USF Libraries
USF Digital Collections

The relationship between caregiving and bereavement

MISSING IMAGE

Material Information

Title:
The relationship between caregiving and bereavement a series of three studies
Physical Description:
Book
Language:
English
Creator:
Burton, Allison
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Caregiving
Bereavement
Social support
Depression
Stressors
Well-being
Dissertations, Academic -- Aging Studies -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Spousal bereavement has been consistently demonstrated in the literature to be one of the most highly stressful experiences in an individuals lifetime. In addition many deaths in the United States are preceded by a period of caregiving, which is also believed to be highly stressful and have a profound impact on bereavement. However the literature has been inconsistent as to the exact nature of the relationship between caregiving and bereavement and there has been some debate as to whether or not positive and negative affect variables are mutually exclusive. This dissertation sought to further address the issue of the relationship between caregiving and the bereavement experience through a series of three studies which utilized information from two datasets. The first was the Changing Lives of Older Couples (CLOC) study, a project that included both pre- and post-loss data.The second was a subset of a dataset that recruited elderly spousal caregivers of terminally ill patients from a large, local non-profit hospice. This dataset included both pre- and post-loss data and included a variety of widely used and validated measures that allowed for the examination of caregiving stressors, appraisals, and social support and their effect on both positive and negative mental health outcomes during bereavement.The first study examined the impact of caregiving on well-being during bereavement, specifically within the domains of psychological, social, and physical health, utilizing both positive and negative affect measures.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Allison Burton.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 106 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001670318
oclc - 62248584
usfldc doi - E14-SFE0001178
usfldc handle - e14.1178
System ID:
SFS0025499:00001


This item is only available as the following downloads:


Full Text

PAGE 1

The Relationship between Caregiving and Be reavement: A Series of Three Studies by Allison M. Burton A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Aging Studies College of Arts and Sciences University of South Florida Major Professor: William E. Haley, Ph.D. Paul B. Jacobsen, Ph.D. Susan C. McMillan, Ph.D. Sandra L. Reynolds, Ph.D. Brent J. Small, Ph.D. Date of Approval: July 18, 2005 Keywords: caregiving, bereavement, social support, depression, st ressors, well-being Copyright 2005 Allison M. Burton

PAGE 2

Dedication This dissertation is dedicated to my Gra nny, Dorothy B. Glover, without whom I could never have gotten to this point in my life. You have been so many things to me: an inspiration, a pillar of suppor t (in countless ways), a role model, and a friend (just to name a few). I have always fe lt like one of the luckiest wo men in the world to have you on my side and cannot thank you enough for all of your love and support. You are an ever-present guiding spirit in my life. I have always believed that I was more like you than anyone else I have ever known, and that idea has helped me to be proud of who I am, to stand up for what I believe in, and to cherish my independence. There were countless times over the years that I ques tioned my decisions and wondered if my priorities were misplaced, and you were alwa ys there exactly when I needed you to put things into perspective and to keep me on the right path. Thank you for always being there for me and for contributing to my life in countless ways. I love you more than I could ever express in words and I sincerel y hope that you are proud of the woman I have become because so much of who I am has been shaped by you.

PAGE 3

i Table of Contents List of Tables…………………………………………………….……………...………..iii List of Figures………………………………………………………………………….…iv Abstract…………………………………………………………………………..………..v Preface…………………………………………………………………………………...vii Chapter One……………………………………………………………………………….1 Introduction………………………………………………………………..............1 Chapter Two……………………………………………………………………...………..6 Abstract……………………………………………………………………………6 Introduction……………………………………………………………….... ……..7 Methods……………………………………………………………………. ……12 Participants………………………………………………….……… ……12 Measures…………………………………………………………… ……14 Pre-loss Caregiving Status……………………………………….14 Demographics and Marital Quality……………………………....15 Well-Being Measures………………………………………. ……16 Psychological………………………………………. ……16 Social Engagement………………………………….……17 Physical Health…………………………………….. ……18 Statistical Analysis Plan……………………………………………. ……18 Results……………………………………………………………………… ……19 Basic Descriptive Analyses………………………………………… ……20 Mixed Model Analyses………………………….……………………….20 Psychological…………………………………………………….20 Social Engagement……...…………………………………..……21 Physical Health……………………………………….……. ……23 Discussion…………………………………………………………………..........25 Chapter Three…………………………………………………………………….....……28 Abstract…………………………………………………………………………..28 Introduction………………………………………………………………………29 Methods………………………………………………………………….… ……31 Participants……………………………………………………………….31

PAGE 4

ii Measures………………………………………………………….... ……34 Pre-loss Caregiving Status…………………...…………….. ……34 Bereavement Trajectories………………………………….. ……35 Statistical Analysis Plan………………………………………….… ……36 Results……………………………………………………………………………37 Discussion………………………………………………………………..………39 Chapter Four.……………………………………………………………………..... ……45 Abstract…………………………………………………………………………..45 Introduction………………………………………………………………………46 Stress Process Models and th eir Application in Caregiving and Bereavement………………………………………………...47 The Current Study………………………… …………………………….52 Methods………………………………………………………………….……….54 Participants……………………………………………………………….54 Data Collection Procedure……………………………………………….55 Attrition from Time 1 to Time 2…………………………………....……56 Measures…………………………………………………………...…….57 Caregiver Demographics………………………………………...57 Caregiving Stressors……………………………………………..57 Appraisals………………………………………………………..57 Social Resources…………………………………………………58 Well-Being……………………………………………………….59 Statistical Analysis Plan………………………………………….… …....60 Results……………………………………………………………………………62 Demographics and Basi c Descriptive Analyses…………………………62 Univariate Analyses……………………………………………………...64 Regression Models……………………………………………………….66 Discussion………………………………………………………………..………67 Limitations……………..………………………………………………..71 Future Directions………………………………………………………..72 Chapter Five.……………………………………………………………………..............74 Conclusions………………………………………………………………………74 Limitations…………………………………………………………………...…..77 Future Directions………………………………………………………...............78 References………………………………………………………………………………..81 Appendix...…………………………………………………………………….…… ……89 Appendix A: Curriculum Vitae.………………………………...……….………90 About the Author…………………………………………………………………End Page

PAGE 5

iii List of Tables Table 1. Demographic Characteristics of the Sample by Caregiving Status……… ……14 Table 2. Summary of Mi xed Model Analyses…………...………………………... ……24 Table 3. Demographic Characteristics of the Sample by Caregiving Status…….... ……34 Table 4. Results of the Chi-Square Analysis for Bereavement Trajectories by Caregiving Status……………………………………………….. ……38 Table 5. Demographic Charact eristics of the Sample………………...............................62 Table 6. Means and Standard Deviations for all Study Variables………..……..………63 Table 7. Correlation Matrix for All Study Variables……………………………………65 Table 8. Regression Model fo r Post-loss Depression ……………....……….…..………67

PAGE 6

iv List of Figures Figure 1. Depression change as a func tion of Time and Caregiving Status…….………21 Figure 2. Frequent Contact with Chil dren and Grandchildren change as a function of Time and Caregiving Status…………………..……..… ……22 Figure 3. Positive Support from All Sources change as a function of Time and Caregiving Status……………………………………………............23

PAGE 7

v The Relationship between Caregiving and Be reavement: A Series of Three Studies Allison M. Burton ABSTRACT Spousal bereavement has been consistently demonstrated in the literature to be one of the most highly stressful experiences in an individual’s lifetime. In addition many deaths in the United States are preceded by a period of caregiving, which is also believed to be highly stressful and have a prof ound impact on bereavement. However the literature has been inconsistent as to th e exact nature of the relationship between caregiving and bereavement and there has been some debate as to whether or not positive and negative affect variables are mutually exclusive. This dissertation sought to further address the issue of the relationship between caregiving and the bereavement experience through a series of three studies which utilized information from two datasets. The first was the Changing Lives of Older Couples (CLOC) study, a project that include d both preand post-lo ss data. The second was a subset of a dataset that recruited elderl y spousal caregivers of terminally ill patients from a large, local non-profit hospice. This dataset include d both preand post-loss data and included a variety of widely used a nd validated measures that allowed for the examination of caregiving stressors, apprai sals, and social suppor t and their effect on both positive and negative mental hea lth outcomes during bereavement.

PAGE 8

vi The first study examined the impact of caregiving on well-being during bereavement, specifically within the doma ins of psychological, social, and physical health, utilizing both positive and negative a ffect measures. The second addressed how varying levels of car egiving experience were relate d to previously established bereavement trajectories (Boerner, Wortman, & Bonanno, 2005; Bonanno, Wortman, Lehman, Tweed, Haring, et al., 2002), and the last one examined the mechanisms by which caregiving had an effect on bereavement outcomes, specifically focusing on stressors, appraisals, and social support. This dissertation consists of five chapters. Chapter 1 is an introduction to the caregiving and bereavement literature, Chapters 2-4 describe the series of three studies conducted, and Chapter 5 discusses the overall c onclusions as well as future directions for research.

PAGE 9

vii Preface I want to thank the following people, without whom this dissertation and my professional growth during my years in the Ph.D. in Ag ing Studies program would not have been possible: William E. Haley, Ph.D.: Thank you for all of your patience, support, and guidance throughout my time in the Ph.D. program. I could never have made it to this point without your mentorship. You took on the arduou s task of helping me develop both as a researcher and a professional and for that I will be forever grateful. Brent J. Small, Ph.D.: You have been a truly outstand ing mentor and advisor and I cannot thank you enough for your patience during the countless consultations, particularly throughout this dissertation process. Ther e were so many times that “statistical crises” aros e and your knowledge and guidance we re the only things that kept me from giving up. Paul B. Jacobsen, Ph.D., Susan C. McMillan Ph.D., and Sandra L. Reynolds, Ph.D.: A very special thank you to all of my committee members for such exceptional guidance. Each of you showed extraordinary patience w ith me throughout this process and I feel honored to have had such distinguished resear chers willing to devote their very valuable time to this committee. Cathy L. McEvoy, Ph.D.: I am very grateful to you fo r providing the perspective that I often was unable to obtain on my own. You ha ve had an immeasurable effect (note the proper usage) on me by helping me to learn no t to take everything so seriously and that making mistakes is perfectly acceptable as long as you are wi lling to learn from them and work harder the next time. The students in our program are very lucky to have you. Faculty and Staff at the Sc hool of Aging Studies: The School of Aging Studies has been a constant source of funding for me a nd I am so appreciative for having been given the opportunity to teach. It ha s been such an honor to learn from and be exposed to such exceptional faculty members throughout my year s in the doctoral pr ogram and I can only hope that I am able to find a job in a departme nt that compares. I also want to thank Gail, Amy, Pam, Rosa, Nancy, and Lydia for helpi ng me on countless occasions while enrolled in this Ph.D. program; if I thanked each of you individually every day for the rest of my life it would never be enough.

PAGE 10

viii The Center for Hospice, Palliative Care, and End of Life Studies, LifePath Hospice and Palliative Care, and The Hospice of the Florida Suncoast: I want to thank these three organizations for their financial and re search support throughout my years at USF. I have been so fortunate to work with such professional organizations and am grateful for the invaluable opportunities each of th ese institutions provided to me. Arthur M. Guilford, Ph.D.: Thank you so much for agreeing to chair my defense. On a more personal note, I want to thank the people who have provided me with so much love and support throughout my graduate school career and who have played such instrumental roles in shaping my growth as an individual. I know I have not been the easiest person to deal with at times, but I hope each of you knows how much you are loved and appreciated by me, even when I am too wrapped up in my own stresses to express it. All of you mean more to me then I could ever possibly explain. My Family: Being so far away from you all of the time is harder than I have ever really expressed. I regret missing so many of the fa mily moments over the years, but hope that all of you understand why I have chosen this path in life. I love all of you very much and know beyond a shadow of a doubt that there is no way I could be where I am today without having received so much love, suppor t, and guidance from all of you. I cannot possible acknowledge everyone in dividually, so please don’t ta ke it personally if you are not mentioned by name. Mom, you have been my best friend, biggest cheerleader, conscience, and inspiration my entire life. I have always been proud of the fact that I ca n talk to you about anything and that our relationship has matured through the years into something more meaningful and respectful than simply a “mother-daughter” bond. All I have ever wanted is for you to be proud of me and I hope I have been able to live up to all of your expectations. Dad, I will never be able to thank you e nough for all the help that you have given me throughout the years. I know that at time s a more “traditional” daughter would have made your life easier but I think it was appa rent from a pretty young age that that was never going to happen. I hope that our relations hip will continue to grow as we both get older and that you always know how important you are to me. Jung: There truly are no words. You have si ngle-handedly endured every step of this Ph.D. program and this dissertation with me (with only minor complaining) and there is no possible way that I could have made it to this point without your friendship, guidance, perspective, wisdom, love, and support. You are a brilliant an d amazing woman and I love and respect you more than I could ever possibly express. I ha ve known from very early on in our friendship that you are going to do remarkable things with your life and I sincerely hope that I am always there to share them with you. Thank you for all of the memories (both good and bad), for the integral role you played in helping me to become a better person, and most importantly for bei ng the best friend that I have ever known.

PAGE 11

ix Chris: You came into my life under the most memorable of circumstances and have been there for me during what I can only describe as the most complicated and “schizophrenic” part of my life so far. You have put up with more then you should ever have had to and I don’t even know how to be gin to thank you for the hours that you spent listening to me these last months. Ther e are not that many people who could have endured my mood swings or tolerated my (sometimes) co mpletely uncontrollable emotions and, from my perspec tive, the fact that you did sp eaks volumes about the depth of your character. I only hope that I can so meday be there for you the way that you have been for me. Linda: The closest thing to a big sister that I have ever known. You have been there for me from the very beginning to offer poignant advice, a compassionate ear, or a much needed laugh. You are truly one of the few people that I have come across in my life that I would call remarkable and I am so proud to count you as one of my close friends. Thank you so much for always being there exactly when I needed you and for providing such a wise and clear viewpoint through some of my most perplexing experiences the last five years. As you have said to me many times, it is rare to encounter someone who approaches the world the same way that you do, and that when it does happen, you should hold on to that friendship. I genuinely hope that we are able to do that. My Friends: You know who you are. I can’t th ank all of you enough for the great memories. Each of you has contributed in bo th small and large ways to making my time in Tampa unforgettable. Thank you for shari ng part of your lives and your talents with me and for always creating an environmen t where I was allowed to unapologetically express who I really am, both good and bad. I have felt very lucky to have been surrounded and supported by such intelligent, passionate, independe nt, interesting, and complex people. Each one of you has taught me important lessons about myself, life in general, and how the real world works and I am so grateful for that. My Students: Teaching has been such a rewarding experience for me, and I feel as though I have been fortunate enough to enc ounter some of the brightest and most inspirational undergrads that USF has to offe r. Thank you all for allowing me to be comfortable in the classroom, for always laughing at my mistakes, and for sharing so many of your life stories with me. There are so many of you that are going to do amazing things with your lives and I can only hope th at you learned as much from me as I did from all of you.

PAGE 12

1 Chapter One Overview Introduction The loss of spouse is one of the most st ressful events that an individual may encounter over the course of his or her lifetime (Stroebe & Stroebe, 1987). Spousal bereavement has been demonstrated in previous research to have a profound effect on an individual’s well-being, incl uding increased incidence of depression, decreased life satisfaction and social activities, and physical health decrements (i.e., Bonanno, Notarius, Gunzerath, Keltner, & Horowitz, 1998; Carr, House, Wortman, Nesse, & Kessler, 2001; Davis & Nolen-Hoeksema, 2001; Folkma n, 2001; Schulz & Beach, 1999; Stroebe, Stroebe, Abakoumkin, & Schut, 1996). Little is known about how family caregiving experiences affect the course of bereavement, although it ha s been well established that certain types of deaths, such as suicide or homicide, are particularly stressful and that unexpected death may be particularly di fficult for survivors (Carr et al., 2001). Extensive and stressful periods of in formal caregiving often precede spousal bereavement for older adults who die after a chronic illness (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Minino & Smith, 2001) Caregiving experiences often place family caregivers, particularly spouses, at risk for a variety of mental and physical health

PAGE 13

2 decrements; some of these documented effects include increased incidence of depression, increased time for wound healing, and increase d chance of mortality (i.e., Kiecolt-Glaser, Marucha, Malarkey, Mercado, & Glaser, 1996; Pinquart & Sorensen, 2003; Schulz & Beach, 1999; Schulz, O’Brien, Bookwala, & Fleissner, 1995). A review article that systematically a nd comprehensively examined the literature on the effects of bereavement after family caregiving came to the conclusion that caregiving researchers often do not carry out their research long e nough after the death of the care recipient to get a clear picture of the role th is experience plays in shaping the bereavement experience. In addition many bereavement researchers to date have not obtained enough information about pre-death ci rcumstances to fill in this gap either (Schulz, Newsom, Fleissner, Decamp, & Nieboer, 1997). This review (Schulz et al., 1997) notes that two opposing hypotheses have shaped the research on caregiving and bereavement, the first that the reduction in caregiving burden following the loss of a spouse results in improvements in mental and physical health outcomes, and the second that a depletion of resources results fro m prolonged exposure to caregiving stressors leaving the surviving spouse more vulner able to negative outcomes during widowhood. This review paper also identified a vari ety of methodological weaknesses in prior caregiving and bereavement research. Th ese methodological weaknesses included crosssectional design, inappropriate comparison gr oups, small sample sizes, no control group, inconsistency in outcomes measures, and a focus on very specific subsets of the caregiving population (i.e., Bass & Bowma n, 1990; Bodnar & Kiecolt-Glaser, 1994; Cleiren, van der Wal, & Deikstra, 1988; George & Gwyther, 1994; McHorney & Mor, 1988; Mullan, 1992; O’Bryant, 1991; Skaff, Pearlin, & Mullan, 1996). In addition

