EFFECTS OF SOCIAL NETWORKS ON THE HEALTH
OF FAMILY CAREGIVERS IN LATER LIFE:
A CROSS-SECTIONAL STUDY
By
Lihua Huang

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Social Work
2011

ABSTRACT
EFFECTS OF SOCIAL NETWORKS ON THE HEALTH
OF FAMILY CAREGIVERS IN LATER LIFE:
A CROSS-SECTIONAL STUDY
By
Lihua Huang

The purpose of this study was to better understand the social value of family caregiving in
later life by examining the effects of social networks and family caregiving on the health of older
adults aged 65 and over. Based on positive gerontology, the study asked the overarching research
question: “Are there significant mean differences in health for different levels of social networks
and for different family caregiving status after controlling for demographic factors?” In this
study, the independent variables were three-dimensional social networks and family caregiving
status. The dependent variables were levels of self-rated health, physical function, absence of
chronic diseases, and mental health.
This was a cross-sectional, explanatory survey study. It drew a probability sample from a
regional representative population of older adults aged 64 and above in three Michigan
communities. The survey resulted in a response rate of 36.6% and a final sample of 358
community-dwelling older adults: 136 were caregivers, and 222 were non-caregivers. The
questionnaire contained a total of 58 quantitative measures and two open-ended questions. Five
standardized measures used in the study had high internal consistency. They were Bakas
Caregiving Outcomes Scale, Lubben Social Networks Scale, Berkman-Syme Social Networks
Index, Life Events Survey, Instrumental Activities of Daily Livings, and Geriatric Depression
Scale.

Multiple data collection and data analysis methods were used to enhance the validity of the
results. Between-methods triangulation primarily included a survey instrument that integrated
open-ended questions and a consequent combination of descriptive statistics and content
analysis. Within-methods triangulation included multiple scales used for the social networks
construct as well as the health construct.
Multivariate Analyses of Variance and Covariance were performed. Results indicated
significant mean differences in the health for social networks and family caregiving
independently. Specifically, high levels of functional social networks significantly predicted
better self-rated health and higher physical function. The study also found that family caregiving
significantly affected the health of the older family caregivers. Results highlighted that being
family caregivers in later life significantly predicted lower levels of physical function and mental
health, but social networks positively affected well-being of family caregivers in later life.
The nature of cross-sectional study and a regional representative sample limited
generalizability of the results to populations beyond the study population. However, several
implications were identified regarding methodology, research, and practice. The results
suggested the functional dimension was the most important factor in social network analysis, but
there was not similar prior research to which a comparison could be made. More inquiry is
needed to assess conceptualization and measurement of social networks in social work and
gerontology. Although the study made efforts to combine quantitative and qualitative methods,
two open-ended questions were not adequate to fully understand the social networks of older
family caregivers. More research is needed to collect social network-specific qualitative data to
obtain in-depth understandings regarding health effects of social networks among older family
caregivers.

Copyright by
Lihua Huang
2011

To my mother 鲍永年(Bao Yongnian)

v

ACKNOWLEDGMENTS

This dissertation work could not have been accomplished without the love, support and
encouragement from many individuals who have shared my joy and the struggle of my journey
as a doctoral student at the School of Social Work, Michigan State University. I want to express
my sincere gratitude for their close support and genuine attentiveness.
I wish to first acknowledge the ongoing support, encouragement, and mercy of my
dissertation committee chairperson, Dr. DeBrenna L. Agbényiga, and committee members, Dr.
Rena D. Harold, Dr. Amanda T. Woodward, and Dr. Steven J. Gold. Very special thanks go to
Dr. Agbényiga for her guidance, advice, and inspiration, which directed me through every step of
the way, no matter whether it was an exciting discovery or a bumpy road. I extend gratitude to
Dr. Harold, who started this journey with me in 2006 and has since shared her steadfast support
and wisdom, which helped me persevere at times I could not see the end of the tunnel. I have
been fortunate to have had much support and direction from Dr. Woodward, who has
unreservedly contributed her knowledge and skills to my gerontological inquiry. I also wish to
say a special thank you to Dr. Gold, who has provided support that has enhanced my scientific
discovery and moved forward toward the dissertation completion.
I wish to acknowledge the support and encouragement of Dr. Gary Anderson and every
faculty member I have worked with at the School of Social Work. The high standard and quality
of work I have witnessed here at the School has made me grow to be a better social work scholar.
I wish to thank all of the older adults who so graciously volunteered to participate in my
study. They shared their experience of aging, family caregiving, and social relations. I appreciate
their respect of family caregiving in later life, even when they were not involved in caregiving at
the time of the study.

vi

I wish to acknowledge the unconditional love and support of my dear family and friends.
You have provided immeasurable emotional support and I will be forever grateful. Thank you,
Steve, for making it possible to pursue my dream and for accompanying me to work hard and to
have a beautiful life. Thank you to my daughter Binyan, who brought the joy and the color of
being a mother to my life while becoming an independent and excellent student yourself. I want
to delicate my dissertation study to my mother, Bao Yongnian, who has taught me persistence
and excellence. I am saddened by the fact that my father, Huang Chenen, cannot share this
moment with me. All praise is due to my family and friends in China and in the United States. I
do not list all of your names, but you know so well that your faith in me and your unwavering
support make everything different in a wonderful way.

vii

TABLE OF CONTENTS

LIST OF TABLES

x

LIST OF FIGURES

xi

CHAPTER ONE: INTRODUCTION
Statement of problem
Trends of family caregiving in later life
Theoretical framework
Purpose of the study and research questions
Conceptual definitions

1
2
8
11
14
15

CHAPTER TWO: LITERATURE REVIEW
Caregiving networks
Network characteristics of older family caregivers
Health effects of social networks
Limitations of existing literature

20
21
23
28
38

CHAPTER THREE: METHODS
Operational model
Research design
Data collection procedure
Survey instrument and measures
Data management
Data analysis

45
45
49
51
55
65
66

CHAPTER 4: RESULTS
Descriptive statistics and preliminary analyses
Results from mixed methods
Effect modeling

70
70
89
98

CHAPTER 5: SUMMARY AND DISCUSSION
Summary of the study
Interpretation of the results
Limitations
Implications and recommendations
Conclusion

109
109
117
124
129
136

viii

APPENDICES
A:
Initial IRB Application Approval
B:
Invitation Letter
C:
Informed Consent
D:
Questionnaire
E.
Demographic Comparison of Research Sites

138
139
142
145
149
168

REFERENCES

170

ix

LIST OF TABLES

Table 2.1 Research Gaps: Population, Sampling, and Measures of Social Networks
and Health

40

Table 3.1 Demographic Characteristics of the Sample

52

Table 3.2 Summary of Variables by Name, Instrument, and Psychometric Properties

57

Table 3.3 Family Caregiving Measures

58

Table 3.4 Levels of Alcohol Use Based on the Number of Standard Drinks Consumed

63

Table 3.5 Research Questions and Data Analysis Strategies

68

Table 4.1 Demographic Characteristics of the Sample by Caregiving Status

71

Table 4.2 Descriptive Statistics of Social Network Variables by Caregiving Status

73

Table 4.3 Changes in Social Networks by Caregiving Status

79

Table 4.4 Frequency of Health Behavior and Treatment Use

82

Table 4.5 Co-residency by Caregiving Status

83

Table 4.6 Descriptive Statistics of Health Variables by Caregiving Status

87

Table 4.7 Analysis of Variance for Health between Family Caregivers and Non-caregivers

88

Table 4.8 Mean Scores of Changes in Family Caregiving Experience in Later Life

92

Table 4.9 Correlation Matrix for Major Variables of Interest

101

Table 4.10 Multivariate Statistics for Social Network and Family Caregiving on Health

103

Table 4.11 Adjusted and Unadjusted Mean Differences for Health by Social Networks
Category and Caregiving Status

105

Table 4.12 Multiple Comparison of Health by Social Networks and Family Caregiving

106

Table A1 Demographic Comparison of Research Sites

168

x

LIST OF FIGURES

Figure 3.1 Operational Model of Health Effects of Social Networks among Older
Family Caregivers

48

Figure 3.2 The Final Sample of Family Caregivers

54

Figure 4.1 LSNS Frequency among Caregivers

77

Figure 4.2 LSNS Frequency among Non-caregivers

77

Figure 4.3 Average Social Network Structural Characteristics by Caregiving Status

78

Figure 4.4 Functional Social Networks by Caregiving Status

84

Figure 4.5 Ecomap of Social Networks of Family Caregivers

85

Figure 4.6 Coding Paradigm of Perceived Family Caregiving Experience in Later Life

93

xi

CHAPTER ONE: INTRODUCTION

The health care system in the United States is increasingly dependent on sustainable and
quality family caregiving (Calasanti & Slevin, 2001; Feinberg & Newman, 2006; Goodman,
Potts, & Pasztor, 2007; Heller, Caldwell, & Factor, 2007; Talley & Crews, 2007). Available data
have shown that the number of older adults who provide unpaid care to their family members,
friends, and neighbors across the life span is growing in recent decades (McGarry & Arthur,
2001; Moren-Cross & Lin, 2006; Pinquart & Sörensen, 2006). The 2000 U.S. Census
documented approximately 2.4 million grandparents raising 4.5 million grandchildren younger
than age 18, a 30% increase from 1990 (Simmons & Dye, 2003; Hayslip Jr. & Kaminski, 2005).
Over 15% of 4.7 million persons with intellectual disabilities received care from older family
caregivers in 2006, an estimated 10–20% increase in demand for residential services over the
past three decades (Braddock, 2002; Braddock, Hemp, & Rizzolo, 2008). In 2007, older adults
accounted for about 30% of 34 million family caregivers, and on average they provided 20 hours
of care per week, increased from an estimated 25.8 million family caregivers and 17.9 care hours
weekly in 1997 (AHRQ, 2001; Houser & Gibson, 2008; Arno, Levine, & Memmott, 1999).
Thompson (2004) depicted the critical position of family caregiving in the health care structure
as 78% of community-based long-term care recipients receive care exclusively from unpaid
family and friends, while 14% receive some combination of family care and paid assistance, and
only 8% rely on formal care alone.
The intensified demand has raised great concerns about sustainability of family caregiving in
later life and quality of life among older family caregivers. Efforts have been made to better
understand needs of older family caregivers (Beach, Schulz, Williamson, Miller, Weiner, &
Lance, 2005; Teschendorf, Schwartz, Ferrans, O’Mara, Novotny, & Sloan, 2007; Weeks,
1

Nilsson, Bryanton, & Kozma, 2009). It is evident that a better understanding of economic,
physical, mental, social, cultural, and spiritual needs of older family caregivers could greatly
contribute to the sustainability of family caregiving and quality of life of older family caregivers
(Carpentier & Ducharme, 2007; Krause, 2009; Netto, Jenny, & Philip, 2009; Sim-Gould &
Martin-Matthews, 2007; Traustadóttir, & Sigurjónsdóttir, 2008).
Statement of problem
Despite an increasing dependence on older family caregivers in overall health care system, a
risk perspective has exclusively dominated research on family caregiving in later life. It is
evident that research has disproportionally focused on negative health outcomes of family
caregiving burden, increased risk for mortality, decreased immunity function, cardiovascular
disease, depressive symptoms, and potentially harmful caregiver behavior and elder abuse
(Kiecolt-Glaser, Preacher, MacCallum, Atkinson, Malarkey, & Glaser, 1991, 1996; Schulz &
Beach, 1999; Beach et al., 2005; Nakanishi, Hoshishiba, Iwama, Okada, Kato, & Takahashi,
2009). Pinquart and Sörensen (2003) reviewed 228 family caregiving studies in 1986–2006.
They suggested that 151 studies assessed caregiving burden, stress, and strain, and 120 studies
measured depression, and that only 28 studies assessed uplifts of caregiving such as satisfaction,
enjoyableness, and perceived gains. Additionally, older family caregivers have been treated as
co-users of services or “hidden patients” (Fengler & Goodrich, 1979; O’Mara, 2005).
Furthermore, this disease-focused research has pathologized older family caregivers by
categorizing them with the diseases that their care recipients have (Alwin, Öberg, & Krevers,
2010; Gideon, 2007; Jankauskiene, Lesauskaite, & Naumaviciene, 2007).
A review of family caregiving literature suggested that family caregiving research has
identified problems with aging and family caregiving as the following: (1) Increasing old age

2

dependency ratio: The emphasis on old age dependency ratio refers to older adults as dependents
of the family and the society; (2) Disease and disability of older adults were identified as key
reasons for family caregiving and care burden: By focusing on one or many elder care-related
physical, functional, cognitive, or mental ills, family caregiving studies detailed the dependency
of older adults and referred to older adults as “patients” (Eaves, 2006; Gilley, McCann, Bienias,
& Evans, 2005; Kim & Schulz, 2008); and (3) Cost of institutional care: Research has
consistently estimated the total cost of aging and subsequent demands for institutionalization.
Using U.S. Census data, Schneider and Guralnik (1990) made a compelling argument that the
aging of America has a direct impact on escalating health care costs. This argument provided a
framework for the cost of illness studies of dementia (McDaid, 2001; Guentin, Riedel-Heller,
Luppa, Rudolph, & König, 2010). For instance, Harrow et al. (2004) estimated the total annual
cost per care receiver as $23,436 for informal caregiving and $8,064 for formal service in 1997
dollars, based on a national sample of 1,200 family caregivers for people with Alzheimer’s
disease.
Epistemologically, a pathology-oriented framework pays little attention to conditions and
characteristics that contribute to a full understanding of older Americans and family caregiving
in later life. Positive gerontology in this study highlights four counterpoints of a pathological
framework. First, the majority of older Americans, including the oldest old, increasingly remains
independent and remains in the home and community as long as possible (Raphael & Cornwell,
2008). Secondly, a complete picture of 21st century older Americans must include the fact that
older adults have become increasingly active actors in the workplace, civic engagement, and
private relations (Hinterlong, Morrow-Howell, & Rozario, 2007). Older adults are not only care
receivers, but they are also care providers to people across all age groups, particularly older

3

family members and friends suffering geriatric diseases and/or disabilities, adult children or
siblings with chronic disease and disability, and grandchildren under age 18 whose birth parents
are absent from parental responsibilities (Brown et al., 2009; Fujiura, 2010; Standing, Musil, &
Warner, 2007). Third, family caregiving delays or reduces institutionalization of dependent older
adults or disabled people (Van Houtven & Norton, 2008). In the case of grandparents raising
grandchildren in skipped generational households, family caregiving avoids the involvement of
the foster care system. Equally important, it is generally agreed that family caregiving provides
care that is of high quality, consistent with care recipients’ preferences, has better care outcomes,
and costs less (Raphael & Cornwell, 2008). Finally, family caregiving helps reduce overall
health care costs. Arno et al. (1999) estimated that the economic value of family caregiving was
equivalent to approximately 18% of the total national spending on health care. Harrow et al.
(2004) estimated family care cost per care recipient with Alzheimer’s disease was almost three
times the cost of formal services. In 2007, the annual economic value of family caregiving was
estimated at $375 billion, which far exceeded public expenditures for home health and nursing
home care combined and approached total expenditures for the Medicare program (Houser &
Gibson, 2008; Robison, Fortinsky, Kleppinger, Shugrue, & Porter, 2009).
Since a pathological framework often appears in scientific language and in forms of policy
documents, research and human services of family caregiving in later life are constructed and
organized by what we have known through pathological lenses rather than by what little is
known about the productive roles of older family caregivers or positive features of family
caregiving in later life (Gopalan & Brannon, 2006; Zarit & Femia, 2008). What literature has
best documented and what we know best can be described as (1) the notion of “the adverse
physical and psychological health effects of family caregiving” is well established (Vitaliano,

4

Zhang, & Scanlan, 2003; Fortinsky, Tennen, Frank, & Affleck, 2007; Pinquart & Sörensen,
2003, 2007); and (2) family caregivers, especially older family caregivers, are at higher risk for
physical and psychosocial health problems than noncaregivers (Beeson, 2003; Carter & Clark,
2005; Pinquart & Sörensen, 2003).
The current study observed that incomplete knowledge about health outcomes of family
caregiving in later life has provided biased evidence that contributed to public meanings of being
old and being an older family caregiver. This observation is supported by strong evidence that
family caregiving in later life is not the single determinant of negative health outcomes among
older family caregivers (Letiecq, Bailey, & Kurtz, 2008; Pinquart & Sörensen, 2007).
Psychology has well established that psychological states contribute to physical outcomes.
Negative psychological states such as grief, stress and sense of powerlessness more likely lead to
negative physical and psychological health outcomes, while positive psychological states such as
love and gratitude more likely result in positive health outcomes (Edwards & Cooper, 1988;
Fredrickson, 2004; Ong, Bergeman, Bisconti, & Wallace, 2006). Family caregiving research can
minimize biases by examining positive aspects of family caregiving in later life beyond its risks,
stress, and negativity.
Evidence from recent quantitative and qualitative data has suggested that older family
caregivers recognize social, psychological, cultural, and spiritual values of caregiving
(Jankauskiene et al., 2007; Netto et al., 2009). A five-year longitudinal study of 166 caregiverdementia patient dyads found that higher levels of satisfaction with social support networks
among dementia family caregivers predicted fewer depressive symptoms and higher levels of life
satisfaction (Clay, Roth, Wadley, & Haley, 2008). Based on a random sample of 339 CanadianChinese family caregivers, Lai (2010) examined the effects of filial piety on the appraisal of

5

caregiving burden and suggested that filial piety as a cultural belief served as a protective
function to reduce the negative effects of stressors and to enhance the positive effect of appraisal
factors on caregiving burden. A strength-based gerontological approach has called for
advancement of our understanding of family caregiving experience in later life, with the focus on
its social recourses and positive outcomes.
Based on the Changing Lives of Older Couples sample, Brown, Nesse, Vinokur, and Smith
(2003) found that mortality was significantly reduced for individuals who reported providing
instrumental support to friends, relatives, and neighbors, and individuals who reported providing
emotional support to their spouse. This finding suggested that helping behavior and social
contact influence health and longevity. Using a national sample of 1,229 older dementia
caregivers, Tarlow, Wisniewski, Belle, Rubert, Ory, and Gallagher-Thompson (2004) tested a
newly developed measure for the positive aspects of caregiving and found that caregivers
perceived their caregiving as providing them with a variety of positive and satisfying
experiences. These caregivers frequently reported that caregiving made them feel important and
good about themselves. Most caregivers also reported that caregiving enabled them to develop a
more positive attitude toward life and strengthened their relationships with others. Based on three
empirical studies on the role of positive emotions in the stress process in late adulthood, Ong et
al. (2006) found that the occurrence of daily positive emotions served to moderate stress
reactivity and mediate stress recovery and suggested that over time, the experience of positive
emotions functioned to assist highly resilient individuals in their ability to recover effectively
from daily stress (Ong et al., 2006).
It is observed that the pathological family caregiving research failed to identify factors that
contributed to positive and negative health outcomes among older family caregivers, which

6

further contributed to drawing clear conclusions about the nature and the prevalence of negative
health outcomes (Pinquart & Sörensen, 2003). This observation is supported by the fact that
sources of stress and depressive symptoms existing among family caregivers are complicated
(Carter & Clark, 2005; Gopalan & Brannon, 2006; Kiecolt-Glaser et al., 2003; MontoroRodríguez, Kosloski, Kercher, & Montgomery, 2009). Without consideration of the roles of
socioeconomic and psychological factors in the family caregiving context, any attempt to
generalize particular health effects of caregiving stress would be unscientific. Given, Sherwood,
and Given (2008) addressed the sense of lack of preparation as a source of stress for family
caregivers of people with chronic illnesses and multiple comorbid conditions. They proposed
knowledge and skills that gerontological social work should possess to prepare caregivers.
Feelings of unpreparedness as well as a sense of lack of mastery, grief, loss, prior conditions,
“patient suffering” or certain lifestyle could lead to stress among family caregivers (Jansen et al,
2007; Neundorfer, McClendon, Smyth, Strauss, & McCallum, 2006; Noyes et al., 2010; Roepke
et al., 2009; Schulz et al., 2007). Failing to control for unmeasured confounding variables has
been a major threat to the validity of inferences made about health effects of family caregiving.
Family caregiving literature generally considers caregiving as a predictor of social isolation
among older family caregivers. However, recent evidence has presented different aspects of
social positions of older family caregivers and has demonstrated the importance of including
various social, psychological, and cultural resources in studying family caregiving in later life
(Phillips & Christ, 2008; Sanders, 2007). Observations about positive features of family
caregiving in later life can provide opportunities to examine conditions and characteristics that
contribute to a full understanding of strengths, capacities, and strongest qualities of older family
caregivers. On the other hand, positive gerontology of family caregiving is relatively new and

7

less documented. Using four waves of the Asset and Health Dynamics (AHEAD) cohort of the
Health and Retirement Study, Brown et al. (2009) recently reexamined the relationship between
caregiving and mortality. They reached the conclusion that caregiving behavior was associated
with decreased mortality risk rather than increased mortality risk, but a generalizable conclusion
between caregiving and mortality of caregivers was far from being reached. More empirical
evidence is needed to test positive gerontological assumptions, especially what and how social
relations play roles in the health of older family caregivers when they become more emphasized
in family relationships and mastering social networks (Brown et al., 2009; Carpentier &
Ducharme, 2007; Gergen & Gergen, 2003).
Trends of family caregiving in later life
Traditionally, caregiving was a natural component of the family system and community
identity in the United States as well as in the rest of the world (Ruggles & Browner, 2003;
Holzman, 1998; Stack, 1974; Whyte, Alber, & Geissler, 2004). Unpaid informal caregiving was
no longer the only social protection as profound sociodemographic changes occurred beginning
in the late 19th century (Elmore & Talley, 2009). The most notable shifts were population aging,
urbanization, household living arrangement, increasing number of women in the paid workforce,
and the establishment of public protections and advocacy systems (Glasgow, 2000; Alwin,
Converse, & Martin, 1985; Costa, 1999; Haber & Gratton, 1994).
In 1900, 4.1% of 76 million Americans were age 65 and over, and 57% of them lived in a
multigenerational family household and shared the responsibility and protection of kinship
networks (Ruggles, 2003). In 1910, only 20.6% of elderly whites lived alone or with a spouse
(Ruggles, 1994). Over the last century, people age 65 and above were the most rapidly growing
segment of the national population. In 2010, the number of older Americans grew to 40.3 million

8

Americans, or 13.0% of the total population in the United States (US Census, 2011). More
importantly, population aging was concurrent with industrialization. As early as 1910, Addams
had recognized that industrialization had deconstructed the traditional family support system for
children, the disabled, and the aged in rural American communities (Addams, 1910). Assistance
and protection provided by kin and neighbors were not sufficient. This was due to declining
multigenerational households and unavailability of family care resources (Ruggles, 1994;
Chappell & Blandford, 1991). Multigenerational households were no longer a norm. More
people lived alone or lived in nuclear households than before. In 2008, 50.5% of older adults
lived alone or with a spouse (US Census, 2010), a near 30% increase from 1910. All these
changes have gradually reshaped family caregiving context and family caregiving itself
(Szinovacz & Davey, 2008; Elmore & Talley, 2009; Copen, 2009; Weeks et al., 2009).
Family caregiving, a public agenda
As Addams (1910) advocated and fought for, a new protection and advocacy system was
established to respond to a faded traditional social support system. Caregiving increasingly
become a public agenda. Social Security, Medicare, and other preventions were enacted to insure
economic security and health of older adults and their families. To date, family caregivingspecific federal initiatives have provided a necessary policy environment for family caregivers
(Elmore & Talley, 2009; Leos-Urbel, Bess, & Geen, 2002).
The National Family Caregiver Support Program, the Aging and Disability Resource
Centers, the Lifespan Respite Care Act of 2003 (H. R. 6893) and the Fostering Connections to
Success and Increasing Adoption Act of 2008 (P. L. 110- 351) were established to deliver direct
services for family caregivers. For example, the Older American Act Amendments of 2000 (P.L.
106-501) created a federal caregiving program called the National Family Caregiver Support

