time, and the veteran race/ethnicity*time interaction. The dependent variabl es will be veteran Geriatric Depression Scale, veteran Functional Independence Measure (activities of daily living), and veteran Frenchay Activities Index (quality of life). In either of these analyses, if the effect of race/ethnicity on the

PAGE 37

37 dependent vari ables is significant, it will indicate that the dependent variables differ in a multivariate manner by race/ethnicity. If the effect of time on any of the dependent variables is significant, it will indicate that the dependent variables change in a multiva riate manner over time. And finally, if the effect of the race/ethnicity*time interaction on any of the dependent variables is significant, it will indicate that the dependent variables change in a multivariate manner differentially over time according to race/ethnicity. If the omnibus F statistics from the independent variables in the RMANOVAs is significant, follow up Bonferroni corrected t tests will be employed to determine the precise locations of mean differences among the different racial/ethnic grou ps or over time regarding caregiver mental health and functioning of veterans with stroke. For the RMANOVAs, we used the statistical software G*Power to compute the sample size we would need in order to have sufficient statistical power. With a mixed mode l RMANOVA, a sample size of 79 dyads would be needed to detect a large (Cohen, 1988) multivariate effect. As stated above, at 12 months post discharge, through attrition, the sample size reached 99 dyads. Therefore, if the omnibus multivariate effect is la rge, the current sample size will be sufficient. However, an RMANOVA with the current sample size would not uncover many medium sized effects, unless they were approaching a large sized effect. Therefore, follow up t tests may need to be performed even in the absence of an omnibus RMANOVA effects, but these t tests would be performed with Bonferroni corrections in order to control for family wise error.

PAGE 38

38 Canonical Correlations The second data analysis will be three canonical correlation models run independe ntly for dyads based on caregiver racial/ethnic identification. In other words, one canonical correlation model will be run for dyads with a White caregiver, one model for dyads with a Latino/Puerto Rican caregiver, and one model for dyads with a Black car egiver. The purpose of these three models is to see whether the broad correlations between caregiver mental health and the functioning of veterans with stroke (care recipients) at each of the three points in time (1 6 and 12 monts post discharge, denot ed in Figure 7 below by correlations x, y, and z, respectively) differ according to the race/ethnicity of the caregiver. Figure 3 1 Visual d epiction of the c anonical c orrelation m odel a pplied s eparately for Black, Latino/Puerto Rican, and White c aregiv ers

PAGE 39

39 A canonical correlation uses shared variance from two sets of variables (in this case, caregiver mental health variables and care recipient functioning variables at each point in time) to create a correlation coefficient (r) indicating shared variance between the two sets (correlations x, y, and z). A canonical correlation is similar to a traditional Pearson r, but is calculated from two latent roots of shared variance (one root from each variable set) as opposed to two observed variables. In the above figure, amalgamations of latent, shared patterns of variance (shown in circles) are derivations of the two sets of measured variables (shown in boxes) at each point in time. A canonical correlation analysis produces a number of canonical correlations equa l to the number of measured variables in the smallest variable set. In the current analyses, both variable sets at each point in time contain three measured variables, so each canonical correlation analysis will produce three canonical correlations. Each c anonical correlation produced in a single analysis is comprised of variance unique from the variance within the other canonical correlations in that same analysis. The first canonical correlation is always the largest in magnitude, with each subsequent can onical correlation decreasing in magnitude and in statistical significance. In the analyses shown in Figure 7, the focus will be only on the first (and largest) canonical correlation, because a goal of this research is to simplify the complex web of interc onnections between the veteran with stroke and caregiver variable sets and to increase generalizability to other settings by focusing only on relatively large effects. Though power testing for canonical correlations is not as straightforward as in general linear models like RMANOVA, Mendoza, Markos, and Gonter (1978) have suggested that relatively small sample sizes of approximately 50 are sufficient to detect

PAGE 40

40 strong effects in canonical correlations. In fact, with small sample sizes, canonical correlation s are a much preferred method to other similar methods such as structural equation modeling which are more heavily dependent on large samples. It is important to note that the canonical correlation model shown in the figure above is not a structural equati on model or a factor analysis, so there are no indices of fit and no focus on overall model fit. Weston, Gore, Chan, and Catalano (2008) cautioned against using structural equation models in rehabilitation research with samples of fewer than 200 participan ts. Due to attrition, the current sample reached 99 dyads at 12 months post discharge, so correlations can answer with fewer participants many of the same modeling qu estions even for canonical correlations, bivariate correlations will be used to examine the connections between individual caregiver variables and individual veteran variables within racial/ethnic group and within each point in time, as outlined in the above model.

PAGE 41

41 CHAPTER 4 RESULTS Hypotheses 1a and 2a It was hypothesized that Black caregivers would have better mental health than White caregivers across measures of burden, depression, and coping (Hypothesis 1a) and that White caregivers would have better mental health that Latino/Puerto Rican caregivers It was also hypothesized that the mental health of Black caregivers would improve at a more rapid rate than th at of White caregivers, which would improve at a more rapid rate than that of Latino/Puerto Rican caregivers (Hypothesis 2a). A repeated measures multivariate analysis of variance (RMANOVA) was performed in which the independent variables of interest were caregiver race/ethnicity, time, and the caregiver race/ethnicity*time interaction. The dependent variables were caregiver Geriatric Depression Scale (GDS; depression), caregiver Sense of Competence Questionnaire (SCQ; burden), and caregiver Sense of Cohere nce (SOC; coping). The Box M test for homogeneity of the variance covariance matrices across design cells was non significant, Box M = 129.69, F (90, 8318.63) = 1.18, p = .116, indicating multivariate support for the homogeneity of variance assumption. Add itionally, homogeneity of variance assumption. Taken together, these findings suggest that a more liberal estimate of F Th e overall omnibus RMANOVA revealed no statistically significant effects for caregiver race/ethnicity, F (2, 87) = .14, p = .87, 2 F (2, 86) = 1.67, p = .194, 2 = .037, or the caregiver race/ethnicity*time interaction, F (4, 172) = 1.11, p = .355, 2 = .025. Accordingly, neither

PAGE 42

42 Hypothesis 1a nor 2a was supported. As strictly an exploratory analysis, follow up Bonferroni corrected t tests for caregiver race/ethnicity were run within each dependent measure at each point in time. These t tests showed no statistically significant differences on the GDS, SCQ, or SOC scores among W hite, Latino/Puerto Rican and Black caregivers at any point in time. See Table 4 1 for means and standard deviations of all caregiver measures from the current study broken down by time and racial/ethnic group. Table 4 1 Caregiver s cale A lphas and m eans and s tandard d eviations by r acial/ e thnic g roup over t ime 1 Month Measure White ( n = 43) Latino/PR ( n = 49) Black ( n = 25) 7.02 (5.81) 7.49 (5.97) 4.84 (6.05) 50.14 (10.34) 51.92 (10.29) 52.72 (12.56) 70.65 (14 .08) 70.86 (13.34) 72.76 (17.34) 6 Month Measure White ( n = 38) Latino/PR ( n = 44) Black ( n = 21) 6.21 (5.44) 6.48 (5.45) 5.52 (6.17) 51.34 (10.02) 53.16 (12.12) 51.57 (13.72) 70.47 (13.16) 68.95 (13.26) 73.81 (14.34) 12 Month Measure White ( n = 34) Latino/PR ( n = 41) Black ( n = 20) 6.97 (5.51) 5.54 (5.50) 3.45 (4.41) 50.74 (11.07) 52.32 (10.72) 51.00 (12.19) 71.00 (12.26) 69.07 (11.91) 70.90 (13.74) In Table 4 1 several visual trends in racial/ethnic differences emerged, though these trends should be interpreted with extreme caution due to statistical non significance both in the overall RMANOVA and in the exploratory follow up t tests. Black caregivers seemed consistently to report lower depression than caregivers from the other two racial/ethnic groups did at each of the three time points. Similarly, Black caregivers appeared to report better coping at 1 and 6 months than did caregivers from the other two racial/ethnic groups. Conversely, in Table 4 1 Latino/Puerto Rican caregivers had the lowest coping and highest burden out of the three racial/ethnic