PAGE 14

3 studies addressing this issue wh ich were published subsequent to the Schulz et al. (1997) review paper also possess some of the same methodological issues (Park & Folkman, 1997; Robinson-Whelen, Tada, MacCallum, McGuire, & Kiecolt-Glaser, 2001). Bereavement researchers who have attempted to examine death context variables have largely utilized two cat egories: expected and une xpected deaths (i.e., Carr et al., 2001; Hill, Thompson, & Gallagher, 1988; L undin, 1984; Reed, 1998). Overall results from these studies indicate that individuals who experience the loss of a loved one due to an unexpected death are more likely to expe rience higher levels of depression and an increased difficulty in adjusting to the death of a loved one than thei r counterparts. This categorization ignores the great variability among individuals who are included in the expected death group, ranging from no persona l experience with pr oviding care to highly stressful, full-time caregiving. In order to address the variability in caregiving, researchers have used a variety of cla ssification methods, speci fically breaking down groups using disease type, durati on, severity of impairment, extent of caregiving, or level of perceived stressfulness. A prospective, population-based sample was used in a recent study to examine changes in self-reported and objective mental and physical health outcomes after bereavement among elderly in dividuals who were involved in varying levels in the caregiving expe rience (Schulz, Beach, Lind, Mar tire, Zdaniuk, et al., 2001). This study used three groups, non-caregivers caregivers repor ting no strain, and caregivers reporting strain, to examine change s in depression, antid epressant medication use, health risk behaviors, and weight am ong for individuals who had been bereaved for varying lengths of time. This study found that depression scores remained high for the strained caregiving group from pre-loss to pos t-loss; however they increased for the other

PAGE 15

4 two groups. In addition while the non-car egiver and low strain caregiver groups experienced little change in health risk behaviors during the st udy period, the strained caregiver group experienced significant improvements in this domain following the deaths of their spouses. The researchers came to the overall conclusion that variations in the caregiving experience play a significant role in determ ining the mental and physical health outcomes of bereaved spouses. The previous study by Schulz and his co lleagues (2001) addressed an important area of the caregiving and bereavement litera ture that merits further attention: improvement. More specificall y, the idea that improvement may occur in certain areas of well-being following the loss of a spouse for hi ghly stressed caregivers. As previously noted, the researchers of th is study found that, when compared with behavior while caregiving, individuals demonstrated significan t improvements in health risk behaviors following the death. Consistent with th ese results, declines in depressive symptomatology by three-months post-loss we re found in a recent study of dementia caregivers, and by one year post-loss these i ndividuals reached levels lower than those reported while actively caregiving (Schulz, Mendelsohn, Haley, Mahoney, Allen, et al., 2003). In addition, Schulz et al. (2003) repor ted that 72% of these bereaved dementia caregivers considered the death of their relative to be at least somewhat of a relief. The current series of studies is design ed to further address the issue of the relationship between caregiving and the be reavement experience by using information from two datasets. The first is the Cha nging Lives of Older Couples (CLOC) study, a project that includes both preand post-lo ss data. The comprehensive nature of the CLOC dataset allows for the examination of a wide variety of variab les, including but not

PAGE 16

5 limited to caregiving characteristics, personality traits, bereavement trajectories, social interaction, depression, sociodemographics, self-rated physical health, and positive wellbeing, and their relationship to involvement in the caregiving experience. The second is a subset of a dataset that recruited elderly spousal caregivers of termina lly ill patients from a large, local non-profit hospice. This dataset includes both preand post-loss data and includes a variety of widely used and validat ed measures that allow for the examination of caregiving stressors, appraisals, and soci al support and their eff ect on both positive and negative mental health outcomes during bereavement. The first study examines the impact of caregiving on well-being during bereavement, specifically within the doma ins of psychological, social, and physical health, utilizing both positive and negative affect measures. Th e second in this series of three studies addresses how varying levels of caregiving experience are related to previously established bereavement traj ectories (Boerner, Wortman, & Bonanno, 2005; Bonanno, Wortman, Lehman, Tweed, Haring, et al ., 2002), and the last one examines the mechanisms by which caregiving has an effect on bereavement outcomes, specifically focusing on stressors, apprai sals, and social support.

PAGE 17

6 Chapter Two Bereavement after caregiving or unexpected death: Effects on elderly spouses Abstract Responses to bereavement have been shown to vary depending on whether death is expected or unexpected, and on the nature of family caregiving experiences, but no previous research has examined these factor s simultaneously. To address these issues, we utilized prospective data on bereavement from 193 partic ipants in the Changing Lives of Older Couples (CLOC) study, who were asse ssed both before their loss and at 6 and 18 months after the death. Participants w ho experienced either unexpected loss, or expected loss without caregiving, with lo w-stress caregiving, or with high-stress caregiving completed measures of psychologi cal, social, and health functioning on each occasion. Using mixed model analyses, the results showed that unexpected death was associated with marked increases in depres sion, while the nature of caregiving did not affect the trajectory of any of the psychol ogical well-being measures. All groups except highly stressed caregivers show ed improvements in social activity and support after bereavement, suggesting that highly stressed ca regivers may be at an increased risk for social isolation during bereavement. Thus experiencing an unexpected death may put

PAGE 18

7 bereaved spouses at risk for depression, while high stress caregiving may lead to problems with social isolation. Introduction One of the most stressful events that an individual may encounter over the course of his or her lifetime is th e death of a spouse (Stroebe & Stroebe, 1987). Previous research has demonstrated that spousa l bereavement has a profound effect on an individual’s well-being, incl uding increased incidence of depression, decreased life satisfaction and social activities, and physical health decrements (i.e., Bonanno, Notarius, Gunzerath, Keltner, & Horowitz, 1998; Carr, House, Wortman, Nesse, & Kessler, 2001; Davis & Nolen-Hoeksema, 2001; Folkma n, 2001; Schulz & Beach, 1999; Stroebe, Stroebe, Abakoumkin, & Schut, 1996). While it has been well established that certain types of deaths, such as suicide or homici de, are particularly st ressful and unexpected death may be particularly difficult for surviv ors (Carr et al., 2001), little is known about how family caregiving experiences a ffect the course of bereavement. Spousal bereavement frequently follows extensive and stressful periods of informal caregiving for older adults who die after ch ronic illness (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Minino & Smith, 2001) Family caregivers, particularly spouses, are at risk for a variet y of mental and physical heal th decrements as a result of their caregiving experience; some of th ese documented effects include increased incidence of depression, decreased immune sy stem response, and increased chance of

PAGE 19

8 mortality (i.e., Kiecolt-Glaser, Marucha, Malarkey, Mercado, & Gl aser,1996; Pinquart & Sorensen, 2003; Schulz & Beach, 1999; Schulz, O’Brien, Bookwala, & Fleissner, 1995). A review article that systematically a nd comprehensively examined the literature on the effects of bereavement after fam ily caregiving concluded that caregiving researchers often do not conti nue their research long enough af ter the death to get a clear picture of the role this e xperience plays in shaping the bereavement experience; in addition many bereavement researchers to da te have not obtained enough information about pre-death circumstances to fill in th is gap (Schulz, Newsom, Fleissner, Decamp, & Nieboer, 1997). This review (Schulz et al., 1997) notes that research on caregiving and bereavement has been based on two opposing hyp otheses, the first that a depletion of resources results from prolonged exposure to caregiving stressors leaving the surviving spouse more vulnerable to negative outcom es during widowhood, and the second that the reduction in caregiving burden following the lo ss of a spouse results in improvements in mental and physical health outcomes. This review paper also identified methodological weaknesses in prior caregivi ng and bereavement research including cross-sectional design, inappropriate comp arison groups, small sample sizes, no control group, inconsistency in outcomes measures, and a focus on very specific subsets of the caregiving population (i.e., Bass & Bowma n, 1990; Bodnar & Kiecolt-Glaser, 1994; Cleiren, van der Wal, & Deikstra, 1988; George & Gwyther, 1994; McHorney & Mor, 1988; Mullan, 1992; O’Bryant, 1991; Skaff, Pearlin, & Mullan, 1996). In addition studies addressing this issue wh ich were published subsequent to the Schulz et al. (1997) review paper also possess some of the same methodological issues (Park & Folkman, 1997; Robinson-Whelen, Tada, MacCallum, McGuire, & Kiecolt-Glaser, 2001).

PAGE 20

9 Bereavement researchers who have attempted to examine death context variables have largely limited these to two categories: expected and unexpected deaths (i.e., Carr et al., 2001; Hill, Thompson, & Gallagher, 1988; Lundin, 1984; Reed, 1998). Taken together, results from these studies indi cate that individuals who experience an unexpected death are more like ly to experience higher leve ls of depression and an increased difficulty in adjusting to the death of a loved one than thei r counterparts. This distinction ignores the great variability am ong individuals who are categorized into the expected death group, ranging from no persona l experience with pr oviding care to highly stressful, full-time caregiving. Caregiving researchers have addressed variability in caregiving in many ways, specifically by breaking down groups using disease type, duration, severity of impairment, extent of caregiving, or level of perceived stressfulness. One recent study utilized a prospective, popul ation-based sample in order to examine changes in self-reported a nd objective mental and physical health outcomes after bereavement among elderly individuals who we re involved in varying levels in the caregiving experience (Schulz, Beach, Lind, Martire, Zdaniuk, et al., 2001). This study examined changes in depression, antidepressant medication use, health risk behaviors, and weight among three groups: noncaregiver s, caregivers repor ting no strain, and caregivers reporting strain for individuals who had been bereaved for varying lengths of time. The findings from this study indicated th at depression scores remained high for the strained caregiving group thr oughout the study; however they increased for the other two groups. In addition the strain ed caregiver group experienced significant improvements in health risk behaviors following the death of their spouses, whereas the other two groups experienced little change duri ng the study period. Overall the researchers concluded that

PAGE 21

10 the variations in the caregiving experience play a significant role in determining the mental and physical health outcomes of bereaved spouses. The previous Schulz et al. (2001) stu dy addressed an additional area of the caregiving and bereavement literature that meri ts further attention: that improvement may occur in certain areas of well-being following the loss of a spouse for highly stressed caregivers. As previously noted, in this study the resear chers found significant improvements in health risk behaviors during bereavement when compared with behavior while caregiving. Consistent with these re sults, another recent study found that dementia caregivers showed declines in depressive symptomology by three-months post-loss and by one year post-loss they reached levels lower than those reported while actively caregiving (Schulz, Mendelsohn, Haley, Ma honey, Allen, et al., 2003). Schulz et al. (2003) also reported that among these bereaved dementia caregivers, 72% considered the death of their relative to be at least somewhat of a relief. The current study is designed to further address the issue of the relationship between caregiving and the bereavement e xperience by using information from the Changing Lives of Older Couples (CLOC) st udy, a project that includes both preand post-loss data. This project will go beyond th e examination in previous CLOC analyses (Carr et al., 2001) of expected versus une xpected deaths and utilize four groups: unexpected death, expected death with no careg iving, expected death with low levels of stress associated with the caregiving experien ce, and expected death with high levels of stress associated with the caregiving experien ce. The comprehensive nature of the CLOC dataset allows for the examination of a wide variety of variables, including but not limited to social interaction, depression, sociodemographics, self-rated physical health,

PAGE 22

11 and positive well-being, and their relationship to invol vement in the caregiving experience. For the current study, this dataset has sufficient sample size and variability to allow for the examination of the impact of caregiving on well-being during bereavement, specifically within the domains of psychologica l, social, and physical health. In addition this dataset allows for the examination of positive and negative affect measures as separate constructs and they will be analyzed as such. There is some debate in the literature at this time as to whether or not positive and negative affect variables are mutually exclusive and the current study ex amines this idea with in the context of bereavement (Larsen, McGraw, & Cacioppo, 2001; Reich, Zautra, & Davis, 2003; Russell & Carroll, 1999); in addition there are a handful of studies to date that have examined this concept in the caregiving literat ure, but the findings are not clear (Chappell & Reid, 2002; Lawton, Moss, Kleb an, Glicksman, & Rovine, 1991). Despite the important recent findings in this area, there are several important gaps in the literature on caregiving and bereavem ent. While a number of studies have contrasted either expected versus unexpected deaths, or low or high stress caregiving, no study to date has included all relevant groups: unexpected death, expected death without caregiving, expected death with low-stress caregiving, and expected death with highstress caregiving. In addition, studies in this area to date have examined a limited range of potential outcome variables, primarily focu sing on depression and physical health. It is particularly important to include not only measures of negative affect, such as depression, but also measures of posi tive affect and social engagement. Based on the prior literature we hypothesized that hi ghly stressed caregivers would show improvements in well-being at the eighteen-month follow-up, particularly

PAGE 23

12 within the domain of negative psychological affect. We also predicted that spouses experiencing unexpected deaths would significan tly increase in depression scores at both post-loss time points, while both no caregiv ing and low stress caregiving groups would show either stability or temporary worsen ing at the short-term follow-up. For the measures within social engagement, we pred icted that highly stress ed caregivers would show the lowest levels of so cial interaction and satisfacti on with support when compared with the other three groups, due to prev ious findings showing that caregiving is associated with decreased social suppor t and engagement (Robinson-Whelen et al., 2001). In addition we expected to see the highest decrements within the domain of physical health, particularly self-rated health, to be for the unexpected death group and the highly stressed caregivers. Methods Participants The current study utilized the data ob tained from bereaved individuals who participated in the Changing Lives of Olde r Couples (CLOC) study, a prospective study of a two-stage area probability sample of 1532 married men and women from the Detroit Standardized Metropolitan Statistical Area. In order to be eligible for the study, individuals had to meet the following criter ia: English-speaking, married, residing in a household in which the husband was at least 65 years of age, noninstitutionalized, and able to participate in a two hour face-to-face initial interview (University of Michigan,

PAGE 24

13 n.d.). In addition, the researchers who desi gned the study oversampled women in order to obtain the highest number of bereaved spouses possible during the study period. Baseline interviews for the CLOC study we re conducted between June 1987 and April 1988, and approximately 68% of individuals contacted participat ed in the initial interview, a response rate that is consistent with other studied conducted in the Detroit area (see the University of Michigan CL OC website for additional details). Researchers utilized the daily obituar ies in the Detroit-area newspapers and monthly death record tapes provided by the st ate of Michigan in order to identify the initial participants who had experienced the loss of a spouse. The National Death Index (NDI) was used to confirm the deaths and to obtain the causes of death. Of the initial sample, 316 individuals lost their spouse during the study period and were asked to participate in a follow-up interview six months after the loss of thei r spouse. The final sample for the current study was 193 (61% of the original dataset) with the primary reasons for non-participati on being ill health or death and refusal. The mean age of this sample was 69.48 (s .d.=6.90) at the baseline interview and 88.08% (n=170) were female. In addition, 160 participants identified themselves as Caucasian American (84.22%) and 30 as Af rican American (15.78%) and had on average 11.48 years of education. The four groups di d not differ significantly on any of these demographic characteristics, as shown in Tabl e 1. For the purposes of this study, the two cases that reported death as a result of murder suicide, or homicide have been excluded because of the unique attributes associated w ith a loss of this nature (Sheskin & Wallace, 1976).

PAGE 25

14 A number of papers have been publis hed utilizing the CL OC dataset (Bonanno, Wortman, Lehman, Tweed, Haring, et al., 2002; Carr, House, Kessler, Nesse, Sonnega, et al., 2000; Carr et al., 2001); however to date we have not able to identify any that examined caregiving characteristics and how they affect the bereavement experience. Table 1. Demographic Characteristics of the Sample by Caregiving Status Measures Pre-loss Caregiving Status The four groups for this study were chosen based upon their response to three questions: the first asking whether the spouse died of either an ongoing serious Unexpected Death No Caregiving Low Stress Caregiving High Stress Caregiving N 37 65 43 48 Age, mean years (SD) 70.14 (5.79) 70.49 (6.94) 69.74 (5.72) 67.58 (8.39) % Female 89.20 87.70 83.70 89.60 % White 70.30 90.80 83.70 83.30 Education, mean years (SD) 11.43 (3.03) 11.28 (2.46) 11.40 (2.74) 11.73 (2.44) Marital Quality, z-score (SD) -.26 (1.21) -.34 (1.33) .02 (0.93) -.17 (0.91)

PAGE 26

15 condition, an accident, murder, or suicide, or an unspecified cause indicated as not being an ongoing serious condition, the second asking those whose spouses died of a serious ongoing condition whether or not they provided care for their spouse 6 months prior to death, and the third asking individuals who re ported providing care fo r their spouses how stressful they found the experience to be. Fo r the question regardi ng stressfulness of the caregiving experience, partic ipants were asked to provi de a response based upon a 5point Likert-type scale with responses ranging very stressful to not at all stressful. Individuals who responded to this question as either very stre ssful or quite stressful were included in the high stress group and indivi duals who responded as somewhat stressful, not too stressful, or not at all stressful are included in the low stress group. Participants were divided into the following four categ ories: individuals whose spouses died unexpectedly (n=37), individuals whose spous es died from a serious ongoing medical condition but did not provide ca re six months prior to th e death (n=65), individuals whose spouses died from a serious ongoing medical condition who di d provide care six months prior to the death and reported low leve ls of stress associated with the caregiving experience (n=43), and individuals whose s pouses died from a serious ongoing medical condition who did provide care si x months prior to the death and reported high levels of stress associated with the caregiving experience (n=48). Demographics and Marital Quality Demographic and descriptive variables, specifically age, gender, race/ethnicity, education, and number of hours per week spen t in the caregiving role, were assessed by self-report. Marital quality (alpha=.85) was determin ed based upon a 4 item composite