9

Program. The Older American Act Amendments of 2006 (P.L. 109-365) established three
amendments: (1) family caregivers of persons with Alzheimer’s disease or a related dementia
may be served regardless of the age of the care recipient with dementia; (2) grandparents and
other relative caregivers providing care to children under age 18 years may receive services at 55
years of age and older; and (3) grandparent or relative caregivers providing care for adult
children with a disability between 19 and 59 years of age may be served under the program.
The federal family caregiving agenda has employed policy approaches beyond direct
services. They include financial incentives and compensation, consumer-directed approaches,
and employment-based mechanisms (Elmore & Talley, 2009). They foster community-based
long-term care for the disability and the aging communities as well as kinship child care (Kearns
& Andrews, 2005; Berrick, 2008). Ongoing extensions of federal initiatives have led predictions
that the art and science of caregiving has made a meaningful transition (Whittington, 2010).
Family caregiving, an aging experience
The Older American Act Amendments of 2006 specifically respond to the fact that an
increasing number of older adults has been involved in family caregiving. The prevalence of
family caregiving among older adults suggests that a better understanding of family caregiving
should become a gerontological social work mission. It is simply because family caregiving has
become a part of the aging experience not only for older care receivers but also for all older
adults.
Social capital, a continuity of family caregiving
The prevalence of family caregiving in later life addresses the continuity of family
caregiving, which is often ignored when focus has been on changes and differences from an
historical perspective (Uhlenberg & Cheuk, 2008). Although family protection is no longer a

10

solo social protection mechanism, individuals and families still use their social capital to provide
much needed family care. Kin, neighbors, and friends play important roles as much as policy
measures and professional care providers (Keating & Dosman, 2009). Existence and use of
social capital in contemporary American society partially explain why an estimated 12.1% of
non-institutionalized Americans with a disability were able to enjoy community living (US
Census, 2010). Although institutionalized long-term care has been a hot topic since the 20th
century, the fact is that the majority of older Americans, with or without disabilities, are aging in
place with their family and community. According to the 2010 U.S. Census, only 4.9% of older
Americans lived in nursing homes and other group facilities (US Census, 2010). Even for
institutionalized older adults, family caregiving continues as a complement to professional
services (Chappell, & Blandford, 1991; Yeh, Wierenga, & Yuan, 2009).
Theoretical framework
Positive gerontology has fundamentally shaped and defined this study. This theoretical
framework is derived from two lines of inquiry—critical gerontology and positive psychology
(Estes, 1999; Calasanti & Slevin, 2001; Ong & Bergeman, 2004). Primarily based on critical
theory and political economy, critical gerontology emphasizes the broad implications of political,
economic, and social relations for the aging and for treatment of older adults in the society. It
illuminates topics such as the social construction of caregiving and the impacts of the current
politics of retrenchment on minority older adults.
A positive psychology emphasis is based on the fact that the disease model does not move
psychology closer to the prevention of physical disease, mental illness or social problems.
Seligman and Csikszentmihalyi (2000, p. 7) declared that: “Psychology is not just the study of
pathology, weakness, and damage; it is also the study of strength and virtue.” In its quest for

11

what is best within individuals and institutions, positive psychology does not base its positive
orientation on wishful thinking (Seligman & Csikszentmihalyi, 2000). Rather, positive
psychology has accumulated enormous empirical studies on how positive features promise to
improve quality of life and prevent pathologies (Gable & Haidt, 2005; Salove, Rothman,
Detweiler & Steward, 2000; Vaillant, 2000).
Positive gerontology challenges the assumption that aging is a life stage of decline,
disengagement, and disease. It critically examines the socially constructed dependence status of
older adults and the pathology model of gerontology. It commits to change the course of aging in
more positive ways (Calasanti & Slevin, 2001; Holstein & Minkler, 2007; Minkler & Estes,
1999). As Gergen and Gergen (2003) proposed, positive gerontology reconstructs the discourse
of aging as a time to return to self-growth and relationships. By adapting positive gerontology in
research on family caregiving in later life, this study rejects the notion that caregiving can be
isolated from larger political, social or economic realities. It emphasizes the relevance and
significance of social relations for understanding how family caregiving in later life is socially
constructed and processed in particular social contexts. Two propositions that redefine family
caregiving in later life are emerging.
First, under the framework of positive gerontology, this study conceptualizes family
caregiving in later life as a socially constructed product. My assumption implicit in socially
constructed caregiving is that family caregiving in later life is shaped by social structure and by
those social factors that influence the individual’s position in that structure. More specifically,
this study analyzed social factors such as age, gender, race, ethnicity, and social relations in
questions such as how older family caregivers engage in caregiving in contemporary America,
and what family caregiving means to older caregivers and their social network members directly

12

or indirectly involved in family caregiving. It is expected that the meaning of caregiving for a
white, middle-class married man is different from the caregiving experience of an AfricanAmerican single woman, a Latino immigrant woman in a three-generational family, or other
older adults in other social contexts. People in caregivers’ social networks constantly and deeply
form the caregiving experience not only through resources but also through interpretation.
Identifying characteristics of older family caregivers’ everyday social networks and the ways that
everyday social networks link to family caregiving in later life, will help to understand how
family caregiving in later life as older caregivers and their social network members see it, which
may contribute to the demystification of family caregiving in later life.
Second, contrasting with ageism assumptions, this study moved the research focus to a
positive vision of family caregiving in later life (Hunt, 2003; Johnson-Dalzine, 2007). My
argument in positive family caregiving is that the availability and sustainability of family
caregiving as well as quality of life among older family caregivers may rely on their strengths
and capacity. Gerontology is not just a branch of human medicine concerned with illness of older
adults. It is also about better understanding of strengths and growth among older adults. From
positive gerontology perspectives, research on family caregiving in later life can contribute to
identifying and nurturing older family caregivers’ strongest qualities and can empower them to
find niches in which they can best live out these strengths (Seligman & Csikszentmihalyi, 2000).
Empirical evidence indicates that family caregiving involves both positive and negative features.
Positive features include positive emotions, beliefs, and attitudes like reward, gratitude,
satisfaction, closeness, and growth (Brown, 1993, 2007; Grant, 1998; Robertson, Zarit, Duncan,
Rovine, & Femia, 2007). When investigating the relations between psychosocial predictors and
the course of illness in men with HIV infection, Taylor, Kemeny, Reed, Bower and Gruenewald

13

(2000) suggested that psychological beliefs such as meaning, control, and optimism act as
resources, which may not only preserve mental health in the context of traumatic or lifethreatening events, but be protective of physical health as well.
Purpose of the study and research questions
The debate on health outcomes of family caregiving presents the core proposition of
contemporary geriatrics that shapes older adults’ images and the social environments in which
older adults live. There has been a consistent focus on health costs for older family caregivers.
An emerging understanding of family caregivers, however, argues that values and gains are also
evident in being part of family caregiving experiences and that social networks link to health
outcomes of caregivers (Gerard, Landrey-Meyer, & Guzell, 2006; Goodman et al, 2007; Netto et
al., 2009; Nijboer, Triemstra, Tempelaar, Sanderman, & van den Bos, 1999; Nijboer, Triemstra,
Tempelaar, van den Bos, & Sanderman, 2001; Pinquart & Sörensen, 2006; Yap, Luo, Ng,
Chionb, & Lim, 2010).
The purpose of this study was to better understand the social value of family caregiving in
later life by examining effects of social networks and family caregiving in later life on the health
of older adults aged 65 and over. Drawing theoretical support from positive gerontology and
methodological efforts of social networks analysis (Cohen & Wills, 1985; Kawachi & Berkman,
2001; Minkler, Driver, Roe, & Bedeian, 1993), this study expected to not only revisit the
significance of social networks in family caregiving in later life, in particular, and in positive
gerontology, in general, but also called for strength-based geriatric social work research and
health care policies.
In line with existing research on social networks and the health of family caregivers in later
life, the study examined the following three research questions:

14

Question 1: What are the demographics, social network characteristics, and health status of
older family caregivers, compared to other older adults who do not provide family caregiving?
Question 2: How do older family caregivers perceive family caregiving in later life?
Question 3: Are there significant mean differences in health for different levels of social
networks and for different family caregiving status after controlling for demographic factors?
Conceptual definitions
Based on positive gerontology, the current study identified four key concepts respectively—
older family caregivers, family caregiving, social networks, and health—are the key aspects of
the study.
Older family caregivers
In the current study, older family caregiver was defined as any community-dweller who is
aged 65 and over and has provided unpaid care for a family member, friend, or neighbor with
chronic illness, disability, or other dependent characteristic on a regular basis in the last 12
months. As a concept of positive gerontology, the notion of older family caregivers in this study
described caregiving as community-based caregiving, intergenerational caregiving, and family
caregiving. As community-based caregiving, the notion of “older family caregiver” highlights
the social and physical location of the caregiver, no matter whether the family care receiver is
community-dwelled or institutionalized. Literature has suggested that the subjective
responsibility of caregiving remains with family caregivers after family care receivers were
institutionalized (Hibbard, Neufeld, & Harrison, 1996). In intergenerational caregiving, three
groups of older family caregivers are commonly differentiated in research: caregivers of older
family members suffering geriatric diseases and/or disabilities, caregivers of adult children or
siblings with chronic disease and/or disability, and caregivers of grandchildren under age 18

15

whose birth parents are absent from parental responsibilities (Greenberg, Seltzer, & Greenley,
1993; Standing et al., 2007). Family caregivers in later life include primary caregivers, secondary
caregivers, and other caregivers with different degrees of care involvement. Penrod, Kane, Kane,
and Finch (1995) noticed that approximately 7% of the 242 primary caregivers reported caring
for the family member alone. Tolkacheva, van Groenou, Boer, and Tilburg (2011) argued that
larger caregiving network size, more shared care tasks in the network, and longer periods of
shared care contributed to caregivers’ lower care burden.
Family caregiving
An older family caregiver is defined as anyone involved in unpaid care for a kin or a non-kin
person. The current study highlighted family caregiving in later life as a social event in which
social network members collectively contributed to care responsibility. For gerontological social
workers, the social meaning of family caregiving contains much potential for resources to be
discovered and utilized for better understanding and better services.
It is essential for family caregiving research to define what constitutes care. Literature has
contributed to a hierarchy of care that embraces a range of family caregiving activities (Nolan,
Keady, & Grant, 1996; Ekwall, Sivberg, & Hallberg, 2004). At the base of this care hierarchy,
“personal care” refers to informal help with instrumental activities for daily living. On the next
level of the care hierarchy, “instrumental care” is usually defined as informal and
nonprofessional help with personal activities for daily living. Although personal care and
instrumental caregiving are two dimensions universally accepted by family caregiving literature,
what constitutes instrumental caregiving has remained controversial. The majority of geriatric
caregiving literature most commonly referred to it as care for activities for daily living or
instrumental activities for daily living, especially Activities of Daily Living (ADLs) or

16

Instrumental Activities of Daily Living (IADLs) (Katz, Ford, Moskonitz, Jackson, & Jaffe, 1963;
Lawton & Brody, 1969). Others argued that a geriatric perspective of care dimensions could not
reflect a broader notion of family caregiving, which involves child-specific care as well as
disabled advocacy (Oktay, 2006). Some researchers have looked beyond personal care and
instrumental care in the hierarchy of care. They addressed the importance of care dimensions
such as preservative care, anticipatory care, preventive care, supervisory care, protective care,
emotional care, reconstructive care, and reciprocal care (Nolan et al., 1996; Oktay, 2006). Based
on a sample of 783 older family caregivers, Ekwall et al. (2004) found only a small percentage of
the sample provided help with personal care, and the hierarchy of caregiving was described as
anticipatory, preventive, supervisory care, instrumental care, personal care, reconstructive care,
and reciprocal care. Using gender analytical lenses, Campbell and Martin-Matthews (2003), on
the other hand, categorized care activities as traditional gendered care and gender neutral care.
For the purpose of the current study, care was defined from five dimensions: personal care,
instrumental care, emotional care, informational care, and advocate care. This definition included
care from caregivers’ daily activities to macro influence. It covered not only the scope of family
caregiving research, but also the range of gerontological social work interventions.
Social networks
In this study, social networks were defined as older family caregiver-centered
multidimensional social relationships that cut across traditional kinship, residential, and class
groups (Barker, 2002; Berkman, Glass, Brissette, & Seeman, 2000; Bott, 1957; Chatters, Taylor,
& Jackson, 1985). Social networks is a measurable term for the total of social connections that
most often consist of the caregiver’s spouse or partner, friends, neighbors, relatives, and
members from religious or spiritual and community organizations (Berkman & Syme, 1979;

17

Lubben, 1988). More specifically, social networks are considered from three distinct dimensions.
First, the structural dimension comprises its size, composition, frequency, density, strength of
ties, reachability, and reciprocity of the network relationship. Secondly, the content dimension
consists of information, values, and behaviors conveyed by network members. Finally, the
functional dimension refers to the network’s support, conflict, and social control. A supportive
social network could be described in terms of connectedness and social integration. Egocentric
social network connectedness includes seven network properties: marital status, co-residence,
network size, volume of social interaction with network members, closeness to network
members, network composition, and network density. Social integration includes three network
properties: religious participation, organized group involvement, and volunteering.
As Carpentier and Ducharme (2003) pointed out, a theoretic and empirical distinction should
be made between social network and social support. The notion of a social network calls
attention to social contexts of social relations among network members and is defined by its
structural parameters. On the other hand, the notion of social support emphasizes interpersonal
help among network members and is defined by helping behavior from one network member to
another to meet physical, psychological, and social needs (Scott & Roberto, 1985). While social
network analysis examines how social relations in the caregiver network function through
information, values, and behaviors, it targets social support as well as social conflict and social
control in which gerontological social work researchers identify and advocate for the strongest
qualities of the caregiver and her/his networks.
Health
Health was defined as highly self-rated health, a high level of physical function, absence of
chronic diseases, and absence of depressive symptoms. Social networks analysis of health in

18

later life concentrates on the structure and composition of the networks, the resources that are
embedded within the networks, and their influences on the health of actors in the networks
(Berkman et al., 2000; Moren-Cross & Lin, 2006). In this study, the health effects of social
networks were linked to actual and perceived physical and mental health of older caregivers. The
health effects of social networks contributed to the perceived value of older family caregivers.

19

CHAPTER TWO: LITERATURE REVIEW

It is evident that social networks are crucial to understanding how the needs of older adults
are met, and how older Americans achieve positive physical and psychological health outcomes
(Christakis & Fowler, 2007; Cornwell & Waite, 2009; Lubben & Gironda, 2003; Newsom,
Mahan, Rook, & Krause, 2008). Literature has argued that strong social networks are
significantly associated with low mortality, high self-rated health, and high levels of cognitive
functioning among older patients (Blazer, 1982; Crooks, Lubben, Petitti, Little, & Chiu, 2008;
Lyyra & Heikkinen, 2006), and that social support networks play a preventive role in chronic
diseases, cognitive decline, and mental diseases (Beeson, 2003; Boden-Albala et al., 2005;
Cornwell, Schumm, Laumann, & Graber, 2009; Crooks et al., 2008; Nijboer et al., 2001; Schulz
& Beach, 1999; Weiss, 1973). Less is known about family caregivers’ social relations and health
effects of their personal social networks (Edwards, Higgins, & Zmijewski, 2007). Attempts to
examine the intersection among social networks, health, and family caregiving in later life have
been rare.
Methodologically, there has been inconsistency in available literature as to how to
conceptualize social networks and the mechanisms through which social networks affect healthy
aging. For instance, when Pinquart and Sörensen (2007) explained why a meta-analysis of 176
studies indicated weaker associations of social support with physical health than with burden and
depression, they considered that most social support instruments were assessing emotional
support, which was more likely to correlate with psychological than physical health. While many
have referred social networks to social network structure, even simply to size of social networks,
an effort to integrate the multidimensional concept of social networks has been persistent since
the 1970s (Berkman & Syme, 1979; Cornwell et al., 2009; Ekwall, Sivberg, & Hallberg, 2005).
20

Inconsistency in the understanding of social networks in family caregiving literature has made it
even more difficult to determine which social relations older family caregivers have in everyday
life, whether or not these relations are different from those of non-caregiving older adults, and
how social networks affect health of older family caregivers.
Review of the literature on health effects of social networks of family caregiving in later life
suggested that social network analysis of health and family caregiving is an emerging social
work approach. Gerontological social work has used social network analysis to study caregiving
networks, the role of social support and social capital, and the influence of personal networks on
health behavior and information transmission, particularly on help-seeking behavior. However,
most network investigations have centered on the elder patient rather than the family caregiver
(Barrett & Lynch, 1999; Keating & Dosman, 2009). It was also notable to mention that many
recent available family caregiver network data have come from Canada and Europe (Carpentier
& Ducharme, 2005, 2007; Lai, 2010; Tolkacheva et al., 2011). In limited caregiver network
literature, social network analysis investigated social network structural characteristics and the
relationships in the network to determine how social ties influence care burden and stress, health,
and the organization of resources of family caregivers.
Caregiving networks
The current study distinguished social networks of family caregivers and family caregiving
networks. Social networks include all network members with whom family caregivers have
connections and interactions, regardless of whether or not they were active connections.
Caregiving networks are also referred to as “helper networks” (Chatter et al., 1985) and “care
networks” (Fast, Keating, Otfinowski, & Derksen, 2004; Keating & Dosman, 2009; Keating,
Otfinowski, Wenger, Fast, & Derksen, 2003). Caregiving networks, on the other hand, only

21

focus on those who are involved in family caregiving in the social networks of family caregivers.
Keating and colleagues (Fast et al., 2004; Keating & Dosman, 2009; Keating et al., 2003) made a
similar distinction between support networks and care networks and called care networks a
functional subset of a support network. In other words, a personal social network served as a
pool of caregiving assets, and a caregiving network was an activated care resource within the
network. Research and policy could not assess care resources only by size of personal social
networks or support networks of family caregivers.
Existing literature on caregiving networks attempted to move research attention from primary
caregivers and care receiver dyad to informal caregiving networks. This was a critical shift for
family caregiving research because it challenged the assumption that family caregiving was
strictly a one-to-one relationship and that the primary caregiver provided a lone and continuous
care, isolated from his or her everyday social relationships (Traustadóttir & Sigurjónsdóttir,
2008). Efforts have been made to understand the collective contribution of the family caregiving
network to the care receiver as well as to the primary caregiver (Keating et al., 2003; Ray &
Street, 2005). Research has demonstrated the key to quality and sustainable family caregiving is
its collective nature. More specifically, caregiving networks comprise primary caregivers,
secondary caregivers, and helpers of caregivers and it has become part of caregiving networks
research (Tennstedt, McKinlay, & Sullivan, 1989; Connidis, 2001; Sims-Gould & MartinMatthews, 2007).
Literature on the dynamics of caregiving networks reveals that networks were not merely a
static social reality. Two themes were particularly important for family caregiving research: the
changing caregiving networks and the impact of relational influences within networks. Drawing
data from the Longitudinal Study of Aging (LSOA), Peek, Zsembik, and Coward (1997)

22

described changes in caregiving networks of older adults over a two-year period and patterns of
network evolution, which were characterized by both stability and change. Carpentier and
Ducharme (2005) further investigated support network transformations in the first stages of
family caregiving for individuals with dementia. Although this case study could not be
generalized to all family caregivers across the life span, it observed two meaningful networks in
transition: an adjusted caregiving network that better served the family caregiving function, and a
diminished network that exhausted the spousal caregiver. This observation suggested that our
understanding and utilization of caregiving networks could have meaningful impact on
caregiving capacity and sustainability as well as health of caregivers. Jewell and Stein (2002)
demonstrated how parents of individuals with severe mental illness could influence network
members, siblings in this case, to be involved in caregiving and caregiving plans.
Network characteristics of older family caregivers
Available literature on family care networks has been most interested in structural properties
of social networks in relation to the health outcomes of family caregivers. Although social
network analysis has investigated network characteristics beyond network structure, the two
network properties that have caught most research attention are size and composition of social
networks among older family caregivers.
Network size
Network size of older family caregivers holds special importance in gerontology. Literature
has reached a general consensus that decreased social participation is a common early response
to caregiving, which results in feelings of isolation, loss, and frustration (Hibbard, Neufeld, &
Harrison, 1996; Carpentier & Ducharme, 2003, 2005; Szinovacz & Davey, 2006, 2007). Based
on qualitative data of 30 elderly men caregivers for spouses with cognitive impairment, Russell

23

(2004) described radical changes in social networks of retired male family caregivers as a shift
from a visible and interactive world to an invisible and isolated one.
Across studies, the majority of family caregivers reported having two to three people in their
networks who were immediate family members, distanced family members, friends, or
neighbors. In a sample of 242 Medicare beneficiaries from 52 hospitals in Pittsburgh, Houston,
and Minneapolis/St. Paul, Penrod et al. (1995) documented an average of 2.33 social network
members per primary caregiver, with an average of 2.1 network members who shared care
responsibilities, and .25 and .69 network members assisting financially and with decisions,
respectively. When exploring network characteristics of 602 Dutch caregivers, Tolkacheva et al.
(2011) indicated a caregiving network size of 2.76 in the sample. However, other studies
reported a larger network size. Greenberg et al. (1993) reported a network size of 7.9 among 313
aging mothers of adults with disabilities. More recently, Haley, Roth, Howard, Monika, and
Safford (2010) reported a similar network size (7.2) in a sample of 767 spouse caregivers for
whom the authors examined the relationship between caregiving strain and stroke risk.
Although sampling and other study characteristics might contribute to the difference in the
reported sizes of caregiver networks, inconsistency in defining networks as well as family
caregiving activities have excluded or included some network members in family caregiving
studies (Nolan et al., 1996; Ekwall et al., 2004; Oktay, 2006). In a pioneer caregiver network
study, Chatter et al. (1985) innovatively portrayed patterns of informal helper networks of
elderly blacks. It argued that the majority of the sample (56%) reported a network size of three.
Since the study only measured care help in the instrumental dimension, the actual size of social
networks among family caregivers would be larger when emotional help is included. More
recently, Phillips and Crist (2008) found that over a one-year period network sizes for caregivers

24

in both Mexican American and non-Hispanic white groups remained the same, while changes
happened within network throughout the year of caregiving: loss of about three network
members and gain of three new members.
Even though inconsistency exists in caregiver network size, literature has shown that family
caregiving in later life, in general, is not isolated. According to Chatters et al. (1985) and Penrod
et al. (1995), both national and regional representative data reported a low percentage of
caregivers (7%) for family members with post-acute care with no help from anyone else.
Chatters et al. (1985) further examined the differences in caregiver network size and found that
sex, marital status, and region were significant predictors of helper network size, with older
women having larger networks than older men, married persons having significantly larger
network sizes than widowed and never married persons, and persons in the South having larger
size networks than people in the North Central and Northeast regions.
Network composition
Caregiving network composition was another area of interest in social network analysis of
family caregiving in later life. However, how network composition was structured varied across
studies. Sociology and anthropology tended to categorize network composition as kin and nonkin networks. Social capital theory tended to make distinctions between bonding social capital
and bridge social capital. In the case of gender studies, literature sometimes used the feminist
perspective to define network clusters. For instance, in order to explore the association of gender,
sibling network composition, and patterns of parent care, Coward and Dwyer (1990) separated
sibling network composition into three types: only-child networks, single-gender networks, and
mixed-gender networks. Through this approach, the study found that daughters from mixed-