PAGE 43

43 groups at 6 and 12 months, though all of these effects failed to reach statistical significance. Ther e were no visual trends in racial/ethnic differences for the rate of change in any caregiver mental health measure over time. Hypotheses 1b and 2b It was hypothesized that Black and Latino/Puerto Rican veterans with stroke would have poorer functioning tha n White veterans across measures of depression, activities of daily living, and quality of life (Hypothesis 1b). It was also hypothesized that the functioning of Black and Latino/Puerto Rican veterans with stroke would improve more slowly than that of Whit e veterans (Hypothesis 2b). An RMANOVA was performed in which the independent variables of interest were veteran race/ethnicity, time, and the veteran race/ethnicity*time interaction. The dependent variables were veteran Geriatric Depression Scale, veteran Functional Independence Measure (activities of daily living), and veteran Frenchay Activities Index (quality of life). The Box M test for homogeneity of the variance covariance matrices across design cells was significant, Box M = 226.21, F (90, 13487.43) = 2.13, p < .001, indicating a multivariate violation of the homogeneity of variance assumption. homogeneity of variance assumption was broken for four dependent variables individually: veteran GDS Time 2, F (2, 95) = 3.22, p = .044; veteran FIM Time 1, F (2, 95) = 3.77, p = .027; veteran FIM Time 2, F (2, 95) = 12.93, p < .001; and veteran FIM Time 3, F (2, 95) = 8.14, p = .001. Taken together, these findings suggest that a mor e conservative estimate of F The overall omnibus RMANOVA revealed statistically significant main effects for veteran race/ethnicity, F (2, 95) = 3.67, p = .029, 2

PAGE 44

44 F (2, 94) = 12.58, p < .001, 2 = .211, though the veteran race/ethnicity*time interaction F (4, 190) = .45, p = .772, 2 = .009. These results fail to support Hypothesis 2b, suggesting that there were no racial/ethnic differences in the rate of change in veteran functioning. However, the up Bonferroni corrected t tests within each dependent me asure (GDS, FIM, and FAI) at each point in time to determine the specific locations of racial/ethnic differences in the functioning of veterans with stroke. These t tests showed no statistically significant differences among the GDS scores of White, Latin o/Puerto Rican and Black veterans with stroke at any point in time. However, when examining a visual depiction of GDS means over time by racial/ethnic group (Table 4 2 ), a fairly clear, though statistically non significant, trend emerged with Black vetera ns experiencing lower GDS scores (lower depression) that the other two racial/ethnic groups at each of the three time points. Table 4 2 Veteran s cale A lphas and m eans and s tandard d eviations by r acial/ e thnic group over t ime 1 Month Measure White ( n = 43) Latino/PR ( n = 48) Black ( n = 27) 9.14 (6.75) 10.23 (7.12) 6.59 (4.89) 112.28 (16.53) 107.92 (18.90) a 116.56 (10.48) a 31.02 (10.21) 25.92 (10.92) a 31.74 (9.25) a 6 Month Measure White ( n = 40) Latino/PR ( n = 45) Black ( n = 24) 9.48 (6.61) 10.11 (7.29) 5.83 (4.47) 115.60 (10.90) a 110.58 (17.96) b 121.79 (5.59) a,b 34.95 (10.64) 28.76 (11.65) a 34.83 (10.65) a 12 Month Measure White ( n = 35) Latino/PR ( n = 41) Black ( n = 22) 8.14 (6.03) 8.34 (6.36) 5.82 (4.11) 119.09 (6.92) 115.22 (15.91) 120.33 (10.34) 35.17 (10.01) 29.85 (12.95) 34.55 (10.87) (Note. Means within a row marked by the same superscript letter were sig nificantly different after Bonfer r oni corrections )

PAGE 45

45 This series of Bonferroni corrected t tests uncovered two racial/ethnic differences on FAI scores. At 1 month, Black veterans had higher FAI scores than Latino/Puerto Rican veterans did, t (79) = 2.94, p d = .67, a medium sized effect. This effect was also mirrored at 6 months, t (71) = 2.52, p d = .43, a small sized effect See Table 4 2 for a visual depiction of these trends. The Bonferroni corrected t tests also uncovere d three racial/ethnic differences on FIM scores. At 1 month, Black veterans had higher FIM scores than Latino/Puerto Rican veterans did, t (79) = 2.46, p d = .58, a medium sized effect. This effect was mirrored at 6 monts, t (71) = 3.14, p d = .84, a large sized effect. And Black veterans also had higher 6 month FIM scores than Whites veterans did, t (67) = 2.77, p d = .73, a medium sized effect. See Table 4 2 for a visual depiction of these trends. Hypothesis 3 It was hypothesized that the canonical correlations between caregiver and veteran variable sets at 1 6 and 12 months post discharge would be larger for dyads with a White caregiver than for dyads with a Black or Latino/Puerto Rican caregiver. Three canonical correlation models were run independently for dyads based on caregiver racial/ethnic identification. Each of these three overall models was comprised of three canonical correlation analyses, one at each point in time. In each analysis, the caregi ver variables were GDS, SCQ, and SOC, and the veteran variables were GDS, FIM, and FAI. Though a canonical correlation analysis with three variables in each set produces three canonical correlations, as outlined in the Data Analysis Plan section of this

PAGE 46

46 ma nuscript, only the first canonical correlation will be reported and interpreted here in order to focus on the strongest patterns of connections between the two variable sets. Comparisons of Canonical Correlations by Race/Ethnicity In order to compare by caregiver race/ethnicity the magnitudes of canonical correlations between caregiver variables and veteran variables, Fisher r to z transformations were performed on each of the nine canonical correlations, producing nine z scores. Within each point in tim e, these z scores were used to calculate three absolute value difference z scores between groups of dyads with a White, Latino/Puerto Rican or Black caregiver. This process generated nine comparisons, three at each point in time, each comparison with its own absolute value difference z score and p value. None of these absolute value difference z scores were statistically significant, all z s < 1.35, all p s > .089, suggesting that none of the canonical correlations between caregiver mental health variables a nd veteran functioning variables differed as a function of caregiver race/ethnicity within each point in time. However, because of the correlations may have gone undetected in the Fisher r to z transformation analysis. For reference, overall canonical correlations collapsing across racial/ethnic groups were calculated at each point in time. The first canonical correlation produced from the Time 1 analysis was .34, 11.6% share 2 (9) = 23.76, p = .005; the first canonical correlation produced from the Time 2 analysis was .33, 10.9% shared 2 (9) = 16.80, p = .052; and the first canonical correlation produced from the Time 3 analysis was .3 2 (9) = 11.04, p = .273. A graph of the first canonical correlation within each racial/ethnic group as well as overall across racial/ethnic grou ps over time appears in Figure 4 1

PAGE 47

47 Figure 4 1 Canonical correlations betw een veteran and caregiver variable sets over time by caregiver race/ethnicity Figure 4 1 suggests that in comparison to dyads with a White or Black caregiver, dyads with a Latino/Puerto Rican caregiver exhibited the smallest canonical correlations between caregiver mental health and veteran functioning. Additionally, dyads with a White caregiver exhibited the largest canonical correlations between caregiver mental health and veteran functioning, except at Time 3 when dyads with a Black caregiver exhibited the highest canonical correlation. These findings are consistent with Hypothesis 3, though the tests of Hypothesis 3 failed to reach conventional levels of statistical significance. Further, all within race/ethnicity canonical correlations were larger than the overall canonical correlations collapsing across race/ethnicity. Dyads with a White Caregiver The first canonical correlation model was run for dyads with a White caregiver. Within this model, the first canonical correlation produced from the Time 1 a nalysis was the first canonical correlation produced from the Time 2 analysis was

PAGE 48

48 the first canonical correlation produced fro m the Time 3 analysis was these correlations are considered large standards, though only the first two correlations reached statistical signific ance. These correlations appear in Figure 4 1 In the Time 1 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that caregiver GDS loaded most highly (.84), whereas the other coefficients were below the conve ntional cutoff of .40. For the veteran variables, veteran GDS loaded most highly (.84) and had the only coefficient above .40. This pattern of shared variance suggests that at Time 1 in dyads with a White caregiver, the variables most highly contributing t o the relationship between caregiver mental health and veteran functioning were caregiver and veteran depression. In the Time 2 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that caregiver GDS loaded most highly (1.12), followed by SCQ (.89) and SOC (.93). For the veteran variables, veteran GDS loaded most highly (.83), followed in magnitude by FAI ( .71). This pattern of shared variance suggests that at Time 2 in dyads with a White caregiver, the variable s most highly contributing to the relationship between caregiver mental health and veteran functioning were all three caregiver mental health indices and veteran depression and quality of life. Though the Time 3 canonical correlation did not reach statisti cal significance due to the small sample size in this analysis, a correlation with a magnitude of .52 is still considered large sized (Cohen, 1988). Accordingly, examining the patterns of loadings still can be illuminating, though interpretations must be m ade with caution, given the