PAGE 27

16 score with higher scores being associated w ith more positive ratings of the marriage. These four questions were asked during the initial interview and included the following items: 1) How much does your spouse make you feel loved and cared for, 2) How much is your spouse willing to listen when you need to talk about your wo rries or problems, 3) Thinking about your marriage as a whole, how often do you feel happy about it, and 4) Taking all things together, how sa tisfied are you with your marriage. Well-Being Measures Psychological Overall psychological well-being was m easured using five items from the Bradburn Affect Balance Scale (ABS; Br adburn, 1969) (alpha=.79). Participants responded using a 5-point Likert-type scale ra nging from strongly disagree to strongly agree to five questions: 1) I was particularly excited or inte rested in something, 2) I was pleased about having accomplished something 3) I felt that things were going my way, 4) I felt proud because someone complimented me on something I had done, and 5) I felt on top of the world. This variable (as well as all of the other indi ces within the CLOC dataset) was converted into z-scores with a m ean of 0 and a standard deviation of 1, with higher scores indicating higher levels of overall psychological well-being. Depression was measured using an elev en item subscale of the Center for Epidemiological Studies-Depression (CES -D; Kohout, Berkman, Evans, & CornoniHuntley, 1993; Radloff, 1977) (alpha=.78) code d using a 5-point Likert-type scale with responses ranging from strongly disagree to strongly agree; as with the full CES-D,

PAGE 28

17 higher values on this scale indicate higher le vels of depressive symptomology. The CESD 11-item scale has been used in other studies, most notably in the Health and Retirement Study (HRS) and the Assets and Health D ynamics of the Oldest-Old (AHEAD) study (HRS/AHEAD Document ation Report, 2000). Self-esteem was a composite of five items within the dataset (alpha=.72). These items included: 1) On the whole I am satisfi ed with myself, 2) At times I think I am no good at all, 3) I wish I could ha ve more respect for myself, 4) All in all I feel that I am inclined to be a failure, and 5) All in all, I feel that I am a person of worth, at least equal with others. All of these items utilized a 5point Likert-type scale with ratings ranging from strongly disagree to str ongly agree, with higher numbers indicating higher levels of self-esteem. Social Engagement Frequent contact with children and gra ndchildren (alpha=.70) was a summation of two questions within the dataset that utiliz ed a 5-point Likert-type scale with higher scores indicating more frequent contact. These two items were : 1) In the past 12 months, how often did you have contact wi th at least one of your chil dren who does not live with you either in person, by phone, or by ma il, and 2) How often do you see your grandchildren. Positive support from all sources (alpha=.66) was a composite variable that included four items with higher scores i ndicating higher levels of positive support and were rated using a 5-point Likert-type scale. Two questions were asked for two groups of

PAGE 29

18 people: children and friends and relatives. The two ques tions were: 1) How much do your ______ make you feel loved and cared for, a nd 2) How much is he or she willing to listen to you when you need to talk about your worries or problems. Physical Health Satisfaction with good health (alpha=.84) was a combination of three items that utilized a 5-point Likert-type scale with higher scores indica ting higher levels of satisfaction. The items included in this variab le were: 1) In gene ral how satisfied are you with your health, 2) How would you rate you r health at the present time, and 3) How much are your daily activities limited in any way by your health or health-related problems. Difficulty with activities (a lpha=.78) utilized four items rated on a 5-point Likerttype scale with higher numbers indicating a higher degree of difficulty with daily activities. The questions were : 1) How much difficulty do you have bathing yourself, 2) How much difficulty do you have climbing a few flights of stairs because of your health, 3) How much difficulty do you have walking se veral blocks because of your health, and 4) How much trouble would you have doing heavy work around the house such as shoveling snow or washing walls because of your health. Statistical Analysis Plan First measures including race/ethnicit y, gender, age, education, and marital quality were examined in order to determine if any of these differed in our four groups of

PAGE 30

19 interest. In addition number of hours per week in the caregiving role was examined for the high and low stress caregiver groups in or der to determine whethe r or not there were quantitative differences in their caregiving experiences. In addition to describing the sample and the groups, these analyses allowed us determine if it was necessary to include any covariates in the analyses. In order to further examine the relati onship between caregiving status and the domains of well-being, mixed model designs were used for all seven of our outcome variables. The procedures used for these analyses are detailed in Singer and Willett (2003). Specifically, three random effects models were run for each of our outcome measures. Model one was an unconditional means model. This model examined whether a change or time component needed to be inco rporated into the model. Model two was an unconditional growth model that specified chan ges in each of the outcome variables as a function of linear terms (pre-lo ss, 6 months post-loss, and 18 months post-loss). This model examined whether a group component needed to be inco rporated into the model. Model three was a growth model that examin ed the effects of group status (unexpected loss, expected loss with no caregiving, expect ed loss with low stress caregiving, expected loss with high stress caregiving) on each of the outcome variables. This model examined whether an interaction component needed to be incorporated into the model. If a significant interaction was present, separate models were run for each of the groups, examining change over time, to determine th e nature of longitudi nal changes across the groups. Results

PAGE 31

20 Basic Descriptive Analyses The initial set of analyses determined th at there were no significant differences (p>.05) between the four groups of interest on any of the demographic characteristics, specifically age, gender, race/ethnicity, and e ducation, or marital quality (Table 1). The analyses did not include any of the demographic characteristics or marital quality in subsequent analyses as covariates. In order to gain a better understanding of the differences between the high stress caregiver group and low stress caregiver group, we looked at the mean number of hours per week in the caregiving role these indivi duals reported during the six months prior to their spouses’ deaths. The low stress caregiver group indicated that they spent an average of 55.19 hours per week and the high stress car egiver group indicated that they spent an average of 70.48 hours per week on caregivi ng-related tasks (t=2.02, df=85, p=.048). These findings illustrate that both of the caregiving groups dedicated significant amounts of time to the caregiving tasks, but that there were quantitative differences in their caregiving experience. Mixed Model Analyses Psychological Mixed model analyses revealed no si gnificant (p>.05) time, group, or group by time interaction effects for either of the pos itive affect variables, overall psychological well-being and self-esteem.

PAGE 32

21 For depression we found both a signifi cant time effect and a significant group by time interaction. Table 2 (see page 24) displays the parameters of the random effects models for depression scores as a function of time and caregiving status. In addition Figure 1 (see page 21) shows that depre ssion increased significantly over time for the unexpected death group (estimate=.146, s.e.=.04 2, p=.001) and remained stable for the high stress (estimate=-.043, s.e.=.035, p=.234), low stress (estimate=.032, s.e.=.041, p=.441), and non-caregiving groups (estimate=-.016, s.e.=.031, p=.620). Figure 1. Depression change as a func tion of Time and Caregiving Status Social Engagement As shown in Table 2 (see page 24), for frequent contact with children and grandchildren we found a significant time effect and a significant group by time Depression-0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 42 months pre-loss 6 months post-loss 18 months post-lossz-score Unexpected Death N o Caregiving Low Stress Caregiving High Stress Caregiving

PAGE 33

22 interaction. As shown in Figure 2 (see page 22), the unexpected death (estimate=.127, s.e.=.045, p=.007) and no caregiving (estim ate=.089, s.e.=.024, p<.001) groups showed significant improvements in their frequency of contact with children and grandchildren over time, while the low stress caregiver group showed a trend towards this pattern (estimate=.046, s.e.=.024, p=.069); howev er the high stress caregiver group (estimate=.039, s.e.=.036, p=.282) remained relatively stable on this measure at both the 6-month and 18-month follow-up points. Figure 2. Frequent Contact with Children a nd Grandchildren change as a function of Time and Caregiving Status Positive support from all sources demons trated a significant time effect and a trend (p=.070) toward a significant group by time interaction according to the mixed model analyses, as shown in Table 2. In a ddition, Figure 3 (see pa ge 23) shows that the Frequent Contact with Children and Grandchildren-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 42 months pre-loss 6 months post-loss 18 months post-lossz-score Unexpected Death N o Caregiving Low Stress Caregiving High Stress Caregiving

PAGE 34

23 unexpected death (estimate=.103, s.e.= .028, p=.001), no caregiving (estimate=.088, s.e.=.023, p<.001), and low stress caregivi ng (estimate=.096, s.e.=.032, p=.005) groups showed significant improvement s in their perceptions of positive support at both the 6month and 18-month follow-up points; however the high stress caregiver (estimate=.023, s.e.=.027, p=.408) group remained relatively stable over time. Figure 3. Positive Support from All Sources ch ange as a function of Time and Caregiving Status Physical Health Neither of the health indicator variab les in this study, satisfaction with good health and difficulty with activities, was s hown as having significant time effects, group effects, or group by time inte raction effects in the mixed model analyses. Positive Support from All Sources0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 42 months pre-loss 6 months post-loss 18-months post-lossz-score Unexpected Death N o Caregiving Low Stress Caregiving High Stress Caregiving

PAGE 35

24 Table 2. Summary of Mixed Model Analyses Model 1 Estimate(SE) p-value Model 2 Estimate(SE) p-value Model 3 Estimate(SE) p-value Depression Intercept .16 (.06) .007 .17 (.06) .006 .22 (.16) .152 Time .02 (.02) .306 .14 (.05) .004 Group -.02 (.06) .711 Group x Time -.05 (.02) .008 Residual Variance .72 (.05) .48 (.05) .48 (.05) Variance Intercept (UN 1,1) .55 (.08) .55 (.08) Variance Slope (UN 2,2) .03 (.01) .03 (.01) Frequent Contact with Children and Grandchildren Intercept .24 (.04) <.0001 .28 (.04) <.0001 .35 (.10) .001 Time .08 (.02) <.0001 .15 (.04) <.0001 Group -.03 (.04) .452 Group x Time -.03 (.01) .034 Residual Variance .46 (.03) .17 (.02) .17 (.02) Variance Intercept (UN 1,1) .27 (.03) .27 (.03) Variance Slope (UN 2,2) .04 (.01) .04 (.004) Positive Support from All Sources Intercept .30 (.07) <.0001 .34 (.06) <.0001 .54 (.16) .001 Time .08 (.01) <.0001 .14 (.04) <.0001 Group -.08 (.06) .174 Group x Time -.02 (.01) .070 Residual Variance .40 (.03) .26 (.03) .26 (.03) Variance Intercept (UN 1,1) .63 (.08) .62 (.08) Variance Slope (UN 2,2) .02 (.004) .01 (.004)

PAGE 36

25 Discussion Results support the importance of ex amining circumstances of death and caregiving when studying bereavement. The fi ndings from this study demonstrate that the idea of stability following the loss of a spouse can be seen in multiple domains of bereavement outcomes, particularly positive affect, self-esteem, satisfaction with health, and difficulty with physical activities. Both of our positive psychological outcome measures showed no change over time, indicatin g that all four of our groups maintained stable levels of positive aff ect and self-esteem throughout th e study period and adding to the literature which suggests that positive and negative affect variables are separate and distinct concepts and should be examined as such in future research. Findings from the literature on positive and negative affect variables have demonstrated that these two concepts are not opposite sides of the same c onstruct and that each has a distinct set of correlates and predictors (i.e., Lawton 1983; Diener & Emmons, 1984; Watson & Tellegen, 1985). In addition, contrary to our hypotheses, bereavement showed no impact on self-reported heath or difficulty with activities in our sample. For depression, the unexpected deat h group was associated with worsening depression after bereavement, while the other groups remained stable over time. The finding that spouses who experienced an unexpected death showed high levels of depression at both the 6 month and 18 mont h post-loss points was consistent with previous research. Our hypot hesis that highly stresse d caregivers would show improvements in well-being was not supported; instead, stability was found for all of the groups with an expected loss. Thus our resu lts suggest that, for predicting changes in

PAGE 37

26 well-being after bereavement, unexpected deat h was a more important predictor than the nature of the caregiving experiences. Both of our social well-being outcome m easures, frequent contact with children and grandchildren and positive support from all sources showed marked increases by 6 months post-loss and these increases remained steady at the 18 month follow-up point for the unexpected death, no caregiving, and low stress caregiving groups. The increases were found on both an objective and subjectiv e measure within the social well-being domain, indicating that not only does the amount of actual contact in crease but also the survivor’s satisfaction with the quality of that contact. Highly stressed caregivers, however, showed no increase over time, suggesti ng that they may be at risk for social isolation during bereavement. We have found no previous studies th at demonstrate this effect. Highly stressed caregivers may be particularly likely to experience role engulfment (Aneshensel et al., 1995; Pearlin, Mu llan, Semple, & Skaff, 1990) and thus to have difficulty retaining soci al roles while caregiving; our results suggest that these effects may continue after bereavement. Similar to the resource depletion hypothesis discussed by Schulz et al. ( 1997), caregivers who become disengaged socially may find it difficult to develop such relationships afte r bereavement. Zettel and Rook (2004) found that rekindling of old ties wa s the most common way that widows altered their social networks after bereavement; such rekindling may be difficult if, for example, highly stressed caregivers have experienced a lack of support or negative support, which is associated with stressful caregiv ing (Robinson-Whelen et al., 2001). Overall our findings highlight the importance for continued research that systematically examines the relationship between caregiving and bereavement using

PAGE 38

27 sophisticated methodology and prospective data with multiple post-loss follow-up interviews. The impacts of bereavement vary according to both whether or not a loss is expected and the nature of car egiving, with effects varying across domains of well-being. Since psychological interventions for gr ief and bereavement have generally had disappointing effects (Schut, Stroebe, van den Bout, & Te rheggen, 2001), it is important to more fully understand how bereavement puts some widows and widowers at high risk, and to understand which consequences of be reavement may be most appropriate as targets of interventions.

PAGE 39

28 Chapter Three Patterns of caregiving and bereavement trajectories Abstract Prior literature has identified five core bereavement trajectories: common grief, chronic grief, chronic depression, improveme nt during bereavement, and resilience. Participants (n=175) in the Changing Live s of Older Couples study who experienced the death of a spouse were categorized into f our groups: unexpected loss, expected loss without caregiving, or caregiving with either high or low stress. Participants completed measures of depressive symptoms pre-loss and at 6 and 18 months post-loss, and were assigned to trajectory patt erns using previously pub lished decision rules. The correspondence between caregivi ng status and bereavement tr ajectories was examined using chi-square analysis. Results showed that unexpected death was associated with higher prevalence of chronic grief, highly stre ssed caregiving was associated with higher rates of chronic depression, and low stress car egivers exhibited distributions similar to those of the non-caregiving group. Across all groups the high est number of participants fell into the resilient categor y. The results suggest that further examination of the complicated relationship between caregi ving and bereavement is warranted, and advantages of trajectory analysis are discussed.

PAGE 40

29 Introduction Spousal bereavement is one of the most st ressful events that an individual may encounter over the course of his or her lifetime (Stroebe & Stroebe, 1987). Previous research has demonstrated that the deat h of a spouse has a profound effect on an individual’s well-bei ng. Some of these documented eff ects include increased incidence of depression, decreased life satisfaction and social activ ities, and physical health decrements (i.e., Bonanno, Notarius, Gunzerat h, Keltner, & Horowitz 1998; Carr, House, Wortman, Nesse, & Kessler, 2001; Davis & Nolen-Hoeksema, 2001; Folkman, 2001; Schulz & Beach, 1999; Stroebe, Stroebe, Abakoumkin, & Schut, 1996). While it has been well established in the bereavement literatu re that certain types of deaths, such as suicide or homicide, are pa rticularly stressful and th at unexpected death may be particularly difficult for the su rvivors (Carr et al., 2001), ther e is little known about how family caregiving experiences aff ect the course of bereavement. One relatively recent development in the literature is the identification of bereavement trajectories. Two recent studi es by Bonanno and his colleagues (Boerner, Wortman, & Bonanno, 2005; Bonanno, Wortman, Lehman, Tweed, Haring, et al., 2002) have identified five core bereavement traject ories, based on prosp ective data collected both pre-loss and at three time points post-lo ss: common grief, ch ronic grief, chronic depression, depressed-improved, a nd resilience. Common grief fo llows the pattern that is traditionally considered to be a “normal” grieving pattern: low levels of depression preloss, followed by a spike in depressive symp tomatology at short-term follow-up, and then

PAGE 41

30 a return to baseline levels of depression at long-term follow-up point s. Chronic grief and chronic depression are both characterized by consistently high levels of depression following the loss of a spouse, however they are differentiated by the level of depression pre-loss; individuals who exhi bit chronic grief show low le vels of depression pre-loss, while individuals who exhibit chronic depression show high levels at this time point. The depressed-improved trajectory is noted by hi gh levels of depression pre-loss followed by low levels of depressive symp tomatology at all post-loss fo llow-up points; the resilient group shows low levels of depression at all ti me points, both prior to and after the death of a spouse. For more detailed information on these trajectories a nd how the researchers developed them, please see Bonanno et al. (2002). Spousal bereavement frequently follows info rmal caregiving for older adults who die after chronic illness, and this caregiving experience is often extensive and stressful (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Minino & Smith, 2001). Family caregiving experiences, particularly for spous es, often place individuals at risk for a variety of mental and physic al health decrements; some of these documented effects include increased incidence of depression, slower wound hea ling, and increased mortality (i.e., Kiecolt-Glaser, Maruch a, Malarkey, Mercado, & Glas er,1996; Pinquart & Sorensen, 2003; Schulz & Beach, 1999; Schulz, O’Brien, B ookwala, & Fleissner, 1995). While the previous study by Bonanno and his colleagues (2002) examined level of caregiving as a context variable as part of the larger st udy, it was not a major focus and our goal is to expand on this prior research and attempt to gain a better understanding of how caregiving experiences lead to divergent bereavement trajectories.