25

gender networks reported significantly higher levels of the association of gender and patterns of
parent care than daughters from only-child networks and single-gender networks.
Literature most commonly employed sociological approaches when differentiating network
composition and concluded that networks composition of older adults could be characterized as
family-based, friend-based, diverse, and limited. Chatters et al. (1985) found that the majority of
the sampled elder blacks (56%) had immediate family only networks, one-third had mixed
networks with immediate family members and others, and the minority (11%) had relatives/nonkin networks. The study also found that lack of close kin network members affected helper
network composition, from few kin members to more distant relative and non-kin members, but
did not affect the size of helper networks. Since the study only measured care help in the
instrumental dimension, the actual size and composition of social networks among family
caregivers would be different when emotional support was included in helper network
composition. Fiori, Smith, and Antonucci (2007) derived six social network types from the
Berlin Aging Study data. Family based and friend based networks emerged as two basic network
compositions in all six network types. Using a social capital framework, Keating and Dosman
therefore called family members and friends “the building blocks of social capital” in care
network of frail seniors (2009).
Literature explained the reason why family members and friends were the building blocks of
elderly network composition from sociological approaches. A preference order in the choice of
support in caregiving networks is always for the closest available person, cascading through
spouse, children, other relatives, and non-kin friends and neighbors (Chatters et al., 1985;
Carpentier & Durchame, 2003). Although parents, spouses, and children have been the core of
family caregiving networks, recognition and inclusion of siblings, in-laws, friends, and neighbors

26

in caregiving networks were critical developments in understanding care availability and
capacity (Barker, 2002; Fast et al., 2004; Lohrer, Lukens, & Thorning, 2007). Weeks et al.
(2009) investigated future plans of parental caregivers for adults with intellectual disabilities and
suggested that siblings and other family members were reported as the top candidates of informal
caregivers in parents’ plans after they died. Based on a random sample of 23,588 from the 2001–
2007 Southern Australians Health Omnibus Survey, Burns, Abernethy, Leblanc and Thomas
(2011) identified older women acting as friends and neighbors caring for the frail and old. The
study related the contribution of friend caregivers to sustainable community living, which was
measured by the increased likelihood of the care receiver being able to stay at home.
Social capital theory offered another view of network compositions of family caregiving in
later life. Social capital approaches categorized network members as homogenous groups and
heterogeneous groups (Keating & Dosman, 2009). According to social capital approaches,
homogenous groups had strong and intimate ties, which were well suited to providing social and
psychological supports to meet group members’ everyday needs. In the past, family groups were
typical homogenous groups. They shared a strong sense of identity and reciprocal services. They
spent high amounts of time together and exhibited emotional intensity and mutual confiding. Due
to homogeneity, close-kin might have been less likely to have new and alternative access for
family caregiving. While homogenous groups provided intense levels of family care over time,
they also may have mitigated against information exchanges and formal community services
(Furstenberg, 2005; Zacharakis & Flora, 2005). In contrast, heterogeneous groups showed more
potential for connecting caregiving networks with external assets as well as service information.
Therefore, they were better suited to bringing new network assets, including instrumental and
informational resources. Non-kin network members, such as friends and neighbors, were typical

27

heterogeneous groups in the network. They often moved between groups and were more likely to
provide links to external support. In short, homogenous groups and heterogeneous groups in a
caregiving network functioned as two distinct social capitals: bonding social capital and
bridging social capital of family caregiving. As Putnam summarized, these two social capitals
reflected various needs of family caregiving in later life: “Bonding is good for getting by, but
bridging is crucial for getting ahead” (Putnam, 2000, p.23). Obviously, a resilient and resourceful
social network should have diverse social capital to meet emotional, instrumental, and
informational needs of family caregivers.
Carpentier and Ducharme (2009) offered empirical evidence that demonstrated how changes
in network composition carried significant impact on network care capacity and the well-being
of caregivers for family members with Alzheimer’s disease. In the first case, the caregiver’s
network size had remained at eight over nearly two years. The primary caregiver gained one
heterogeneous network member while he lost his mother as a close tie due to the development of
her Alzheimer’s disease. The new friend functioned as a confidant who provided information
about the disease and related services, something which his family members were not able to
offer. In the second case, the size of the network of a wife caregiver had been downsized from
seven to four over a 19-month period while the network composition was changed from two
groups (family and friends) to one group (family). Lack of diversity in network composition
jeopardized the sustainability of family caregiving, even though the caregiver wanted to believe
that she could cope with caregiving challenges without outside assistance (Carpentier &
Ducharme, 2005).
Health effects of social networks

28

Prior literature has suggested a general consensus that social networks were associated with
physical and mental health outcomes. Supportive and diverse social networks were significantly
associated with high self-rated health, high levels of physical functioning, and low vulnerability
to mental disorders among older family caregivers (Bergman & Haley, 2009; Brown et al., 2003;
Crooks et al., 2008; Robison et al., 2009). Early applications of social networks in gerontological
studies have claimed that social networks had a direct and independent effect on health in later
life (Berkman, 1984, 1986; Berkman & Syme, 1979). More recent empirical evidence has
explored the association between family caregiving in later life, social networks, and health.
Attempts have been made recently to understand how social networks influence physical and
mental health outcomes, including the relationship between health and social network aspects
other than network size and/or composition (Carpentier & Ducharme, 2003; 2005; Newsom et
al., 2008). Available empirical evidence indicates that the nature of connectedness and social
integration, as well as information, values, and behaviors that flow in the network, might explain
physical and mental health of older family caregivers as network size, strength of ties,
reachability, and reciprocity of the network relationship might do (Gerard et al., 2006; Kim &
Knight, 2008; Sit, Wong, Clinton, Li, & Fong, 2003).
Social network literature has indicated mixed results about conventional assumptions that
caregiving restricts the social networks of family caregivers and that caregivers experience
higher rates of physical, mental, and emotional health problems than non-caregivers. Using a
large and representative sample in Australia, Edwards et al. (2007) discovered that family
caregivers for individuals with disabilities reported a low level of face-to-face social contact with
friends or relatives not living with them. On the other hand, this study also reported that the selfreported hours of family caregiving did not associate with face-to-face social contact among

29

family caregivers for individuals with disabilities, nor were they associated with their need for
more face-to-face social contact with friends or relatives outside of the household. Drawing
postal survey data from a probability sample of 4,278 older adults aged 75 years and over in
Sweden, Ekwall et al. (2005) found that caregivers had larger social networks and were healthier
than non-caregivers. Meanwhile, results from a quasi-experimental study conducted in the U.S.
Midwest documented that spousal caregivers of persons with Alzheimer’s disease presented
greater loneliness and depression, and that loneliness significantly contributed to the explanation
of caregivers’ depression (Beeson, 2003). This study reported that loneliness was predictive for
spousal caregivers of those with Alzheimer’s disease, explaining 49% of the total variance.
Comparing fathers of adult children with schizophrenia to age-matched peers who did not have
children with schizophrenia, Ghosh and Greenberg (2009) indicated that caregivers showed a
pervasive pattern of poorer mental health and perceived their health as worse. Some explained
that different care conditions and other complicated factors predicted the different associations
between family caregiving, social networks, and health. On the basis of a random stratified
sample in Italy, Magliano et al. (2006) compared social networks of family caregiving for
relatives with schizophrenia and physical illnesses. It suggested that social support is
significantly lower for caregivers of relatives with schizophrenia than caregivers of relatives with
physical illness and general population.
Available literature has proposed a diverse explanation of the association between social
lives and health of older family caregivers. Greenberg, Seltzer, Krauss, and Kim (1997) argued
that effects of social networks are not uniform on psychological well-being across groups of
aging family caregivers. Their study indicated that, although aging mothers of adults with mental
illness had smaller social support networks than aging mothers of adults with mental retardation,

30

social support network was a more prominent predictor of changes in care burden and depressive
symptoms in mothers of adults with mental illness. Magliano et al. (2006) confirmed that the
more support the caregivers perceived in a crisis from their social networks, the better their
mental health. More specifically, the study measured social support as the perceived crisis
support, the perceived routine support, the received crisis support, and the received routine
support. It suggested that the perceived crisis support might provide caregivers with a sense of
security to mobilize psychological and physical resources to handle the crisis, and that the
perception might enable caregivers to receive advice, aid, and affect through their social network
to produce positive psychological states, including a sense of purpose, belonging, and
recognition of self-worth.
Structure dimension
Literature has debated whether family caregivers have a smaller or larger social network
than non-caregivers, but generally it has agreed on a significant association between small social
networks and low levels of physical and mental health among older caregivers. Greenberg et al.
(1997) found evidence that certain social network ties predicted declining well-being and
increasing subjective burden. Weeks et al. (2009) reported that when thinking of future plans for
the adult child with intellectual disabilities, the first concern of the aging mother was the care
capacity within her personal social network. Aschbrenner, Greenberg, Allen, and Seltzer (2010)
indicated that a higher number of confidants were positively associated with personal gains
among older parents of adults with serious mental illness. Parents who provided more assistance
with activities of daily living to their adult children reported higher levels of gains. Lee et al.
(2006) assessed direct and indirect effects of social support on health of family caregivers for
individuals with schizophrenia. The results of the study suggested that perceived social support

31

was positively related to the health of caregivers, while received crisis support had a positive
effect on the caregivers when the care receiver had severe symptoms. Ekwall et al. (2005) found
that compared to non-caregivers, caregivers have had not only a higher chance of having a
confidant to trust (96.4% of caregivers vs. 94.9% of non-caregivers), but also a larger size of
networks (58.5% of caregivers and 52.0% of non-caregivers with three or more persons in their
social network). The authors also reported that caregivers had feelings of loneliness less often
than non-caregivers and that loneliness and weak social network are significantly associated with
low mental quality of life among caregivers as well as in the total sample. Another crosssectional study of 64 family caregivers for individuals with dementia suggested that social
interactions with family and non-family social network members have differential effects on the
emotional well-being of caregivers (Gideon, 2007). More specifically, higher levels of emotional
support from family might tend to reduce overall caregiver emotional distress, while emotional
support from non-familial sources might bolster caregivers’ optimistic outlook on the experience
of caregiving. Rozario, Chadiha, Proctor, and Morrow-Howell (2008) examined the influence of
social resources on depression symptoms among 358 urban African-American wives and
daughters as primary caregivers. The authors measured social resources by social participation,
satisfaction with family functioning, and perceived instrumental support. Multivariate regression
indicated that for the older group, satisfaction with family functioning was the only social
support variable that was significantly predictive of depressive symptoms, a different pattern of
association between social resources and depressive symptoms from the younger group.
However, the cross-sectional nature of the study limited its ability to determine whether
depressive symptoms were a function of social resources or vice versa (Rozario et al., 2008).
Function dimension

32

Family caregiving literature has long been interested in supportive networks and their health
implications. Ample empirical evidence has documented that higher levels of perceived and
actual social support was associated with better health, especially fewer depressive symptoms
among spousal caregivers, parental caregivers and grandparents raising grandchildren (Gerard et
al., 2006; Greenberg et al., 1997; Williams, Williams, Zimmerman, Munn, Dobbs, & Sloane,
2008). Sit et al. (2004) categorized social support into four functional dimensions. Their crosssectional study found that family caregivers for a relative with stroke received more emotional
support and social companionship than tangible support and informational support. It
documented that tangible support and social companionship along with physical independency of
care receivers explained 36% of variance in the psychosocial health of 102 family caregivers.
Clay et al. (2008) provided longitudinal evidence that higher levels of satisfaction with social
support were associated with fewer depressive symptoms and increased levels of life satisfaction
for African American and white dementia caregivers over a five-year period.
A review of literature on social networks and health of older family caregivers has found an
emerging research interest in effects of conflict networks, including positive impact of conflict
networks. Some have developed a social network model that employs ecomap to visually present
supportive and tensioned relationships between caregivers and network members (Neufeld &
Kushner, 2009; Ray & Street 2005; 2011). Neufeld and Kushner (2009) conducted an
ethnographic study of 39 male primary caregivers for a relative with dementia. They emphasized
that observation of family caregivers’ experience about supportive and non-supportive
interpersonal interactions could sensitize social network members as well as professionals to the
context of male caregivers’ experience. In a longitudinal sample of 351 adults and older adults in
Hong Kong, Fung, Yeung, Li, and Lang (2009) reported positive effects of negative social

33

exchanges: more negative social exchanges were associated with more positive change in
emotional closeness over a two-year period, even after statistically controlling for social support
and structural characteristics of the sample and their social partner. Drawing data from the Late
Life Study of Social Exchanges, a national longitudinal study of 666 older adults, Newsom et al.
(2008) provided opposite evidence with more specific measures. They demonstrated that higher
levels of stable negative social interactions were significantly predictive of lower self-rated
health, greater functional limitations, and a higher number of health conditions over two years
after controlling for initial levels of health and sociodemographic variables.
Although literature has not uniformly used social network approaches to explore functional
social networks of family caregivers, it has moved empirical inquiry further by employing more
cause-specific network measures. Among these measures, co-residence and proximity have been
used to explore health effects of social networks. Others have explored social participation, civic
engagement, and other variables and their health outcomes. Some of these variables were
originally included in the Berkamn-Syme Social Network Index (SNI) (Berkamn & Syme, 1979).
Among 12 social relationships assessed by SNI, marital status, employment status, religious
group membership, organizational participation, and volunteering were often ignored by social
network literature in geriatrics, social work, and other related fields. Ekwall et al. (2005)
indicated that caregivers in later life were more likely to live together with someone (74.1%)
than non-caregiving older adults (34.5%), which contributed to higher self-rated health status and
less frequency of the intense feeling of loneliness among family caregivers. Using secondary
data from the 1996 Family Caregiver Survey, Chou, Yeung, and Chi (2001) investigated the
effect of physical distance between caregivers and care receivers in family caregiving
experience, including caregiving strains. Their study indicated different results: caregivers

34

sharing the same household with care receivers reported higher levels of physical, emotional, and
financial strain than other caregivers. It further explained that impact of distance remained after
controlling demographic characteristics of family caregivers in this representative community
sample of 1,509 caregivers.
Drawing data from the 1998 and 2000 waves of the Health and Retirement Study, Choi, Burr,
Mutchler, and Caro (2007) used a volunteerism framework to investigate volunteer activities
among family caregivers. They found that spousal caregiving was not significantly associated
with the likelihood of formal or informal volunteering for men; however, female caregivers were
found to be less likely than non-caregivers to have engaged in formal or informal volunteering. It
was noticed that the health effect of volunteering was not in the scope of the study.
Content dimension
Empirical evidence about health outcomes of the content dimension of social networks
among family caregiving has been rare, although literature commonly has recognized the
importance of health behavior, shared social values, and access to information about services to
family caregiving in later life as well as to health of caregivers. More specifically, literature has
suggested that health behavior, information channeling, and service use have had direct or
indirect effects on health of older adults (Christakis & Fowler, 2007; Lee et al., 2006).
Traditional social network analysis has studied how the ego’s health behavior and health
were influenced by alters in her or his networks. In the Alameda study, Berkman and Syme
(1979) found that shared norms around health behaviors within social networks might be
powerful sources of mutual influence with direct consequences for the behavior of networks
members. They especially examined alcohol and cigarette consumption, health care use, and
dietary patterns. In the same line, Christakis and Fowler (2007) observed a densely

35

interconnected social network of 12,067 people over 32 years and found that obesity spread
through social ties. The explanation was that socially close network members shared social
norms about the acceptability of obesity and related negative health behaviors (Smith &
Christakis, 2008).
Despite proposals for healthy aging and social networks (Berkman et al., 2000; Carpentier &
Ducharme, 2007), literature has failed to provide strong empirical evidence about content
network characteristics and the association between health and the content dimension of social
networks among older family caregivers. For instance, when Castro et al. (2007) conducted an
epidemiologic survey study on rural family caregivers and health behaviors it was the first
empirical research that focused on information about health behaviors, and self-care habits
among rural caregivers. It found that family caregivers reported lower fruit intake, more walking
for exercise, and more advice about stress, fruits, and vegetables than non-caregivers, while the
two groups did not differ in smoking, dietary fat, and obesity.
Literature recently has paid more attention to the association between information shared in
social networks and health of family caregivers. Sit et al. (2004) found that not only was
informational support the most lacking area within caregivers’ social networks, but also
information provided by professionals was not effectively retained. When explaining indirect
effects of social support on health of family caregivers for individuals with schizophrenia, Lee et
al. (2006) explained that social support might provide information about the disease and clinical
knowledge about services, which might alleviate caregivers’ mental burden and benefit their
mental health. Caldwell (2008), on the other hand, documented that poorer access to health care
contributed to poorer mental and physical health among female family caregivers of adults with
developmental disabilities.

36

There has been lack of consistency across studies about how to posit the content dimension
of social networks in conceptual and operational models of family caregiving in later life. For
example, Sit et al. (2004) and others identified information as one of the functional dimensions
of social support. Carpentier and Ducharme (2003; 2005), on the other hand, treated information
as one of the network contents that circulated in social networks and shaped caregivers’ health.
Other examples included health effects of help-seeking behaviors and service use. Available
evidence has suggested low rates of service use among older family caregivers. Keith, Wachwe,
and Collins (2009), for instance, found that among 224 studied family caregivers, the mean score
of utilization of 19 formal elder care services was 2.11 and that family member influence
attributed to caregiver resistance to service use.
However, some have argued that debates over health and social network dimensions could be
meaningless or lead to prejudice and cultural incompetence if social contexts were not
considered in conceptual and operational models (Castro et al., 2007; Hayslip et al., 2006;
Letiecq et al., 2008; Neufeld, Harrison, Stewart, & Hughes, 2008; Phillips & Crist, 2008). This
argument emphasized that the meaning of certain health behavior and information about certain
health conditions and services might vary by gender, racial-ethnic identity, cultural-social values,
socioeconomic boundaries, or other social contexts (Kim & Knight, 2008; Kohn & Smith, 2006;
Valle, 2004). Umberson and Montez (2010) claimed that social relationships, measured by
network size and quality of relationship within the network, have short- and long-term effects on
health. They also stressed the importance of social variation in the link between social
relationships and health outcomes. Evidence provided by Christakis and colleagues (Christakis &
Fowler, 2007; Smith & Christakis, 2008) supported the argument that the meaning of specific
health behaviors varied within social contexts.

37

In summary, a review of literature indicated that knowledge about social networks and health
of older family caregivers has been developing over the last two decades, especially in
conceptual modeling and methodological inquiry (Ekwall et al., 2005; Keating & Dosman,
2009). Efforts have been made to provide empirical evidence from specific social network
characteristics to cultural variations to health effects of social networks (Caldwell, 2008; Fast et
al., 2004; Newsom et al., 2008; Phillips & Crist, 2008; Ray & Street, 2011), which will prove
valuable for further research.
Limitations of existing literature
Despite progress in understanding the relevance of social networks to healthy aging of family
caregivers over the last few decades, many questions remain unanswered. Little is known about
social influence on health of individual older family caregivers at empirical levels, including
non-primary caregivers. Attempts to examine the intersection between social networks and
health of older family caregivers have been scarce. Available empirical evidence was based on
description of structural network characteristics, which was derived from insufficient
understanding of the multidimensional nature of social networks. Inconsistency in operational
modeling to assess health, social networks, and family caregiving have contributed to limitations
in generalization of findings from large representative samples. I describe each of these research
gaps briefly below.
Caregiving network
While literature has come to recognize family caregiving as a collective event, what we know
about social networks and their health effects on family caregiving in later life has been
narrowed, for the most part, to primary caregivers, particularly family primary caregivers for
older adults. There has been lack of evidence about social network characteristics and their

38

health effects among older adults caring for adult individuals with disabilities and grandparents
raising grandchildren. As table 2.1 suggests, little is known about siblings, in-laws, friends,
neighbors, and other underrepresented social ties in the caregiving network and their caregiving
experience.
Multidimensional social networks of family caregivers
Many family caregiving studies have failed to capture the multidimensional nature of social
networks of older family caregivers. Studies often solely focus on one social network dimension.
Social network structure and social support networks were used interchangeably with social
networks. Evidence about content and functional dimensions of social networks was rare. Even
in the structural dimension, we knew very little about social network structure of non-primary
family caregivers, for instance, friend or neighbor caregivers. It was also noticed that research
has contained little agreement about size, composition, frequency, strength, reachability, and
reciprocity of social networks, social support networks, and caregiving networks.
Sampling
In their meta-analysis of 176 family caregiving studies, Pinquart and Sörensen (2007) found
that only 13% of studies used probability samples while 87% of studies were based on
convenience samples. Table 2.1 indicates that only four (22%) studies used probability sampling
in this review of a total of 18 published empirical studies on social networks and health of family
caregivers. It was noticed that among reviewed studies, only nine studies drew samples from
American populations, and two of them used probability sampling.

39

Table 2.1
Research Gaps: Population, Sampling, and Measures of Social Networks and Health

Study
Beeson,
2003

Country
U.S.

Population
49 spousal caregivers
and 52 non-caregiving
spouses

Sampling
Convenient sampling

Social Networks
Relational deprivation;
loneliness measured by the
UCLA Loneliness Scale

Health
Depression
measured by CESD

Carpentier
&
Ducharme,
2005

Canada

2 family caregivers of
person with dementia

Non-probability
sampling; case study

Relational structure: size,
density, and homogeneity;
content of interaction:
attitudes, values, and
behaviors; and functional
properties: social support
and conflict.

N/A

Carpentier
&
Ducharme,
2007

Canada

49 older family
caregivers of persons

Convenient sampling;
interviews

Network size through a
name generator; informal
and formal

N/A

Castro et
al., 2007

U.S.

147 family caregivers
for a frail older relative

Probability sampling;
regional behavioral risk
factor survey

Physical activity, nutrition,
tobacco use, preventive
health care behaviors

BMI; self-rated
health

Edwards
et al.,
2007

Australia

998 paid primary family
caregivers for a person
with disability

Convenient sampling;
interview the Families
Caring for a Person with
a Disability Study
(FCPDS)

Levels of face-to-face social
contact outside of the
household

N/A

40

Table 2.1 (cont’d)

Study
Keith et
al., 2009

Country
U.S.

Population
224 primary family
caregivers for community
elderly with dementia

Sampling
Probability
sampling; regional

Social Networks
Family norms that influence on
primary

Kim &
Knight,
2008

U.S.

87 pair Korean American
caregivers and noncaregivers

Convenient
sampling; interview
survey

The quantity and quality of
informal instrumental and
emotional social support

Purposeful
sampling;
ethnography

Size, composition, strength and N/A
direction of social ties, change
over time, and the presence of
non-supportive and supportive
interactions with kin and friends
as well as professionals

148 couples of cancer
patients and primary
caregivers

Convenient
sampling; interview
survey

Social support measured by the
Social Support List of
Interactions

Depression
measured by the
CES-D

358 aging African
American primary
caregivers

Convenient
sampling; interview

Social participation satisfaction
with family functioning,
perceived instrumental support,
and availability of secondary
informal help

Depressive
symptoms
measured by CESD

Newfeld
&
Kushner,
2009

Canada

Nijboer,
2001

Netherland

Rozario et
al., 2008

U.S.

41

Health
N/A

Physical health
measured by selfreported general
health, blood
pressure, and
cortisol levels

Table 2.1 (cont’d)

Study
Valle,
2004

Country
U.S.