PAGE 49

49 statically non significant result. In the Time 3 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that caregiver GDS loaded most highly (.44), followed in magnitude by SOC ( .42). For the veteran variables, veteran GDS loaded most highly (.73), followed by FAI ( .45). This pattern of shared variance suggests that at Time 3 in dyads with a White caregiver, the variables most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver depression and sense of coherence and veteran depression and quality of life. The standardized canonical coefficients for the first canonical correlation at Time 1, 2, and 3 in dyads with a White caregiv er appear in Figure 4 1 Table 4 3 Standardized canonical c oefficients over time in dyads with a White c aregiver Caregiver Variables Time 1 Time 2 Time 3 SOC 0.18 0.93 0.42 SCQ 0.06 0.89 0.34 GDS 0.84 1.12 0.44 Veteran Variables Time 1 Time 2 Time 3 FIM 0.20 0.29 0.16 FAI 0.17 0.71 0.45 GDS 0.84 0.83 0.73 ( Note. Coefficients in bold tended to cluster together consistently over time and will be the focus of discussion ) Dyads with a Latino/Puerto Rican Caregiver The second canonical correlation model was run for dyads with a Latino/Puerto Rican caregiver. Within this model, the first canonical correlation produced from the Time 1 analysis was .42, 18.0 2 (9) = 13.17, p = .155; the first canonical correlation produced from the Time 2 analysis was .45, 20.0% shared 2 (9) = 12.20, p = .202; and the first canonical correlation produced from the Time 3 analysis was 2 (9) = 7.78, p = .557. All of these correlations are considered medium sized effects according to

PAGE 50

50 these correlations are mediu m sized and account for quite a bit of shared variance (all at or above 14.7%), it may be informative to examine the patterns of loadings, albeit cautiously. These correlations are graphically displayed in Figure 4 1 In the Time 1 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that SOC loaded most highly ( 1.19), followed by GDS ( .92). For the veteran variables, FAI loaded most highly (1.15), followed by GDS (.49). This pattern of shared variance suggest s that at Time 1 in dyads with a Latino/Puerto Rican caregiver, the variables most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver sense of coherence and depression and veteran quality of life and depression. In the Time 2 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that caregiver SOC loaded most highly (1.34), followed by GDS (.48). For the veteran variables, FAI had the highest loading ( 1. 23), and none of the other loadings surpassed .40. This pattern of shared variance suggests that at Time 2 in dyads with a Latino/Puerto Rican caregiver, the variables most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver sense of coherence and depression and veteran quality of life. In the Time 3 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that caregiver SOC loaded most highly (1.28), follow ed in magnitude by GDS (.79). For the veteran variables, FAI loaded most highly ( 1.29), followed in magnitude by GDS ( .55) and FIM (.45). This pattern of shared variance suggests that at Time 3 in dyads with a Latino/Puerto Rican caregiver, the variables

PAGE 51

51 most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver sense of coherence and depression and all indices of veteran functioning. The standardized canonical coefficients for the first canonical c orrelation at Time 1, 2, and 3 in dyads with a Latino caregiver appear in Figure 4 1 Table 4 4 Standardized canonical coefficients over time in d yads with a Latino/Puerto Rican c aregiver Caregiver Variables Time 1 Time 2 Time 3 SOC 1.19 1.34 1.28 SCQ 0.25 0.28 0.36 GDS 0.92 0.48 0.79 Veteran Variables Time 1 Time 2 Time 3 FIM 0.16 0.41 0.45 FAI 1.15 1.23 1.29 GDS 0.49 0.09 0.55 ( Note. Coefficients in bold tended to cluster together consistently over time and will be the focus of discussion ) Dyads with a Black Caregiver The third canonical correlation model was run for dyads with a Black caregiver. Within this model, the first canonical correlation produced from the Time 1 analysis was 2 (9) = 8.36, p = .499; the first canonical correlation produced from the Time 2 analysis was 2 (9) = 9.08, p = .430; and the first canonical correlation produced from the Time 3 analysis was .58, 33.1% shared 2 (9) = 8.69, p = .466. The Time 1 correlation is considered a medium sized effect, and the Time 2 and 3 correlations are considered large sized effects (Cohen, 1988), though none reached statistical significance. The patterns of loadin gs within these correlations will be examined and interpreted cautiously. These correlations are graphically depicted in Figure 4 1

PAGE 52

52 In the Time 1 canonical correlation, the standardized canonical coefficients for the caregiver variables showed that GDS l oaded most highly (1.76), followed by SOC (1.67). For the veteran variables, FIM loaded most highly ( .73), followed by FAI ( .57), and GDS ( .53). This pattern of shared variance suggests that at Time 1 in dyads with a Black caregiver, the variables contr ibuting the most to the relationship between caregiver mental health and veteran functioning were caregiver depression and sense of coherence and all the indices of veteran functioning. In the Time 2 canonical correlation, the standardized canonical coeffi cients for the caregiver variables showed that caregiver SCQ loaded most highly ( .97), followed by GDS (.41). For the veteran variables, GDS had the highest loading ( 1.22), followed by FIM ( .80). This pattern of shared variance suggests that at Time 2 i n dyads with a Black caregiver, the variables most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver burden and depression and veteran activities of daily living. In the Time 3 canonical correla tion, the standardized canonical coefficients for the caregiver variables showed that SCQ loaded most highly ( 1.10), with no other coefficients surpassing .40. For the veteran variables, GDS loaded most highly ( .77), followed in magnitude by FAI (.52). T his pattern of shared variance suggests that at Time 3 in dyads with a Black caregiver, the variables most highly contributing to the relationship between caregiver mental health and veteran functioning were caregiver burden and veteran depression and qual ity of life. The standardized canonical coefficients for the first canonical correlation at Time 1, 2, and 3 in dyads with a Black caregiver appear in Table 6.

PAGE 53

53 Table 4 5 Standardized canonical coefficients over time in d yads with a Black c aregiver Caregiv er Variables Time 1 Time 2 Time 3 SOC 1.67 0.31 0.02 SCQ 0.06 0.97 1.10 GDS 1.76 0.41 0.27 Veteran Variables Time 1 Time 2 Time 3 FIM 0.73 0.80 0.16 FAI 0.57 0.02 0.52 GDS 0.53 1.22 0.77 ( Note. No coefficients te nded to cluster together consistently over time, so none were bolded or will be the focus of discussion ) See Table 4 6 for a correlation matrix of all variables from the current study collapsed across racial/ethnic groups.

PAGE 54

54 Table 4 6 C orrelation coeffic ients among all v ariables Vet: FAI 1 Vet: FAI 2 Vet: FAI 3 Vet: FIM 1 Vet: FIM 2 Vet: FIM 3 Vet: GDS 1 Vet: GDS 2 Vet: GDS 3 Cg: GDS 1 Cg: GDS 2 Cg: GDS 3 Cg: SCQ 1 Cg: SCQ 2 Cg: SCQ 3 Cg: SOC 1 Cg: SOC 2 Cg: SOC 3 Vet: FAI 1 1 Vet: FAI 2 .76 ** 1 Vet: FAI 3 .74 ** .83 ** 1 Vet: FIM 1 .56 ** .52 ** .49 ** 1 Vet: FIM 2 .44 ** .57 ** .54 ** .74 ** 1 Vet: FIM 3 .41 ** .46 ** .47 ** .66 ** .82 ** 1 Vet: GDS 1 .30 ** .28 ** .30 ** 28 ** .32 ** .22 1 Vet: GDS 2 .34 ** .38 ** .40 ** .31 ** .46 ** .32 ** .67 ** 1 Vet: GDS 3 .25 .24 .34 ** .25 .33 ** .24 .61 ** .71 ** 1 Cg: GDS 1 .20 .16 .17 .28 ** .24 .00 .19 .23 .18 1 Cg: GDS 2 .12 .13 .18 .28 ** .20 .13 .10 .16 .08 .79 ** 1 Cg: GDS 3 .01 .03 .11 .17 .15 .10 .16 .15 .11 .67 ** .74 ** 1 Cg: SCQ 1 .02 .02 .07 .05 .03 .01 .20 .21 .14 .35 ** .29 ** .24 1 Cg: SCQ 2 .08 .16 .13 .02 .08 .01 .02 .2 5 .10 .31 ** .34 ** .26 .66 ** 1 Cg: SCQ 3 .09 .23 .23 .20 .23 .14 .16 .30 ** .25 .30 ** .30 ** .33 ** .65 ** .75 ** 1 Cg: SOC 1 .03 .05 .07 .12 .07 .07 .19 .16 .10 .72 ** .69 ** .56 ** .50 ** .37 ** .30 ** 1 Cg: SOC 2 .03 .03 .00 .13 .08 .01 .13 .10 .09 .67 ** .73 ** .61 ** .45 ** .46 ** .37 ** .73 ** 1 Cg: SOC 3 .02 .05 .02 .17 .13 .06 .10 .10 .16 .55 ** .57 ** .52 ** .38 ** .48 ** .47 ** .59 ** .692 ** 1 ( Note = p < .05. ** = p < .01 )