PAGE 42

31 The current study is designed to furthe r address the issue of the relationship between caregiving and the bereavement e xperience by using information from the Changing Lives of Older Couples (CLOC) st udy, a project that includes both preand post-loss data. This project utilized four groups: unexpected death, expected death with no caregiving, expected death with low levels of stress associated with the caregiving experience, and expected death with high leve ls of stress associated with the caregiving experience and examined how th ese varying levels of caregiv ing experience were related to previously established bereavement traj ectories. Based on the prior Bonanno et al. (2002) study and the caregiving and bereavemen t literatures, we exp ect to see a higher number of individuals who e xperience an unexpected death to experience chronic grief, attributable to the circumstances surrounding the loss. In addition, prior research has shown that individuals who report low levels of stress while caregiving experience grief reactions similar to those of individuals who are not involved in caregiving activities, and we expect similar results in the current st udy (i.e., Schulz, Beach, Lind, Martire, Zdaniuk, et al. 2001). As for highly stressed caregivers, we expect to see a hi gher number of these individuals experience chroni c depression, based upon prior li terature (Schulz et al. 2001). Methods Participants

PAGE 43

32 The current study utilized the data ob tained from bereaved individuals who participated in the Changing Lives of Olde r Couples (CLOC) study, a prospective study of a two-stage area probability sample of 1532 married men and women from the Detroit Standardized Metropolitan Statistical Area. In order to be eligible for the study, individuals had to meet the following criter ia: English-speaking, married, residing in a household in which the husband was at least 65 years of age, noninstitutionalized, and able to participate in a two hour face-to-face initial interview (University of Michigan, n.d.). In addition, the researchers who desi gned the study oversampled women in order to obtain the highest number of bereaved spouses possible during the study period. Baseline interviews for the CLOC study we re conducted between June 1987 and April 1988, and approximately 68% of individuals contacted participat ed in the initial interview, a response rate that is consistent with other studied conducted in the Detroit area (see the University of Michigan CL OC website for additional details). Researchers utilized the daily obituar ies in the Detroit-area newspapers and monthly death record tapes provided by the st ate of Michigan in order to identify the initial participants who had experienced the loss of a spouse. The National Death Index (NDI) was used to confirm the deaths and to obtain the causes of death. Of the initial sample, 316 individuals lost their spouse during the study period and were asked to participate in a follow-up interview six and eight een months after the loss of their spouse; 195 individuals (62%) agreed to participate in both of these interviews. For the purposes of this study, the two cases that reported deat h as a result of murder, suicide, or homicide have been excluded because of the unique attrib utes associated with a loss of this nature (Sheskin & Wallace, 1976). An additional r eason for non-inclusion in this study was the

PAGE 44

33 classification of a participant into a bereavement trajectory that accounted for less than 5% of the dataset. The final sample for the current study was 175 (55% of the original dataset) with the primary reasons for non-pa rticipation being ill health or death and refusal. The mean age of this sample was 69.99 (s .d.=6.79) at the baseline interview and 87.40% (n=153) female. In addition, 149 ( 85.10%) of these participants identified themselves as Caucasian American and 26 (14.90%) as African American and had on average 11.57 years of educati on. The four groups did not di ffer significantly (p>.05) on any of the demographic characteristics except for race/ethnicity, as shown in Table 3. Chi-square analyses revealed a significant, non-random distribution of participants into the four groups utilized in this study by r ace/ethnicity ( 2=8.419, df=3, p=.038). African Americans who participated in this study were more likely to have a spouse who died unexpectedly and less likely to have a spous e who was diagnosed with a serious ongoing condition but did not require care than their Caucasian American counterparts.

PAGE 45

34 Table 3. Demographic Characteristics of the Sample by Caregiving Status Measures Pre-loss Caregiving Status The four groups for this study were chosen based upon their response to three questions: the first asking whether the spouse died of either an ongoing serious condition, an accident, murder, or suicide, or an unspecified cause indicated as not being an ongoing serious condition, the second asking those whose spouses died of a serious ongoing condition whether or not they provided care for their spouse 6 months prior to death, and the third asking individuals who re ported providing care fo r their spouses how stressful they found the experience to be. Fo r the question regardi ng stressfulness of the Unexpected Death No Caregiving Low Stress Caregiving High Stress Caregiving N 35 59 40 41 Age, mean years (SD) 70.37 (5.78) 70.59 (6.99) 69.95 (5.84) 68.83 (8.13) % Female 88.60 88.10 85.00 87.80 % White 74.30% 94.90% 85.00% 80.50% Education, mean years (SD) 11.48 (2.68) 11.42 (2.25) 11.67 (2.58) 11.76 (2.50) Baseline CES-D, mean (SD) -.31 (.72) .14 (1.14) -.03 (.99) .11 (1.10)

PAGE 46

35 caregiving experience, partic ipants were asked to provi de a response based upon a 5point Likert-type scale with responses ranging very stressful to not at all stressful. Individuals who responded to this question as either very stre ssful or quite stressful were included in the high stress group and indivi duals who responded as somewhat stressful, not too stressful, or not at all stressful are included in the low stress group. Participants were divided into the following four categ ories: individuals whose spouses died unexpectedly (n=35), individuals whose spous es died from a serious ongoing medical condition but did not provide ca re six months prior to th e death (n=59), individuals whose spouses died from a serious ongoing medical condition who di d provide care six months prior to the death and reported low leve ls of stress associated with the caregiving experience (n=40), and individuals whose s pouses died from a serious ongoing medical condition who did provide care si x months prior to the death and reported high levels of stress associated with the caregiving experience (n=41). Bereavement Trajectories The bereavement trajectories utilized in this study are identical to those identified in studies published by Bonanno and his colleag ues with the CLOC da taset (Boerner et al. 2005; Bonanno et al., 2002). They developed these trajectories using scores on the Center for Epidemiologic Studies-Depressi on (CES-D; Radloff, 1977) scale obtained at three time points: approximately 3 years pr e-loss, and 6 and 18 m onths post-loss. The first step in developing these trajectories involved categoriz ing participants as either having high or low pre-lo ss depression (using the 80th percentile as the cut-point); next change scores were calculated for two time points: pre-loss to 6 months post-loss and 6

PAGE 47

36 months to 18-months post-loss. These change scores were then used to develop three categories: a grief reaction wa s defined as a depression increa se relative to pre-loss by 1 standard deviation (SD) or better, improved functioning was defined as a depression decrease by greater than 1 SD, or no change if the depressi on scores remained constant or varied by less than 1 SD. The SD was defi ned separately for the high and low pre-loss depression groups. For participants with hi gh pre-loss depression, .88 standard units was used to define meaningful change, and for participants with low pre-loss depression, .57 standard units was used. In addition, in orde r to compensate for the fact that pre-loss depression scores tended to cluster around th e mean, a grief reaction was assigned when post-loss depression scores incr eased to greater than the 50th percentile for the overall dataset (N=1532, z=0). The change scores for each of the two time points were combined to create eight possible bereavem ent trajectories, and only those that were exhibited by at least 5% of the sample were used. The five patterns used in the final analysis included 185 participants and fell into the following categ ories: 1) common grief, n=18 (10.3%); 2) depressed-improved, n= 20 (11.4%); 3) resili ent, n=92 (52.6%); 4) chronic grief, n=30 (17.1%); 5) chronic de pression, n=15 (8.6%). (see Bonanno et al., 2002 for additional details). Statistical Analysis Plan Chi-square analyses were conducted in order to assess the relationship between caregiving group and bereavement trajectories. In order to gain a bett er understanding of the relationship between thes e two variables, significant chi-square analyses were

PAGE 48

37 followed up with a comparison of distribu tions for all groups using the group of individuals whose spouses died of a serious ongoing condition but did not provide care as a referent group. Results The chi-square analysis of this da ta revealed a significant, non-random distribution across the five bereavement patterns ( 2=23.43, df=12, p=.024). Table 4 shows the results of the chi-s quare analysis and the varying distributions of bereavement trajectories for our four caregiving groups. Using the group of individuals whose spouses died of a serious ongoing condi tion but did not provide care as a referent group, we can see that there are distinct variations in the di stributions. It is impor tant to note that we obtained distributions similar to those reported by Bonanno et al. (2002) however there are some differences in the overall distributions, attributable largely to the slight variation in the sample.

PAGE 49

38 Table 4. Results of the Chi-Square Analys is for Bereavement Trajectories by Caregiving Status Common Grief DepressedImproved Resilient Chronic Grief Chronic Depression Total Unexpected Death Actual Expected % Group % Trajectory 2.0 3.6 5.7% 11.1% 0.0 4.0 0.0% 0.0% 17.0 18.4 48.6% 18.5% 13.0 6.0 37.1% 43.3% 3.0 3.0 8.6% 20.0% 35.0 35.0 100.0% 20.0% No Caregiving Actual Expected % Group % Trajectory 10.0 6.1 16.9% 55.6% 10.0 6.7 16.9% 50.0% 28.0 31.0 47.5% 30.4% 6.0 10.1 10.2% 20.0% 5.0 5.1 8.5% 33.3% 59.0 59.0 100.0% 33.7% Low Stress Caregiving Actual Expected % Group % Trajectory 4.0 4.1 10.0% 22.2% 5.0 4.6% 12.5% 25.0% 22.0 21.0 55.0% 23.9% 7.0 6.9 17.5% 23.3% 2.0 3.4 5.0% 13.3% 40.0 40.0 100.0% 22.9% High Stress Caregiving Actual Expected % Group % Trajectory 2.0 4.2 4.9% 11.1% 5.0 4.7 12.2% 25.0% 25.0 21.6 61.0% 27.2% 4.0 7.0 9.8% 13.3% 5.0 3.5 12.2% 33.3% 41.0 41.0 100.0% 23.4% Total Actual Expected % Group % Trajectory 18.0 18.0 10.3% 100.0% 20.0 20.0 11.4% 100.0% 92.0 92.0 52.6% 100.0% 30.0 30.0 17.1% 100.0% 15.0 15.0 8.6% 100.0% 175.0 175.0 100.0% 100.0%

PAGE 50

39 When comparing the unexpected death group to the referent group, we see that individuals whose spouses died unexpectedly are less likely to experience a common grief (5.7% compared with 16.9%) or depr essed-improved (0.0% compared with 16.9%) bereavement reaction, but are more likely to experience a chroni c grief reaction (37.1% compared with 10.2%). When comparing the low stress caregiving group to the referent group, we see that the low stress caregivers ex perience a very simila r distribution to the non-caregiving group, with only slight differen ces across the five tr ajectories. When examining the differences between the hi gh stress caregivers and the non-caregiving groups, we see that the high stress caregivers are less likely to expe rience a chronic grief reaction (4.9% compared to 16.9%), more likel y to fall into the re silient category (61.0% compared to 47.5%), and slightly more lik ely to experience chronic depression (12.2% compared to 8.5%). Discussion The results from this study were cons istent with those in the prior work by Bonanno and his colleagues (2001) and also the hypotheses based upon prior caregiving and bereavement literature as it relates to depression outcomes for individuals who experience varying levels of stress while ca regiving (Schulz et al., 2001). Our sample does reiterate the idea that individuals who experience unexpected losses are more likely to experience chronic grief, and this is not surprising given the difficult circumstances surrounding these losses. However, our results also indicate that almost half (48.6%) of the spouses who experienced an unexpected loss fell into the resilient category. This is a

PAGE 51

40 finding that deserves attenti on and warrants further investig ation, particularly aimed at gaining a better understanding of the protective factors that account for low levels of depression both at short and long term bereavement follow-up points. Our hypothesis that individuals who reported a low level of stre ss associated with the caregiving experience would show similar bereavement pa tterns to those who did not provide care was confirmed by the analys is, while there were differences when comparing the highly stressed caregivers to the non-caregiving group. The highly stressed caregivers were sli ghtly less likely than non-caregivers to experience a common grief reaction which we did not hypothesize, and slightly more likely to experience chronic depression, which we did hypothesize. Interestingly, however, even though the highly stressed caregivers were more likely to experience chronic depression, the actual percentage was quite low (12.2 %). These findings are consis tent with the literature and indicate that the subjective a ppraisal of the stressfulness of the caregiving experience has an effect on mental health outcomes (Schul z et al., 2001), but also leaves us with a number of questions about the type of grie f experienced by the majority of the widowed spouses. As with the unexpected death group, the bereavement pattern that was demonstrated by the highest percentage of individuals in each caregiving group was resilience. Approximately 55.0% of low stress caregivers and 61.0% of high stress caregivers demonstrated consistently low leve ls of depression both pr e-loss and at both short and long term bereavement follow-ups. This indicates that the majority of individuals who experience spous al loss (52.6% of our sample) do not show low levels of depression either pre-loss or during ber eavement, regardless of pre-loss caregiving experience, and also leads researchers to fu rther examine the indivi dual characteristics

PAGE 52

41 that enable widowed spouses to exhibit this me ntal health outcome. It is also important to note that although the descriptive analys es showed that African Americans who participated in this study were more likel y to have a spouse who died unexpectedly and less likely to have a spouse who was diagnosed with a serious ongoing condition but did not require care than their Caucasian American counterparts, we were unable to interpret these results due to the small number of Afri can American participants. However the fact that we did find this difference suggests that future research needs to include a diversity of participants in order to examine the differing caregiving experiences of multiple racial and ethnic groups. It should be noted that the approach used to classify participants into bereavement trajectories utilized a conser vative method that may potentia lly bias the sample towards resilience. Future research could attemp t to replicate the cr eation of bereavement trajectories by using slightly le ss conservative criteria to denote change. One possible approach that may be potentially utilized is prediction analys is (Hildebrand, Laing, & Rosenthal, 1977). This type of analyses was used by Whitlatch, Zarit, and von Eye (1991) to reanalyze data from a previous study examining the efficacy of caregiver interventions (Zarit, Anthony, & Boutselis, 1987) and involves the comparison of observed frequencies of outcome scores with predicted outcomes that take into account the initial levels of dependent measures. This rean alysis found that a systematic evaluation of initial levels of dependent measures in the participants is important in detecting change. Although this was outside the scope of this work, we believe that it does need to be incorporated into future res earch. A second potential approach is the use of a relatively new procedure that is designed to isolate di stinct trajectories and fit a

PAGE 53

42 mixture model to calculate the probability of membership in each latent class for each participant (Jones, Nagin, & Roeder, 2001). The be nefit of this approach is that it would allow the development of the trajectorie s to be data-driven and it may produce trajectories that are different from the one s developed by Bonanno and his colleagues. One possible area for future research is th e utilization of the revised stress process model recently published by Folkman (2001) to attempt to gain a be tter understanding of how individuals use various coping mechanis ms which result in either positive or negative mental health bereavement outcomes. This model includes increased attention to issues such as the use of meaning-based coping and the benefits of caregiving which need to be addressed in future research and could potentially offer insight into the mechanisms which allow individuals to adjust to spousal loss in a healthy way. In addition, the continued examination of the role of social support and its possible protective factors against ne gative mental health outcomes both during caregiving and following the loss of a spouse would be hi ghly beneficial both for researchers and practitioners (Robinson-Whelen, Tada, MacCa llum, McGuire, & Kiecolt-Glaser, 2001; Zettel & Rook, 2004). Prior literature has show n that highly stressed caregivers may be particularly likely to experience role e ngulfment (Aneshensel et al., 1995; Pearlin, Mullan, Semple & Skaff, 1990) and thus to ha ve difficultly retaining social roles while caregiving. Similar to the re source depletion hypothesis (S chulz, Newsom, Fleissner, Decamp, & Nieboer, 1997), caregivers who become disengaged socially may find it difficult to develop such relationships afte r bereavement. Future research that systematically examines the maintenance of social roles and social support both during caregiving and subsequent loss a nd their impact on mental hea lth is needed in order to

PAGE 54

43 address these issues. There is also a growing interest and body of literature that focuses on the positive aspects of caregiving and ad aptation to bereavement, and the findings from our study contribute to the ideas presente d in a recent study which also suggests that this is a very important direction to be ta ken (Boerner, Schulz, & Horowitz, 2004). This study found that higher levels of post-loss de pression and grief were associated with higher levels of caregiver benefit (as m easured by an 11-item scale assessing the caregiver’s mental-affective stat e in relation to the caregiving experience), and that this relationship was particularly st rong for grief. The authors of this study concluded that individuals who experience more positive aspects of the caregiving experience may have a more difficult adjustment period due to th e simultaneous loss of both a family member and a meaningful role and suggested that fu ture research focus on the whole picture of the caregiving experience. A significant portion of the picture of th e relationship between caregiving and bereavement is not captured by focusing only on negative outcomes, su ch as depression, and researchers need to expa nd their outcome variables to include positive measures, such as enjoyment of caregiving and empowerment, in order to begin to fill in these gaps. In order for effective interventions to be developed which decrease negative outcomes during bereavement, researchers need to have a clearer picture of all of the pre-loss characteristics that contribute to mental health outcomes during bereavement. Overall the results from our study highlight the importance for continued research on the relationship between caregiving and bereavement that focuses on a variety of both positive and negative outcome domains. The fi ndings related to the high prevalence of a resilient bereavement trajectory, regardless of caregiving status, is promising both for

PAGE 55

44 widowed spouses and practitioners, however it should lead us to further examine the individual characteristics of th ese resilient widows in order to gain a better understanding of the underlying mechanisms.