Population
89 dementia caregivercare recipient pairs (39
Latino and 50 EuroAmerican)

Sampling
Convenient sampling;
interview

42

Social Networks
Composition and size of
potential and utilized social
network; 6 network helpseeking behaviors.

Health
N/A

It is not surprising that findings of health effects of social networks of family caregiving in
later life vary depending on study characteristics. For instance, gender differences were larger in
probability samples than convenience samples, where they were often biased toward caregivers
at high risk. Unfortunately, non-probability sampling prevented the majority of social network
studies from inferences about health effects of social networks among older family caregivers.
Although small, purposive samples of ethnic groups provided valuable information about
populations who were often invisible and challenging to recruit, small and non-random drawn
samples made them difficult to make inferences about the characteristics of the population.
Future research requires probability samples with established instruments and measures to test
health effects of older family caregiver-specific social networks.
Measures across studies
The examination of social networks and health among older family caregivers has been
hampered by the different measures being used to assess social networks and their health
outcomes across studies. As table 2.1 indicates, nine relevant studies have investigated the
association between social networks and health, but there was no study that used measures of
social networks and health that were identical with any others. The only exception was that five
studies used the 20-item CES-D to measure depressive symptoms of family caregivers among six
studies that examined health effects of social networks in the last two decades. Although CES-D
is not specifically designed for the aging population, its wide recognition in gerontology makes it
worth considering for future research on social networks and health of older family caregivers in
the United States.
In summary, future research can add new empirical evidence to knowledge advancement in
social networks and health of older family caregivers. Specifically, research emphasizing the

43

following four aspects can contribute to our new understanding of social value of family
caregiving in later life: (1) identifying contributions of all older informal caregivers to quality
and sustainable family caregiving; (2) recognizing multidimensional nature of social networks;
(3) using probability sampling in the United States; and (4) contributing to new understanding of
the effects of such social networks on physical and mental health of all informal caregivers.

44

CHAPTER THREE: METHODS

The purpose of this study was to better understand the social value of family caregiving in
later life by examining the effects of social networks and family caregiving in later life on health
of older adults. This chapter describes the operational model of the study as well as methods
employed in its implementation. First, the health effects model of social networks and family
caregiving in later life is discussed. This chapter then specifies the research design, survey
instrument and measures, and data management and analysis strategies. Population, sampling
and human subject protection are addressed.
Operational model
Emerging awareness of inquiry into the health effects of social networks in later life was one
of the most significant scientific advances in the field of gerontology at the turn of the 21st
century. Various models have provided empirical evidence of the complicated relationships
between social networks, health outcomes, and sociodemographic factors of older adults (Cantor,
1991; Sims-Gould & Martin-Matthrews, 2007). Together they have contributed to our
conceptual and methodological understanding of social networks and health among older family
caregivers.
Originating in Moreno’s sociograms (Moreno, 1953), social network analysis models have
investigated the relationships among social entities. By focusing on relational patterns and their
implications for family and community, social network models have provided theoretical
definition and explicit measures of social structural properties (Bott, 1957; Wasserman & Faust,
1995). Since the 1980s, gerontological social work has applied personal, egocentric network
models in family caregiving to trace social resources of older adults through kinship and
community networks (Chatter et al., 1985; Penrod et al., 1995; Tolkacheva et al., 2011).
45

Carpentier and Ducharme (2003) gave particular attention to a three-dimensional social
network model. This model distinguished and integrated caregiver network structure, content,
and function. The authors articulated the explanatory power of the network model as centered on
the associations between the structure of the network and the content of the ties. Social ties
between the actors created information channels through which knowledge, opinions, and norms
flowed, which conceptualized social networks in a way that linked formal and informal support
networks of family caregivers (Carpentier & Ducharme, 2003). This multidimensional model
indicated that social networks of family caregiving involved transformation in which networks
regulated the behavior of their members so that formal and informal supports were
complementary to each other (Carpentier & Ducharme, 2003; 2005; 2007).
Berkman et al. (2000) proposed a main effect model of social networks that integrated
macro-social context of social networks and micro-psychosocial mechanisms. This model was
rooted in community-based studies of health effects of social networks in the 1970s and 1980s
(Cassel, 1976; Cohen & Wills, 1985; Kawachi & Berkman, 2001). Evidence from these
longitudinal studies confirmed that social ties were significant predictors of lower risk of
mortality and mental illness for older adults, independent of age, gender, race, baseline health
status, and health behaviors (Berkman, 1984, 1986; Seeman, Kaplan, Knudsen, Cohen, &
Guralnik, 1987). It further found that ties with close friends and/or relatives assumed greater
importance for older adults than was true for their younger counterparts. This main effect model
employed Burt’s definition of network models so as to embed egocentric networks in the
network structure. Despite its theoretic ambition and statistical infeasibility, this overarching
model confirmed that the health effect of social networks could not be assessed without
consideration of complex relationships among social networks, family caregiving, and

46

demographic factors (Berkman et al., 2000). The authors made efforts to examine the influences
that structural networks and functional networks had on physical and mental health of
individuals at the behavioral level.
The multidimensional social networks model and the main effect model have contributed to
our conceptual and methodological understanding of social networks and health among older
family caregivers. In the current study, several important aspects from each model were
integrated into an operational model, as shown in figure 3.1.
The core of this operational model is that older adults do not perform informal caregiving in
isolation, but they are embedded in a network of social relations (Granovetter, 1985). The model
conceptualizes social networks as older adult-centered three-dimensional social relationships that
affect their physical and mental health. Much effort has been made in this model to investigate
how the individual older adult (ego) exchanges services and information with his or her network
members (alters) and how the ego benefits from resources in the network as well as negotiates
and challenges social norms and values of the social network (Granovetter, 1973). The model
considers age, gender, race, and SES as social-structural conditions of social network influences.
Methodologically, this operational model simplifies the main effect model in the way that
multidimensional social networks and their health effects are statistically measurable. This model
allows the study to determine patterns of network structural characteristics among older family
caregivers and non-caregiving older adults, as well as other network factors that predict physical
and mental health outcomes.

47

Figure 3.1
Operational Model of Health Effects of Social Networks and Family Caregiving in Later Life

Social-Structural
Conditions
Age
Gender
Race
SES
Social Network Structure
• Component
• Size
• Frequency
• Strength of ties
• Reachability
• Reciprocity
• Function
Social Network Function

Health

• Marital status
• Co-residence
• Religious/spiritual participation
• Organization participation
• Group membership
• Volunteering

• Self-rated health
• Physical health
• Level of physical function
• Number of chronic diseases
• Depressive symptoms

Content
• Behavior
• Information
• Value and Expectations
Family Caregiving
in Later Life

Adapted from Carpentier and Ducharme (2003) and Berkman et al. (2000).
48

Research design
In this population-based survey study, a cross-sectional, explanatory survey study design was
used to better understand social values of family caregiving in late life by examining the effects
of social networks and family caregiving in later life on the health of older adults, aged 64 and
over. The data collection method for this study was a paper-pen mail survey. It was administered
in three Midwest communities from July to September 2010. The decision to use the selfadministered mail survey as the data collection mode was based on coverage consideration.
Research has observed that mail surveys have higher response rates than surveys using a webbased mode, postcard mode, or a web-postcard mixed mode (Kaplowitz, Hadlock, & Levine,
2004). With the increase of households with unlisted telephone numbers and cell-phone-only, a
mail survey became a better option to cover more older adults, especially when computer-based,
large-scale mailing lists were available (Dillman, Smyth, & Christian, 2009). Although research
has studied web-based family caregiver support programs (Glueckauf, Ketterson, Loomis, &
Dages, 2004; Marziali & Donahue, 2006), response to an Internet survey would still be limited
by relatively low rates of access to a computer, Internet service, and high-speed services among
older adults (Czaja & Rubert, 2002; Dillman et al., 2009).
A mailing list of 6,514 non-institutionalized, community-dwelling older adults aged 64 and
above in the three research sites was provided by USADATA, a computer-based mailing list
provider. USADATA used cut-off ages with two-year intervals. Between two closest cut-off
ages, 64 and 66, I choose 64 as the cut-off age for this sample. Using stratified random sampling
strategies, I randomly selected 1,000 potential subjects from a total of 6,514 older adults from
urban, suburban, and rural areas. Of the 1,000 individuals contacted, 366 completed and returned
the questionnaire, which yielded a 36.6% response rate. Two participant selection criteria were

49

community dwelling and age of 64 and above. Of 366 respondents, 358 met inclusion criteria
and were considered as eligible participants. Six respondents were excluded due to being under
64 years old. Two institutionalized respondents were also excluded from the study. The
questionnaire consisted of variables of interest in social networks, health status, health care,
family caregiving experience, and demographics. The questionnaire is presented in appendix D.
I analyzed the study population using descriptive and multivariate analyses to examine health
effects of social networks among older family caregivers. Variables measuring age, gender, race
and ethnicity, and SES of older adults were included. Group comparisons were implemented to
identify whether caregivers and non-caregivers had different patterns in social networks and
health.
Research sites
This study was conducted in Michigan. Three communities were selected as the research
sites for the purpose of maximum representation in geographic regions, residence types,
population aging rates, and racial and ethnic diversity. According to the U.S. Census definition,
Site 1 Ypsilanti Township (YT) was categorized as a metropolitan area, Site 2 Grand Rapids
(GR) suburban, and Site 3 Powers Village (PV) rural. They were located in Southeast,
Southwest, and Upper Peninsula regions of Michigan, respectively. The communities’ aging
rates ranged from 6.1% to 27.5%, and rates of non-white population ranged from 2.0% to 34.7%.
The demographic comparison of three research sites is presented in Appendix E.
Stratified random sampling
Based on the mailing list of 6,514 community-dwelling older adults aged 64 and above, I
used Excel 2007 to generate the random numbers for each research site and then sorted the
numbers from the smallest to the largest. In order to ensure the most representative data available

50

that included diversity in race, ethnicity, and residence type, I equally selected the first 15% of
the random numbers from Site 1 and Site 2, then oversampled Site 3 by 35% of the random
numbers, which created 1,000 potential subjects: 553 in Site 1, 407 in Site 2, and 40 in Site 3.
The stratified random sampling process applied in this study resulted in a response rate that
varied between 33.1% and 45%, with Site 1 the lowest and Site 3 the highest.
Participants
Community-dwelling older adults aged 64 and above completed and returned 358 mail
survey questionnaires. The mean age of the sample was 73.71, ranged from 64 to 95; 95.3% aged
65 and above, and 25.5% aged 80 and above. The sample was 51.4% female, 85.2% nonHispanic white, and 51.7% urban older adults. Other demographic factors—marital status,
education, employment status, and income—are reflected in table 3.1.
Four self-identified groups emerged from a total sample of 358 participants: 82 family
caregivers without receiving care (22.9%); 32 caregivers who also received care (8.9%); 21 care
receivers who did not provide care (5.9%); and 210 older adults who neither gave nor received
care (58.7%). There were 13 participants (3.6%) did not answer the self-identified family
caregiving status question.
Data collection procedure
The survey was 12 pages long and used a 14-point font size. It included brief instructions for
completion and two open-ended questions. The pretest estimated that on average it would take
about 15 minutes to complete the survey. A multiple contact strategy resulted in a total of 366
returned questionnaires, with an estimated 33% increase following the postcard contact.

51

Table 3.1
Demographic Characteristics of the Sample (n = 358)

Demographics

(n)

(%)

16
237
86

4.7
69.7
25.5

169
179

48.6
51.4

296
34
2
6
9

85.2
9.9
.6
1.7
2.6

16
181
70
80

4.6
52.2
20.3
22.9

14
72
111
150

4.0
20.9
31.9
43.2

267
14
72
12

73.0
3.9
19.8
3.3

Age
<65
≥ 65–79
≥ 80
Gender
Male
Female
Race/Ethnicity
White/Non-Hispanic
Black
Native American
Multiracial
Hispanic
Marital Status
Single
Married/Live with Partner
Separated/Divorced
Widowed
Education
≤ High School
High School
Some College
≥College
Employment Status
Retired, not Work or Schooling
Retired, Working or Schooling
Working or Schooling
Homemaking
52

Table 3.1 (cont’d)

Demographics

(n)

(%)

78
161
68

25.4
52.4
22.2

185
155
18

51.7
43.2
5.1

Income
<$25,000
$25,000–$74,000
≥$75,000
Residence
Urban
Suburban
Rural
Recruitment
Based on social exchange theory, I used seven recruitment strategies to influence response
rates identified by Edwards et al. (2002) and Dillman et al. (2009): personalized questionnaire,
personalized consent letter, multiple contact, first class postage, providing stamped return
envelope, a small token, and providing incentive without condition. First, I sent a personalized
invitation letter to 1,000 selected subjects to introduce the research study and to pre-notice the
upcoming questionnaire. Three days after the invitation letter I sent the survey package to 1,000
subjects. The survey package included the informed consent, a small token of a one-dollar bill,
the questionnaire, and a pre-stamped return envelope. A week later, I sent a thank-you postcard
to all subjects. The thank-you postcard was served as a thank-you note, a reminder, and notice of
availability of replacement questionnaires. Of 366 returned questionnaires, 103 were received
after the postcard was sent. I received 19 phone messages and one email to request replacements.
I also received five phone calls and four emails to clarify eligibility criteria and sampling
strategies. Another 24 returned questionnaires resulted from these phone calls and emails. Figure
3.2 shows the final sample of family caregivers in the study.
53

Figure 3.2
The Final Sample of Family Caregivers (n = 358)
Mailing List
n = 6,514
Randomly stratified
Potential Participant Pool
n = 1,000

Returned Survey
n = 366

Exclusion
Under Age 64
n=6
NonCaregiver

n = 358

Non-community Dwelling

n = 222

Final Sample

n=2

Perceived Caregiver Status
n = 114
Final Sample of Family Caregivers
Caregivers
Actual Caregiver Status

n = 136

n = 136

For Grandchildren
n = 15

For
Older
Adults
n = 87

54

For Adult Child
with Disability
n = 34

Protection of human subject
An informed consent letter was sent to the randomly selected 1,000 survey subjects. The
consent letter used Institutional Review Board (IRB) language, including an explanation of the
purpose of the research, procedures used, the time required, the extent of confidentiality, any
potential risks and benefits, and the condition that the study would involve entirely voluntary
participation. The Michigan State University Institutional Review Board (MSU IRB) for the
Protection of Human Subjects had reviewed and approved the proposal, including the study
design and procedure, the survey questionnaire, and the informed consent letter. All documents,
from the consent letter to the questionnaire, used a 14-point font size. Using the Word
Readability Test, the Flesch-Kincaid grade levels of the consent forms and the survey instrument
were rated at 5.5.
Additionally, USADATA as a mailing list provider identified older adults aged 64 and above
as one of its protected vulnerable groups. Before obtaining the mailing list, I processed its
protection procedure to ensure that no older adult’s interest was violated.
Survey instrument and measures
The survey instrument consisted of demographic background, parameters of personal social
networks, and parameters of physical and mental health. This survey instrument had been
developed and pretested. It contained a total of 58 quantitative measures, seeking to obtain
quantitative data regarding health effects of social networks among older family caregivers. All
variables were self-reported. Multiple indicators were used to measure independent variables of
social networks and family caregiving in later life as well as dependent variables of health. Table
3.2 indicates internal consistency of standardized instruments used in the study (Emlet, 2006;
Bakas & Champion, 1999).

55

Family caregiving measures
Family caregiving measures identified caregiving status, the relationship between caregiver
and care recipient, the primary reason for caregiving, duration of care, care intensity, care
distance, and gain of caregiving. Family caregiving measures are presented in table 3.3.
A single-item question was used to investigate perception of family care status: “Do you
consider yourself a caregiver for or a care receiver from a loved one (a family member, relative,
friend, neighbor or other)?” The response options were caregiver, care receiver, both, and
neither. A dichotomous item was created to indicate that the respondent identified himself or
herself as a family caregiver (1) or a non-family caregiver (0). To cross examine family
caregiving status, respondents were also asked to answer questions as to whether he or she has
regularly helped a loved one with 13 caregiving activities over the past 30 days. A summed score
of caregiving ranged from 0 to 13. A caregiving score of 1 or greater indicated a caregiving
status in this study.
A comparison was drawn between perceived family caregiving status and actual family
caregiving status to explore perception of family caregiving among older adults. A dummy
variable was created to measure family caregiving status. Those who helped one or more
caregiving activities were coded 1, and all others were coded 0. In this study, this variable was
the base of group comparison between family caregivers and non-caregiving older adults.
Caregivers were asked the number of care receivers to whom he or she has provided family
care. They were also asked the reason for the care and the duration of care. The reasons for care
included absence or unavailability of birth parents of grandchild, chronic illness, and disability.
Since caregivers might care for more than one family member, relative, friend, or neighbor, three
care types were not applicable for percentage calculation.

56

Table 3.2
Summary of Variables by Name, Instrument, and Psychometric Properties

Variable

Description

Type

Family
Caregiving

Caregiving status;
13-item caregiving activities of
daily living 10-item Bakas
Caregiving Outcomes Scale
(Bakas & Champion, 1999)

IV

18-item Lubben Social
Networks Scale (Lubben &
Gironda, 2003), 9-item Social
Networks Index (Berkman &
Syme, 1979), and 10-item Life
Events Survey (Holmes &
Rahe, 1967).

IV

Health
Behavior

4-item

Info
Channel

Instrument

Range

Scale

N

α

Perception of caregiving
13 basic activities of daily
living
Changes in caregiving
experience

Categorical
0–1
.84
0–13 Ordinal
.77
-30–30 Ordinal

Naturally occurring personal
social networks related to kin,
friends, neighbors, church, and
community organizations and
other informal social groups

0–90

Ordinal

.82

354

0–12

Ordinal

.66

310

IV

Nutrition intake, absence of
alcohol misuse, absence of
smoking, and physical exercise

0–4

Ordinal

2-item human and health
service information channel

IV

Information from social
networks and professionals

0–1

Ordinal

Physical
Function

IADL (Lawton & Brody, 1969).

DV

A score of 1 and lower
indicates high daily function

0–7

.80

334

Depression

15-item Geriatric Depression
Scale (Yesavage et al., 1983).

DV

A score of 5 or above suggests
mild to severe depression

0–1

.84

352

Social
Networks

57

346
143
100

200

.33

303

58

Table 3.3
Family Caregiving Measures

Measures

n

%

Perceived Caregiving Status (n = 345)
Yes
No

114
231

33.0
67.0

136
222

38.0
62.0

82
29
14

64.6
22.7
11.7

61
61
5

50.0
50.0
3.9

15
95
34

N/A
N/A
N/A

59
24
15
15

52.2
21.2
13.3
13.3

Caregiving Status (n = 358)
Yes
No
Number of People Care for (n = 127)
1
2
3–5
Solo Caregiver (n = 122)
Yes
No
≥ 1 hour
Reason for Care (n = 136)
Grandparents raising grandchildren
Elder care
Adult child with disability
Weekly Care Hour (n = 113)
<9
9–20
21–40
>40

58

Table 3.3 (cont’d)

Measures

n

%

11
8
30
23
23
26
2

8.9
6.5
24.4
18.7
18.7
21.1
1.6

30
17
53

30.0
17.0
52.0

Length of Care (n = 123)
< 6 month
6–11 month
1 or 2 years
3 or 4 years
5–9 years
≥ 10 years
Other
Bakas Caregiving Outcomes Scale (n = 100)
<0
0
>0

Care distance between the care receiver(s) and the caregiver was measured by co-residence
(0), nearby (<60 minutes), and long-distance (≥1 hour). Care intensity was defined by care hours
last week. Care duration was measured by the shortest duration (<6 months) to the longest (≥10
years).
Bakas Caregiving Outcomes Scale-10. In this study, the 10-item Bakas Caregiving Outcomes
Scale (BCOS-10) assessed personal changes, changes in relationship, and higher level gains of
family caregivers. The instrument used a 7-point Likert scale from -3 (changed for the worst) to
3 (changed for the best). A summed score ranged from -30 to 30. A score of 0 indicated no
change since serving in the family caregiving role. A score <0 suggested negative changes in
caregiving experience. A score >0 indicated gains in caregiving. Higher positive scores denoted
higher gains in caregiving.

59

Social networks measures
Social networks of community-dwelling older adults aged 64 and above were measured by
three dimensions within social networks: structure, function, and content.
Structural dimension. In this study the Lubben Social Networks Scale-18 (LSNS-18) and
Life Events Survey-Short Form (SF-LES) were used to measure the structural dimensions of
social networks. They identified participants’ network size, composition, strength of ties,
reciprocity, and change in networks over time. The LSNS-18 contained three 6-item subscales.
They extensively measured the size, frequency, intimacy, and reciprocity of personal social
networks related to kinship, friendship, and neighborhood (Luden & Gironda, 2003). The scale
utilized a 6-point Likert-type scale, ranging from 0 to 5 to measure. It provided a summed score
ranging from 0–90. Individuals with LSNS scores below 36 were considered socially isolated
while LSNS scores above 54 were socially highly connected (Lubben, 1988; Emlet, 2006).
Meanwhile, this cross-sectional study used a revised SF-LES in an effort to examine changes in
social networks over the last two years. The respondents were asked to identify the number of
major life events which occurred during the previous two years. Ten life events were categorized
as marriage, family and friends, work, finance and residence, health, and learning.
Functional dimension. Based on research findings regarding the relationships between health,
religion/spirituality-based networks, and community participation, a revised Berkman-Syme
Social Networks Index (BSNI) was used to measure social networks related to marital status,
quantity and quality of close relationships with relatives and friends, religious/spiritual
community membership, and community organization and group membership. Based on the
marital status question, a dichotomous item of marital status was created to assess whether the
respondent was 1 (currently married) or 0 (currently not married). One dichotomous item was

60

used to assess whether the respondent had attended religious services in the last four weeks, with
1 (yes) and 0 (no). Three questions were asked to assess intimate relationships with relatives and
friends: (1) How many friends do you feel at ease with that you can talk about private matters?
(2) How many relatives do you feel at ease with that you can talk about private matters? and (3)
How many relatives and friends do you see or hear from at least once a month? The responses to
these questions were rated on a 6-point scale: (1) 0, (2) 1, (3) 2, (4) 3–4, (5) 5–9, and (6) 9 and
more. Answers of 0–2 then were re-coded as 0, 3–4 re-coded as 1, and other re-coded as 2. Three
dichotomous items were used to identify social and civic engagement of the respondents in
organizational participation, group membership, and volunteering other than religious
membership. If the respondent answered yes to any of questions about organization or group
membership or volunteering, the respondent was scored 1. Nine items were summed to create
BSNI scores, which ranged from 0 (least connected) to 12 (most connected). Marital status was
weighted as 0 (no) or 2 (yes).
Content dimension. Information channel and health behavior measures were used to measure
the content dimension of the respondents’ social networks. The respondents who identified
themselves as family caregivers were asked what information channels they used to obtain
information about human and social services for family caregivers. Options listed for
information channels were social network members, human and health professionals, and media.
Health behavior measures consisted of exercise, nutrition, absence of alcohol misuse, and
absence of cigarette smoking. One dichotomous question was asked to measure whether the
respondent exercised regularly. The respondent was then asked the type of exercise and the
frequency of exercise he or she did each week. Four-item questions were used to assess nutrition
intake. A single-item question asked the respondents, “How many full meals do you eat daily?”