PAGE 55

55 CHAPTER 5 DISCUSSION Overview of Findings Out of the three racial/ethnic groups, Latino/Puerto Rican veterans with stroke consistently showed the lowest functioning, and their caregivers showed the poorest mental health, though the mental health effects did not reach statisti cal significance. Consistent patterns emerged over time in the relationships between caregiver mental health and functioning of individuals with stroke, though the patterns differed as a function of race/ethnicity. The following sections discuss the findin gs from this study as relating to each hypothesis. Caregivers: Hypotheses 1a and 2a Hypothesis 1a did not garner statistical support: Black caregivers did not have significantly better mental health than White caregivers, who did not have significantly bet ter mental health than Latino/Puerto Rican caregivers across measures of burden, depression, or coping. In fact, there were no statistically significant racial/ethnic differences among any of these measures at any point in time, even in exploratory analyse s. However, when examining directionality of statistically non significant mean differences, Black caregivers seemed to report lower depression (Geriatric Depression Scale; GDS) than the other two racial/ethnic groups did at each of the three time points. This statistically non significant finding is consistent with a number of previous studies, which have found that African American dementia caregivers exhibit less depression than White dementia caregivers ( Connell & Gibson, 1997; Covinski et al., 2003; Fa rran, Miller, Kaufman, & Davis 1997; Janevic & Connell, 2001). Also, in the current study, a statistically non significant trend emerged, suggesting that Black caregivers may have

PAGE 56

56 had better coping (Sense of Coherence; SOC) at Times 1 and 2 than did the o ther two appeared to fare better than the mental health of members of other groups, though this finding must be interpreted with caution because it was not statistically sig nificant. However, these trends are consistent with the published literature on the mental health of African American caregivers. On the other hand, statistically non significant trends suggested that Latino/Puerto Rican caregivers had the lowest coping (S OC) and the highest burden (Sense of Competence Questionnaire; SCQ) out of the three racial/ethnic groups at Times 2 and 3. These trends are consistent with previous research suggesting that Puerto Rican individuals have often displayed extremely strong lo yalty in providing support for sick or hurt family members (Delgado & Tennstedt, 1997; Snchez Ayendz, 1998; Villarreal, Blozis, & Widamen, 2005; Zea, Quezada, & Belgrave, 1994; Zsembik & Bonilla, 2000) and perhaps as a result, face substantial role engul fment, depression, and personal role strain (Dilworth Anderson, Williams, & Gibson, 2002). As with Hypothesis 1a, Hypothesis 2a did not garner statistical support: The mental health of Black stroke caregivers did not improve at a significantly more rapid rate than that of White caregivers, which did not improve at a significantly more rapid rate than that of Latino/Puerto Rican caregivers There were no statistically significant racial/ethnic differences in the rate of change in any mental health measure o ver time, and even a visual inspection for statistically non significant trends revealed no consistent patterns. Similarly, no statistically significant effect of time emerged, suggesting that caregiver mental health remained constant over time. There was likely

PAGE 57

57 too much variability within racial/ethnic groups over time, even to identify racial/ethnic trends in the current study. Veterans with Stroke: Hypotheses 1b and 2b Hypothesis 1b found partial support: Latino/Puerto Rican veterans with stroke exhibite d the poorest functioning of the three racial/ethnic groups. At Times 1 and 2, Latino/Puerto Rican veterans had lower quality of life (FAI) and activities of daily living (FIM) than Black veterans did. Latino/Puerto Rican were also consistently approximately lower than or equal to that of White veterans, though these differences did not reach statistical significance. In terms of directionality of statistically non significant mean differences, Latino/Puerto Rican veterans consistently reported the highest depression (GDS) out of the three racial/ethnic groups at each point in time. This pattern of significant stroke disparities for Latino individuals conforms to that shown in the literature (Ayala et al., 2001; McGruder, M alarcher, Antoine, Greenlund, & Croft, 2004; Scheinart et al., 1998). However, the functioning of Black veterans with stroke did not conform to Hypothesis 1b or the published literature in that Black veterans exhibited the highest functioning out of the th ree racial/ethnic groups. Black veterans had the lowest depression over time, though again this trend did not reach statistical significance and should be interpreted with caution. This finding may be understood by looking to Knight heories on depression in African Americans, including the Americans may underreport their depression or focus on the positive aspects of stress. Also, Black veterans i n the current study may have exhibited less depression because overall their activities of daily living and quality of life were better than that of

PAGE 58

58 Latino/Puerto Rican veterans generally and better than White veterans on the Time 2 FAI. Interestingly, th is set of findings for Black veterans with stroke runs contrary to much of the previous literature which has found that African Americans with stroke experienced more strokes and recovered more slowly from stroke than White individuals ( Ayala et al., 2001; McGruder, Malarcher, Antoine, Greenlund, & Croft, 2004; Scheinart et al., 1998). One likely interpretation for this finding is that African Americans are more likely to be discharged to nursing homes than are individuals from other racial/ethnic groups (O nukwugha & Mullins, 2007). Because discharge to a nursing home was an exclusion criterion in the current study, the sample may have had an overrepresentation of healthier Black veterans, which also could have accounted for the statistically non significant trend of lower depression levels for Black veterans and caregivers in the sample. Hypothesis 2b was not supported in that the functioning of Black and Latino/Puerto Rican veterans with stroke did not improve more slowly than that of White veterans. There were no statistically significant racial/ethnic differences in the rate of change in any veteran functioning measure over time, and even a visual inspection for statistically non significant trends revealed no consistent patterns. Unlike with caregiver me ntal health variables, a statistically significant main effect of time emerged for veteran functioning variables, suggesting that veteran functioning changed over time. Inspection of the graphs suggests that both veteran FAI and FIM scores increased over t ime, indicating improvement in quality of life and in activities of daily living across all racial/ethnic groups.

PAGE 59

59 Hypothesis 3 Hypothesis 3 was not supported statistically, and t he racial/ethnic differences in the broad correlations between caregiver men tal health and veteran functioning at Times 1, 2, and 3 did not reach statistical significance. However, examining the magnitude of differences in correlations provides some support for Hypothesis 3, though the analyses were underpowered. At Times 1 and 2, the sizes of these correlations were exactly as predicted, as dyads with a White caregiver exhibited the largest associations between caregiver mental health and veteran functioning, followed in magnitude by dyads with a Black caregiver and then dyads wit h a Latino/Puerto Rican caregiver. At Time 3, the pattern in magnitude reversed for dyads with White and Black caregivers, when dyads with a Black caregiver exhibited the largest association, followed by dyads with a White caregiver and then dyads with a L atino/Puerto Rican caregiver. Again, it is important to emphasize that these differences, though fairly consistent over time, were not statistically significant and should be interpreted with caution. Taken together, this pattern of results in the current study suggests that in dyads with a White or Black caregiver, caregiver mental health and veteran functioning were closely intertwined. This same magnitude of association did not emerge in dyads with a Latino/Puerto Rican caregiver. One promising potentia l explanation for this finding is that in the current sample, dyads with a Latino/Puerto Rican caregiver had a larger caregiving network than did dyads with a White or Black caregiver. Hinojosa, Zsembik, and Rittman (2009) have argued that as the number of informal caregivers increases in a family, burden on the primary caregiver may decrease, a viewpoint that may help illuminate the smaller correlations in the current study for dyads with a Latino/Puerto

PAGE 60

60 Rican caregiver. Unfortunately, the current data can not empirically test this possible explanation because the racial/ethnic differences in correlation magnitude were not statistically significant, even before network size is employed as a covariate. As a result, tests of this possible explanation must the focus of future research. The magnitudes of canonical correlations collapsed across caregiver race/ethnicity were consistently smaller and at times much smaller than the canonical correlations within caregiver race/ethnicity. This pattern suggests that dif ferences among racial/ethnic groups served to mask important relationships between caregivers and veterans with stroke within racial/ethnic groups across time. An important implication of this finding is that critical knowledge regarding the connections be tween caregiver mental health and veteran functioning is lost when researchers and clinicians ignore race/ethnicity, instead of looking within and between racial/ethnic groups. A monolithic racial/ethnic model does not seem to hold, and race/ethnicity is l ikely a key variable in how caregiver mental health and veteran functioning reciprocally influence each other. A first step in stroke rehabilitation can involve culturally sensitive mental health interventions for stroke caregivers. These interventions cou ld improve the way that caregivers within each racial/ethnic group provide informal care and ultimately contribute to more effective rehabilitation of veterans with stroke, especially in ethnic minority veterans facing stroke disparities. As a result, it m ay be useful to draw a number of concrete, culturally specific implications for stroke rehabilitation from the current findings and previous research. Implications: Dyads with a White Caregiver In the current study, dyads with a White caregiver exhibited a very strong connection between caregiver mental health and veteran functioning, especially with