PAGE 56

45 Chapter Four Predictors of well-being in bere aved former hospice caregivers: The role of caregiving stressors, appraisals, and social resources Abstract The current literature on caregiving a nd bereavement indicates that the relationship between these two common life events is complex and needs to be further studied in order to gain a more comprehensiv e understanding of thei r interaction. In the current project, 50 spouses of hospice patients were assessed while caregiving and at an average of 4 months after the death on a variety of measures, including caregiving stressors, appraisals, social resources, and well-being. A stress process model was utilized in order to examine which pre-lo ss factors were associated with post-loss depression, life satisfaction, and grief in hi erarchical multiple regression models. Our results indicated that caregi ver appraisals (stressfulness of functional impairment and positive aspects of caregiving) were not predictors on any of our well-being outcomes. However, fewer months caregiving was a si gnificant predictor of both higher depression and grief post-loss. In additi on, lower levels of social activit ies, smaller social networks, and satisfaction with social support were significantly associated with higher post-loss depression. Results support both the res ource depletion and anticipatory grief

PAGE 57

46 hypotheses, and suggest that sh ort-term bereavement outcomes are largely independent of factors that predict well-bei ng while caregiving. Future st udies should address whether long-term bereavement outcomes differ by base line caregiving charac teristics to guide intervention research. Introduction The death of a spouse is one of the most stressful events that an individual may encounter over the course of his or her lifetime (Stroebe & Stroebe, 1987). Previous bereavement research has demonstrated that the loss of a spouse has a profound effect on an individual’s well-being, in cluding increased incidence of depression, decreased life satisfaction and social activities, and deteri oration of physical h ealth (i.e., Bonanno, Notarius, Gunzerath, Keltner, & Horowitz 1998; Carr, House, Wortman, Nesse, & Kessler, 2001; Davis & Nolen-Hoeksema, 2001; Folkman, 2001; Schulz & Beach, 1999; Stroebe, Stroebe, Abakoumkin, & Schut, 1996). For older adults who die af ter a chronic illnes s, spousal bereavement frequently follows extensive and stressful periods of informal caregiving (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Minino & Smith, 2001). Family caregiving, particularly for spouses, places an individual at risk for a variety of mental and physical health decrements as a result of the caregivi ng experience. Some of the health effects that have been documented in the literature include increased incidence of depression, slower wound healing, and increased mortalit y (i.e., Kiecolt-Glaser, Marucha, Malarkey,

PAGE 58

47 Mercado, & Glaser,1996; Pinquart & Sorense n, 2003; Schulz & Beach, 1999; Schulz, O’Brien, Bookwala, & Fleissner, 1995). Hospice caregivers may be at particular risk for these mental and physical health declines, as they are often highly engr ossed in the caregiving experience while simultaneously facing the loss of a loved one. One recent study comparing hospice caregivers to non-caregivers found that spousal caregivers of hospice patients spent an average of approximately 97 hours/week in the caregiving role and were at an increased risk for depression, decreased life satisf action, and health problems compared with demographically matched non-caregiving controls (Haley, LaMonde, Han, Narramore, & Schonwetter, 2001). Stress Process Models and their Applic ation in Caregiving and Bereavement Stress process models are increasingl y used for studying both caregiving and bereavement (i.e., George, 1990; Haley, LaMonde, Han, Burton, & Schonwetter, 2003; Pearlin, Mullan, Semple, & Skaff, 1990; St roebe & Schut, 1999). This theoretical framework posits that the relationship between a stressful life event and the effect of that event on an individual’s well-be ing is a function of the unique interaction of the appraisal of the stressor and the resources the individua l has available. A pplied to the study of caregiving and bereavement, stress process th eories suggest that well-being outcomes, such as depression or life satisfaction, are not so lely based on an individual’s exposure to caregiving stressors such as functional impairments or duration of caregiving. Each individual appraises the stressfu lness of the situation as well as examines the availability of resources, such as social support, and the co mbination of all of these factors affect the

PAGE 59

48 degree to which that stressor affects the caregi ver’s well-being. This theoretical approach can be helpful in examining two opposing hypot heses in the caregiving and bereavement literature: resource depletion a nd relief. Resource depletion theory is based on the idea that a depletion of resources results fro m prolonged exposure to caregiving stressors leaving the surviving spouse more vulner able to negative outcomes during widowhood. Alternatively, relief theory states that th e reduction in caregiv ing burden following the loss of a spouse results in improvements in me ntal and physical health outcomes (Schulz, Newsom, Fleissner, Decamp, & Nieboer, 1997). The first theory is based on the ideas that resources, such as social support, buffer the effects of stressful life events, and that the demands of caregiving lead to a depletion of these resources. Thus, when the death occurs, the bereaved caregiver is at risk for increased distress due to a depletion of resources over time. The second hypothesis is based on the idea th at the loss of the burden of caregiving also means the removal of an important individual stressor and that this change in circumstances may lead to improvements in bereavement outcomes. In addition to the resource depletion and relief hypot heses, stress process models can also be used to examine anticipatory grief. The an ticipatory grief hypothesis is based on the idea that individuals who have time to psychologi cally prepare themselves for the loss of a loved one may be at a decreased risk for diffi culties after the death. It is believed by some researchers that this theory may be pa rticularly important w ithin the context of caregiving and bereavement, although the research to date is inconclusive (Schulz et al., 1997). The studies using stress process models to examine well-being outcomes during caregiving have been inconclusive about the effect of objective measures of caregiving

PAGE 60

49 stressors (Beery, Prigerson, Bi erhals, Santucci, Newsom, et al., 1997; Given & Given, 1996; Haley & Bailey, 1999; Kurtz, Kurtz, Given, & Given, 1995; Sales, Schulz, & Biegal, 1992). However the findings related to subjective appraisals are more consistent and suggest that caregivers are more likely to be depressed if they appraise caregiving tasks as stressful, feel incapable of fulfilling their caregiving roles adequately, or are unable to find meaning or satisfaction in their role as caregiver (i.e., Folkman & Moskowitz, 2000; Haley & Bailey, 1999; Ober st, Gass, & Ward, 1989; Weitzner, Haley, & Chen, 2000). Taken together, the research examining the role of social support finds that caregivers with larger so cial networks, increased particip ation in social activities, and higher perceived satisfaction w ith social support are less like ly to experience distress (i.e., Northouse, 1988; Schulz et al. 1995; Weitzner et al., 2000). However, there is some question as to whether caregivers who are highly involved in their caregiving roles are able to maintain adequate levels of so cial support and social interaction. Research suggests that highly stressed caregivers may be particularly likely to experience role engulfment (Aneshensel et al., 1995; Pearlin et al., 1990) and thus to have difficulty retaining social roles while caregiving. While there is a fairly extensive body of literature which examines the impact of stressors, appraisals, and social resour ces on mental and physical health during caregiving, very few studies have been conducte d to date examining the impact of these variables on bereavement outcomes. The rese arch to date that examines objective measures of caregiving stressors and th eir impact on well-being outcomes during bereavement has been sparse. Of the studies that are available, most utilized one postloss interview, which took place between four a nd six months after the loss. Collectively,

PAGE 61

50 the findings from these studies (discussed below in greater detail) found that longer duration of caregiving, negative appraisals of the caregiving experience, and lower social resources were related to poorer well-bei ng outcomes during bereavement. The one study related to caregiving stressors found that longer duration of caregiving was correlated with higher levels of relief experienced by the ca regivers following the death (Cleiren, van der Wal, & Diekstra, 1988). It is possible that this finding relating longer duration of caregiving with highe r levels of relief is evidence of anticipatory grief; however, the findings in this area have not been definitive (Schulz et al., 1997). The handful of studies that have been conduc ted related to caregiving appraisals and bereavement outcomes have focused on the stre ssfulness of the caregiving experience. Overall these studies suggest that caregiver dissatisfaction with caregiving abilities and higher levels of stressfulness of the caregivi ng experience are related to higher levels of depression and distress during bereavement (B ass & Bowman, 1990; Cleiren et al., 1988; McHorney & Mor, 1988). One study (discussed in detail below) that focused on positive aspects of caregiving found that higher preloss caregiving perceptions of benefit from caregiving was associated with higher levels of post-loss depressi on and grief (Boerner, Schulz, & Horowitz, 2004). Social resource variables are the most commonly examined in this body of literature, with the findings suggesting that higher levels of social resources and satisfaction with these resource s while caregiving are associated with better bereavement outcomes (Bass & Bowman, 1990; Cleiren et al., 1988; McHorney & Mor, 1988). There have been two studies conducte d to date that included extensive information about circumstances during care giving, and utilized longer term follow-up

PAGE 62

51 points during bereavement. The first study examined caregiving stressors and their relationship to bereavement outcomes one y ear post-loss and found that the higher the amount of assistance with functional impairme nts while caregiving, the higher the levels of depression post-loss (Mullan, 1992). To our knowledge, there are no studies currently available examining appraisals of the careg iving experience and th eir relationship to bereavement utilizing prospective, long term follow-up points. For the domain of social resources, one relatively small qualitative study that conducted unstructured interviews with bereaved caregivers on an average of 18 months post loss (range was 3 months to 4 years) found results similar to those prev iously discussed. Individuals who reported higher levels of social support while careg iving also reported easier adaptation to bereavement (Sankar, 1991). Pre-loss depression is genera lly believed to be signifi cantly associated with postloss depression and grief and th is association has been seen in previous caregiving and bereavement literature (Boerner et al., 2004). Researchers need to place more emphasis on studying the relationship betw een caregiving context and bereavement outcomes. A clearer picture of the pre-loss characteristics that affect post-loss well-being is needed so that they can be used to develop effective interventions that can then be implemented with at risk caregivers. To our knowledge, only one study has been pu blished to date that focuses on the positive aspects of caregiving as an appraisal va riable within the context of adaptation to bereavement. The findings from this recent st udy suggest two things; the first that this is a very important direction to be taken, and th e second that this rela tionship may be more complex than it seems from a theoretical st andpoint (Boerner et al., 2004). This study

PAGE 63

52 found that higher levels of post-loss depressi on and grief were associated with higher levels of pre-loss caregiver benefit (as measured by questions pertaining to positive emotions related to the caregiving experience), and that this relations hip was particularly strong for grief. The authors of this st udy concluded that indi viduals who experience more positive aspects of the caregiving expe rience may have a more difficult adjustment period due to the simultaneous loss of both a family member and a meaningful role and suggested that future research focus on a mo re comprehensive picture of the caregiving experience. In addition this study highlighted the concept that grief and depression are outcomes representing different facets of ber eavement with unique pathways and need to be addressed as such in future research. One area of the literature that is particularly important in helping both practitioners and researchers to gain a better understandi ng of the relationship between caregiving and bereavement is the examination of multiple measures of well-being outcomes. There is some debate in the literature at this time as to whether or not positive and negative affect variables are mutually exclusive. As previously discussed, bereavement researchers are increasingly aw are of the differences between grief and post-loss depression and the need to address th em as distinct facets of well-being during bereavement. The current study will add to this growing body of literature within the context of bereavement by examining the re lationship between caregiving stressors, appraisals of those stressors, and the availabil ity of resources, and how they interact with multiple outcome measures, both positive and negative. The Current Study

PAGE 64

53 The current study addresses the extent to which stressors, appraisals, social support, and well-being while caregiving pred ict well-being during bereavement. More specifically, we were interested in the fo llowing questions: How do caregiving stressors (duration of caregiving, hours/ week caregiving, and functiona l impairments), appraisals (stressfulness of functional impairments and positive aspects of caregiving), and objective (size of social netw ork, overall social support, num ber of visits and social activities) and subjectiv e (satisfaction with social support) measures of social support while caregiving affect post-loss depression, life satisfaction, and grief in bereaved former hospice caregivers? In order to addr ess this question, we used a subset of a dataset that recruited elderly spousal caregiv ers of terminally ill patients from a large, non-profit hospice. This dataset includes both preand post-loss data and includes a variety of widely used and validated meas ures that allow for the examination of caregiving stressors, appraisals, and social support and their effect on both positive and negative well-being outcomes during bereavement Drawing on prior literature and the te nets of stress process theory, we hypothesized that higher pre-loss levels of bot h depression and life sa tisfaction would be associated with higher post-loss levels of both depression and life satisfaction. As far as the relationship between caregiv ing stressors and our three outcome measures, resource depletion theory predicted that higher pre-lo ss levels of caregiving stressors (hours/week caregiving, duration of careg iving, and ADL/IADL impairment) would be associated with higher levels of depre ssion and grief and lower levels of life satisfaction post-loss after controlling for baseline. However, th e anticipatory grief hypot hesis predicted that individuals who experienced longer duration of caregiving would be associated with

PAGE 65

54 lower levels of grief and depr ession post-loss. Within the domain of appraisals, resource depletion hypothesis suggested that higher pre-loss levels of perceived stressfulness of the caregiving experience and lower positive aspects of caregiving would be associated with higher levels of post-loss depression a nd grief and lower levels of post-loss life satisfaction after controlling for baseline well -being measures. Alte rnatively, the relief hypothesis predicted that higher levels of perceived levels of stressfulness of caregiving and higher levels of positive aspects of caregiving would be associated with a greater sense of relief post-loss, partic ularly higher levels of life sa tisfaction and lower levels of depression and grief after controlling for ba seline well-being measur es. In relation to social resources, the resource depletion hypothe sis predicted that af ter controlling for baseline well-being measures, higher levels of objective (size of so cial network, overall social support, number of visits and social activities) and subjective (satisfaction with social support) levels of social support pre-loss would be asso ciated with lower levels of post-loss depression and grief and higher le vels of post-loss life satisfaction. Methods Participants Spousal caregivers of terminally ill older patients diagnosed with either end-stage dementia or lung cancer were recruited as pa rt of a larger resear ch project conducted by the LifePath Hospice of Tampa, Florida, and the faculty at the School of Aging Studies at the University of South Florida, Tampa, Flor ida. The current project involves a followup of these participants. Full details regarding the recruitment of the original sample are

PAGE 66

55 provided in another paper (Haley et al., 2001). Briefly, the or iginal project involved the recruitment of eighty (N=80) spousal caregiv ers over the age of 50 who were caring for terminally ill dementia or lung cancer hospice patients from LifePath Hospice of Tampa, Florida, patient records. All patients we re diagnosed by their family physicians with either end-stage dementia (N=40) or l ung cancer (N=40) and we re certified by their family physicians as having six (6) months or le ss of life expectancy. To be eligible for the research project, elderly spousal caregi vers had to meet the following eligibility criteria: (a) confirmation that they were ei ther the husband or wife of the patient; (b) confirmation that they were the primary car egiver for the patient; and (c) confirmation that they were able to understand th e spoken and written English language. Data Collection Procedure Each eligible caregiver was contacted via telephone by a trained research assistant and asked if he or she would be interested in participating in the research project. Those who agreed were interviewed in their homes by a trained research assistant. The appointments were usually scheduled within two weeks of the pa tient’s admission to hospice. The caregivers received no financial compensation for their participation in the research project. The current project utilized fi fty (50) of the original ei ghty (80) elderly spousal caregivers of either end-stage dementia (N= 21) or lung cancer (N=29) patients. For the current sample, the mean length of time betw een initial interview and the death of the patient was approximately four (4) months, however this varied by patient diagnosis

PAGE 67

56 (mean of 145 days for dementia patients and 57 days for cancer patient s). It is important to note that an average of four months of stay in hospice is atypical and the current mean length of stay for most hospice patients is 55 days with a median of 22 days (National Hospice and Palliative Care Organization, n.d.). The bereaved caregivers were recruited in the same manner as described previ ously and were agai n administered a comprehensive structured personal interview and questionnaire in their homes, usually within four and one half (4 ) months fo llowing the date of the patient’s death. The minimum time elapsed from the patient death to the caregiver bereavement interview was one month and the maximum time elapsed fr om the patient deat h to the caregiver bereavement interview was ten months. Attrition from Time 1 to Time 2 The average amount of time that elapsed between the Time 1 (T1-Pre-loss) Interview and the Time 2 (T2-Post-loss) In terview was eight (8) months. The minimum amount of time that elapsed between the T1 and the T2 interviews was approximately two (2) months and the maximum amount of time that elapsed between the T1 and T2 interviews was approximately two (2) years. During this period of time, thirty (30) participants were lost from the study (13 decl ined participation, 9 m oved out of the area, and 8 were not recertified by LifePath Hospi ce). The participants who remained in the study did not differ significantly (p>.05) from those who were not included in the followup interview on any of the demographic character istics or baseline variables used in this study.

PAGE 68

57 Measures Caregiver Demographics Caregiver demographic and descriptive information, specifically age, gender, education, and race/ethnicity, were assessed through caregiver self-r eport during the preloss interview. Caregiving Stressors To assess caregiving stressors, inform ation was gathered during the pre-loss interview on a variety of indi cators including: dur ation of caregiving in months, hours of caregiving per week, patient self-care probl ems, and patient diagnosis. Duration of caregiving and hours per week caregiving were assessed using self-repo rt. Patient selfcare problems (alpha=.89) included impairments in both functional abilities (Activities of Daily Living; ADL; Katz, Ford, Moskowitz, Ja ckson, & Jaffe, 1963), such as bathing and dressing, and higher-level functioning abil ities, such as managing finances and completing household chores (Independent Ac tivities of Daily Living; IADL; Lawton & Brody, 1969), with higher scores indicating mo re self-care problems. Patients included in this study were diagnosed with either de mentia or lung cancer, and were identified as such when initially recruited into the study. Appraisals Stressfulness of ADL and IADL impairme nts (alpha=.87) was assessed during the pre-loss interview using a 0 to 3 Likert-type scale, with responses ranging from “not at all

PAGE 69

58 stressful” to “very stressful” for each of the functional abilities (Katz et al., 1963; Lawton & Brody, 1969). Stressfulness scores were adjusted for the number of problems experienced by the patient, so scores refl ect the average subjective stressfulness of problems, independent of how many problem s the caregiver faced, with higher score indicating more stressful appraisals. Positive aspects of caregiving (alpha =.80) were assessed during the pre-loss interview using an 11-item self-report measur e developed by Schulz and his colleagues (Schulz, Newsom, Mittelmark, Burton, Hirsch & Jackson, 1997). Each item begins with “Providing help to (care recipient) has…” a nd contains specific it ems such as “given more meaning to my life” and “made me feel usef ul.” The participants were asked to rate each item using either a “yes” or “no” re sponse, with higher scores indicating more caregiving benefit. Social Resources Social activities (alpha=.62) were assessed during the pre-loss interview using the 16 item Multilevel Assessment Inventory (MAI; Lawton, Moss, Fulcomer & Kleban, 1982), which utilizes a 5 point Likert-type scale with responses ranging from “0” indicating “never” and 5 indicating “12 or more times” to examine level of caregiver participation in social activities during th e previous month. Some of the items included in this measure are eating at a restaurant with friends and relatives, attending club activities, and doing volunteer work, with higher scores indicating higher number of social activities. Size of the social network (alpha=.74) was assessed during the pre-loss interview