61

The response options were 1–3 and other. The respondents also were asked three dichotomous
questions about whether or not they ate or drank three types of dairy products and food within
the measured unit and period.
Three questions were asked to identify levels of alcohol drinking within the last month.
Alcohol drinking was defined as consuming alcoholic beverages such as beer, wine, or liquor. A
dichotomous item was used to identify whether the subject ever drank alcoholic beverages. The
responses were 0 (no) and 1 (yes). Two follow-up questions were asked of those who answered
“Yes.” The first question asked the frequency of drinking: “In the last four weeks, on average,
how many days per week have you had any alcohol to drink?” The second question asked the
quantity of drinking: “In the last four weeks, on the days you drank, about how many drinks did
you have?” As table 3.4 shows, absence or presence of alcohol misuse was determined based on
the level of drinking, with response options ranging from 0 to 4 (0=None, 4=Heavy, and 5=Very
heavy). The drinking level 4 was defined as presence of alcohol misuse, and all lower levels
were indicated as absence of alcohol misuse.
Absence of smoking was assessed by two dichotomous questions: “Have you ever smoked
cigarettes?” and “If yes, do you smoke cigarettes now?” Presence of smoking was identified
when both answers were 1 (yes). Absence of smoking then was defined as either never smoked
or currently not smoking.
Health measures
Health of older family caregivers was measured by self-rated health, self-reported level of
physical function, total number of 13 chronic diseases, and perceived depressive symptoms. The
first health measure was self-rated health that was scored on a 5-point Likert scale, based on the
question: “In general, would you say your health is excellent, good, neutral, poor, or very poor?”

62

Table 3.4
Levels of Alcohol Use Based on the Number of Standard Drinks Consumed

Number of Standard Drinks a Session
Frequency

2

3

4

5

None Very light Very light

Light

Moderate

Heavy

Heavy

2–4 times None Very light Very light
a month

Light

Moderate

Heavy

Heavy

Heavy

Very heavy

≤1 a
month

0

<1

1

2–3 times None
a week

Light

Light

Moderate Moderate

≥4 a week None

Light

Light

Moderate

Heavy

Very heavy

Very heavy

Daily

Light

Moderate

Moderate

Heavy

Very heavy

Very heavy

None

Source: Adapted from Moore et al., 1999; Berks & McCormick, 2008.
Revised Instrumental Activities of Daily Living (IADLs-R). IADLs-R was used to assess
physical function of participants. Seven dichotomous items identified were difficulty of driving,
grocery shopping, hot meal preparation, making phone calls, taking medicine, conducting
housework, and money management (Lawton & Brody, 1969). Reversed items were summed to
produce the IADLs-R scale (alpha=.80). The IADLs-R scale indicated a high daily function and
absence of dependency when it was 6 and greater.
Absence of chronic conditions. Thirteen dichotomous items were used to assess whether or
not a doctor had ever told the respondent that he/she had the following conditions: high blood
pressure, diabetes, cancer, chronic lung disease, congestive heart failure, heart attack, stroke,
arthritis or rheumatism, hip problem, falls, persistent back pain, persistent headaches, and severe
fatigue or exhaustion. The reversed items were summed to produce a scale of absence of chronic

63

conditions (alpha=.65). When the IADLs-R scale and the absence of chronic conditions were
summed, the physical health scale measured physical health of the respondents.
Geriatric Depression Scale-15. The Geriatric Depression Scale (GDS) was designed
specifically for geriatrics. It measures mood, cognitive complaints, and social behavior
(Yesavage et al., 1983). The GDS-15 allowed the study to examine a key concern of mental
health in family caregiver studies. The GDS-15 in this study contained 15 dichotomous items,
each scored 0–15. Responses of 0 (no) suggested absence of depressive symptomatology, while
1 (yes) suggested presence of depressive symptomatology. There were five reversed items in
which 0 (yes) indicated absence of depressive symptomatology and 1 (no) indicated a presence
of depressive symptomatology. The GDS-15 scores were reflected by the number of 1 (yes)
responses. The cut-off score for clinical purpose was 5. A score greater than 5 points was
suggestive of depressive symptoms. Scores higher than 5 points warranted a follow-up interview.
A score greater than 10 almost always indicated depression.
Demographic measures
Large studies have supported the argument that health effects of social networks vary by age,
gender, race, and SES factors. In this study, the demographic variables of age, gender, race, and
SES were used as control variables. Covariance variables are presented in table 3.1.
The participants were asked their age: “What is your date of birth?” Three age groups were
created by ranging respondents’ age: under 65, 65 to 79, and 80 and above. One single-item
question was used to identify gender of the respondent: “What is your gender?” The responses
were 1 (male) and 2 (female). A dichotomous item was used to identify whether or not the
respondent was Hispanic: “Are you Hispanic or Latino?” The responses were 0 (no) and 1 (yes).
The respondent also was asked, “What is your race?” The responses were coded as 0

64

(Caucasian/White, non-Hispanic), 1 (Black/African American), 2 (American Indian and Alaska
Native), 3 (Asian), 4 (Native Hawaiian and other Pacific Islander), and 5 (Multiracial). In this
study, a single-item question was used to assess the level of socioeconomic status: “What is the
highest education you have completed?” The responses were coded from 0 (none) to 7 (post
master’s).
Data management
Data management strategies were used to ensure voluntary participation, anonymity, and
confidentiality. A one dollar bill was included in the survey package, independent of any
commitment on the part of the receiver to participate in the study. This was an effort to not
influence participants with the small token. All multiple contacts were sent to 1,000 potential
participants. Although this action increased the cost of the survey, it ensured anonymity because
no identification of the participants was associated with the returned questionnaires in any way.
The completed survey questionnaires were mailed to me in a pre-addressed, stamped envelope.
As questionnaires were received by me, each was given an identification number and coded
according to the research site. The research site of all returned questionnaires was identified from
the zip code on the postal stamp outside of the return envelope. The sending date was observed
and recorded to monitor the return rates. Eleven participants put his or her return address on the
outside envelope. All return envelopes were shredded after the research site information was
recorded. The questionnaire data were entered and stored by me in a secure computer, for which
a password was required. The returned questionnaires were stored in a locked cabinet. Access to
the computerized data and cabinet-stored data was limited to me, the dissertation committee
chair, and potentially the university institutional review board for the protection of human
subjects.

65

Questionnaires were reviewed for quantitative and qualitative responses. Qualitative
responses included additional notes attached to the questionnaires. All quantitative and
qualitative data were entered into the PASW Statistics 18. Quantitative data cleaning was
performed for eligibility and missing data. Frequencies were examined to screen for unusual
values that would indicate incorrect data entry.
Data analysis
The unit of analysis was “individual older adult.” In order to test the hypothesized
operational model as presented in figure 3.1, I conducted descriptive statistical analyses, simple
and multiple regression analyses, and content analysis to answer three research questions.
Qualitative data were transformed from a PASW-18 dataset to a Word document. I reviewed and
evaluated the data for content analysis. Emergent themes were identified during the data
collection period.
Quantitative data were analyzed for caregiver and non-caregiver demographic characteristics,
network characteristics, and health. For question one—“What are the demographics, social
network characteristics, and health status of older family caregivers, compared to other older
adults who do not provide family caregiving?”—I split data by caregiving status comparison
between older family caregivers and non-caregiving older adults. Frequencies and measures of
central tendency were obtained for demographic and social network characteristics and health
status of older family caregivers and non-caregiving older adults. One-way analyses of variance
(ANOVA) were performed to test whether there were significant differences in structural social
networks and content social networks between caregivers and non-caregivers. F values were
used for six social network structural variables and the total LSNS scores independently. Chi-

66

squared tests were used to assess the relationships between functional social networks and family
caregiving in later life.
A mixed methods analysis was used to answer question two: “How do older family
caregivers perceive family caregiving in later life?” A descriptive statistical analysis of BCOS in
family caregiving was conducted to observe quantitative data on changes in family caregivers. A
content analysis was conducted to review and code qualitative data on how family caregivers
interpreted their caregiving experience. Positive and negative changes were triangulated between
BCOS scores and emerging themes to gain both statistically significant and in-depth
understanding of perception of family caregiving using older family caregivers’ own words.
Question three—“Are there significant mean differences in health for different levels of
social networks and for different family caregiving status after controlling for demographic
factors?”—was addressed through bivariate correlation analyses, Multivariate Analysis of
Variance (MANOVA), and Multivariate Analyses of Variance and Covariance (MANCOVA).
First, bivariate correlation analyses were used to examine the relationship between social
networks, family caregiving, health, and demographics of participants. Caregiving statuses were
entered as dummy variables as the baseline category. Second, MANOVA and MANCOVA were
performed to evaluate the effect model of social networks on health of family caregivers in later
life if the assumption of homogeneity of regression slopes was assumed. To test the null
hypothesis in MANCOVA that the population mean vectors are equal, the research question
three was further broken down into three questions, as table 3.5 presented. Finally, a series of
univariate ANOVA was performed to test whether there were significant mean differences for
each health measure due to social networks and family caregiving after removing the effects of
age, gender, race, and SES.

67

Table 3.5
Research questions and data analysis strategies

Research Question

Data Analysis Strategy

1. What are social network characteristics of older family

Descriptive statistics

caregivers, compared to other older adults who do not

ANOVA

provide family caregiving?

F value

2. How do older family caregivers perceive family
caregiving in later life?

Descriptive statistics
Content analysis
Ecomap

3. Are there significant mean differences in health for

Bivariate correlation analysis

different levels of social networks and for different

MANOVA

family caregiving status after controlling for

MANCOVA

demographic factors?
3.1. Are there significant mean differences in the
combined health, as measured by self-rated health,
physical function, absence of chronic diseases, and
mental health, for the different levels of social networks
after removing the effect of age, gender, race, and SES?
3.2. Are there significant mean differences in the
combined health, as measured by self-rated health,
physical function, absence of chronic diseases, and
mental health, for family caregivers and non-caregivers
after removing the effect of age, gender, race, and SES?
68

Table 3.5 (cont’d)

Research Question

Data Analysis Strategy

3.3 Is there a significant difference for social networks
and family caregiving on the combined health, as
measured by self-rated health, physical function,
absence of chronic diseases, and mental health, for
family caregivers and non-caregivers after removing
the effect of age, gender, race, and SES?

69

CHAPTER FOUR: RESULTS

Results of the study are illustrated in three parts, organized by three research questions.
Research question one is addressed using results from descriptive analyses, ecomap, and oneway ANOVA analyses. Question two is answered through results from triangulation between
quantitative and qualitative data of changes in caregiving experiences. Question three is
answered through results from effect models of social networks and family caregiving on health
of older adults before and after controlling for age, gender, race, and SES.
Descriptive statistics and preliminary analyses
Tables 4.1 through 4.6 show descriptive characteristics of 136 older family caregivers and of
222 non-caregiving older adults, answering research question one of this study: “What are the
demographics, social network characteristics, and health status of older family caregivers,
compared to other older adults who do not provide family caregiving?”
Demographics
Table 4.1 reflects four age structural characteristics. The sample had an average age of 73.71.
Caregivers were slightly older (72.1%) than non-caregivers (68.6%), and 24.8% of caregivers
were oldest old, which was 37.2% of oldest old in the sample. The majority of older adults in
both groups were between 65 and 79 years of age. Among all caregivers, male caregivers shared
nearly half of family caregiving (45.7%) with female caregivers (54.3%). In the total sample,
48.6% were male, and 51.4% were female, with an estimated gender ratio of 96. Gender ratio
among caregivers dropped to 83, while it increased to 102 among non-caregivers. Table 4.1 also
shows that older adults across racial and ethnic groups and SES experienced family caregiving
with distinctive characteristics.

70

Table 4.1
Demographic Characteristics of the Sample by Caregiving Status (n = 358)

Variable

Caregiver

Non-caregiver

Total

n = 129

n = 210

Mean = 73.71

<65

4
(3.1)

12
(5.7)

16

65–79

93
(72.1)

144
(68.6)

237

≥80

32
(24.8)

54
(25.7)

86

n = 129

n = 219

Male

59
(45.7)

110
(50.2)

169

Female

70
(54.3)

109
(49.8)

179

n = 130

n = 217

White

108
(83.1)

188
(90.9)

296

Non-white

22
(16.9)

29
(9.1)

51

n = 130

n = 217

≤High school

40
(30.8)

46
(21.2)

86

Some college

36
(27.7)

75
(34.6)

111

≥Bachelor’s degree

54
(41.5)

96
(44.2)

150

Age

Gender

Race

Education

71

Table 4.1 (cont’d)
Variable

Caregiver

Non-caregiver

n = 120

n = 189

<$25,000

40
(33.3)

38
(20.1)

78

$25,000–49,999

32
(26.7)

62
(32.8)

94

$50,000–74,999

25
(20.8)

43
(22.8)

68

≥$75,000

23
(19.2)

46
(24.3)

69

n = 130

n = 217

Married

77
(59.2)

104
(47.9)

181

Not married

53
(40.8)

113
(52.1)

166

n = 136

n = 222

Urban

68
(50.0)

117
(52.7)

185

Suburban

57
(41.9)

98
(44.1)

155

Rural

11
(8.1)

7
(3.1)

18

Co-residency

n = 136

n = 205

Live alone

41
(30.2)

81
(39.5)

122

Not live alone

95
(69.8)

124
(60.5)

212

Income

Marital status

Residence

Note. Statistics presented in the table are numbers and percentages.
72

Total

In contrast to non-caregivers, caregivers were more likely non-white (16.9 % vs. 9.1%) and
low household income (33.3% vs. 20.1%). Caregivers were less educated than non-caregivers,
with a higher rate of high school education or less and a lower rate of higher education.
Caregivers were lower in the levels of the household income, with one-third at the level lower
than $25,000, in comparison to one-fifth of non-caregivers at the same level.
Descriptive analysis found that caregivers were more likely married and not living alone than
non-caregivers, with 52.1% of not married non-caregivers and 40.8% of not married caregivers,
and 39.5% of non-caregivers living alone and 30.2% of caregivers living alone. Among married
respondents, 12.7% of non-caregivers co-resided with people other than their spouse, while this
number was 9.1% among married caregivers.
Social network characteristics
Table 4.2 presents descriptive characteristics of three-dimensional social networks of
caregivers and non-caregivers. Overall, no statistically significant difference was observed in
social networks between caregivers and non-caregivers.
Table 4.2
Descriptive Statistics of Social Network Variables by Caregiving Status

Dimension

Caregiver

Non-caregiver

Total

n = 135

n = 221

n = 343

LSNS

46.90

47.44

47.24

0–90

LSNS-Relative subscale

20.61

20.13

20.31

0–30

LSNS-Friend subscale

16.84

16.99

16.93

0–30

LSNS-Neighbor subscale

11.48

11.16

11.28

0–30

Structural

73

Range

Table 4.2 (cont’d)

Dimension

Caregiver

Non-caregiver

Total

Range

Size

9.61
(3.42)

10.00
(3.56)

9.85

0–15

Frequency

9.10
(3.13)

9.33
(3.38)

9.24

0–15

Strength

6.44
(3.29)

6.02
(3.22)

6.18

0–15

Help

7.93
(3.43)

8.09
(3.59)

8.03

0–15

Reachability

7.72
(3.72)

7.74
(3.44)

7.73

0–15

Reciprocity

5.49
(3.02)

5.41
(2.92)

5.44

0–15

Health behavior

2.81
(.98)

2.99
(.91)

2.93
(.94)

0–4

Service use

.73
(.44)

.69
(.46)

.71
(.47)

0–1

Information

n = 32

n=4

Network numbers

.55
(.51)

.50
(.58)

0–1

Professionals

.69
(.47)

.50
(.58)

0–1

BSNI

6.28
(2.81)

6.13
(2.94)

Co-residence

.59
(.49)

.48
(.50)

Content

Functional

74

6.17
(2.89)
.
52
(.50)

0–12

0–1

Table 4.2 (cont’d)

Dimension

Caregiver

Non-caregiver

Total

Range

Religious/spiritual participation

.61
(.49)

.61
(.49)

.61
(.50)

0–1

Organization participation

.38
(.49)

.45
(.50)

.42
(.49)

0–1

Group membership

.52
(.50)

.58
(.50)

.56
(.50)

0–1

Volunteering

.50
(.50)

.60
(.49)

.56
(.50)

0–1

Note. Statistics presented in the table are means and standard deviations.
Table 4.2 reports descriptive statistics of other structural social network variables. Caregivers
and non-caregivers had LSNS frequency scores of 9.10 and 9.33, respectively. They often saw or
heard from relatives, friends, and neighbors with whom they had the most contact. Caregivers
had higher scores of LSNS strength (6.44) than non-caregivers (6.02). This suggests that
caregivers had about six or more confidants with whom they could talk about private matters.
LSNS scores for the number of relatives, friends, and neighbors they could call for help were
7.93 for caregivers and 8.09 for non-caregivers. Both groups perceived there were 6–9 relatives,
friends, and neighbors whom they could call upon for help. Data revealed that caregivers
perceived their network members and confidants had a reachability of “sometimes” or “often”
when caregivers had an important decision to make. Reciprocity scores of 5.49 among caregivers
and 5.41 among non-caregivers showed that both groups felt their network members “seldom” or
“sometimes” talked to them about their important decisions. Generally, descriptive statistics
indicated that network size of caregivers was 9–12, but care network size was smaller, at the
level of 6 or more. Frequency of network contact was “often,” but frequencies of reachability
75

and reciprocity were only “sometimes.” In comparison to non-caregivers, caregivers had slightly
lower scores in network size, frequency, help, and reachability, and slightly higher scores in
network strength and reciprocity.
Table 4.2 also reports that caregivers had a moderately higher BSNI score (6.28) than noncaregivers (6.13), indicating a relatively higher degree of overall social function. Table 4.2 also
indicates that higher percentages of non-caregivers had engaged in organizations, groups, and
volunteering than caregivers, while, on average, both had exactly the same degree of religious
participation. In line with table 4.1, table 4.2 further suggests that caregivers (mean=.68) were
more likely to not live alone than non-caregivers (mean=.61).
Structural social networks. Figures 4.1 and 4.2 show the overall differences of structural
social networks between caregivers and non-caregivers. They indicate that caregivers have a
slightly lower LSNS score (46.90) than non-caregivers (47.44). An estimated 32% of caregivers
and non-caregivers were found to have LSNS scores greater than 54; 23.1% of caregivers had an
average LSNS score lower than 36, 2.7% higher than non-caregivers. Data indicate caregivers
generally presented a similar level of social connectedness as non-caregivers while a higher risk
of social isolation among family caregivers was observed.
A close look at three LSNS subscales in figure 4.3 reveals that caregivers and non-caregivers
had a similar sized group of relatives, friends, and neighbors in their social networks. Among
three network components, relatives were scored highest, with 20.61 for caregivers and 20.13 for
non-caregivers, and neighbors were scored lowest with 11.48 for caregivers and 11.16 for noncaregivers. While LSNS subscale scores of relatives and neighbors were slightly higher among
caregivers than non-caregivers, LSNS subscale scores of friends were slightly higher among
non-caregivers (16.99) than caregivers (16.84).

76

Figure 4.1
LSNS Frequency among Caregivers ( = 134)
(n

Note. Mean = 46.90, SD=16.00. Dotted lines are LSNS cut
cut-off points.
Figure 4.2
caregivers (n
LSNS Frequency among Non-caregivers ( = 216)

cut-off points.
Note. Mean = 47.77, SD=15.41. Dotted lines are LSNS cut

77

Descriptive data analysis found that caregivers and non caregivers had network size LSNS
non-caregivers
scores of 9.61 and 10.00, respectively. Every caregiver and non caregiver, on average, had 9–12
non-caregiver,
9
relatives, friends, and neighbors in his or her network. This network size had taken into
consideration changes within networks during the prior two years. Table 4.3 shows changes in
ration
social network structure in the last two years between caregivers and non caregivers. It indicates
non-caregivers.
that caregivers generally experienced more radical changes in social networks during the past
networks
two years: 42.2% of caregivers and 32.5% of non caregivers had lost an immediate family
non-caregivers
member; 62.6% of caregivers and 51.2% of non caregivers had lost one or more close friends.
non-caregivers
Meanwhile, higher percentages of caregivers gained new family members (63.7%) and had
new
started school, training, or other learning (12.8%) than non caregivers (45.8% and 10.6%,
non-caregivers
respectively).
Figure 4.3
Average Social Network Structural Characteristics by Caregiving Status

78

Note. All variables recoded on 15-point scale.
Table 4.3
Changes in Social Network Structure by Caregiving Status

Change

Caregiver

Non-caregiver

No

70
89.7

125
89.3

Yes

8
10.3

15
10.7

No

52
57.8

104
67.5

Yes

38
42.2

50
32.5

No

37
37.4

79
48.8

1

39
41.5

63
38.9

2

10
10.6

13
8.0

3–6

8
8.5

7
4.3

No

33
36.3

83
54.2

Yes

58
63.7

70
45.8

Loss of spouse

Loss of an immediate family member

Loss of a close friend

Gain in new family members

79

Table 4.3 (cont’d)

Change

Caregiver

Non-caregiver

No

68
87.2

126
89.4

Yes

10
12.8

15
10.6

No

53
63.9

101
68.2

Yes

30
36.1

47
31.8

No

55
64.7

110
77.5

Yes

30
35.3

32
22.5

No

64
83.1

121
89.6

Yes

13
16.9

14;
10.4

No

70
92.1

127
94.1

Yes

6
7.9

8
5.9

Beginning school, training, or other learning

Change in number of family get-togethers

Change in social activities

Marriage

Marriage separation or divorce

80

Table 4.3 (cont’d)

Change

Caregiver

57
62.0

Yes

124
89.9

14
18.7

Legal issue of a network member
No

55
35.0

61
81.3

Yes

102
65.0

39
46.4

Diagnosis of a severe disease of a loved one
No

19
13.6

45
53.6

Yes

121
86.4

35
38.0

Spouse major illness
No

Non-caregiver

14
10.1

Note. Statistics presented in the table are numbers and percentages.
Content social networks. In this study, health behavior, service use, and information channel
were measured to examine the content of social networks of the sample. Descriptive data
analyses indicated that, except for one caregiver and two non-caregivers, all respondents had
health insurance. A lower percentage of caregivers (27.4%) reported to have four health
behaviors—namely, nutrition intake, physical exercise, absence of alcohol misuse, and absence
of smoking—than non-caregivers (34.4%). Of caregivers who had health condition(s), 26.5% did
not undergo treatment. In comparison, this number was 31.0% in the non-caregiver group. Table
4.4 presents more details about this comparison.

81

Table 4.4
Frequency of Health Behavior and Treatment Use

Variable

Caregiver

Non-caregiver

(M=2.81)

(M=2.99)

4

37
27.4

75
34.4

2–3

86
63.7

132
60.6

0–1

12
8.9

11
5.0

(M=.73)

(M=.69)

No

35
26.5

66
31.0

Yes

97
73.5

147
69.0

Health behavior

Treatment use

Note. Statistics presented in the table are numbers, percentages, and means.
In addition, 20 caregivers (14.8%) used human and health services for family caregivers in
the past year. The other 115 caregivers either did not use family caregiving services (85.2%) or
did not answer the question about family caregiving service utilization (20.7%). Fifteen
caregivers (11.0%) never knew family caregiving services were available. Of caregivers who
used family caregiving services, 39.9% reported that they had obtained information about family
caregiving services through both professionals and social network members; 32.3% of caregivers
reported that they had obtained information exclusively from social network members; and
37.8% had obtained information exclusively from professionals.