PAGE 61

61 depression. A fairly consistent pattern emerged in canonical correlation loadings for dyads with a White caregiver in that across all three time points, caregi ver depression and veteran depression loaded most highly, suggesting that these two variables held the strongest connection as compared to other indices of caregiver mental health and veteran functioning. Depression in White caregivers may elicit depressio n in veterans with stroke or vice versa, slowing rehabilitation. As compared to other racial/ethnic groups, mental health interventions for White caregivers may produce some of the strongest positive effects on informal care, and as a result on stroke reh mental health and the functioning of individuals with stroke. Cognitive behavioral therapy for White caregivers may be especially effective for reducing depression. Also, because White care givers tend to have a small caregiving network, averaging only slightly more than one caregiver, encouraging White caregivers to enlarge their and generally improve informa l care. Implications: Dyads with a Latino/Puerto Rican Caregiver In the current study, dyads with a Latino/Puerto Rican caregiver exhibited a looser connection between caregiver mental health and veteran functioning This could be a result of the fact that in the current sample, dyads with a Latino/Puerto Rican caregiver had larger caregiver networks than did dyads with a Black or White caregiver. Despite this looser connection, as with dyads with a White caregiver, a consistent pattern emerged in canonical correlation loadings for dyads with a Latino/Puerto Rican caregiver in that across all three time points, caregiver coping and depression and veteran quality of life loaded most highly. This consistent pattern suggests that the

PAGE 62

62 component of stroke which m ost heavily affected the mental health of Latino/Puerto Rican caregivers (and vice versa) was reduced functioning of the veteran with stroke in higher order quality of life beyond the necessary skills for basic activities of daily living, such as domestic life, leisure/work, and outdoor activities. As compared to other racial/ethnic groups, mental health interventions for Latino/Puerto Rican caregivers could focus more on role engulfment and personal role strain. In the current sample, Latino/Puerto Rican c aregivers reported poorer mental health than other groups, though the effects fell short of statistical significance. Depression in Latino/Puerto Rican caregivers combined with reduced meaning making abilities may limit quality of care, impacting veteran q uality of life and slowing rehabilitation. Mental health interventions that specifically target the ways in which Latino/Puerto Rican caregivers create meaning out of the caregiving process and interpret the limitations in veteran quality of life may help reduce caregiver depression and improve the quality of the informal care they provide. Implications: Dyads with a Black Caregiver I n the current study, dyads with a Black caregiver exhibited a moderately strong connection between caregiver mental health a nd veteran functioning, though no consistent pattern emerged in canonical correlation loadings over time. As compared to the other racial/ethnic groups, Black caregivers and veterans seemed to fair the best across most indices of caregiver mental health an d stroke functioning. Given previously published research, these findings were likely accurate for caregiver mental health, though they did not reach statistical significance, probably because of the limited number of Black caregivers in the current study. However, these findings were likely inaccurate for veteran functioning because of the probable inadvertent over sampling of

PAGE 63

63 higher functioning Black veterans with stroke in this study that resulted from differential discharge to nursing homes as a functio n of race/ethnicity. This over sampling issue points to the absolute necessity of considering external and systemic third variables when understanding disparities in stroke or when performing rehabilitation counseling with Black caregivers and veterans. I n Black caregivers and veterans, the effects of social and economic forces (such as systemic racism, socio economic status, caregiver health, and access to care) may be especially pronounced and create significant racial/ethnic disparities, for example, in the likelihood of being discharged to a nursing home. In the current study, these disparities and third variables may have obscured a valid interpretation of the connections between caregiver mental health and veteran functioning. This issue also points t o the importance of rehabilitation counseling for Black caregivers incorporating a systems perspective and helping caregivers examine the impact of various social forces on their mental health and on the care they provide. General Implications Stroke reha bilitation interventions often target only veterans with stroke, rather than the caregiver/veteran dyad, approaching rehabilitation perhaps with only half of the potential power and effectiveness that they could. Findings from previous studies (e.g., Perri n et al., 2008; Perrin et al., 2009) and from the current study have identified very close links between the mental health of stroke caregivers and the functioning of veterans with stroke, suggesting that a first step in stroke rehabilitation could involve attending to the needs of the primary stroke caregiver. Early psychosocial interventions, especially for at risk stroke caregivers, potentially could be less expensive and more

PAGE 64

64 effective than dealing with complications resulting from inadequate, ineffecti ve, or diminished informal caregiving. The findings from the current study help illuminate the relationships between the mental health of stroke caregivers and the functioning of veterans with stroke both within and between racial/ethnic groups. These con nections exist differentially for dyads with Black, Latino/Puerto Rican, or White caregivers, and as a result, the findings from this study may be a first step in creating culturally sensitive mental health interventions for caregivers of veterans with str oke. Because major disparities exist in the prevalence and course of stroke among African American, Latino, and White veterans, it is of extreme importance to understand the patterns of relationships among caregiver mental health and the functioning of vet erans with stroke differentially within each ethnic group. The functioning of veterans with stroke is intimately linked to the mental health of stroke caregivers, so culturally sensitive interventions for caregivers could improve the way that caregivers w ithin each ethnic group provide informal care and ultimately contribute to more effective rehabilitation of veterans with stroke, especially ethnic minority veterans. Specifically, findings from the current study are consistent with the use of several form s of psychosocial interventions for stroke caregivers, and the treatment modality should depend on the specific needs of the caregiver/care recipient dyad. Visser Meily et al. (2005) reviewed a number of studies on interventions for stroke caregivers and f ound four primary categories of interventions: specialist services, psycho education, counseling, and social support by peers. Visser Meily et al. (2005) suggested that of the four predominant types of interventions, counseling programs had

PAGE 65

65 the most positi ve outcomes, especially those involving active problem solving strategies. For example, Grant, Elliott, Weaver, Bartolucci, and Newman Giger (2002) gave stroke caregivers a 3 hour social problem solving training program two days before hospital discharge, followed by scheduled telephone contacts. The caregivers in the intervention group had several positive treatment outcomes relative to controls: improved problem solving skills, greater caregiver preparedness, less depression, improved vitality, better soc ial functioning, and higher mental health, though there were no differences in burden. mind, o ne potential intervention could involve cognitive behavioral counseling for ca regivers burden and depression and leading to improved informal care and better stroke recovery of individuals with stroke. Cognitive behavioral counseling is well suited to help s troke caregivers who provide care for low functioning veterans improve their sense of coherence by refining meaning making strategies to ward off burden and depression. Good caregiver mental health in turn could lead to better informal caregiving and ultim ately to more effective stroke rehabilitation, based on the causal feedback loop theory articulated by Perrin et al. (2008). In addition to providing psychotherapeutic interventions, creating interventions that increase knowledge of stroke caregivers about their vulnerability to depression and to other reductions in their mental health could facilitate self care and appropriate treatment seeking. Many stroke caregivers have reported feeling inadequately informed because they lack basic information about str oke, strategies to care for individuals with

PAGE 66

66 Rodgers, Bond, & Curless, 2001; Smith, Kerr, Langhorne, & Lees, 2004; Wiles, Pain, Buckland, & McLellan 1998). Stroke careg ivers have indicated interest in knowing about the psychological sequelae and prognoses of stroke ( Hanger, Walker, Paterson, McBride, & Sainsbury, 1998) and have reported the desire to acquire skills to assist in daily care, mobility, and toileting, as wel l as the desire to assist in taking blood pressure and conducting glucometer checks (Hartke & King, 2003; Wiles et al., 1998). These interests are even keener in some racial/ethnic groups than others. Puerto Rican stroke caregivers living in Puerto Rico re ported a greater need for health information about stroke compared with those living on the U.S. Mainland (Hinojosa & Rittman, 2007). An intervention that provides caregivers with health education on maintaining and improving their mental health may be hel pful and especially desired by caregivers from specific racial/ethnic groups. Limitations and Future Directions A primary limitation to the current study is that the majority ( 87.1%) of stroke caregivers were women, and the majority ( 98.4%) of veterans wit h stroke were men. The patterns of connections between caregiver mental health and functioning of individuals with stroke found in the current study may not generalize to populations with more gender balance, because male and female gender roles may have d ifferent effects on mental health and on stroke functioning. It is notable that Tiegs et al. (2006) found no significant differences in the functioning of individuals with stroke based on the gender of the caregiver, but female caregivers in their study ex hibited higher levels of depression, burden, and consequences to their personal life than did male caregivers. As a result, the patterns of connections in the current study should be applied with