PAGE 70

59 using an 8 item measure from the revised Lubben Social Network Index (Lubben, 1988). Two types of questions are included in this measure: number of contacts and frequency of contact. Participants were asked to list the number of re latives and friends they hear from or feel close to using a 6 point Like rt-type scale with resp onses ranging from 0 indicating “zero” and 6 indicating “nine or more.” For questions addressing the frequency of contact with a close friend or relative, the same response range was used, with responses indicating “less than monthly” to “daily.” Higher scores on this measure indicated a larger social network. Social support (alpha=.76) and satisfacti on with social support (alpha=.63) were assessed during the pre-loss interview using a 14 item scale by Krause and BorawskiClark (1995). Specifically, to assess social support, participan ts were asked to rate how often others have provided ta ngible, emotional, and inform ational support during the past month using a 4 point Likert-type scale with responses ranging from 1 indicating “never” and 4 indicating “often.” Sa tisfaction with soci al support was assessed by participant ratings for each of the three types of social support using the same response range with responses indicating “not at all” to “very” satisfied. For both of these measures, higher scores indicate higher levels of received social support and satisfaction with social support. Well-Being Depression (alpha=.81) was assessed during both the pre-loss and post-loss interviews using the 20 item Center for Ep idemiological Studies-Depression Scale (CESD; Radloff, 1977). Participan ts rated how often they experienced symptoms associated

PAGE 71

60 with depression, such as restless sleep, poor a ppetite, and feeling lonely, over the past week using a 4 point Likert-type scale with responses ranging from 0 indicating “rarely or none of the time (less than once a day)” to 3 indicating “most of the time (5-7 days)” with four items, such as enj oying life, feeling happy, and f eeling hopeful about the future reverse coded. Higher scores on the CESD indicate higher frequency/severity of depressive symptomatology. Life satisfaction (alpha=.76) was as sessed during both pre-loss and post-loss interviews using the 13 item Life Satisf action Index-Z (LSI-Z; Wood, Wylie, & Sheafor, 1969). Participants were asked to indicate if they agreed, disagreed or were uncertain about items related to general st atements about life, such as “I am just as happy as when I was younger” and “I have made plans for things I’ll be doing a month or year from now.” In addition five items in this measure, in cluding “this is the most hopeless time of my life” and “most of the things I do are boring or monotonous ,” were reverse-coded. Higher scores on the LSI-Z indicate highe r levels of overall life satisfaction. Grief (alpha=.88) was measured during th e post-loss interview using 13 items from the Texas Revised Inventory of Grief (TRI G; Faschingbauer, Zisook, & DeVaul, 1987). This scale assessed symptoms associated with separation distress, such as “sometimes I very much miss the person who died” and “I am unable to accept the death of the person who died,” using a 5 point Likert-type scal e with responses rangi ng from 1 indicating “completely false” to 5 indicating “completely true.” Higher scores on this scale indicate higher levels of grief. Statistical Analysis Plan

PAGE 72

61 The first step was to examine correlations between the demographic characteristics, specifically age, gender, education, and race/ethnicity, and the three outcome measures, depression, life satisfacti on, and grief, to determine if any were needed as covariates in the regression anal yses. Due to the small sample size, only demographic characteristics and pre-loss ca regiving well-being va riables that were significantly correlated (p<.05) with the dependent variab les were utilized in the regression analyses. The second step was to conduct univari ate analyses in order to get an understanding of which variable s were significant predictors of postloss depression, life satisfaction, and grief. The domains of variable s used for this step of the analysis include pre-loss caregiver well-being, ca regiving stressors, appraisals, and social resources. This step allowed us to determine which predic tors to use in the regression models. The third step was to utilize three hierarchical multiple regression models (one for depression, one for life satisfaction, and one fo r grief) using the significant predictors from the previous analyses in order to gain a better understanding of the relationships between pre-loss characteristics and their e ffect on post-loss well-being outcomes. In addition, using this rationale for selection of predictor variables allowed us to limit the number of predictors in or der to accommodate the limited sample size. The order of entry for the regression models was as fo llows: (1) any necessary covariates (as determined by step one) and pre-loss careg iver well-being measures, (2) objective measures of caregiving, (3) s ubjective appraisals of the car egiving experience, and (4) objective and subjective meas ures of social support.

PAGE 73

62 Results Demographics and Basic Descriptive Analysis As shown in Table 5, the mean age of the sample was 72.84 at the pre-loss interview and predominantly female. In addition, these spousal hospice caregivers identified themselves as primarily Caucasian American (n=39), with 3 participants identifying themselves as African America n, 7 as Hispanic American, and 1 as Asian American (for the purpose of the statistical analyses, race/ethnicity was examined as a dichotomous variable, with individuals categoriz ed as either White or Non-White). Table 5 also indicates that the e ducation level of the particip ants in this study varied. Table 5. Demographic Charact eristics of the Sample Age, mean years (SD) 72.84 (10.21) % Female 80.00 % White 78.00 Education % <8th grade % High School/Some College % College Degree 34.00 52.00 14.00

PAGE 74

63 In order to gain a bette r understanding of the caregiv ing stressors, appraisals, social resources, and well-being of our pa rticipants, Table 6 shows the means and standard deviations for all variables utilized in the univariate and regression analyses. Table 6. Means and Standard Deviations for all Study Variables Variable Actual Range Mean SD Well-Being Measures Pre-loss Depression Pre-loss Life Satisfaction Post-loss Depression Post-loss Life Satisfaction Post-loss Grief 0-60 0-26 0-60 0-26 1-5 16.42 15.58 23.28 14.38 3.81 9.17 5.67 10.11 5.47 .79 Caregiving Stressors Duration (months) Hours/week ADL/IADL Impairments Diagnosis n/a n/a 0-14 n/a 25.18 94.74 11.28 n/a 32.86 70.10 3.33 n/a Appraisals Stressfulness of ADL/IADL Impairments Positive Aspects of Caregiving 0-3 0-11 .80 8.18 .74 2.90 Social Resources Social Activities Social Network Social Support Satisfaction w/ Social Support 0-80 0-40 11-44 3-12 9.36 25.54 26.52 10.36 6.12 6.92 6.12 1.76

PAGE 75

64 The initial set of analyses indicate d that only one correla tion was significant between the caregiver demographics and our three outcome measures (post-loss depression, post-loss life satisfaction, and postloss grief): race/ethnicity and post-loss depression (see Table 7). This significant ne gative correlation indi cates that Non-White hospice caregivers report higher levels of depressive symptomatology in our sample. Univariate Analyses The results of the univariate correlational analysis for all of the variables included in this study are displayed in Table 7. The results of these analyses indicated that lower baseline depression, fewer months caregiving, lower levels of soci al activities, smaller size of social network, and lower satisfaction with social support were all significantly correlated with higher post-l oss depressive symptomatol ogy. Neither of our pre-loss measures of appraisal of the caregivi ng experience, stressfulness of ADL/IADL impairments and positive aspects of caregi ving, was significantly correlated with postloss depression. Higher levels of satisfaction with soci al support while caregiving was the only variable significantly correlate d with higher levels of life sa tisfaction post-loss. None of the caregiving stressor, appraisal, or obj ective social resource measures, or life satisfaction pre-loss were significantly corr elated with life sati sfaction post-loss. (Follow-up regression analysis was not conducte d for post-loss life satisfaction due to the fact that only one variable was significantly correlated with this outcome variable.)

PAGE 76

Table 7. Correlation Matrix for all Study Variables p <.05 ** p <.01 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Caregiver Demographics 1. Age 2. Gender 3. Education 4. Race/Ethnicity 1.00 -.10 1.00 .21 -.28** 1.00 .20 .22 .16 1.00 -.10 .11 -.20 -.07 .02 -.17 .31 .25 .12 .04 .32* .10 .51** .18 -.08 -.05 -.39* .01 .06 -.39** .17 -.12 -.20 -.01 -.09 -.16 .07 -.09 -.15 .03 .01 -.08 .01 .03 .26 .22 -.29* .20 -.14 .22 -.06 .41** -.11 .18 .22 .16 -.04 .22 -.13 -.04 -.21 -.41** .18 -.08 .03 .28 -.27 .02 -.28 -.27 Caregiving Stressors 5. Duration (months) 6. Hours/week 7. ADL/IADL Impairments 8. Diagnosis 1.00 -.15 1.00 .30 .04 1.00 -.48** .323* -.52** 1.00 -.08 .31* .26 .18 .14 .11 .08 .08 .08 -.26 .20 -.14 -.03 .01 -.30* .26 .12 -.02 .01 .06 .07 -.08 -.18 -.03 -.15 .06 .18 .14 .15 -.10 .15 .06 -.31* .05 -.17 .08 .25 -.01 .02 -.04 -.35* .13 -.31* .26 Appraisals 9. Stressfulness of ADL/IADL Impairments 10. Positive Aspects of Caregiving 1.00 .28* 1.00 .03 -.21 -.11 .23 -.15 .30* -.20 .23 .15 -.37** -.11 .19 .10 -.01 -.01 -.01 .16 .11 Social Resources 11. Social Activities 12. Social Network 13. Social Support 14. Satisfaction w/ Social Support 1.00 .18 1.00 -.07 .54** 1.00 .17 .42** .53** 1.00 -.20 -.26* .24 -.46** .33* .12 .07 .28 -.28* -.33* -.25 -.41** .09 .22 .17 .31* -.27 .08 -.05 -.07 Well-Being Measures 15. Pre-loss Dep 16. Pre-loss LS 17. Post-loss Dep 18. Post-loss LS 19. Post-loss Grief 1.00 -.46** 1.00 -.29* -.13 1.00 -.24 .22 -.60** 1.00 .20 .01 .61** -.41** 1.00 65

PAGE 77

66 Fewer months in the caregiving role and lower number of ADL/IADL impairments were significantly correlated with grief in the correlational analysis. None of the baseline measures of well-being, apprai sal variables, or social resource measures were significantly correlated with grief in our sample. Regression Models The regression model with post-loss depr ession as the dependent variable is shown in Table 8. In the final regressi on model, being non-White and having fewer months in the caregiving role were associ ated with higher post-loss depression. The block of social resource va riables accounted for significan t variance, but none of the individual variables reached statistical significance. Th e model accounted for 39% of variance in post-loss depression. The model with grief as the dependent variable was examined using linear regression due to the fact that the two variables that were significantly correlated with grief were both from the same block of the proposed hierarchi cal regression model (caregiving stressors). The tw o variables utilized in this regression model were duration of caregiving and ADL/IADL impairments. This block of variables accounted for 17.2% of the variance in the model. When both va riables were considered simultaneously the number of ADL/IADL impairments was not a ssociated with higher levels of grief postloss (B=-.05, =-.23, p=.111), but fewer months in th e caregiving role was associated with higher levels of grief (B=-.01, =-.29, p=.04).

PAGE 78

67 Table 8. Regression Model for Post-loss Depression Variable B R2 R2 Step 1: Demographics and Pre-loss Well-Being Race/Ethnicity (Non-White) Depression -6.52 .07 -.27* .06 .23* Step 2: Caregiving Stressors Duration of Caregiving -.08 -.25* .29* .06* Step 3: Social Resources Social Activities Social Network Satisfaction with Social Support -.20 -.22 -1.29 -.12 -.15 -.22 .39* .10* *p<.05 Discussion Although there has been a growing interest in th e expansion of both the caregiving and bereavement literature to include both positive and negative assessment measures as both predictor and outcome variables, this literature is still relatively small, particularly when it comes to the relationshi p between these two life events. This study sought to identify how pre-lo ss caregiving experiences were associated with post-loss depression, life satisfaction, and grief usi ng a stress process theory framework in a sample of bereaved former hospice caregivers.

PAGE 79

68 Both the univariate and regression anal yses highlighted the importance of studying multiple domains of well-being outcome s, as the significant associations and predictors for depressi on, life satisfaction, and grief showed great vari ability. The results of the univariate analyses indicated that being non-White, having lower baseline depression, spending fewer months caregiving, having lower leve ls of social activities, having smaller size of social network, and ha ving lower satisfaction with social support were all significantly correlated with highe r post-loss depressive symptomatology. Once these variables were placed into the regression model, being non-White, fewer months caregiving, and having lower levels of social resources remained as significant predictors. The findings from our analyses were consistent with the anticipatory grief hypothesis that longer duration of caregiving was relate d to better well-being outcomes during bereavement and the resource depletion hypothesi s that higher levels of social resources would be associated with lower depression pos t-loss in our sample. However, neither the resource depletion or relief hypothesis related to the role of apprai sals of the caregiving experience was supported in this study. Intere stingly, pre-loss depression was not very highly correlated with post-lo ss depression in our study. Prior studies have found relative consistency in depression scores over time wh ile an individual is still in the caregiving role and much higher correlations between pr e-loss and post-loss depression (Boerner et al. 2004; Goode et al., 1998), however the latter finding has not always been consistent (Bonanno et al., 2002). It is likely that the s hort-term nature of our bereavement followup was important in this finding as prior re search has shown that it is common for grieving persons to be acutely distressed regard less of their baseline levels of depression, particularly in the short te rm (Raphael, Minkov, & Dobson, 2001).

PAGE 80

69 Given the findings from our sample as th ey relate to depres sion, it may actually be the concept of anticipatory grief that is playing the most important role. Individuals who had been providing care for a longer pe riod of time exhibited lower levels of depressive symptomatology in our sample sugge sting that on an individual level, having time to adjust to the inevitable loss followi ng a debilitating illness may lead to better outcomes during bereavement. The ideas that individuals experien ce a sense of relief following the loss of a spouse, particularly after a peri od of extensive caregiving, (Mullan, 1992; Schulz, Mendelsohn, Haley, Mahoney, Allen, et al., 2003) and that unexpected death leads to higher le vels of depression (Carr et al. 2001) have both been demonstrated in the bereavement literature. One of the strengths of our study is the tremendous variability in duration of caregiv ing. Only 32% of our sample had been providing care to their spouse for six months or less and 48% had been involved in caregiving for more than a year. It is possibl e that the inclusion of such a wide range of caregiving durations allowed us to see the eff ects of anticipatory grief. Effects such as those related to relief may be mo re evident on longer term follow-up. Due to the small sample size and the dive rsity in the non-White participants, it is beyond the scope of this study to be able to thoroughly interpret the finding that race/ethnicity was associated with post-l oss depression. Howeve r the strength of the relationship between these two variables suggests that future attention needs to be paid to the differing effects of bereavement on multiple racial and ethnic groups. One study that examined reactions of caregivers of Alzheime r’s disease patients to the loss of the care recipient using both White and African Am erican participants found that African American caregivers reported less acceptanc e of the relative’s death, greater perceived

PAGE 81

70 loss, and less anticipatory grief (Owen, G oode, & Haley, 2001). The authors of this study suggested that ethnic differenc es in bereavement merit future research, and our finding related to the association be tween race/ethnicity and depres sion offer credence to that suggestion. One additional recent study that examined the effects of widowhood on wellbeing found that African American and Ca ucasian American widows reported similar levels of overall grief, yearning, intrusiv e thoughts, shock, depressive symptoms, and anxiety (Carr, 2004). This study did find differences betw een these two groups on two outcomes, anger and despair, with African Am ericans reporting significantly lower levels compared to Caucasian Americans. Future studies should include large enough samples to examine differential effects of bereavem ent on distinct racial /ethnic groups, since cultural variables appear to have a marked impact on end-of-life issues (Kwak & Haley, in press). The second outcome variable examined in this study was life sa tisfaction. None of the hypotheses relating to car egiving stressors or apprai sals were supported by the research findings, and interes tingly only satisfaction with so cial support was significantly associated with post-loss life satisfaction. These findings lend support to the idea that positive and negative well-being outcome measures are not mutually exclusive and need to be examined independently of each other in order for researchers and practitioners to begin to understand the complex relations hip between pre-loss characteristics and bereavement. Our results related to grief following th e loss of a spouse indicated that fewer months in the caregiving role and lowe r number of functional impairments were significant predictors of higher levels of grief. These findi ngs were both consistent with

PAGE 82

71 and distinct from those seen in the depressi on model. The univariate analyses revealed that none of the appraisal or social resource variables was si gnificantly asso ciated with grief, contrary to the tenets of resource de pletion and relief hypotheses. As discussed above, it is possible that for our sample, the f actor that most protect ed an individual from experiencing a higher grief reaction was the chan ce to prepare him or herself for the loss. In addition, caregivers who were exposed to higher levels of functional impairments experienced less grief, suggesting that ha ving to provide assistance to a spouse on everyday tasks such as dressing, toileting, and bathing, may have also been a factor in aiding these individuals in adju sting to the loss with less dist ress. The fact that social resources were not significant in the grief analyses as th ey were in depression lends support to the belief by many re searchers that grief and depr ession are distinct concepts that represent different facets of bereavement a nd need to be examined in the research as such (Boerner et al. 2004; Bonanno, Wortman, Lehma n, Tweed, Haring, et al., 2002; Prigerson & Jacobs, 2001). Current bereavem ent researchers have differentiated between grief and depression in the development of bereavement trajectories (Bonanno et al., 2002), and have also begun to identify symptoms of “traumatic grief” that are distinct from those of depression-related depression and anxiety (see Prigerson & Jacobs, 2001). Limitations Our study had a number of limitations that warrant discussion, including a single, relatively short-term follow-up, and the sma ll sample size. Although the use of both preloss and post-loss data allowed us to examine predictors, the post-loss interview was conducted in a relatively short amount of time after the death (approximately 4 months).