82

Table 4.4 reports means of health behavior and service use between caregivers and noncaregivers. Caregivers reported an average of 2.81 health behaviors, in comparison to 2.99
reported by non-caregivers. In contrast, caregivers appeared to have a slightly higher service use
rate (0.73) than non-caregivers (0.69) when treatments for their health conditions were
considered.
Table 4.5
Co-residency by Caregiving Status (n = 358)

Number of Co-residents

Caregiver

Non-caregiver

0

41
(32.0)

80
(39.2)

1

67
(52.3)

107
(52.5)

2

15
(11.7)

14
(6.9)

3–5

4
(3.9)

3
(1.3)

Note. Statistics presented in the table are numbers and percentages.
Functional social networks. A revised Berkman-Syme Social Network Index (BSNI) was
used to investigate functional social networks. Table 4.5 presents numbers of co-residents among
caregivers and non-caregivers. Caregivers were less likely than non-caregivers to live alone and
more likely to live with two and more people than non-caregivers. Figure 4.4 breaks down the
BSNI scores into four levels. It suggests that non-caregivers presented higher percentages at both
the most functional level (31.67%) and the least functional level (6.79%) than caregivers
(30.37% and 5.19%, respectively). It also showed that although caregivers presented a lower risk

83

of being least functioning, 29.63% of them were at the medium low functional level, indicating
marginal risk of social dysfunction.
Figure 4.4
Functional Social Networks by Caregiving Status (n = 356)
40
34.81

35
30
25

29.63

30.37

20
15
10
5

5.19

0
Most Function

Medium High

Medium Low

Caregiver

Least Function

Non-caregiver

Note. Statistics presented in the table are percentage.
Results from social network analyses answer research question one: “What are social
network characteristics of older family caregivers, compared to other older adults who do not
provide family caregiving?” Using ecomap, figure 4.5 summarizes key social network
characteristics of family caregivers in this study. It indicates that a typical family caregiver in the
sample had a range of 9–12 members in his or her social networks, consisting of relatives,
friends, and neighbors, with medium high social functioning. More than a quarter of caregivers
practiced health behaviors, and the majority of them used treatments when they had health
conditions. They obtained treatment information from both network members and professionals,
with a preference for professionals.

84

Figure 4.5
Ecomap of Social Networks of Family Caregivers in Later Life (n = 136)
Relative
Relative

Friend
Relative

Friend

Neighbor

Friend

Friend
Relative

Friend

Caregiver

Co-resident

Neighbor

Relative

Neighbor
Relative

Neighbor
Relative
Friend

Note.

Professional

strong reciprocity reciprocity strong reachability connected

profession

Health status
Table 4.6 reports results from descriptive statistics. It reveals that the majority of family
caregivers reported excellent or good health, high physical function, and good mental health.
85

Almost one in five family caregivers rated their global health as excellent, more than half of
caregivers identified their health as good, 75.4% reported high daily function, and 76.9%
reported being mentally healthy. Results, however, also suggest that higher percentages of
caregivers were at risk for problems associated with physical function, diseases, and depression;
for example, 15% of caregivers reported their health as poor or very poor, which was higher than
non-caregivers (8.3%). Results from other health indicators confirm this finding. In comparison
to non-caregivers, higher percentages of caregivers were found to have some degree of
dependency and total dependency. Similarly, data analysis found that 84.6% of caregivers
reported that they had two or more chronic conditions, 10.4% higher than non-caregivers. It also
found that 17.2% of caregivers had GDS scores of 6–10, and 5.9% had scored greater than 10.
Both were higher than non-caregivers (10.7% and 3.8%, respectively).
An Analysis of Variance (ANOVA) was conducted to further examine whether there were
significant health differences between family caregivers and non-caregivers. Table 4.7 provides
the results of ANOVA. It indicates significant health differences between two comparison
groups, especially in terms of physical function and absence of chronic diseases (p<.001).
In summary, results from caregivers’ demographics, social network characteristics, and
health outcomes provide rich information about social networks, health, and family caregiving in
later life. Family caregivers presented strengths in all three dimensions of social networks.
Compared to non-caregivers, family caregivers showed a higher level of social function, stronger
and reciprocal relationships with family, neighbors and friends, and a higher rate of service use.

86

Table 4.6
Descriptive Statistics of Health Variables by Caregiving Status

Health

Caregiver

Non-caregiver

Total

Excellent

26
(19.5)

41
(19.0)

67
(19.2)

Good

74
(55.7)

133
(61.5)

207
(59.3)

Neutral

13
(9.8)

33
(15.3)

40
(13.2)

Poor or very poor

20
(15.0)

9
(4.2)

29
(8.3)

≥6

101
(75.4)

202
(93.1)

303
(86.3)

2-5

27
(22.1)

10
(4.6)

37
(10.6)

≤1

6
(4.5)

5
(2.3)

11
(3.1)

≤1

22
(16.4)

56
(25.8)

78
(22.2)

≥2

112
(84.6)

161
(74.2)

273
(77.8)

Self-rated health

IADLs-R

Number of chronic conditions

87

Table 4.6 (cont’d)

Health

Caregiver

Non-caregiver

Total

≤5

103
(76.9)

183
(85.5)

286
(82.2)

6-10

23
(17.2)

23
(10.7)

46
(13.2)

>10

8
(5.9)

8
(3.8)

16
(4.6)

GDS

Note. Statistics presented in the table are numbers and percentages.
Table 4.7
Analysis of Variance for Health between Family Caregivers and Non-caregivers

Health Outcomes

df

Between

Mean
Square

F

p

groups

Self-rated health

1

2.73

3.85

.05

Physical function

1

34.41

19.25

.000

Absence of chronic diseases

1

107.13

22.39

.000

Mental health

1

105.48

10.11

.002

Within

groups

Self-rated health

347

246.21

Physical function

349

623.95

Absence of chronic diseases

349

1669.95

Mental health

346

3608.51

88

Comparisons also suggest that higher percentages of family caregivers were at risk of social
isolation, absence of health behaviors, and least social function. Descriptive statistics and
preliminary analyses demonstrate the differences in demographic characteristics and health status
between family caregivers and non-caregivers.
Results from mixed methods
Quantitative and qualitative data were obtained through use of a standardized instrument,
BCOS, in family caregiving and two open-ended questions to address research question two in
the study: “How do older family caregivers perceive family caregiving in later life?” A total of
99 participants rated the BCOS scale to indicate changes they had experienced since serving in
the family caregiving role. A total of 56 participants answered two open-ended questions.
Responses from open-ended questions triangulated results from quantitative analysis of BCOS
scores to ensure validity and reliability of data. The results of mixed methods presented below
include either quantitative or qualitative analysis or a combination of both data analyses.
Results from quantitative analysis
Table 4.8 details results from mean tests of the total BCOS scores. The total BCOS scores
ranged from -23 to 30, with an overall mean BCOS score of 4.47. It compares BCOS scores to
examine whether or not family caregiving experience in later life was different for different
geographic residences. Urban caregivers reported the biggest range of caregiving experience,
from -23 to 30, representing a broad range of experience, from very negative change to most
positive change, as result of caregiving. Rural caregivers reported no change or positive changes
due to family caregiving. Suburban caregivers reported changes ranging from -8 to 29, a
narrower range, leaning more toward positive changes.

89

An ANOVA test indicates that changes in family caregiving experience between three types
of residence were not statistically different, with F=2.67 at p=.08. Overall change and changes in
three resident areas were positive, with mean scores of the sample and three sub-samples above
0, ranging from 2.95 in the urban area to 13.60 in the rural area. When specific BCOS items were
examined, the only negative item was “My time for family activities” in the urban area.
Results from qualitative analysis
To triangulate results from quantitative analyses, the qualitative data were analyzed using
content analysis techniques. The raw data were obtained from narrative responses to the two
open-ended questions. Open coding was used to conceptualize the raw data. The raw data were
documented in a Word file for a line-to-line coding. 21 codes were identified, for they repeatedly
occurred across lines and cases. A code was identified when it occurred equal to or more than
two times across cases. Axial coding was then performed to review the initial codes from open
coding. 12 themes emerged in the axial coding. Three categories were detected from themes.
Figure 4.6 shows the coding process and coding paradigm.
Caregiving. In responding to two open-ended questions “How does caregiving make you
feel?” and “In general, how has your life changed as a result of family caregiving?” caregivers
described their caregiving background, especially care relations, care demand, care activities,
and care receiver’s health condition. Caregivers reported the distance to their care receivers.
They explained the years of caregiving and weekly care hours. Some mentioned caregiving
activities related to IADLs and ADLs. One wife listed care activities as lifting her husband and
bowel excrement cleaning. One reported that “I manage finances for my mother-in-law who
lives in a nursing home.”

90

Care activities and health conditions of the care receiver were often talked about
simultaneously. It was noted that among six respondents who discussed care activities, two
articulated the importance of emotional care. One emphasized, “I care for my life partner
emotionally.” One husband saw himself as “more of an emotional supporter than caregiver. I
try to keep my wife's mood high.” It is because he identifies “(…) memory loss and emotional
stress are the biggest (concern) of the carer so far.”
Qualitative data reveal that in the majority of cases family caregivers were solo caregivers or
permanent caregivers of care receivers. In some cases they were partners of formal health care
providers. A wife caregiver took care of her husband before he went to a nursing home, and
between nursing home stays: “He was kicked out (of) the last one because he is an extremely
difficult patient.”
Caregivers recognized the importance of family caregiving for community living in later life.
A daughter caregiver articulated this best: “She is 94 and living independently with support from
us. She wants to stay in her home, and we are happy to help her realize this.” She went on to say
“Life is wonderful! I have the time and enjoy (thank God!) doing all that I want and have to do.”
A long-term caregiver signified the meaning of caregiving in his life: “Magnificent part of life, I
care for my life partner emotionally. Huge part of my and our life together.” He went on to
criticize society for overlooking caregiving: “I realize how much one benefits from caregiving,
and how truly little society cares about such things in an emotionally, meaningful way.”

91

Table 4.8
Mean Scores of Changes in Family Caregiving Experience in Later Life (n = 99)

Variable

Urban

Suburban

Rural

Total

2.95
(10.33)

5.46
(10.25)

13.60
(13.58)

4.47
(10.63)

My self-esteem

.56
(1.28)

.71
(1.09)

1.20
(1.10)

.65
(1.20)

My physical health

.14
(1.53)

.31
(1.38)

1.00
(1.41)

.25
(1.46)

My emotional well-being

.38
(1.51)

.62
(1.35)

1.40
(1.52)

.52
(1.46)

My ability to cope with stress

.25
(1.60)

.26
(1.14)

1.40
(1.52)

.32
(1.44)

My future outlook

.12
(1.49)

.67
(1.45)

1.40
(1.52)

.40
(1.51)

My relationship with family

.90
(1.56)

.86
(1.59)

2.00
(1.41)

.95
(1.57)

My relationship with care recipient(s)

.81
(1.65)

1.23
(1.67)

2.40
(1.34)

1.07
(1.68)

My time for family activities

-.20
(1.63)

.25
(1.24)

.80
(1.92)

.02
(1.53)

My relationship with friends

.29
(1.37)

.75
(1.16)

1.40
(1.34)

.53
(1.31)

My time for social activities with friends

.00
(1.65)

.53
(1.44)

.60
(1.95)

.24
(1.59)

Caregiving outcome score

Caregiver well-being

Family relations

Relations with friends

Note. Standard deviations are in parentheses.
92

Figure 4.6
Coding Paradigm of Perceived Family Caregiving Experience in Later Life
Community Living

Nursing Home

Formal Health Care

Child Day Care
Group Home
Kin

Care Relations
Non-kin
ADL/IADLCare
Caregiving

Care Activities
Emotional Care
Recipient Health
Condition

Care Load

Care Demand

Care Length
Care Dislantance
Obligation

Rationale for Care
Social
Network

Giving

Family Network Structure

Reciprocity

Care Network
Changed Social Life
Change in Social Life
Better Relationships
Exhaustion

Change

Emotional
Experience

Stress
Lack of Sense of Contriol
Loss & Grief
Positive
Feelings

Positive Effect

Growth
93

It was important to discover what caregivers articulated as the reason why they engaged in
family caregiving in later life. The top three rationales were obligation, giving, and reciprocity.
Among 12 respondents who rationalized caregiving as obligation, three respondents used “duty,”
two used “family obligation,” and another two used “responsible” to state the reason for caring
for a family member. Similar phrases used include “I have to” and “It is the right things [thing]
to do for those you love.” A long-term caregiver described his spousal caregiving experience as,
“Complete, I feel like I have a reason for being alive.” Thirteen respondents rationalized their
caregiving as giving. A long-distance caregiver said: “I do like to help or give.” Many indicated
that they felt good to help. Six caregivers explained caregiving in later life as reciprocity. A male
caregiver for his ill parent declared, “It is time to return the love and care that was given to me!”
An 84-year-old husband stated the rationale for his caregiving was to “repay my wife.” A longterm caregiver described her caregiving as “—helping my husband who helps me.” A solo
family caregiver explained the reason she has cared for her mother and step-father: “I feel they
raised me and I have to take care of them.” A retired math teacher who cared for her mother with
the help of siblings said: “She has always been there for her children, and now it is our turn!” It
was noticed that the reciprocity theme did not emerge exclusively from spousal caregiving and
adult child caregiving. A care receiver articulated reciprocity in family caregiving: “My support
network and I exchange giving/receiving help from each other.” An 86-year-old, life-long
caregiver described giving and receiving in her life: “Makes me happy to help others!
Appreciation for help I received after a knee replacement.”
Social networks. Two open-ended questions did not specifically mention social networks of
caregivers, but social networks were described by respondents when their family caregiving
experience was presented. An 86-year-old wife caregiver mentioned her parents and parents-in-

94

law, neighbors who lived alone or needed assistance, and others. The retired teacher mentioned
her mother, siblings, and grandchildren in her social networks. A newly widowed woman
described her grief related to a smaller network size due to the loss of her mother and husband.
One explanation for this phenomenon is that social networks were essential in family caregiving
and that caregivers recognized their significance from both positive and negative approaches.
Four respondents repeatedly described care networks a total of six times. An urban woman
constructed a care network of five siblings, with one sister living in town, one sister living
outside of the town, one sister living out of state, and two working brothers. She detailed how
she and her sister in town shared the care hours, how her sisters out of town provided longdistance care to their mother, and how her brothers and other sisters provided respite care to her
and the sister in town on weekends. Although conflicts in care networks were only mentioned
twice, conflicted relationships stood out for the strong emotion they carried. A veteran discussed
stress caused by network conflicts: “Stress (one of three siblings has own agenda—will not listen
to others.)” A woman was angry that “no family members lifted a finger to help me.”
Changes. It was evident that caregivers indicated many changes as a result of family
caregiving. A widow who had cared for a friend for over five years said, “I had changed my life
very, very much.” Changes in social life, emotional experience, and positive effects were
reported. The theme of changes in social life emerged from two codes—changed social life and
better relationships. Qualitative data suggest that caregivers’ social lives changed in two
directions. Three family caregivers stressed that their social lives were limited due to caregiving.
A daughter caregiver for her older parents cried out she was “tired and alone!” A female
caregiver for her co-resident parent reported “Social life has suffered—don’t get any time away.”
A 66-year-old grandmother caring for three school-age grandsons also mentioned that “social

95

activities (are) limited.” Others suggested caregiving led to greater connectedness with a greater
society. An older mother of two adult daughters with intellectual disabilities reported her high
involvement in disability advocacy because she made caregiving her cause. She exclaimed, “We
are very active in disability ministry, in disability advocacy, and disability housing. We seek to
raise awareness of the value of persons with disabilities, especially intellectual disabilities.” A
71-year-old volunteer legal guardian and caregiver also made caregiving a social cause.
More respondents reported better relationships with the care receiver and their family. A 75year-old man who has been caring for his ill parent and disabled adult child mentioned
caregiving made him “closer to family members!” A husband caregiver said that caregiving
made him have “more time with my spouse.” A 90-year-old husband was thankful “to appreciate
more and love more the person-spouse… share so much time together.” A sibling caregiver also
expressed: “I love my brother even more.”
Strong emotions were expressed by caregivers. Anger and guilt were displayed in narratives,
but they were not repeated across cases or within cases. Four emotions repeatedly occurred:
exhaustion, stress, lack of sense of control, and loss and grief. Eight respondents used “tired,”
“overwhelmed,” “lack of sleep,” or “physically taxing.” Two said they “sometimes” experienced
tiredness. Others did not specify the frequency of exhaustion. Stress was another emotion
frequently described by respondents. “Stress,” “frustrated,” “worried,” and “concerned” were
used to express caregivers’ feelings. Lack of sense of control was described four times in
qualitative data. A son caregiver described how his life had changed as a result of caregiving:
“There is not as much personal time.” A daughter caregiver expressed: “My life is no longer my
own.” Caregivers also described their loss and grief over death and sickness. An 86-year-old
woman expressed that she was “concerned for person’s suffering.” A bereaved woman who had

96

lost her mother and husband shared her grief: “Sad @ [at] times when I knew death was closer
each day.” A mother described that she was “distressed (because) my 2 (two) daughters are very
sick.”
Despite strong emotional experiences, respondents overwhelmingly articulated the positive
effects family caregiving had on their lives. Caregivers described related positive feelings and
personal growth related to caregiving: 21 respondents used words like “happy,” “feel good,”
“great,” satisfied,” “positive,” “better,” and others 31 times. Night caregivers detailed their
growth. An 81-year-old caregiver reported that caregiving resulted in him “being more
responsible, being more on schedule.” A great-grandmother said that co-caregiving for her two
great-grandchildren “keeps me active and unselfish with my money, living space, and time.”
Similarly, a son caregiver expressed that caregiving made him understand life and love more. He
spoke of the “Chance to discover how fragile we are! Show and practice more love!” A female
caregiver successfully managed to “serve as a volunteer legal guardian and caregiver for family
members and friends” while she underwent chemo therapy, which facilitated her growth.
Triangulation
Results of qualitative and quantitative data point to the same conclusion: Older caregivers
have made positive changes as a result of family caregiving experience. BCOS scores indicated
the highest gained items were the relationship with the care recipient, the relationship with
family, self-esteem, the relationship with friends, and emotional well-being. From the qualitative
data analysis, the codes with the highest frequencies (13–31 responses) were positive feelings,
giving, obligation, better relationships, and growth. As Jick (1979) argued, triangulation of
multiple methods was largely a vehicle for cross validation. The convergence of the results from

97

BCOS scores and two open-ended questions provided a more valid conclusion about changes in
caregivers’ well-being and relationships.
While BCOS scores yield comparable data with existing family caregiving studies (Bakas &
Champion, 1999), qualitative data collected by two open-ended questions added rich and
authentic details to the study. Codes and themes provided insights as to how caregivers
interpreted the meanings of caregiving, which explained changes in caregivers’ well-being and
relationships. This shed light on altruism, obligation, and reciprocity and the positive feelings
they brought to the caregiving experience. Qualitative data not only reached the same
conclusions about changes in caregivers as quantitative data indicated, but also added a deep
understanding of these changes in caregivers’ own words. First, negative emotions like anger,
guilt, and distress were rare but powerful. Gerontological social work must cautiously address
their harmful effects. Secondly, active social engagement with family and society co-existed with
lack of sense of control. The mother of adult daughters with severe intellectual disabilities
exemplified the best of caregivers, who take control over health conditions and contributed to a
greater cause.
Effect modeling
For the purpose of testing the effect model, table 4.9 displays results from bivariate
correlation analyses among the variables included in the effect model, including covariance
variables, age, gender, race, and SES. The table reveals that there were significant correlations
between family caregiving and three of four health measures, structural social networks and two
health measures, content social networks and two health measures, and functional social
networks and three health measures. Specifically, family caregiving was negatively associated
with physical function, absence of chronic diseases, and mental health, with p<. 001. All three

98

dimensions of social networks were positively and significantly associated with self-rated health
(p<.001). Structural social networks were positively associated with mental health (p<.05).
Content social networks were positively associated with absence of chronic diseases (p<.05).
Functional social networks were positively associated with physical function and mental health
(p<.001).
Table 4.9 also indicates that SES was the most significant factor for health of older adults in
the sample. It was positively associated with social networks and self-rated health, and was
negatively associated with physical function and lack of chronic diseases. Age was negatively
associated with self-rated health and positively associated with physical function and absence of
chronic diseases. Gender was negatively associated with functional social networks and
positively associated with absence of chronic diseases. Race was not significantly associated
with any health indicator. The model then answered: “Are there significant mean differences in
the combined health, as measured by self-rated health, physical function, absence of chronic
diseases, and mental health, for different levels of social networks between family caregivers and
non-family caregivers, before and after removing the effects of age, gender, race, and SES?”
To test null hypotheses in multivariate analysis, a preliminary MANOVA found that the
Box’s Test was significant (p<.001). The Pillai’s Trace then was used for MANCOVA results.
The preliminary MANOVA observed that social networks and family caregiving interaction was
not significant (p=.31), indicating the main effect of social networks and family caregiving could
be interpreted accurately by examining the F ratio, p values, and effect sizes for the associated
test statistic.
Results from multivariate analysis are reported in table 4.10. Overall, the table reveals that
both social networks (p<.001) and family caregiving (p<.001 or .05) in later life significantly

99

affected health of the older adults in the study. The significance of health effects of social
networks did not change after controlling for age, gender, race, and SES independently or
combined, while the significant level of effect of family caregiving was reduced from p<.001 to
p<.05 when age of family caregivers was controlled or when the effects of the combined age,
gender, race, and SES were removed.
2

Table 4.10 further indicates that the effect model in MANOVA has a Partial η of .68,
p<.001. When controlling for age, gender, race, and SES, the effect models in MANCOVA have
2

a Partial η ranging from .14 to.67, p<.001. According to Cohen (1988), the benchmark for a
2

2

small effect size is the effect-size measure η =.0099 or Partial η =.01, a medium effect size of
2

2

2

2

2

η =.0588 or Partial η =.09, a large-effect size of η =.1379 or Partial η =.25. A Partial η
ranging from .14 to .67, p<.001 suggests that social networks and family caregiving had

significant effects on health of family caregivers in later life, independently of age, gender, race,
and SES.
Table 10 also reveals that the demographic factors of age, gender, race, and SES both
individually and collectively reduced effects of social networks and family caregiving on health
of older adults in the study, but they only reduced the effect sizes from a large one to a medium
one. Table 4.10 suggests the effect of social networks and family caregiving gender reduced
2

Partial η from .68 to .14 when removing combined effects of age, gender, race, and SES.
To determine whether there were significant mean differences for each of the four measures
of health among different levels of social networks after removing the effects of age, gender,
race and SES, results from descriptive statistics and tests of Between-Subjects effects of
MANCOVA and Post Hoc tests are displayed in table 4.11.

100

Table 4.9
Correlation Matrix for Major Variables of Interest

1

2

3

4

5

6

7

8

1

-.02
.05

.11*

.02

.71**

.08

5

-.11

.18**

.18**

.23**

—

6

-.23**

.12*

.09

.20**

.38**

—

7

-.25**

.06

.12*

.08

.53**

-.40**

—

8

-.17**

.13*

.02

.20**

.36**

-.31**

-.31**

12

—

4

11

—

3

10

—

2

9

—

—

9

.02

-.07

.03

-.13*

-.12*

.24**

.13*

.10

—

10

.05

-.07

.01

-.13*

-.10

.09

.10*

.07

.10

—

11

-.03

-.12*

-.06

-.14**

-.06

-.01

.02

.06

-.10

.10

—

12

-.08

-.32**

-.04

.16**

.13*

.25**

.23**

-.19**

101

-.14**

-.08

-.20**

—

Note. Twelve variables above represent as 1. Caregiving status; 2. LSNS; 3. Health behavior; 4. BSNI; 5. Self-rated health; 6.
IADLs―R; 7. Absence of chronic diseases; 8. GDS―R; 9. Age; 10. Gender; 11. Race; and 12. SES. **p<. 001 (2-tailed). * p<.05 (2tailed).