PAGE 67

67 caution to dyads with a male caregiver and/or a woman with s troke, and future studies should examine these connections in samples with more gender diversity. A second limitation is the confound between race/ethnicity and country of origin in the current study. The sample contained Latino/Puerto Rican caregivers and veterans with stroke who were almost exclusively from Puerto Rico, but the White and Black caregivers and veterans with stroke were exclusively from the U.S. mainland. This meant that race/ethnicity was confounded with living in Puerto Rico or in the U.S. mainland. In this study differences between Latino/Puerto Ricans and other groups may be the result of race/ethnicity or the result of Puerto Rico vs. U.S. mainland, but the two potential sources cannot be teased apart. In addition, the findings in this s tudy may not generalize to Puerto Rican dyads from the U.S. mainland, where different levels of acculturation to American culture exist. So future studies need to examine whether the same findings hold in Puerto Rican dyads in the U.S. mainland and whether the findings vary based on the acculturation level of individuals from those dyads. A third limitation to the current study involves the lack of statistical significance in many of the reported effects. Because this has to do with the limited sample in th e current study, the presentation of results and discussion was framed with a heavy emphasis on effect sizes and practical significance, given previous research findings, as opposed to an emphasis on statistical significance. For example, the power analysi s conducted for the omnibus RMANOVAs suggested that this sample would only detect large sized and some medium sized effects, but no small sized effects. Because of this lack of statistical significance for many of the effects discussed, the findings again must be interpreted with an appropriate degree of caution. Future studies should use larger

PAGE 68

68 sample sizes to determine whether the effects found herein are reliable in other samples and reach statistical significance when analyses have more statistical powe r. A fourth limitation involves the underlying theoretical framework for this study, which is that a feedback loop exists among caregiver mental health, quality of informal care, and stroke rehabilitation. Though this loop has received some tentative supp ort in previous research (e.g., Perrin et al., 2009), the current study still used a correlational methodology, as opposed to an experimental one. Accordingly, futures studies should implement experimental techniques to determine whether culturally sensiti ve caregiver mental health interventions improve quality of care and, as a result, improve stroke rehabilitation in populations facing health disparities in the prevalence and course of stroke. This type of research is vital and warranted given the finding s from the current study and from previous studies (e.g. Perrin et al., 2008, 2009). Conclusions Perhaps the most important implication of this study is that viewing stroke rehabilitation through a color blind lens may mask important components of rehabili tation, because race/ethnicity and culture likely influence the social structures in which individuals with stroke receive care and function. Though caregiver mental health interventions hold promise for improving quality of care and, as a result, improvin g stroke rehabilitation, including an appreciation of diversity in interventions may produce better outcomes. Critical knowledge is perhaps lost regarding the connections between caregiver mental health and the functioning of individuals with stroke when r esearchers and clinicians look only across caregiver race/ethnicity as opposed to also looking within race/ethnicity. To avoid this loss knowledge, future research on clinical interventions should incorporate a greater degree of cultural competence, which if found to be

PAGE 69

69 important in follow up studies, could contribute to reducing the racial/ethnic disparities in the prevalence and course of stroke. S uch assertions await further support from future research.

PAGE 70

70 APPENDIX A GERIATRIC DEPRESSION SCALE Choose the best answer for how you felt over the past week (yes/no) 1. Are you basically satisfied with your life? 2. Have you dropped many of your activities and interests? 3. Do you feel that your life is empty? 4. Do you often get bored? 5. Are you hopeful about the future? 7. Are you in good spirits most of the time? 8. Are you afraid that something bad is going to happen to you? 9. Do you feel happy most of the time? 10. Do you often feel helpless? 11. Do you often get restless and fidgety? 12. Do you prefer to stay at home, rather than going out and doing new things? 13. Do you frequently worry about the future? 14. Do you feel you have more problems with memory than most? 15. Do you think it is wo nderful to be alive now? 16. Do you often feel downhearted and blue? 17. Do you feel pretty worthless the way you are now? 18. Do you worry a lot about the past? 19. Do you find life very exciting? 20. Is it hard for you to get started on new projects? 2 1. Do you feel full of energy? 22. Do you feel that your situation is hopeless? 23. Do you think that most people are better off than you are? 24. Do you frequently get upset over little things? 25. Do you frequently feel like crying? 26. Do you have trou ble concentrating? 27. Do you enjoy getting up in the morning? 28. Do you prefer to avoid social gatherings? 29. Is it easy for you to make decisions? 30. Is your mind as clear as it used to be?

PAGE 71

71 APPENDIX B SENSE OF COMPETENCE QUESTIONNAIRE Response cate gories: 1, disagree very much; 2, disagree; 3, agree; 4, agree very much. responsibilities for my family, my job, etc. have. 7. I feel guilty about my i im/her. needs. ugh time for myself. 20. I feel that my hea d a better relationship. he/she could depend on. he does to annoy me.

PAGE 72

72 APPENDIX C SENSE OF COHERENCE Response keys for all questions except Items 4 and 7 ranged from Very often to Never. For Items 4 and 7, response keys ranged from Like around you? 2. How often has it happened in the past that you were surprised by the behavior of people who you thought you knew well? 3. How often has it happened that people whom you counted on disappointed you? 4. How do you think you are going to feel about the things you will do in the future? 5. How often do you have the feeling that you are being treated unfairly? 6. How often do you have know what to do? 7. How do you feel about the things you do every day? ideas? 9. How often does it happen that you have feelings inside that you would rather not feel? 10. Many people even those with a strong character sometimes feel like losers in certain situations. How often have you felt this way in the past? 11. How often does it happen that you have the f 12. How often do you have the feeling that there is little meaning in the things you do in your daily life?

PAGE 73

73 AP PENDIX D FUNCTIONAL INDEPENDE NCE MEASURE Response categories: 7 point Likert scale indicating the amount of assistance required Self care 1. Eating 2. Grooming 3. Bathing 4. Dressing upper body 5. Dressing lower body 6. Toileting Sphincter control 7. Bladder management 8. Bowel management Transfers 9. Bed/chair/wheelchair 10. Toilet 11. Tub/shower Locomotion 12. Walk/wheelchair 13. Stair s Communication 14. Comprehension 15. Expression Social cognition 16. Social interaction 17. Problem solving 18. Memory

PAGE 74

74 APPENDIX E FRENCHAY ACTIVITIES INDEX Item 1, 2: 0 = Never, 1 = Under once weekly, 2 = 1 2 times a week, 3 = Most days Item 3 10: 0 = Never, 1 = 1 2 times in 3 months, 2 = 3 12 times in 3 months, 3 = At least weekly Item 11: 0 = Never, 1 = 1 2 times in 6 months, 2 = 3 12 times 3 months, 3 = At least weekly Item 12, 13: 0 = Never, 1 = Light, 2 = Moderate, 3 = All necessary Item 14: 0 = None, 1 = 1 in 6 months, 2 = Less than 1 a fortnight, 3 = Over 1 a fortnight Item 15: 0 = None, 1 = Up to 10 hrs/week, 2 = 10 30 hrs/week, 3 = Over 30 hrs/week 1. Preparing main meals 2. Washing clothes 3. Light housework 4. Heavy housework 5. Local shopp ing 6. Social outings 7. Walking outside more than 15 minutes 8. Actively pursuing hobby 9. Driving a care/going on a bus 10. In the last six months how often have you undertaken: 11. Travel outings / car rides 12. Gardening 13. Household / car maintenance 14. Reading books 15. Gainful work

PAGE 75

75 LIST OF REFERENCES American Heart Association (2007). Heart disease and stroke statistics 2007 update: A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation, 115 e69 e171. Anderson C. S., Linto, J., & Stewart Wynne E. G. (1995). A population based assessment of the impact and burden of caregiving for long term stroke survivors. Stroke, 26 843 849. Aneshensel, C. S., Pearlin, L. I., Mullan, J. T., Zarit, S. H., & Whitlatch, C. J. (1995). Profiles in caregiving: the unexpected career San Diego: Academic Press. Antonovsky, A. (1987). Unraveling the mystery of health. How people manage stress and stay well San Francisco: Jossey Bass. Arno, P. S., Levine, C. & Memmott, M. M. (1999). The economic value of informal caregiving. Health Affairs, 18 182 188. Ayala, C., Greenlund, K. J., Croft, J. B., Keenan, N. L., Donehoo, R. S., Giles, W. H., Kittner, S. J., & Marks, J. S. (2001). Racial/ethnic disparities in mortality by stroke subtype in the United States, 1995 1998. American Journal of Epidemiology, 154 1057 1063. Baumgarten, M., Battista, R. N., Infante Rivard, C., Hanley, J. A., Becker, R., & Gauthier, S. (1992). The psychological and physical health of family members caring for an elderly person with dementia. Journal of Clinical Epidemiology, 45 61 70. Bergner, M., Bobbitt, R. A., Carter, W. B., & Gilson, B. S. (1981). The Sickness Impact Profile: development and final revision of a health status meas ure. Medical Care, 19 787 805. Blonder, L. X., Langer, S. L., Pettigrew, L. C., & Garrity, T. F. (2007). The effects of stroke disability on spousal caregivers. Neurorehabilitation, 22 85 92. Borhani, N. O. (1965). Changes and geographic distribution o f mortality from cerebrovascular disease. American Journal of Public Health, 55 673 681. Buck, D., Gregson, B. A., Bamford, C. H., McNamee, P., Farrow, G. N., Bond, J., & Wright, K. (1997). Psychological distress among informal supporters of frail older people at home and in institutions. The Resource Implications Study Group of the MRC Cognitive Function and Ageing Study. International Journal of Geriatric Psychiatry, 12 737 744. Bugge, C., Alexander, H., & Hagen, S. (1999). caregivers. Patient, caregiver, and service factors that affect caregiver strain. Stroke, 30 1517 1523.