PAGE 83

72 The inclusion of at l east one additional longe r-term post-loss data collection point and a control group would have allowed us to ex amine patterns of depr ession, life satisfaction, and grief over time. It is also important to note that although the average follow-up data collection interview was four months after th e loss, there was varia tion in the length of time for each participant (range of 1-10 months). Greater power related to a larger sample size, similar to the one used by Boerner et al. (2004), may have produced additional significant predictors; this dataset contained sufficient sample size to detect predictors with a medium effect size, but wa s insufficient in detecting predictors with a small one (Stevens, 2002). In addition, we were limited in the sophistication of our statistical analyses and were unable to u tilize methodologies such as mediating and moderating analysis and structural equati on modeling, which have been used by other researchers to evaluate stre ss process models of caregiving (Goode, Haley, Roth, & Ford, 1998; Haley, Roth, Coleton, For d, West, et al., 1996). As pr eviously discussed, we were unable to make any inferences about the findi ng related to race/et hnicity and post-loss depression due to both the small sample size overall and the small numbers of individuals from any particular cultural group. Future Directions Taken together the findings from our st udy contribute to the relatively recent literature that suggests that bot h positive and negative measures need to be utilized as both predictor and outcome variables in future research in order to begin to develop a more comprehensive picture of the relations hip between caregiving and bereavement. While our study did not find a relationship be tween positive aspects of caregiving and

PAGE 84

73 bereavement outcomes, either positive or nega tive, a recent paper using a much larger dataset found that this appraisa l measure was significantly as sociated with both grief and depression post-loss (Boerner et al., 2004). One of the main goa ls of this area of research is to aid both researchers and practitioners in identifying factors present prior to the loss of a spouse that place caregi vers at risk for complicatio ns during bereavement. By expanding the domains of variables used in this area of research, it is possible to gain an understanding not only of which factors put an individual at risk for negative outcomes, but also to identify those that are protective f actors that lead to a healthy adjustment to bereavement. The findings from our study s uggest that caregiving appraisals may not play as significant a role in adjustment to bereavement as they do while caregiving. In addition, the need for continued focus on soci al resource variables as potential buffers against negative well-being outcomes during bereavement was supported by our study. Our findings also highlight the importance of including caregiving c ontext variables in future research and the inclusion of indi viduals who have a va riety of caregiving experiences. We found evidence in suppor t of both the resource depletion and anticipatory grief hypotheses, howev er this is still a relative ly new area of research and there is much more to be done before interventions targeti ng at risk caregivers can be developed and implemented in caregiving popul ations. Researchers may also find it beneficial to examine additional caregiving context variables, such as comorbidities, and their relationship to anticipato ry grief in future research. Finally, studies of long-term bereavement outcomes will be necessary to be tter understand the clinical implications of caregiving experiences, since initial grief reac tions may not be pred ictive of clinically significant, long-term adaptation.

PAGE 85

74 Chapter Five Conclusions and Future Directions Conclusions The overall purpose of this dissertation was to add to the current literature on the relationship between caregiving and bereavemen t using a series of three studies. As discussed in each of the three articles, th ere is a large body of literature devoted to caregiving and bereavement respectively, but a relatively limited one that addresses these two life stressors simultaneously using prospective data. In addition there is a relatively new and growing interest in including both posit ive and negative measures in this area of research, as findings from recent studies have shown that only focusing on negative variables has left out a signi ficant portion of the relati onship between caregiving and bereavement. In order to address these gaps in the literature, we used two datasets that contained both pre-loss and post-loss interv iews. The first was the Changing Lives of Older Couples (CLOC) study, which allowed fo r the examination of a wide variety of variables, including but not limited to caregiving characteristics, bereavement trajectories, social interac tion, depression, sociodemographics self-rated physical health, and positive well-being, and their relationship to invol vement in the caregiving

PAGE 86

75 experience. The second was a subset of a da taset collected from a large, local non-profit hospice that recruited elderly spousal caregiver s of terminally ill patients. This dataset included a variety of widely used and validated measures collected both before and after the death of the patient and it allowed us to examine caregivi ng stressors, appraisals, and social support while caregiving and their eff ect on both positive and negative well-being outcomes during bereavement. The first study utilized four groups of bereaved individuals based upon the circumstances associated with their spouses ’ deaths: unexpected death, expected loss without caregiving, low stress caregiving, and high stress caregiving. These individuals were interviewed pre-loss and at 6 and 18 months post-loss using measures of psychological, social, and health functioning. The findings from this study indicated that the unexpected death group experienced marked increases in depression from pre-loss to post-loss, which is consistent with prior re search. In addition, the highly stressed caregivers were the only group not to show improvements within the domain of social engagement following the loss. This findi ng added merit to the theory that highly stressed caregivers may experience role e ngulfment, thus making it more difficult to maintain social support while caregiving and potentially leading to so cial isolation during bereavement. The second study focused on the relations hip between the caregiving groups utilized in the first study and fi ve core bereavement trajectori es previously identified by bereavement researchers: common grief, chr onic grief, chronic depression, improvement during bereavement, and resilience. The result s from this study indicated that there were differences in bereavement trajectories ba sed upon caregiving stat us. The non-caregiving

PAGE 87

76 and low stress caregiving groups exhibited similar distributions across the five trajectories, however unexpected death was associated with higher prevalence of the chronic grief trajectory and highly stressed ca regiving was associated with higher rates of chronic depression. Interestingly, the highest number of participants from each of the four groups fell into the resilient category, indicating that there is a need for research that aims to better understand the underlying mechanisms, such as social support, that protect individuals from negative bereavement outcomes. The last study of this diss ertation used a stress proce ss model framework in order to assess how pre-loss measures of caregivi ng stressors, appraisals, and social resources were related to short term well-being outco mes during bereavement. Interestingly, our study did not find that pre-loss caregiving appr aisals, either positive or negative, were significantly associated with well-being out comes during bereavement. The analyses found that fewer months caregiving and lower le vels of social resour ces were significant predictors of higher levels of depression dur ing bereavement. In addition, two caregiving stressors, fewer months caregiving and lowe r numbers of functional impairments, were associated with higher grief post-loss, while neither the appraisal nor the social resource variables had a significant impact. We also found that only satis faction with social support was a significant predictor of life satisfaction during bereavement. These findings lend support to the resource depleti on hypothesis and the theory of anticipatory grief, while at the same time highlighting th e distinct nature of bereavement well-being measures and the complexity of the relati onship between caregiving and bereavement. Taken together, the findings from this dissertation strongly support the idea that researchers need to continue focusing on a variety of study vari ables in order to

PAGE 88

77 understand the complexity in th e relationship between caregiving and bereavement. In addition social support and social resour ces during caregiving may prove to have a similar buffering effect on well-being outcomes during bereavement as have been demonstrated in the broader caregiving literature, but it is only through future longitudinal studies that researchers will be able to assess this. Our studies have provided support to the resour ce depletion and anticipator y grief hypotheses, and we believe that the inclusion of a wider range of individuals with varying degrees of caregiving experience and duration is necessary in future research in order to continue to test these findings. Limitations While this dissertation did address very important gaps in the literature, there were limitations that warrant discussion. Wh ile the large sample size and collection of both short and long term bereavement followups in the CLOC dataset allowed us to assess changes over time and utilize a wide va riety of variables, we were limited to the measures included by the decisions of the rese archers who originally designed the study. One example was the caregiving measures that we needed to use in order to classify participants into groups for the first two studies were a ssessed during the 6-month postloss interview, a design of the dataset which limited our ability to examine caregiving characteristics as predictor vari ables in regression models. It is also important to note that the nature of the data collection met hod (two hour face-to-face interviews conducted at the location of the participant’s choosing) may bias the sample toward participants

PAGE 89

78 with low depression or resiliency. Alternativ ely, the dataset that was utilized in the third study, while it contained a variety of commonl y used and validated measures, included a relatively small and restrictive sample. The size of the sample did not allow us to utilize sophisticated statistical methodology, such as mediator and moderator analyses or structural equation modeling. We also were unable to use large re gression models and had to be very concise in our statistical analysis plan. An additional limitation was the inclusion of a single, short term bereavement follow-up interview, although this is a fairly common methodology in this area of research. The inclusion of either one or more longer term bereavement follow-up interviews w ould have potentially given us the ability to detect additional predictors and change s over time. Overall this dissertation was subject to the limitations of both primary and secondary data analysis. While the utilization of secondary data limited our vari able selection to thos e measures chosen by other researchers, primary data collection was restricted by attrition rates and the need to rely on community organizations for sample recruitment. Future Directions Drawing on the findings from this diss ertation, future research should focus on the transition between caregiving and bereavem ent and the short and long term effects on surviving spouses. The findings from this dissertation work have confirmed the impression that most researchers and clinicians have that the experience of caregiving and subsequent death both for the patient and the caregiver is exceedingly complex. The primary goal of future research should be to study the nature of the relationship between

PAGE 90

79 caregiving and bereavement from a theore tically and methodologically sound basis, addressing some of the weaknesse s that have been present in prior studies. The research to date has pointed to the need to continue to expand the domains of variables utilized in these studies and also to focus on the role of coping through social support and social networks as potentially having a buffering e ffect on bereavement outcomes. In addition, it is vitally important to de velop an updated longitudinal study that includes widely validated and current measures with multiple interview points both before and after the death. Ideally, a study of this nature would also include pre-loss data from the care recipient in order for assessments of the dyad to also be possible. The first major long term goal that researcher s in this field need to address is the development of an inclusive bereavement th eory that draws on prior findings from a variety of disciplines and incorporates pr e-loss characteristics. The roots of both caregiving and bereavement research lie firmly in psychology and psychiatry, with stress process theory and attachment theory bei ng the two most common theories used. In addition, the Freudian grief wo rk hypothesis and its concepts of the importance of grief work and the detrimental effects of denial have been ingrained in the work of bereavement researchers from the very be ginning, although support for its merits are currently being called into question in the literature (Bonanno & Field, 2001). In an attempt to develop an integrative model of bereavement that focused on the strengths of each of these individual theories, Stroebe and Schut (1999) developed the dual process model of coping with bereavement. While th is model had the adva ntages of utilizing existing research findings, incorporating both male and female coping styles, allowing for pre-loss characteristics and individual di fferences, and incorporating the potential

PAGE 91

80 advantages of the short term use of denial, its lack of specific pa thways and measurable characteristics made it difficu lt for researchers to study. The second major goal of research on th e relationship between caregiving and bereavement is the use of both the presen t and future findings from prospective, longitudinal studies utilizing widely valid ated measures and sophisticated methodology to examine a more comprehensive understand ing of these two stressors and how they relate to each other. It is vitally important to conduct research with results that can be applied within the community and lead to the betterment of the lives of both older adults and their families. The long term goal of re searchers in this field should be to conduct studies that can eventually lead to the development of an intervention for at-risk caregivers that can be implemented and scien tifically tested. Research on bereavement intervention to date has been disappointi ng, and does not support the use of many conventional grief therapies that are be ing commonly promoted and used in the community (Neimeyer, 2000; Schut, Stroebe, van den Bout, & Terheggen, 2001). In order for an appropriate intervention to be designed, researchers need to utilize longitudinal data in order to produce a model for the identification of at-risk caregivers that can be used in a community sample Upon successful id entification of an appropriate sample, an intervention program targeted at lowering depression, increasing life satisfaction, and increasing the quality and nature of social support during both caregiving and bereavement could be implemented in the community.

PAGE 92

81 References 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. Bass, D. M., & Bowman, K. (1990). The transition from caregiving to bereavement: The Relationship of care-related strain and adjustment to death. The Gerontologist, 30, 35-42. Beery, L. C., Prigerson, H .G., Bierhals, A. J., Santucci, L. M., Newsom, J. T., Maciejewski, P. K., Rapp, S. R., Fasi czka, A., & Reynolds, C. F. (1997). Traumatic grief, depression, and caregiving in elderly spouses of the terminally ill. Omega, 35 261-279. Bodnar, J. C., & Kiecolt-Glaser, J. K. (1994) Caregiving depression after bereavement: chronic stress isn’t over when it’s over. Psychology and Aging, 9, 372-380. Boerner, K., Schulz, R., & Horowitz, A. ( 2004). Positive aspects of caregiving and adaptation to bereavement. Psychology and Aging, 19 (4), 668-675. Boerner, K., Wortman, C. B., & Bonanno, G. A. (2005). Resilient or at risk? A 4-year study of older adults who initially showed high or low distress following conjugal loss. Journal of Gerontology: Psychological Sciences, 60B (2), P67-P73. Bonanno, G. A., & Field, N. P. (2001). Examining the delayed grief hypothesis across 5 years of bereavement. American Behavioral Scientist 44 798 816. Bonanno, G. A., Notarius, C. I., Gunzerath, L ., Keltner, D., & Horowitz, M. J. (1998). Interpersonal ambivalence, perceived dyadic adjustment, and conjugal loss. Journal of Consulting and Clinical Psychology, 66, 1012-1022. Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring, M., Sonnega, J., Carr, D., & Nesse, R. M. (2002). Res ilience to loss and chronic grief: A prospective study from pre-loss to 18 months post-loss. Journal of Personality and Social Psychology, 83, 1150-1164. Bradburn, N. M. (1969). The structure of psychological well-being. Chicago: Aldine.

PAGE 93

82 Carr, D. (2004). Black/White differences in psychological adjustment to spousal loss among older adults. Research on Aging, 26 (6), 591-622. Carr, D., House, J. S., Kessler, R. C., Nesse R. M., Sonnega, J., & Wortman, C. (2000). Marital quality and psychol ogical adjustment to widow hood among older adults: A longitudinal analysis. Journal of Gerontology: Social Sciences, 55B (4), S197-207. Carr, D., House, J. S., Wortman, C., Nesse, R., & Kessler, R. C. (2001). Psychological adjustment to sudden and anticipated spousal death among the older widowed. Journal of Gerontology: Social Sciences, 56B (4), S237-S248. Chappell, N. L. & Reid, R. C. (2002). Burden and well-being among caregivers: Examining the distinction. The Gerontologist, 42 772-780. Cleiren, M. P. H. D., van der Wal, J., & De ikstra, R. F. W. (1988). Death after a long term disease: anticipation and outcome in the bereaved. Pharos International, Autumn-Winter, 112-114, 136-139. Davis, C. G., & Nolen-Hoeksema, S. (2001). Loss and meaning: How do people make sense of loss? American Behavioral Scientist, 44 (5), 726-741. Diener, E. & Emmons, R. (1984). The indepe ndence of positive and negative affect. Journal of Personality and Social Psychology, 47 1105-1117. Faschingbauer, T. R., Zisook, S., & DeVaul, R. A. (1987). The Texa s Revised Inventory of Grief. In S. Zisook (Ed.), Biopsychosocial aspects of bereavement (pp. 111124). Washington, DC: Ameri can Psychiatric Association. Folkman, S. (2001). Revised coping theory and the process of bereavement. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 563-584). Cambridge: Cambridge University Press. Folkman, S. & Moskowitz, J. T. (2000). Po sitive affect and the other side of coping. American Psychologist, 55 647-654. George, L. K. (1990). Caregiver stress st udies—There really is more to learn. The Gerontologist, 30, 580-581. George, L. K., & Gwyther, L. P. (1986). Caregiver well-being: a multidimensional examination of family caregivers of demented adults. The Gerontologist, 24, 253259.

PAGE 94

83 Given, B. & Given, C. W. (1996). Family caregiver burden from cancer care. In R. McCorkle, M. Grant, M. Frank-Stromborg, & S. Baird (Eds.), Cancer Nursing: A Comprehensive Textbook (pp. 93-109) Philadelphia: W.B. Saunders. Goode, K. T., Haley, W. E., Roth, D. L., Fo rd, G. R. (1998). Predicting longitudinal changes in caregiver physical and ment al health: A stress process model. Health Psychology, 17 (2), 190-198. Haley, W. E. & Bailey, S. (1999). Research on family caregiving in Alzheimer’s disease: Implications for practice and policy. In B. Vellas & J. L. Fitten (Eds.), Research and Practice in Alzheimer’s Disease (volume 2, pp. 321-332) Paris: Serdi Publisher. Haley, W. E., LaMonde, L. A., Han, B., Bu rton, A. M., & Schonwetter, R. (2003). Predictors of depression and life satisfaction among spousal caregivers in hospice: Application of a stress process model. Journal of Palliative Medicine, 6, 215224. Haley, W. E., LaMonde, L. A., Han, B., Na rramore, S., & Schonwetter, R. (2001). Family caregiving in hospice: Effects on psychosocial and he alth functioning among spousal caregivers of hospice patie nts with lung cancer or dementia. The Hospice Journal, 15 (4), 1-18. Haley, W. E., Roth, D. L. Coleton, M. I., For d, G. R., West. C. A., Collins, R. P., Isobe, T. (1996). Appraisal, copi ng, and social support as me diators of well-being in Black and White Alzheimer’s family caregivers. Journal of Consulting and Clinical Psychology, 64 121-129. Hildebrand, D. K., Laing, J. D., & Rosenthal, H. (1977). Prediction analysis of crossclassification. New York: Wiley. Hill, C. D., Thompson, L. W., & Gallagher, D. (1988). The role of anticipatory bereavement in older women’s adjustment to widowhood. The Gerontologist, 28, 792-796. HRS/AHEAD Documentation Report (2000). Documentation of affective functioning measures in the Health and Retirement Study. Retrieved July 1, 2005 from the Health and Retirement Survey website: http://hrsonline.isr.umich.e du/docs/userg/dr-005.pdf. Jones, B. L., Nagin, D. S., & Roeder, K. (2001). A SAS procedure based on mixture models for estimating developmental trajectories. Sociological Methods and Research, 29 374-393.