102

Table 4.10
Multivariate Statistics for Social Networks and Family Caregiving on Health

Model

2

F

SES (Covariate) (Low vs. middle vs. high)
Social networks
Family caregiving
Social networks • family caregiving

103

.005
.000
.001
.54

.05
.04
.06
.01

.000

.19

.31
.000
.000
.36

.02
.04
.07
.01

.000

.67

.42
.000
.000
.40

.01
.04
.07
.01

.000

.45

2.42
3.01
5.89
1.10

MANCOVA with SES***

.27

65.31

Race (white or non-white)
Social networks
Family caregiving
Social networks • family caregiving

.000

0.97
3.53
5.69
0.04

MANCOVA with race***

.04
.07
.01

165.82

Gender (male or female)
Social networks
Family caregiving
Social networks • family caregiving

.000
.000
.31

1.21
3.23
6.19
0.04

MANCOVA with gender**

3.41
6.71
1.12

18.99

Age (<65 vs. 65–74 vs. ≥75)
Social networks
Family caregiving
Social networks • family caregiving

.68

3.75
3.46
5.02
0.90

MANCOVA with age***

.000

28.75

Social networks (high function vs. non-high function)
Family caregiving (caregiving vs. non-caregiving)
Social networks • family caregiving

Partial η

176.93

MANOVA***

p

.048
.000
.000
.36

.03
04
.07
.01

Table 4.10 (cont’d)

Model

2

F

Age
Gender
Race
SES
Social networks
Family caregiving
Social networks • family caregiving

Partial η

12.31

.000

.14

4.69
0.76
0.32
1.16
2.95
4.89
0.86

MANCOVA with age, gender, race, and SES***

p

.001
.55
.87
.33
.000
.001
.59

.06
.01
.00
.02
.04
.06
.01

Note. The Pillai’s Trace was used because the Box’s Test is significant.*** A large effect size.
Data in table 4.11 suggest that there were significant mean differences in self-rated health
and physical function, absence of chronic disease, and mental health for social networks and
family caregiving by SES category. Specifically, the table reveals that there were significant
mean differences in self-rated health and physical function but not for absence of chronic disease
and mental health among different levels of social networks after removing the effect of SES.
Table 12 presents results from Sheffe’s tests of Post Hoc multiple comparisons of mean
differences in health by social networks. It indicates that self-rated health specified that the least
function level significantly differed from the medium high function level and the most function
levels; that the medium low level differed from the most function level (p<.05); that the medium
high function significantly differed from the least function level (p<.05); and that the most
function level significantly differed from the least function level and the medium low function
(p<.05). Post Hoc tests on mean differences of physical function observed that the most function
level significantly differed from the least function level and the medium low function level
(p<.05).

104

Table 4.11
Adjusted and Unadjusted Mean Differences for Health by Social Networks and Family
Caregiving Category

Self-rated
Health

Physical
Function

Absence of
Chronic Disease

Mental
Health

Least function

.55
.53

.40
.35

.86
.82

.34
.35

Medium low

.31
.30

.19
.17

.79
.79

.21
.20

Medium high

.19
.19

.16
.14

.82
.80

.14
.14

Most function

.09
.09

.07
.06

.73
.72

.15
.14

Caregiver

.24
.19

.09
.07

.73
.74

.19
.14

Non-caregiver

.33
.08

.32
.24

.86
.83

.23
.21

Social Networks

Family Caregiving

Note. Statistics presented in the table are adjusted M and unadjusted M. The covariate in the
model is evaluated at the following values: What is the level of SES? = 2.22
Table 4.12 also provides results from ANOVA tests on mean differences of health between
family caregivers and non-caregivers. It suggests that there were significant mean differences on
physical function (p<.001) and mental health (p<.05) but not on self-rated health and absence of
chronic disease between family caregivers and non-caregivers.

105

This study found that family caregivers and non-caregivers presented significantly different
health outcomes in levels of physical function and mental health and that social network
significantly affected health of family caregivers in later life. Although the combined effect of
social networks and family caregiving was not significant in health of older adults in the sample,
significant mean differences were observed in the combined health for social networks and
family caregiving independently.
Table 4.12
Multiple Comparison of Health by Social Networks and Family Caregiving

Health Measure

Social Networks
Group

Self-rated health

p
2
3
4

.393
.047
.003

2

1
3
4
1
2
4
2
3
4

.475
.226
.023

2

1
3
4

.475
.812
.040

3

1
2
4

.226
.812
.350

p

.047
.232
.438

1

Between groups

.393
.232
.002

3

Physical function

1

Family Caregiving

106

.237

.040

Table 4.12 (cont’d)

Health Measure

Social Networks
Group

Absence of chronic disease

1

p
2
3
4

.475
.226
.023

1
3
4
1
2
4

1

2
3
4

.468
.122
.116

2

1
3
4

.468
.493
.463

3

1
2
4

p

.987
.988
.446

.112
.463
1.000

2

Mental health

Between groups

.953
.988
.617

3

Family Caregiving

.237

.040

Note. The mean difference is significant at the 0.05 level.
MANCOVA tests revealed that different levels of social networks caused different self-rated
health and physical function after controlling for the effect of demographic factors. Specifically,
not only older adults with the least function social networks were significantly lower in self-rated
health than those with the medium high function social networks and the most function social
networks, but also older adults with the medium low function social networks had a significantly
lower self-rated health than those with the most function social networks, and the medium high

107

function social networks predicted a significant higher self-rated health comparison to the least
function social networks.
In summary, this study drew a probability sample of 358 older adults aged 64 and above
from a regional representative population. It provided a comprehensive picture of health effects
of social networks between older family caregivers and non-caregivers. Results of the study are
methodologically reliable and significant because the study’s unique design included an effect
model using multivariate techniques and mixed methods, which integrated a relatively large
sample size, a rigorous quantitative analysis, and in-depth understanding of the phenomenon
under investigation.

108

CHAPTER FIVE: SUMMARY AND DISCUSSION

This chapter provides a summary of the study, an interpretation of the data, and limitations.
Implications and recommendations are also explored for future gerontological social work
research, education, practice, and policy related to family caregiving in later life. Finally,
conclusions are drawn about family caregiving in later life, social networks, and healthy aging.
Summary of the study
The purpose of this study was to better understand the social value of family caregiving in
later life by examining the effects of social networks and family caregiving in later life on the
health of older adults aged 65 and over. In this study, family caregiving in later life was
conceptualized as a social event in which community-dwelling older social network members
contribute to unpaid, informal caregiving of a kin or a non-kin person with chronic illness,
disability, or other dependent characteristics on a regular basis within the last 30 days. Social
networks were defined as older-adult-centered multidimensional social relationships that cut
across traditional kinship, residential, and class groups. Specifically, structure, function, and
content of social relations were integrated into three-dimensional social networks. Health was
defined as high self-rated health, a high level of physical function, absence of chronic diseases,
and a high level of mental health.
Theoretical framework
The significance of this study first comes from its positive gerontology framework, based on
critical gerontology and positive psychology (Minkler & Estes, 1999; Seligman &
Csikszentmihalyi, 2000). Built on prior studies and theory, this study attempted to highlight the
positive effects of social networks and family caregiving on healthy aging among family
caregivers in later life. It expected to find that family caregivers with stronger social networks
109

would be more likely to have higher levels of self-rated health and physical function, lower risk
of disease, and higher levels of mental health after controlling for age, gender, race, and SES.
Methodology
This was a cross-sectional, explanatory study. Data were collected from a population-based
self-administered mail survey during July–September 2011. Stratified random sampling was
used to compile a mailing list of 1,000 older adults aged 64 and above in three distinct Midwest
communities. A multiple contact was employed and resulted in a response rate of 36.6% and a
final sample of 358 community-dwelling older adults.
The questionnaire was developed specifically for this study. It had five components. The first
was comprised of LSNS-18 and a subscale of civic engagement of revised BSNI. The second
was comprised of IADLs, chronic diseases and related treatments, and GDS-15. The third was
related to health insurance coverage, nutrition intake, alcohol drinking, smoking, and exercise.
The fourth was comprised of self-identified family caregiving status, caregiving activities,
service use, BCOS, and two open-ended questions about family caregiving experience. The fifth
was comprised of demographics. The questionnaire presented 58 quantitative questions and twoopen-ended, qualitative questions investigating: (1) age, gender, race, and SES structures of older
family caregivers; (2) structural social networks; (3) changes in structural social networks; (4)
functional social networks; (5) content of social networks; (6) care networks of family caregiving
in later life; (7) perceived family caregiving experience in later life; (8) self-rated health; (9)
physical function; (10) chronic diseases; and (11) mental health of the sample.
An operational model of Health Effects of Social Networks and Family Caregiving in Later
Life was adapted from previous works on family caregiving and social networks in gerontology
by Carpentier and Ducharme (2003) and Berkman et al. (2000). Multivariate models and mixed

110

methods were employed to gain a comprehensive and accurate picture of social values of family
caregiving in later life by examining the observed effects of social networks and family
caregiving in later life on the health of older adults. Quantitative and qualitative data were
collected to gain in-depth understanding of social values of family caregiving in later life
reported by older family caregivers. Efforts were made to assess independent variables of social
networks and family caregiving in later life and dependent variables of health through multiple
indicators. Multivariate analysis allowed the study to inquire into separated and combined effects
of social networks and family caregiving in later life on the health of older adults after
controlling for four key demographic factors.
Research questions
Three research questions were answered in this study. The overarching hypothesis and
research question was to compare the similarities and differences in health due to different levels
of social networks and different family caregiving statuses. The null hypotheses were that there
would be no observed differences among different levels of social networks and between
different family caregiving status on the combined health, as measured by levels of self-rated
health, levels of physical function, levels of absence of chronic diseases, and levels of mental
health. MANOVA and MANCOVA were used to determine whether there were significant
differences in the health of older adults for levels of social networks and different family
caregiving statuses before and after controlling for age, gender, race, and SES. Qualitative
analysis was conducted based on data from two open-ended questions to examine family
caregiving in later life using caregivers’ own words.
The first research question sought demographics, social network characteristics, and health
status of family caregivers and non-caregiving older adults aged 64 and above. Frequency tests,

111

ecomap, and one-way ANOVA were performed to compare similarities and differences of all
independent variables, dependent variables, and covariates between family caregivers and noncaregivers. The second question sought perceptions of family caregivers on family caregiving in
later life through the standardized scale BCOS and two open-ended qualitative questions about
changes in family caregiving. The third question examined the causal relationships that social
networks and family caregiving independently had on the health of older adults through bivariate
correlation analysis and a series of multivariate analyses.
Results
Results of the investigation indicated that both social networks and family caregiving in later
life independently and significantly affected the combined health, as measured by levels of selfrated health, levels of physical function, levels of absence of chronic diseases, and levels of
mental health, after controlling for demographic factors. High levels of social networks
significantly predicted better self-rated health and higher physical function. Being family
caregivers in later life significantly predicted lower levels of physical function and mental health.
More detailed results are presented below for each of the research questions.
Question one was stated as: “What are demographics, social network characteristics, and
health status of older family caregivers, compared to other older adults who do not provide
family caregiving?” The final sample of the study consisted of 358 community-dwelling older
adults aged 64 and above, involving 136 family caregivers and 222 non-caregivers. A total of
114 respondents identified themselves as family caregivers. Another 22 respondents performed
family caregiving activities but did not identify themselves as family caregivers. Among
caregivers, 87 cared for older adults; 34 for adult children with disabilities; and 15 for
grandchildren/great grandchildren.

112

Demographics. Data in this study found that caregivers were slightly older than noncaregivers. The majority of older adults in both groups were between 65 and 79 years of age, and
24.8% of caregivers were among the oldest old. Frequency tests indicated that 45.7% of
caregivers were male, and the gender ratio was higher for caregivers than non-caregivers. The
percentages of non-white older adults with low household income were higher in caregivers than
non-caregivers. Caregivers were less educated than non-caregivers, with a higher rate of high
school education or less and a lower rate of higher education.
Social networks. Surprisingly, data in this study found that among three dimensions of social
networks, the functional social network presented as the most important dimension in social
networks for the combined health of the sample. Results from descriptive statistics uncovered no
significant difference in structural social networks between caregivers and non-caregivers, who
had LSNS scores of 46.90 and 47.44, respectively. An estimated one-third of both groups had
LSNS scores greater than 54, an indicator of socially high connectedness. However, more
caregivers than non-caregivers had average LSNS scores lower than 36, an indicator of risk for
social isolation. Social network structural statistics found that caregivers and non-caregivers each
had a network size of 9–12, consisting of relatives, friends, and neighbors. Caregivers and noncaregivers had a similar number of relatives, friends, and neighbors in their social networks.
Descriptive statistics from SF-LES found that caregivers generally experienced more radical
changes in structural social networks than non-caregivers. Data found that in the two years prior
to the study survey, 42.2% of caregivers and 32.5% of non-caregivers had lost an immediate
family member, and 62.6% of caregivers and 51.2% of non-caregivers lost one or more close
friends. At the same time, more caregivers than non-caregivers had gained new family members
as well as started school, training, or other learning.

113

Results from bivariate correlation analyses suggest that family caregiving in later life was
significantly and positively associated with levels of physical function, absence of chronic
diseases, and mental health, with p<. 001. All three dimensions of social networks were
positively and significantly associated with self-rated health (p<.001). Structural social networks
were negatively associated with mental health (p<.05). Content social networks were
significantly and negatively associated with absence of chronic diseases. Functional social
networks were significantly and negatively associated with physical function and mental health
(p<.001).
Data found a lower percentage of caregivers (27.4%) reported to have four health
behaviors—namely, nutrition intake, physical exercise, absence of alcohol misuse, and absence
of smoking—than non-caregivers (34.4%). Caregivers reported an average of 2.81 health
behaviors, in comparison to 2.99 reported by non-caregivers. By contrast, caregivers appeared to
have a slightly higher service use rate (0.73) than non-caregivers (0.69) when treatments for their
health conditions were considered. A noticeable finding about service use was that only 20
caregivers (14.7%) had used health and human services for family caregivers within the past
year. The other 116 caregivers either did not use family caregiving services (64.0%) or did not
answer the question about family caregiving service utilization (20.6%). Fifteen caregivers
(11.0%) never knew family caregiving services were available. Of caregivers who used family
caregiving services, 39.9% reported that they obtained information about family caregiving
services through both professionals and social network members, 32.3% of caregivers reported
that they obtained information exclusively from social network members, and 37.8% exclusively
from professionals.

114

Descriptive statistics of revised BSNI scores uncovered that caregivers had a moderately
higher BSNI score (6.28) than non-caregivers (6.13), indicating a relatively higher degree of
overall social function. Higher percentages of non-caregivers had engaged in organizations,
groups, and volunteering than caregivers, while on average both had exactly the same degree of
religious participation. Caregivers were less likely than non-caregivers to live alone and more
likely to live with two and more people than non-caregivers. Non-caregivers presented higher
percentages at both the most functional level (31.67%) and the least functional level (6.79%)
than caregivers (30.37% and 5.19%, respectively). It was also noted that although caregivers
presented a lower risk of being least functioning, 29.63% of them were at the medium low
functional level, indicating marginal risk of social dysfunction.
Results from social network analysis indicated that a typical family caregiver in the sample
had a range from 13–20 members in his or her social networks, consisting of relatives, friends,
and neighbors, with medium high social functioning. More than a quarter of them practiced
health behaviors, and the majority of them used treatments when they had health conditions.
They obtained treatment information from both network members and professionals, with a
preference for professionals.
Health. Results from descriptive statistics revealed that the majority of family caregivers
reported excellent or good health, high physical function, and good mental health. Almost one in
five family caregivers rated their global health as excellent, more than half of caregivers
identified their health as good; 75.4% reported high daily function; and 76.9% were mentally
healthy. Results, however, also suggested that higher percentages of caregivers were at risk for
problems associated with physical function, diseases, and depression. Of all caregivers, 15%
reported their health as poor or very poor, which was higher than non-caregivers (8.3%). Of all

115

caregivers, 84.6% reported that they had two or more chronic conditions, 10.4% higher than noncaregivers. Data also found that 17.2% of caregivers had GDS scores of 6–10, and 5.9% had
GDS scores greater than 10; both were higher than non-caregivers (10.7% and 3.8%,
respectively). Results of ANOVA analysis indicated significant health differences between
caregivers and non-caregivers, especially in terms of physical function and absence of chronic
diseases (p<.001).
Question two was stated as: “How do older family caregivers perceive family caregiving in
later life?” Frequency tests of BCOS scores found that 52.5% of caregivers reported positive
changes since assuming the caregiver role. An ANOVA test revealed that positive changes
existed among three different geographic residence types. Qualitative data triangulated results
from quantitative analysis of BCOS scores. Among 21 codes, more codes denoted positive
changes than negative ones, and more respondents repeated positive codes than negative ones.
Question three was stated as: “Are there significant mean differences in health for different
levels of social networks and for different family caregiving status after controlling for
demographic factors?” Multivariate analysis found that the effect model in MANOVA has a
Partial η2 of .68, p<.001 and that social networks and family caregiving interaction was not
significant (p=.31). This result indicated that social networks and family caregiving in later life
independently and significantly predicted the health of older adults, and that the health of the
sample was not caused by the combined effect of social networks and family caregiving in later
life. When controlling for age, gender, race, and SES, the effect models in MANCOVA had a
Partial η2 ranging from .14 to.45, p<.001, suggesting that social networks and family caregiving
independently had significant effects on the combined health of family caregivers in later life,
independent of age, gender, race, and SES. Data revealed that the demographic factors of age,

116

gender, race, and SES—both individually and collectively—reduced effects of social networks
and family caregiving on health of older adults in the study, but they only reduced the effect
sizes from a very large one to a medium one. More specifically, family caregivers and noncaregivers presented significantly different health outcomes in levels of physical function and
mental health. Older adults with the least social function were significantly lower in self-rated
health than those with the medium high social function and the most function social networks.
Older adults with the medium low social function had a significantly lower self-rated health than
those with the most social function, and the medium high social function predicted a
significantly higher self-rated health comparison to the least function social networks.
Interpretation of the results
This study found significant effects of social networks and family caregiving in later life on
the health of older adults in the study. Population-based studies on health effects of social
networks among older family caregivers have been sparse, which made it difficult to compare
results of this study to those from prior studies. Nevertheless, this study added new knowledge to
the family caregiving literature by presenting new empirical evidence on the social realities of
family caregivers and their health outcomes. Within a positive gerontology framework, results
from this study highlighted the positive aspects of social networks of family caregivers in later
life. They can be expressed as: (1) family caregiving as a valuable event in later life, (2)
strengths and resilience of caregivers, and (3) relational aging.
Family caregiving as a valuable event in later life
Previous literature has suggested that the reasons why people were called for family
caregiving were that they were young, healthy, and/or available (Szinovacz & Davey, 2006).
This study provided empirical evidence that family caregiving is a daily event of older adults, no

117

matter whether they are young or old, healthy or not healthy, working or retired, or living close
or living far. It is important to note that inclusion of all informal caregivers other than primary
family caregivers made it possible for this study to capture the meaning of family caregiving in
later life. More importantly, older caregivers interpreted caring for people around them as a
meaningful part of their life, aging, or responsibility, which was undervalued by society from
their perspective. The fact that there were 22 family caregivers in the final sample who did not
identify themselves as family caregivers was further evidence that family caregiving was
undervalued by society.
Re-evaluation of family caregiving challenges ageism by reconstructing the realities of aging
and family caregiving in later life. Data in this study revealed that 38% of older adults in the
sample provided unpaid care to family members, friends, and relatives, a higher percentage than
AARP’s estimates (Armo et al., 1999; Houser & Gibson, 2008). Their care recipients were
across the life span, including older adults, adult children, and grandchildren or great
grandchildren. Older adults at all age groups have had an almost equal chance to become family
caregivers. Of the oldest old in the sample, 37.2% were caring for a family member, friend, or
neighbor within the last 30 days prior to the study. Data on employment and civic engagement
other than informal caregiving further supported the notion of productive aging. Employment
data revealed that 25.28% of older adults in the sample were employed, ranging from full-time to
being a student. In comparison to non-caregivers, the retired rate in the caregiver group was only
2.5% higher than the non-caregiver group; the full-time employment rate was 5.2% lower; and
the part-time employment rate was 1.1% higher.
Strength and resilience of family caregivers

118

The reason that over half of caregivers reported positive changes since assuming the
caregiver role partially rested on the meaning of family caregiving in later life as conveyed in the
study, and partially was the result of caregivers’ personal growth throughout the caregiving
journey. Data in this study provided strong evidence to support positive gerontology, particularly
through the strengths and resilience presented by family caregivers in the study. One caregiver
wrote at the end of the survey: “A lot of water has gone under the bridge—Life goes on.” It
constructed family caregivers’ way of life as well as aging. Although loss and grief were
devastating and care responsibility was demanding, dignity remained, and the best within the
individual caregiver and institutions emerged. Among 136 family caregivers: 35.4% cared for
two or more family members, friends, and neighbors; 43.8% cared for co-residents; 52.3%
provided care to non-co-resident family members, friends, and neighbors; 3.9% were longdistance family caregivers; 50.0% were solo family caregivers; 39.8% were long-term
caregivers; and 26.6% provided more than 20 hours of informal care per week. The resilience of
family caregivers also showed in their general health and positive self-assessed health. Although
caregivers might present significantly low levels of physical function and a high number of
chronic diseases, they were socially connected and adjusted. The majority of family caregivers
reported excellent or good health, high physical function, and good mental health. Almost one in
five family caregivers rated their global health as excellent, more than half of caregivers
identified their health as good, and over three of four reported high daily function and high
mental health.
From a historical perspective, the study further documented the continuity of family
protection in contemporary American society, where greater public protection and sophisticated
institutional intervention have been evident. Findings of the study about the high prevalence of

119

family caregiving in later life supported the argument that the family system continues to be
critical to the well-being of its members. One possible explanation for this might be the high
quality of care provided by social network members. Previous literature has identified the
following qualities in family caregiving: care provision, better care outcomes, respect, low cost,
and accessibility (Dooley, Shaffer, Lance, & Williamson, 2007; Raphael & Cornwell, 2008;
Harrow et al., 2004). This study observed that family caregivers provided extensive care for
family members, friends, and neighbors. Among older family caregivers in the sample, 34.4%
provided care for two or more spouses, parents, adult children, or grandchildren. More than 26%
of older family caregivers spent more than 20 hours a week providing unpaid care within 30 days
prior to the study. Nearly 40% had been family caregivers for five or more years, and about 21%
of them for 10 or more years. Furthermore, findings indicated that family caregiving had
respected and met care recipients’ needs for kinship care and community living. Consequently,
older adults and adults with chronic illness or disabilities often either delayed use or simply did
not use institutionalized long-term care.
Relational aging
Convenience samples and small sample sizes in literature have contributed to inconsistency
about the social consequences of family caregiving in prior literature. Drawing conclusions from
a medium-large-size probability sample and well-thought study design and data analysis
strategies, this study has added compelling evidence that family caregivers in later life do not
present significantly different patterns in all three dimensions of social networks when compared
with non-caregiving older adults.
Structural social networks. Family caregivers presented similar patterns of structural social
networks as non-caregivers. Average network size of 9–12 reported in this study is consistent