PAGE 76

76 Carnwath, T. C., & Johnson, D. A. (1987). Psychiatric morbidity among spouses of patients with stroke. British Medical Journal (Clinical Research Ed.) 294, 409 411. Chumbler, N. R., Rittman, M., Van Puymbroeck, M., Vogel, W. B., & Qin, H. (2004). The sense of coherence, burden, and depressive symptoms in informal caregivers during the first month after stroke. International Journal of Geriatric Psychi atry, 19 944 953. Cohen J. ( 1988 ). Statistical power analysis for the behavioral sciences (2 nd ed.). Hillsdale, NJ: Erlbaum. Connell, C. M., & Gibson, G. D. (1997). Racial, ethnic, and cultural differences in dementia caregiving: Review and analysis. T he Gerontologist, 37, 355 364. Corrigan, J. D., Smith Knapp, K., Granger, C. (1997). Validity of the Functional Independence Measure for persons with traumatic brain injury. Archives of Physical Medicine Rehabilitation, 78 828 834. Covinsky, K. E., Newc omer, R., Fox, P., Wood, J., Sands, L., Dane., K., Yaffe, K. (2003). Patient and caregiver characteristics associated with depression in caregivers of patients with dementia. Journal of General Internal Medicine, 18 1006 1114. Cox, C. (1993). Service nee ds and interests: A comparison of African American and and Related Disorders & Research 8 33 40. Cushman, M., Cantrell, R. A., McClure, L. A., Howard, G., Prineas, R. J., Moy, C. S., Temple, E. M., & Howard, V. J. (2008). Estimated 10 year stroke risk by region and race in the United States. Annals of Neurology, 64 507 513. study of the emotional outcome of people caring for stroke survivors. Stroke, 29 1867 1872. Delgado, M., & Tennstedt, S. L. (1997). Puerto Rican sons as primary caregivers of elderly parents. Social Work 42 125 134. Dilworth Anderson, P., Williams, I. C., & G ibson, B. E. (2002). Issues of race, ethnicity, and culture in caregiving research: A 20 year review. The Gerontologist 42, 237 272. Dodds, T. A., Martin, D. P., Stolov, W. C., & Deyo, R. A. (1993). A validation of the Functional Indpendence Measure and its performance among rehabilitation inpatients. Archives of Physical Medicine Rehabilitation, 74 531 536. Dorsey, M. K., & Vaca, K. J. (1998). The stroke patient and assessment of caregiver needs. Journal of Vascular Nursing, 16 62.

PAGE 77

77 Doty, P. (1986). Family care of the elderly: The role of public policy. The Milbank Quarterly, 64 34 75. Evans, R. L., Bishop, D. S., & Haselkorn, J. K. (1991). Factors predicting satisfactory home care after stroke. Archives of Physical Medicine and Rehabilitation, 72 144 147. Evans, R. L., Hendricks, R. D., Haselkorn, J. K., Bishop, D.S., & Baldwin, D. (1992). American Journal of Physical Medicine & Rehabilitation, 71 135 139. Farran, C. J., Mi ller, B. H., Kaufman, J. E., & Davis, L. (1997). Race, finding meaning, and caregiver distress. Journal of Ageing and Health, 9 316 333 prevention in social work services International Social Work, 41 511 522. Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Mini Mental State. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12 189 98. Granger C. V., & Hamilton, B. B. (1990). UDS Report: The uniform data system for medical rehabilitation report on first admissions for 1990. American Journal of Physical Medicine Rehabilitation, 71 108 113. Grant, J. S. (1999). Social problem solving partnersh ips with family caregivers. Rehabilitation Nursing, 24 254 260. Grant, J. S., Elliott, T. R., Weaver, M., Bartolucci, A. A., & Giger, J. N. (2002). Telephone intervention with family caregivers of stroke survivors after rehabilitation. Stroke, 33 2060 20 65. Grant J. S., Glandon, G. L., Elliott, T., Giger, J. N., & Weaver, M. (2004). Caregiving problems and feelings experienced by family caregivers of stroke survivors the first month post discharge. International Journal of Rehabilitation Research, 27 10 5 111. Grant, J. S., Weaver, M., Elliot, T. R., Bartolucci, A. R., & Giger, J. N. (2004). Sociodemographic, physical, and psychosocial fact ors associated with depressive behavior in family caregivers of stroke survivors in the acute care phase. Brain Inju ry 18 797 809. Hanger, H. C., Walker, G., Paterson, L. A., McBride, S., & Sainsbury, R. (1998). What do patients and their carers want to know about stroke? A two year follow up study. Clinical Rehabilitation, 12 42 54.

PAGE 78

78 Hamilton, B. B., Laughlin, J. A ., Fiedler, R. C., & Granger, C. V. (1994). Interrater reliability of the 7 Level Functional Independence Measure (FIM). Scandinavian Journal of Rehabilitation Medicine, 26 115 119. Hartke, R. J., & King, R. B. (2003). Telephone group intervention for ol der stroke caregivers. Topics in Stroke Rehabilitation, 9 65 81. Hinojosa, M. S., & Rittman M. R. (2007). Informational needs of caregivers of veterans post stroke: A comparison of mainland and Puerto Rican caregivers. Journal of Rehabilitation Research and Development, 44 649 658. Hinojosa, M. S., Zsembik, B., & Rittman, M. R. (2009). Patterns of informal care among Puerto Rican, African American, and white stroke survivors. Ethnicity and Health, 14, 591 606. Hirst, M. (2001). Trends in informal care in Great Britain during the 1990s. Health & Social Care in the Community, 9 348 357. Holbrook, M. S., & Skilbeck, C. E. (1983). An activities index for use with stroke patients. Age and Ageing, 12 166 170. Howard, G., Labarthe, D. R., Hu, J., Yoon, S., & Howard, V. J. (2007). Regional for southern African Americans. Annals of Epidemiology, 17 689 696. Janevic, M. R., & Connell, C. M. (2001). Racial, ethnic, and cul tural differences in the dementia caregiving experience. The Gerontologist 41 334 347. Kinney, J. M., Stephens, P., Ann, M., Franks, M. M., & Norris, V. K. (1995). Stresses and satisfactions of family caregivers to older stroke patients. Journal of Appl ied Gerontology, 14 3 20. Knight, B. G., & McCallum, T. J. (1998). Heart rate reactivity in African American and white dementia caregivers: Report bias or positive coping? Aging and Mental Health, 2 212 221. Knight, B. G., Silverstein, M., McCallum, T. J., & Fox, L. S. (2000). A sociocultural stress and coping model for mental health outcomes among African American caregivers in Southern California. The Journals of Gerontology Series B: Psychological Sciences and Social Science, 55 142 150. Kotila, M. Numminen, H., Waltimo, O., & Kaste, M. (1998). Depression after stroke: Results of the FINNSTROKE study. Stroke, 29 368 372.