PAGE 95

84 Katz, S., Ford, A. B., Moskowitz, R. W., Jack son, B. A., & Jaffe, M. W. (1963). Studies of illness in the aged. The index of ADL: A standardized measure of biological and psychological function. Journal of the American Medical Association, 185, 914-919. Kiecolt-Glaser, J. K., Marucha, P. T., Mala rkey, W. B., Mercado, A. M., & Glaser, R. (1995). Slowing of wound heali ng by psychological stress. The Lancet, 346, 1194-1196. Kohout, F. J., Berkman, L. F., Evans, D. A., & Cornori-Huntley, J. (1993). Two shorter forms of the CES-D depression symptoms index. Journal of Aging and Health, 5 (2), 19-193. Krause, N., & Borawski-Clark, E. (1995). So cial class differences in social support among older adults. The Gerontologist, 35 498-508. Kurtz, M. E., Kurtz, J.C., Given, C. W., Gi ven, B. (1995). Relationship of caregiver reactions and depression to cancer patients’ symptoms functional states and depression-a longitudinal view. Social Science Medicine, 13 138-147. Kwak, J., & Haley, W. E. (in press). Current research findings on end-of-life decision making among racially/et hnically diverse groups. The Gerontologist Lawton, M. P., & Brody, E. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. The Gerontologist, 9, 179-186. Larsen, J. T., McGraw, A. P., & Cacioppo, J. T. (2001). Can people feel happy and sad at the same time? Journal of Personality and Social Psychology, 81 (4), 684-696. Lawton, M. P. (1983). Environmental and othe r determinants of well-being in older people. The Gerongologist, 23, 349-357. Lawton, M. P., Moss, M., Kleban, M. H., Glic ksman, A., & Rovine, M. (1991). The twofactor model of caregiving apprai sal and psychological well-being. Journal of Gerontology: Psychological Sciences, 46 (4), P181-P189. Lubben, J.E. (1988). Assessing social networks among elderly populations. Family and Community Health, 11, 42-52. Lundin, T. (1984). Morbidity following sudden and unexpected bereavement. British Journal of Psychiatry, 144, 84-88. McHorney, C. A., & Mor, V. (1988). Predictors of bereavement depre ssion and its health services consequences. Medical Care, 26 882-893.

PAGE 96

85 Minino, A. M., & Smith, B. L. (2001). Deaths: Preliminary data for 2000. National Vital Statistics Reports, 49 (12). Hyattsville, Mary land: National Center for Health Statistics. Mullan J. T. (1992). The bereaved caregive r: a prospective study of changes in wellbeing. The Gerontologist, 32, 673-683. National Hospice and Palliativ e Care Organization (n.d.). Hospice facts and figures. Retrieved July 1, 2005 from the National Hospice and Palliative Care Organization Web site: http://nhpco. org/files/public/Hos pice_Facts_110104.pdf. Neimeyer, R. (2000). Searching for the mean ing of meaning: Grief therapy and the process of reconstruction. Death Studies, 24 531-558. Northouse, L. L. (1988). Social support in pa tients’ and husbands’ adjustment to breast cancer. Nursing Research, 37, 91-95. Oberst, M. T., Gass, K. A., & Ward, S. E. ( 1989). Caregiving demands and appraisal of stress among family caregivers. Cancer Nursing, 12 209-215. O’Bryant, S. L. (1991). Forewarning of a hus band’s death: Does it make a difference for older widows? Omega, 22, 227-239. Owen, J. E., Goode, K. T., & Haley, W. E. (2001) End of life care and reactions to death in African American and White family car egivers of relatives with Alzheimer’s disease. Omega, 43 (4), 349-361. Park, C. L., & Folkman, S. (1997). Stability and change in psychosocial resources during caregiving and bereavement in pa rtners of men with AIDS. Journal of Personality, 65, 421-447. Pearlin, L. I., Mullan, J. T., Semple, S. J., & Skaff, M. M. (1990). Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist, 30 583-594. Pinquart, M. & Sorensen, S. (2003). Differe nces between caregivers and noncaregivers in psychological health and physic al health: A meta-analysis. Psychology & Aging, 18 (2), 250-267. Prigerson, H. G., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale, consensus criteria, and a preliminar y empirical test. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 563-584). Cambridge: Cambridge University Press.

PAGE 97

86 Radloff, L. S. (1977). The CE S-D scale: A self-report depre ssion scale for research in the general population. Applied Psychological Measures, 1 385-401. Raphael, B., Minkov, C., & Dobson, M. (2001). Psychotherapeutic and pharmacological intervention for bereaved persons. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 563-584). Cambridge: Ca mbridge University Press. Reed, M. D. (1998). Predicting grief symp tomatology among the suddenly bereaved. Suicide & Life-Threatening Behavior, 28, 285-301. Reich, J. W., Zautra, A. J., & Davis, M. ( 2003). Dimensions of af fect relationships models and their integrative implications. Review of General Psychology, 7 (1), 66-83. Robinson-Whelen, S., Tada, Y., MacCallum, R. C., McGuire, L., & Kiecolt-Glaser, J. K. (2001). Long-term caregiving: What happens when it ends? Journal of Abnormal Psychology, 110 (4), 573-584. Russell, J. A. & Carroll, J. M. (1999). On th e bipolarity of positive and negative affect. Psychological Bulletin, 125 (1), 3-30. Sales, E., Schulz, R., & Biegal, D. (1992). Pr edictors of strain in families of cancer patients: A review of the literature. Journal of Psychosocial Oncology, 10 1-26. Sankar, A. (1991). Ritual and dying: a cultural analysis of social support for caregivers. The Gerontologist, 31, 43-50. Schulz, R., & Beach, S.R. (1999). Caregiving as a risk factor for mort ality: The Caregiver Health Effects Study. Journal of the American Medical Association, 282 22152219. Schulz, R., Beach, S. R., Lind, B., Martire, L. M., Zdaniuk, B., Hirsch, C., Jackson, S., & Burton, L. (2001). Involvement in careg iving and adjustment to death in a spouse: Findings from the Ca regiver Health Effects Study. Journal of the American Medical Association, 285, 3123-3129. Schulz, R., Mendelsohn, A. B., Haley, W. E., Mahoney, D., Allen, R. S., Zhang, S., Thompson, L., & Belle, S. H. (2003). End of life care and the effects of bereavement among family caregivers of persons with dementia. New England Journal of Medicine 349, 1936-1942. Schulz, R., Newsom, J. T., Fleissner, K., Deca mp, A. R., & Nieboer, A. P. (1997). The effects of bereavement after family caregiving. Aging & Mental Health, 1 (3), 269-282.

PAGE 98

87 Schulz, R., Newsom, J.T., Mittelmark, M., Bu rton, L., Hirsch, C., & Jackson, S. (1997). Health effects of caregiving: Th e Caregiver Health Effects Study. Annals of Behavioral Medicine, 19, 110-116. Schulz, R., O’Brien, A. T., Bookwala, J., & Fl eissner, K. (1995). Psychiatric and physical morbidity effects of dementia caregivi ng: Prevalence, correlates, and causes. The Gerontologist, 35 771-791. Schut, H., Stroebe, M. S., van den Bout, J., & Terheggen, M. (2001). The efficacy of bereavement interventions: Determining who benefits. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 705-737). Cambridge: Cambridge University Press. Sheskin, A., & Wallace, S. E. (1976). Differing bereavements: Suicide, natural, and accidental death. Omega, 7 (3), 229-242. Singer, J.D. & Willett, J. B. (2003). Applied longitudinal anal ysis: Modeling change and event occurrence. Oxford: Oxford University Press. Skaff, M. M., Pearlin, L. I., & Mullan, J. T. (1996). Transitions in the caregiving career: Effects on sense of mastery. Psychology & Aging, 11, 247-257. Stevens, J. P. (2002). Applied multivariate statistics for the social sciences (4th ed.). London: Lawrence Erlbaum Associates, Inc. Stroebe, W., & Schut, H. (1999). The dual pr ocess model of coping with bereavement: Rationale and description. Death Studies, 23 (3), 197-224. Stroebe, W., & Stroebe, M. S. (1987). Bereavement and health: The psychological and physical consequences of partner loss. New York: Cambridge University Press. Stroebe, W., Stroebe, M. S., Abakoumkin, G., & Schut, H. (1996). The role of loneliness and social support in adjustment to loss: A te st of attachment versus stress theory. Journal of Personality and Social Psychology, 70 (6), 1241-1249. University of Michigan (n.d.). Changing lives of older couples: A multi-wave prospective study of bereavement Retrieved on February 1, 2005 from http://www.cloc.isr.umich.edu Weitzner, M. A., Haley, W. E., & Chen, H. ( 2000). The family caregiver of the older cancer patient. Hematology/Oncology C linics of North America, 14 (1), 269-281.

PAGE 99

88 Wood, V., Wylie, M.L., & Sheafor, B. (1969). An analysis of a short self-report measure of life satisfaction: Correlat ion with rater judgements. Journal of Gerontology, 24, 465-469. Watson, D. & Tellegen, A. (1985). Towa rd a consensual structure of mood. Psychological Bulletin, 98 219-235. Whitlatch, C. J.,Zarit, S. H., & von Eye, A. (1991). Efficacy of interventions with caregivers: A reanalysis. The Gerontologist, 31 (1), 9-14. Zarit, S. H., Anthony, C. R., & Boutselis, M. ( 1987). Interventions w ith caregivers of dementia patients: Comparison of two approaches. Psychology and Aging, 2, 225-232. Zettel, L. A., & Rook, K. S. (2004). Subs titution and compensation in the social networks of older widowed women. Psychology and Aging, 19 (3), 433-443.

PAGE 100

89 Appendix

PAGE 101

90 Appendix A: Curriculum Vitae CURRICULUM VITAE Allison M. Burton Business Residence University of South Florida 9507 West Riverchase Drive School of Aging Studies, MHC 1305 Tampa, FL 33637 4202 E. Fowler Avenue Telephone: (813) 914-8395 Tampa, FL 33620-8100 Telephone: (813) 974-2646 Fax: (813) 974-9754 E-mail: amburton@cas.usf.edu Place of Birth: Houma, LA, USA Education 2000 – present University of South Florida-Tampa Doctoral Candidate Major: Ph.D. in Aging Studies 1995 – 2000 Louisiana State University-Baton Rouge Major: B. S. in Psychology Positions Held 2002 present Graduate Teaching Assistant, School of Aging Studies, University of South Florida 2001 present Graduate Research Assistant, Dr. William E. Haley’s Gerontology Lab, University of South Florida 2003 2004 Graduate Research Assistant, Department of Research, LifePath Hospice and Palliative Care, Tampa, FL 2002 2003 Graduate Research Assistant, Center for Hospice, Palliative Care, and End-of-Life Research, Univers ity of South Florida, including serving as a Research Assistant for LifePath Hospice and Palliative

PAGE 102

91 Care, Inc., Tampa, FL, and Hospice of the Florida Suncoast, Largo, FL Appendix A (continued) 1998 2000 Undergraduate Research A ssistant, Dr. Katie Cherry’s Developmental Psychology Res earch Lab, Louisiana State University Teaching Experience 2003 present Death and Dying, a senior-l evel course, University of South Florida (7 semesters) 2002 Psychology and Aging, a senior-l evel course, University of South Florida (1 semester) Journal Publications Burton, A. M., Haley, W. E., & Small, B. J. (2005). Bereavement after caregiving or unexpected death: Effects on elderly spouses. Under review. Cherry, K. E., Brigman, S., & Burton A. M. (2004). Perceptions of forgetfulness in adulthood Under review. Haley, W. E., LaMonde, L. A., Han, B., Bu rton, A. M., & Schonwetter, R. (2003). Predictors of depression and life satis faction among spousal caregivers in hospice: Application of a stress process model. Journal of Palliative Medicine, 6, 215-224. Haley, W. E., Allen, R. S., Reynolds, S., Chen, H., Burton, A. M., & GallagherThompson, D. (2002). Family issues in end-of-life decision making and end-of-life care. American Behavioral Scientist, 46 (2), 284-297. Book Chapters Burton, A. M., Kwak, J., & Haley, W. E. (2004). Elder Ca regiving. In C. Spielberger (Ed.), Encyclopedia of Applied Psychology (pp. 703-711) San Diego: Academic Press.

PAGE 103

92 Appendix A (continued) Haley, W. E., Burton, A. M., LaMonde, L. A. & Schonwetter, R. S. (2004). Family caregiving issues for older cancer patien ts. In L. Balducci G.H. Lyman, W.B. Ershler, & M. Extermann (Eds.), Comprehensive Geriatric Oncology (2nd ed., pp. 843-852). London: Taylor & Francis. Presentations at Professional Meetings Burton, A. M., Haley, W. E., and Small, B. J. (2004, November). The impact of the caregiving experience on well-b eing during bereavement: From baseline to 18-months post-loss. Po ster session presen ted at the annual meeting of the Gerotological Soci ety of America, Washington, DC. Burton, A. M. (2003, November). Psychological well-being in bereaved spouses: From baseline to 18-months pos t-loss. In W. E. Haley (Chair), The USF Graduate Student Symposium on the Continuum of Health: From Wellness to End-of-Life. Sy mposium presented at the annual meeting of the Gerongological Soci ety of America, San Diego, CA. Burton, A. M. (2003, November). Th im pact of the caregiving experience on psychological bereavement outcomes: From baseline to 18-months post-loss. Poster session presente d at the annual conference for the Center for Hospice, Palliative Care, and End-of-Life Research, Tampa, FL. Burton, A. M., Haley, W. E., & Kwak, J. (2002, November). Psychosocial and satisfaction outcomes in bereaved form er hospice caregivers. Poster session presented at the annual meeting of th e Gerontological Society of America, Boston, MA. Haley, W. E., Burton, A. M., Robb, C., & Schonwetter, R. (2002, February). Depression in bereaved spous al caregivers of hospice pa tients. Presented at the annual meeting of the American Academ y of Hospice and Palliative Medicine, Palm Springs, CA. Burton, A. M., Haley, W. E., Robb, C., & Schonwetter, R. (2001, November). Service utilization in bereaved fo rmer hospice caregivers. Poster session presented at the annual meeti ng of the Gerontological Society of America, Chicago, IL.

PAGE 104

93 Appendix A (continued) Burton, A.M. (2001, April). Research expe riences of a first-year Ph.D. student. In W. E. Haley (Chair), USF Ph.D. in Aging Studies Program: Interdisciplinary Student Research on End of Life Issu es. Symposium conducted at the annual meeting of the Southern Geront ological Society, Lexington, KY. Cherry, K. E., Brigman, S., Burton, A. M., & Roecke, C. (1999, November). Perceptions of forgetfulness in adulthood. Presente d at the eighth annual Cognitive Aging conference, Atlanta, GA. Grants Completed Served as the principal i nvestigator on a pilot grant from the USF Center for Hospice, Palliative Care, and End-of-Life Research examining psychosocial outcomes and service ut ilization among bereaved former hospice caregivers ($12,000), W. E. Haley. Honors and Professional Activities 2005 Awarded the Provost’s Award for Outstanding Teaching by a Graduate Teaching Assistant, University of South Florida 2004 Awarded the Provost’s Commendation for Outstanding Teaching by a Graduate Teaching Assistant, University of South Florida 2003 2004 President of the Student Association for Aging Studies, University of South Florida 2003 2004 Student Representative to the Ph.D. in Aging Studies Committee, University of South Florida 2002 Attended the CLOC Data Release Conference, Institute for Social Research, University of Michigan 2002 2003 University of South Florida Ce nter for Hospice, Palliative Care, and End-of-Life Studies pilot gran t/graduate research assistant recipient

PAGE 105

94 2001 present Member, Geront ological Society of America Appendix A (continued) 2002 present Member, Center for Hospice, Palliative Care, and End-of-Life Studies, University of South Florida, 2000 2001 Institute on Aging Fellowship, University of South Florida, 1998 2000 Active member in Psi Chi, including two years as student representative 1995 LSU Presidential Scholarship, fi ve year full tuition and fees, including on campus employment for research Related Activities 2000 –2002 Patient Volunteer, LifePath Hospice and Palliative Care, Inc., Tampa, FL 1990 – 2000 Volunteer at various lo cal nursing homes and hospitals Ad Hoc Reviews American Behavioral Scientist Death Studies Research Interests and Current Projects End of Life Issues, specifically hospice care, bereavement, and family caregiving; Interventions for family caregivers.

PAGE 106

95 About the Author Allison M. Burton received her Bachelor’s of Science Degree in Psychology from Louisiana State University in May of 2000. She entered the Ph.D. in Aging Studies program at the University of South Florid a in the Fall of 2000 w ith an interest in caregiving and end-of-life. While in the Ph.D. program at the Univer sity of South Florida, Ms. Burton was involved with two local large non-profit hos pices and was employed as a Graduate Teaching Assistant in the School of Aging Studies, primarily responsible for teaching Death and Dying. Ms. Burton co-authored four publications in peer-reviewed journals and two book chapters while enrolled as a stude nt, as well as presenting her research at multiple regional and national conferences.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001670318
003 fts
005 20051216093242.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 051109s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001178
035
(OCoLC)62248584
SFE0001178
040
FHM
c FHM
049
FHMM
090
HQ1061 (Online)
1 100
Burton, Allison.
4 245
The relationship between caregiving and bereavement
h [electronic resource] :
b a series of three studies /
by Allison Burton.
260
[Tampa, Fla.] :
University of South Florida,
2005.
502
Thesis (Ph.D.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 106 pages.
Includes vita.
520
ABSTRACT: Spousal bereavement has been consistently demonstrated in the literature to be one of the most highly stressful experiences in an individuals lifetime. In addition many deaths in the United States are preceded by a period of caregiving, which is also believed to be highly stressful and have a profound impact on bereavement. However the literature has been inconsistent as to the exact nature of the relationship between caregiving and bereavement and there has been some debate as to whether or not positive and negative affect variables are mutually exclusive. This dissertation sought to further address the issue of the relationship between caregiving and the bereavement experience through a series of three studies which utilized information from two datasets. The first was the Changing Lives of Older Couples (CLOC) study, a project that included both pre- and post-loss data.The second was a subset of a dataset that recruited elderly spousal caregivers of terminally ill patients from a large, local non-profit hospice. This dataset included both pre- and post-loss data and included a variety of widely used and validated measures that allowed for the examination of caregiving stressors, appraisals, and social support and their effect on both positive and negative mental health outcomes during bereavement.The first study examined the impact of caregiving on well-being during bereavement, specifically within the domains of psychological, social, and physical health, utilizing both positive and negative affect measures.
590
Adviser: William E. Haley.
653
Caregiving.
Bereavement.
Social support.
Depression.
Stressors.
Well-being.
0 690
Dissertations, Academic
z USF
x Aging Studies
Doctoral.
773
t USF Electronic Theses and Dissertations.
856
u http://digital.lib.usf.edu/?e14.1178