PAGE 79

79 Lawton M. P., Rejagopal, D., Brody, E., & Kleban, M. H. (1992). The dynamics of caregiving for a demented elder among black and white families. Journals of Gerontology, Social Sciences, 47 s156 s164. Mahoney, F. I., & Barthel, D. W. (1965). Functional evaluation: The Barthel Index: A simple index of independence useful in scoring improvement in the rehabilitation of the chronica lly ill. Rehabilitation, 4 61 65. Marn G., & Marn B.V. (1991). Research with Hispanic Populations Newbury Park CA: Sage. Mausbach, B. T., Coon, D. W., Depp, C., Rabinowitz, Y. G., Wilson Arias, E., Kraemer, H. C., Thompson, L. W., Lane, G., & Gallagh er Thompson, D. (2004). Ethnicity and time to institutionalization of dementia patients: A comparison of Latina and Caucasian female family caregivers. Journal of the American Geriatrics Society 52 1077 1084. McGruder, H. F., Malarcher, A. M., Antoine, T. L, Greenlund, K. J., & Croft, J. B. (2004). Racial and ethnic disparities in cardiovascular risk factors among stroke survivors. Stroke, 35 1557 1561. National Alliance for Caregiving/American Association of Retired Persons. (1997). Family caregiving in the U.S.: Findings from a national survey Washington, DC. National Heart, Lung, and Blood Institute. (2004). Report of the Working Group on Research in Coronary Heart Disease in Blacks Bethesda, MD. National Institutes of Health. (1995). Community b ased health care nursing strategies: A report of the NINR Priority Expert Panel Bethesda, MD: National Institute of Nursing Research. from the field. International Jou rnal of Nursing Practice, 10 121 126. Onukwugha, E., & Mullins, C. D. (2007). Racial differences in hospital discharge dispositions among stroke patients in Maryland. Medical Decision Making, 27 233 242. Ottenbacher, K. J., Hsu, Y., Grander, C. V., & F iedler, R. C. (1996). The reliability of the Functional Independence Measure: A quantitative review. Archives of Physical Medicine Rehabilitation, 77 1226 1132. Periard, M. E., & Ames, B. D. (1993). Lifestyle changes and coping patterns among caregivers of stroke survivors. Public Health Nursing, 10 252 256.

PAGE 80

80 Perrin, P. B., Heesacker, M., Stidham, B. S., Rittman, M. R., & Gonzalez Rothi, L. J. (2008). Structural equation modeling of the relationship between caregiver psychosocial variables and functio ning of individuals with stroke. Rehabilitation Psychology, 53 54 62. Perrin, P. B., Heesacker, M., Sberna Hinojosa, M., Uthe, C. E., & Rittman, M. R. (2009). Identifying at risk, ethnically diverse stroke caregivers for counseling: A longitudinal study of mental health. Rehabilitation Psychology, 54 138 149. Rodgers, H., Bond, S., & Curless, R. (2001). Inadequacies in the provision of information to stroke patients and their families. Age and Ageing, 30 129 133. Rombough, R. E., Howse, E. L., Bagg, S D., & Bartfay, W. J. (2007). A comparison of studies on the quality of life of primary caregivers of stroke survivors: A systematic review of the literature. Topics in Stroke Rehabilitation, 14 69 79. Snchez Ayendez, M. (1998). Middle aged Puerto Rica n women as primary caregivers to the elderly: a qualitative analysis of everyday dynamics. In M. Delgado (ed.), Latino elders and the twenty first century: Issues and challenges for culturally competent research and practice (pp. 75 98). New York: Haworth. Scheinart, K. F., Tuhrim, S., Horowitz, D. R., Weinberger, J., Goldman, M., & Godbold, J. H. (1998). Stroke recurrence is more frequent in blacks and Hispanics. Neuroepidemiology, 17 188 198. Schnyder, U., Buchi, S., Sensky, T., & Klaghofer, R. (2000). Coherence: Trait or state? Psychotherapy and Psychosomatics, 69 296 302. Scholte op Reimer, W. J., de Haan, R. J., Pijnenborg, J. M., Limburg, M., & van den Bos, G. A. (1998). Assessment of burden in partners of stroke patients wit h the Sense of Competence questionnaire. Stroke, 29 373 379. Psychiatric and physical morbidity effects of dementia caregiving: Prevalence, correlates, and causes. Gerontologist, 35 771 7 91. Schuling, J., de Haan, R., Limburg, M., & Groenier, K. H. (1993). The Frenchay Activities Index. Assessment of functional status in stroke patients. Stroke, 24 1173 1177. Silliman, R. A., Fletcher, R. H., Earp, J. L., & Wagner, E. H. (1986). Familie s of elderly stroke patients. Effects of home care. Journal of the American Geriatric Society, 34 643 648. experience of caring for stroke survivors, Journal of Advanced Nur sing, 46 235 344.

PAGE 81

81 Stein, P. N., Gordon, W. A., Hibbard, M. R., & Sliwinski, M. J. (1992). An examination of depression in the spouses of stroke patients. Rehabilitation Psychology, 37 121 130. Stineman, M. G., Shea, J. A., Jett, A., Tassoni, C. J., Ott enbacher, K. J., Fiedler, R., Granger, C. V. (1996). The Functional Independence Measure: Tests of scaling assumptions, structure, and reliability across 20 diverse impairment categories. Archives of Physical Medicine Rehabilitation, 77 1101 1108. depression and support as predictors of post stroke depression: a cross sectional survey. International Journal of Nursing Studies, 42 611 618. Thommessen, B., Aarsland, D., Braekhus, A., Oksengaard, A. R., Engedal, K., & Laake, K. (2002). The psychosocial burden on spouses of the elderly with stroke, dementia and International Journal of Geriatric Psychiatry, 17 78 84. Tiegs, T. J., Heesacker, M., K etterson, T. U., Pekich, D. G., Rittman, M. R., Rosenbek, J. C., Stidham, B. S., & Gonzalez Rothi, L. J. (2006). Coping by stroke caregivers: Sex similarities and differences. Topics in Stroke Rehabilitation, 13 52 62. Tooth, L., McKenna, K., Barnett, A., Prescott, C., & Murphy, S. (2005). Caregiver burden, time spent caring and health status in the first 12 months following stroke. Brain Injury, 19 963 974. Triandis, H. C. (1995). Individualism and Collectivism Boulder, CO: Westview Press. Van Puym broeck, M., & Rittman, M. R. (2005). Quality of life predictors for caregivers at 1 and 6 months poststroke: Results of path analyses. Journal of Rehabilitation Research & Development, 42 747 760. Veterans Health Administration. (2003). care enrollment and expenditure projections Office of Policy and Planning, Editor. Villarreal, R., Blozis, S. A., & Widaman, K. F. (2005). Factorial invariance of a pan Hispanic familism scale. Hispanic Journal of Behavioral Sciences 27, 409 425. Visse r Meily, A., van Heugten, C., Post, M., Schepers, V., & Lindeman E. (2005). Intervention studies for caregivers of stroke survivors: A critical review. Patient Education and Counseling, 56 257 267. Vernooij Dassen, M. J. F. J., Persoon, J. M. G., & Felli ng, A. J. A. (1996). Predictors of sense of competence in caregivers of demented persons. Social Science Medicine, 43 41 49.

PAGE 82

82 Weston, R., Gore, P. A., Chan, F., & Catalano, D. (2008). An introduction to using structural equation models in rehabilitation p sychology. Rehabilitation Psychology, 53 340 356. White, C. L., Mayo, N., Hanley, J. A., & Wood Dauphinee, S. (2003). Evolution of the caregiving experience in the initial 2 years following stroke. Research in Nursing and Health, 26 177 189. Wiles, R. Pain, H., Buckland, S., & McLellan, L. (1998). Providing appropriate information to patients and carers following a stroke. Journal of Advanced Nursing, 28 794 801. Williams, I. C. (2005). Emotional health of black and white dementia caregivers: A cont extual examination. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 60 287 295. Wright, L. K, Hickey, J. V., Buckwalter, K. C., Hendrix, S. A., & Kelechi, T. (1999). Emotional and physical health of spouse caregivers of persons with and stroke. Journal of Advanced Nursing, 30 553 563. Yesavage, J. A., & Brink, T. L. (1983). Development and validation of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research, 17 37 49. Zea, M. C., Quezada, T., & Belgrave, F. (1994). Latino culture values: Their role in adjustment to disability. Journal of Social Behavior, 9, 1 16. Zsembik, B. A., & Bonilla, Z. (2000). Eldercare and the changing family in Puerto Rico. Journal o f Family Issues 21, 652 674.

PAGE 83

83 BIOGRAPHICAL SKETCH Paul B. Perrin was a student at the University of Florida since August of 2001. In 2005, he completed a Bachelor of Arts in English and a Bachelor of Science in p sychology. T hat same year, he s tarted hi s doctoral studies in counseling p sychology at the University of Florida He earned a Master of Science in p sychology in 2007 and then completed his doc toral qualifying exams in 2008. He d e fended his dissertation in 2010 and complete d a clinical psychology internship at the Baltimore Maryland Veterans Affairs Medical Center and the University of Maryland School of Medicine. He received his Doctorate of Philosophy in counseling p sychology in 2011