120

with Greenberg et al. (1993) and Haley et al. (2010), and is much larger than most social
networks studies on family caregivers (Penrod et al., 1995; Tolkacheva et al., 2011). Similarly,
caregivers, as non-caregivers, often saw and heard from relatives, friends, and neighbors with
whom they had the most contact. Considering changes in social network structures over the past
two years prior to the study, including higher rates of loss of family members and friends,
network size, and frequency of contact with network members among caregivers, on average, led
to conclusions that caregivers generally presented a similar level of social connectedness as noncaregivers. However, the fact that 23.1% of family caregivers were at risk of social isolation,
2.7% higher than non-caregivers, should be an alarm sign for policy makers and gerontological
social workers. On the other hand, the study confirmed the results from caregiving networks
literature that caregiving networks are smaller than social networks of family caregivers (Fast et
al., 2004; Keating & Dosman, 2009; Keating et al, 2003). Among the 9–12 network members of
caregivers in the sample, only 6–9 were caregiving network members. This explains why
caregivers perceived their network members and confidants “sometimes” or “often” reachable
when the caregiver had an important decision to make. Frequencies of reachability and
reciprocity were reported as “sometimes,” while frequency of network contacts was “often.” The
distinction between social networks and caregiving networks has important implications for
social work practice and policy.
In terms of diversity of social network compositions, conclusions drawn in this study
confirmed that family members, particularly immediate family members, are the core of
caregiving networks in most cases and that, while friends count for the second most important
support for family caregiving in later life, friends are not always reachable in caregiving
networks (Carpentier & Ducharme, 2005; Keating & Dosman, 2009). Data also discovered that

121

neighbors are more reachable in care networks than the literature has tended to recognize. In
comparison to non-caregivers, caregivers rely more on neighbors for help.
One contribution that the current study has made to social networks of family caregiving
research is to demonstrate that caregivers present higher scores of network strength and
reciprocity than non-caregivers, but slightly lower scores of network size, frequency, help, and
reachability. Caregivers had six or more confidants with whom they could talk about private
matters and 6–9 network members they could call on for help. One explanation of these minor
differences, which arise from mixed methods, is that caregivers had experienced more intensive
relationship development with family members and the care recipient than friends. Caregivers
reported better relationships as the result of caregiving.
Functional social networks. Previous literature on social networks of family caregivers has
overwhelmingly addressed the importance of structural social networks for the health of family
caregivers. This study has provided new evidence: Functional social networks presented as the
most influential dimension in social networks to explain health effects of social networks.
Results in this study indicated that caregivers had a relatively higher degree of overall social
function. Civic engagement data suggested that nearly half of caregivers had participated in
organizations, groups and volunteering. Caregivers had lower scores of organization
participation, group membership, and volunteering than non-caregivers, but the differences were
not significant. Although there was no correlation between levels of functional social networks
and different family caregiving status, data nevertheless revealed that caregivers presented a
lower risk of being least functioning than non-caregivers. One explanation of the results about
those functional social networks is the significance of co-residency in health of older adults, in
general, and of family caregivers in later life, in particular. Data suggested that caregivers were

122

less likely than non-caregivers to live alone and more likely to live with two and more people
than non-caregivers.
Two additional findings are worth noticing about functional social networks among family
caregivers. While caregivers presented low risk of social dysfunction with a higher percentage of
having a co-resident presence, more than a quarter of them were at marginal risk of social
dysfunction. Care demands and family conflicts might have been two factors contributing to
caregivers being subject to marginal risk. Qualitative data indicated network conflict was rare but
harmful for relationships and health of caregivers. Quantitative data from SF-LES uncovered that
18.63% of caregivers, verses 10.1% of non-caregivers, reportedly had legal issues with a social
network member during the two years prior to the study. Family conflicts manifested in more
informal forms in caregiving families. Anger and guilt were two topics identified by family
caregivers in the sample. Although they did not emerge as codes due to their infrequency, they
added new evidence to recent literature on prevention of elder abuse in regard to family
caregivers’ harmful behavior (Beach et al., 2005, Cooper, Selwood, Blanchard, Walker, Blizard,
& Livingston, 2009).
Content social networks. Social networks influenced information and behavior shared by
network members. Data revealed that more than a quarter of family caregivers reported having
healthy nutrition intake, regular exercise, and absence of alcohol misuse and smoking. Although
family caregivers used more treatments and services available than non-caregivers, only a very
small percentage of family caregivers used human and health services for family caregivers in
the two years prior to the study. Family caregivers used social networks and professionals to seek
information regarding treatments and human services. Nearly one-third of caregivers solely
relied on their information channel of network members, while a little more than one-third solely

123

relied on professionals. Ability and skills to seek access to services and treatment were
particularly important for the health of family caregivers, since they were more likely than noncaregivers to have had two or more chronic conditions, to have suffered depressive symptoms or
depression, and to have experienced more radical changes in social network structures.
Evidence about older family caregivers’ strength and capacity to live a loving, productive,
social functional, and civic engaged life leads to a better understanding of how sustainable and
quality family caregiving can be possible in the United States at the grassroots level. One of the
stunning results in this study was the small number of family caregivers who had known the
availability of and used human and social services for family caregivers, while family caregivers
trusted professionals for information and access to treatment and services. Although this result
was consistent with previous findings, it has sounded an alarm about the severe disconnection
between service needs and service delivery. Anger, resentment, and other negative emotions
existing among family caregivers were signs that our family caregiving system is overburdened
and overly exploited. It is an ethical issue for social work that a sustainable caregiving system
cannot solely rely on family and community institutions.
Limitations
Although this study was valuable for the investigation of social network factors affecting
healthy aging of older family caregivers, this study had a number of limitations. First, the notion
of positive gerontology suggested that healthy aging among older family caregivers is an
outcome of innate strength and resource. This study has provided important empirical evidence
regarding positive features of family caregiving in later life and positive health outcomes of high
levels of social networks in a final sample of 358 older adults. Threats to external and internal
validity were identified in this study. Key threats included a relatively small sample size, a

124

relatively low response rate, a cross-sectional study, a regional representative sample, and
measurement development.
Sample size and response rate
A probability sample size of 358 drawn from a population of 6,514 older adults in three
research sites could have tolerated an estimated ±.5% margin of error and met data requirements
for multivariate analysis, but it was not qualified for a desirable large sample size for a
population-based study. In the same line of data evaluation, efforts were made to produce results
representative of the general population using stratification by random sampling strategies, but
the mailed survey only achieved a response rate of 36.6%, which created concerns about the
representativeness of the sample respondents and response bias. It is possible that caregivers who
had extremely high or low levels of social networks that provided information on extremely
positive or negative health conditions did not participate in the study. In terms of data collection
methods, the multiple contact strategy certainly increased the response rate. On the other hand, a
12-page-long questionnaire with a 14-point font size was a harmful approach for a higher
response rate. For many respondents as well as potential respondents it could have taken longer
than 15 minutes to complete the questionnaire. It could even have taken longer for those
caregivers who wrote long responses to two open-ended questions.
Cross-sectional study
The cross-sectional study allowed an examination of how properties of social networks were
related to the health of the individual respondent at one point in time. Although efforts had been
made to include non-caregivers as a comparison, results drawn from the study could not lead to a
confident claim about the causal relationships between social networks and health of respondents
as it could have been drawn from longitudinal data. Specifically, this was limiting for at least

125

three reasons. First, it did not allow for tracing social network changes over time among older
family caregivers. An effort was made to investigate changes in the size and composition of
social networks over the last two years using SF-LES. Since SF-LES was not a caregivingspecific measurement, it was limited to linking social network changes with family caregiving.
Second, no data on baseline health status was available to record changes in health over time or
control for previous health status. Third, there were no follow-up data collections that would
have allowed for in-depth explanations regarding health effect questions identified in this study.
Other limitations of the survey study included problems with the sample frame. The mailing
list for stratified random sampling contained 4.2% of incorrect or outdated addresses. Wrong
addresses, death, age, and institutionalized older adults contributed to the low response rate.
Because the sample frame used 64 or 66 as the cut-off age, the study used a sample of 358 older
adults aged 64 and over, which contained 16 participants who were 64 years old. A final sample
that excluded 16 participants under 65 would create a more comparable sample in gerontology
literature. Additionally, the mailing list had a lower percentage access to rural areas than urban
and suburban areas. Even after oversampling rural population, only 15 rural respondents were
included in the study, which was too small to establish enough power for a meaningful
comparison between urban, suburban, and rural residences. It is also important to point out that
subjectivity of data tends to limit reliability of data as well as generalization of the results.
Information collected in the study was retrospective and self-reported by nature. Socially
desirable answers and other subjective biases could have distorted the accuracy of the
information collected in this study. Two open-ended questions gave voice to family caregivers
about their caregiving experience in later life. This method added a new data source and
viewpoint but failed to bring objectivity of measures or accuracy of data to the study.

126

Regional representative sample
Results reported and interpreted in this study were based on data drawn from a Midwest
sample where economic downturn has affected demographics dramatically, including population
decline over the last ten years. It was possible that changes in social network size, composition,
and availability were caused by demographic declines, family financial crisis, or unemployment.
This regional representative sample was limited in terms of population diversity. Even though
efforts had been made to integrate more diverse populations into the sample through carefully
selected research sites and oversampling rural older adults, the percentages of minority, rural
older adults, and people living in poverty were lower than national averages. The results cannot
be generalized in other parts of the country since they do not reflect regional differences.
Measurement errors
Two types of measurement errors might have contributed to threats to internal validity in my
investigation. First, there were measures and items included in the questionnaire that did not
closely tie to research questions and the effect model. For instance, questions related to
frequency of civil engagement, insurance coverage, exercise categories, reason for retirement,
and immigration should have been excluded from the questionnaire. Secondly, there were
measures and items not included in the survey that might have increased internal validity and
consistency between research questions and data collected in the study. For example, the data
only captured the dyad relationships between the participants and their social network members.
It did not provide for an explanation in the relationships between and among social network
members. Therefore, the study only revealed a partial picture of the social location of the
caregivers within their complex social networks. Additionally, previous studies and qualitative
data in this study suggested the importance of emotional support in family caregiving, but the

127

measure for caregiving activities in the questionnaire did not include emotional support. It was
possible that issues in conceptualization and measurement development might have missed
alternative explanations for the relationships between social networks and health of family
caregivers in later life. Most importantly, the study design failed to accurately measure the
content dimension of social networks for caregivers and non-caregivers. Thus, the notion of
three-dimensional social networks could not be fully carried out in the study.
Triangulation
Methodological efforts have been made to enhance validity of the results by using multiple
data collection and data analysis methods, namely, between-methods triangulation and withinmethods triangulation (Jick, 1979). Open-ended questions were used in the survey design for
cross validation about changes of family caregivers’ well-being and relationships. Multiple
scales were used to cross-check for international reliability of the social network construct.
However, the use of mixed methods in the study was not systematic. Between-methods
triangulation was only reflected in research question two about family caregiving perceptions.
Only quantitative approaches were used as a single reference point to health efforts of social
networks among older family caregivers. In short, triangulation was not integrated in the
overarching research question on health effects of social networks among older family
caregivers.
Although it is debatable as to whether mixed methods is more than method, an integrative
mixed methods research can provide multiple reference points and greater accuracy of research
findings about social networks among older family caregivers (Johnson, Onwuegbuzie, &
Turner, 2007; Harrits, 2011). For instance, face-to-face interviews of family caregivers and their

128

network members might collect different kinds of data bearing on health effects of social
networks of family caregivers in later life.
Implications and recommendations
As President Barack Obama recognized in the Presidential Proclamation National Family
Caregivers Month of 2011, family caregivers “sustain American men, women, and children at
their most vulnerable moments, and through their devoted acts, they exemplify the best of the
American spirit” (Obama, 2011). Findings of the study highlighted the perseverance of
caregivers. The findings suggested that social support networks play a vital role in sustaining
older family caregivers in their daily activities as well as in their overall well-being. Empirical
evidence that helps us to recognize and better understand the contributions of these networks is
important for gerontological social work in both community settings and long-term care
facilities. Below, I have identified a number of implications for theory, research, practice, and
policy.
Implications for research
Although a systematic discussion about the roles of older adults was beyond the scope of this
study, findings of the study emphasized the intersections of older adults with retirement,
employment, care work, and civic engagement. The study raised a series of questions about
what roles older adults are playing in the private and public spheres. Of 358 older adults in the
sample, 73% were retired, 24% were working or in school, 38% provided informal caregiving,
and 56% volunteered in civic society (m=.56; SD=.50). An engaging and productive old age
challenges the stereotype that older adults are exhausting fiscal and social capital as dependent
and burdensome to the family system and the society. In light of critical gerontology, this study
calls for a commitment to change the social construct of old age in social work research by

129

promoting a positive gerontological perspective of roles of older adults (Holstein & Minkler,
2007; Minkler & Estes, 1999; Brothers & de Jong Gierveld, 2011).
One specific example would be the roles of older family caregivers in the caregiving system.
The study found that there were older adults who were both care recipients and care providers. It
confirmed previous findings that the culture of aging becomes increasingly complicated in terms
of definitions of family, aging, relationships, and growth (Edwards, 2003; Cohen, 2006; Dunne,
Wrosch, & Miller, 2011). Prior literature has found that families exchange all kinds of support
and services in all types of forms throughout the life course (Fingerman, Miller, & Seidel, 2009;
Moen, 2011). Social work research has yet to answer practical questions such as: What is the
meaning of “family”? How can we professionals make a difference in older adults’ lives? How
do we change caregiving systems to support these individuals? What best-practices are available
to provide quality and sustainable care for care recipients as well as their social network
members? Negative health outcomes of family caregiving in later life indicate the urgency and
significance of these unanswered questions. Gerontological social work researchers have a
professional obligation to find evidence-based practice to better support older adults and their
families to create, navigate, and utilize health care systems and caregiving systems.
This study provided useful findings for future studies regarding family systems, social
networks, and health of older adults. It provided new empirical evidence about positive features
of family caregiving in later life and positive effects of social networks on self-rated health and
physical function of family caregivers in later life. Since the nature of a cross-sectional study
prevented the results of this study from being generalized to older adults from other states or
other countries, retesting the conceptual model and research design might contribute to better

130

understanding of the variations that result in more positive experience for some older family
caregivers but more negative experience for others.
Limitations of this study suggest that future research in quantitative and qualitative methods
or their combination is needed to continue exploration of the health effects of social networks
among older family caregivers. Specifically, a larger and national representative sample of
family caregivers in later life could contribute to validation of the significant differences found in
their social networks and health. Further research is needed to identify whether the findings of
this study are relevant to older family caregivers in other states. Future research could lead to
identifying whether the network effects may vary in relation to the prior health status or the
length of caregiving. Longitudinal studies would be a more suitable methodological strategy.
Social networks are complicated phenomena to conceptualize and measure. More research on
social networks in gerontology, including care networks, would be critical to better understand
the differences in family caregiving experience in later life. This study proposed a threedimensional social network concept and only assessed three dimensions separately. Knowledge
would be enhanced by continuing to combine those social network dimensions into a more
integrated concept, by collecting new data regarding social networks of older family caregivers,
and by applying more social network analysis technology. For example, the revised BSNI used
in this study made significant changes from the original BSNI, although it maintained a total of
12 points and was divided into four functional levels. It weighted marriage status as 0 (no) or 2
(yes) but did not weight as much as BSNI, 0 (no) or 4 (yes). It scored three relationship items as
0 (no relative or friend), 1 (3–4 relatives and/or friends), or 2 (5 and more relatives and/or
friends). I scored three dichotomous items to assess organizational memberships as 0 (no) or 1
(yes). Since reliability of the revised BSNI (α=.66) did not reach the level of high internal

131

reliability (α=.70), retests must be conducted before any confident claims for the revised BSNI
can be made.
Negative emotions and relationship conflicts among network members experienced by older
family caregivers affect the health of caregivers as well as sustainability and quality of family
caregiving. Further research efforts regarding issues related to anger, resentment, or other
understudied emotional aspects within social networks of family caregivers could be useful to
not only better understand social networks and care networks, but also prevent negative relations
and network changes and further facilitate positive changes in caregiving networks. Furthermore,
more research about the downside of social capital in older adults will advance not only our
knowledge of the culture of aging, but our practice toward the well-being of older adults and
their network members (Portes & Landolt, 1996; Carr & Hendricks, 2011)
This multivariate study measured health outcomes of social networks and family caregiving
that included self-rated health, physical function, absence of chronic disease, and mental health.
It observed the inconsistent health effects of family caregiving in later life from quantitative and
qualitative data analyses. One might view this inconsistency as being a result of available health
constructs. In line with previous studies on health constructs, this view emphasizes the
unanswered question about conceptualization and measurement of “health” (Pinquart, 2001).
Attempts have been made to propose new frameworks of health, such as subject well-being
(Cheng et al., 2009; Moor, Zimprich, Schmitt, & Kliegel, 2006), quality of life (Shiovitz-Ezra,
Leitsch, Graber, & Karraker, 2009), or a resource for everyday life (Williamson & Carr, 2009)
and beyond (Raehurn & Rootmen, 2007). However, we still have limited tools to fully observe
the health of older adults as well as their experience and understanding of health. For instance,
this study failed to capture cognitive health of older adults through a self-administered

132

questionnaire. Future geriatric research can further evaluate existing health constructs as well as
explore and develop new health constructs that measure a holistic state of physical, functional,
mental, cognitive, social, and subjective well-being of older adults (WHO, 1946; Pinquart, 2001;
Lee et al., 2010). It also can use multiple data collection models to better measure health of
older adults. This advancement in health measures is expected to contribute to the health of older
adults and to policy and practice that promote the health of older adults (Moor et al., 2006;
Rochat, et al., 2010).
Implications for practice
Social networks of family caregivers in later life and their health effects are geriatric issues
that have several implications for social work practice, with special reference to life course
oriented social work practice, from gerontological social work to the child welfare system.
Evidence regarding family caregiving in later life urges social workers and related professionals
to reappraise the role of family and state in the health care system. Since the turn of the 21st
century, many debates and thoughts have occurred about a new partnership between informal
family caregiving and formal service provision (Zarit et al., 1992; Sundström, Johansson, &
Hassing, 2002; Bonsang, 2009). Litwin and Attias-Donfut (2009) found a shared,
complementary relationship between formal and informal care in France and Israel. The fact that
older family caregivers in the current study articulated their needs for family support and
professional services justifies the need for both formal and informal care. It is crucial to note that
need for formal services includes a greater governmental understanding and support of informal
caregiving.
On the other hand, findings about low utilization of family caregiving services should urge
gerontological social workers to reexamine current family caregiving programs and program

133

delivery systems. New information and educational programs are needed to disseminate related
policy and programs. It was evident in this study that every older adult has the chance to become
a family caregiver. Social workers have professional responsibilities to prepare the public, both
older adults and their families, about family caregiving, including informal caregiving and its
economic and social value.
Social workers may also want to pay closer attention to the effects of social networks for
family caregiving in later life as well as health behaviors and service utilization among older
family caregivers. For those caregivers with small available networks, information about the
health effects of supportive social networks can raise awareness of the importance of social
capital in aging. As regards caregivers with homogeneous networks, referral to informal and
formal supportive family caregiving programs could expand their perspective on network
resources. Harmfulness of network conflict in family caregiving expressed in this study also calls
for social workers to provide preventive intervention designed for family relations in caregiving
situations. Special attention needs to be paid to interventions that protect the social network
members from negative peer pressure and other downsides of social capital.
This study has indicated the significant effects of demographic characteristics—especially
age, gender, race, and SES—on the health of family caregivers in later life. Gerontological social
workers need to develop culturally sensitive family caregiving interventions to assist sustainable
and quality of family caregiving in diverse family and community groups. It is particularly
crucial to understand in the social and professional environment that family caregiving is tied to
there have been so many stigmas and stereotypes that it can be difficult to discuss family
caregiving beyond narrow interest areas, for instance, care burden or dementia care.
Gerontological social workers have special knowledge about family caregiving in later life. They

134

may help the public and other professionals realize the significance and value of family
caregiving. The fact that older family caregivers have so much faith in professional knowledge
might require social work and other professions to reach a whole new level of professional
practice.
Implications for policy
Community-based caregiving has been the center of American health care for decades. By
definition, family caregiving involves unpaid, community-dwelling, non-professionals, and
ranges from primary family caregivers to any network member providing certain care to a family
member, friend, or neighbor. This study confirmed that older adults have taken caregiving
responsibility to an extensive degree, no matter the proportion of older adults involving in unpaid
caregiving, the scope of their caregiving, or the length of caregiving. This study suggests the
need to use family caregiving in later life as a public policy analysis tool in future public policy
considerations. For example, legal status of family caregivers has not been changed in the
American public policy arena even though the importance of family caregiving has been evident
(Heller et al., 2007; Murray, 2008). As the data confirmed in this study, some family caregivers
did not identify themselves as caregivers as a result of this policy culture. Legal rights of family
caregivers involve many policy issues, including family leave, housing policies, and various
eligibility issues.
Recent policy changes have addressed issues related to grandparents raising grandchildren.
For example, the Fostering Connections to Success and Increasing Adoption Act of 2008
supports relative caregivers of children in foster care and improves incentives for adoption (P. L.
110―351). Obviously, this amendment focuses on grandparents raising grandchildren or other
relative caregivers of children. Additional people who provide care or who are able to provide

135

care are continuously excluded from policy design and service delivery. For example, siblings
have the longest-lasting relationship among older adults, and they often assume caregiving roles
without being recognized by society and/or their family. The focus of current federally funded
family caregiving support is limited to older adults and their parents and excludes sister and
brother caregivers. Future family caregiving policy should involve siblings more in policy
development, programmatic planning, services, supports, care planning, and discussion of rights
and supports of individuals with disability. More generally, social workers, including
gerontological social workers, should partner with family caregivers to re-conceptualize care
work in later life in public policy to advocate for and promote policy change (Daly, 2001;
Hinrichsen, 2010).
Conclusion
This study employed the positive gerontological framework to investigate the social value of
family caregivers in later life. A positive psychological framework was used to highlight
meaning and values of late family caregiving beyond caregiving burden and stress, with special
emphasis on strengths and capacities of older family caregivers and their social networks. By
defining family caregiving as unpaid, community-based, intergenerational caregiving for any kin
or non-kin person, this study departed from family caregiving literature that separated older
family caregivers by the informal care recipient’s disease, disability, or problem (Baker et al.,
2010; Dillenburger & McKerr, 2009; Standing et al., 2007). This study recognized family
caregivers’ commonalities in their contributions to both the care recipient and society, no matter
whether they were spousal caregivers of patients with Alzheimer’s disease, parental caregivers of
adult children with physical or mental disabilities, or grandparents raising grandchildren.
Although the scope of this study was limited to a certain level of analysis to examine the health

136

effects of social networks among older family caregivers, results indicated that positive
gerontology added new evidence to the study of family caregiving in later life.
Methodologically, the study investigated the health effects of social networks among older
family caregivers through a multi-site probability survey study with quantitative and qualitative
components. The study will contribute to geriatric social work research in two ways. First, a
multi-site community-based survey has allowed the research to reach older family caregivers
who have not previously utilized services. Three research sites maximized the comparisons of
diversity in family caregiving in later life in terms of race, ethnicity, and locale. Secondly, social
networks analysis is an emerging arena in family caregiving studies. The notion of “family
caregiving network” deconstructed the presumption that caregiving was an isolated and lonely
act. It suggested that family caregiving involved a web of social relations in which interactions
among different network members are sources of support and conflict, and strengths and
vulnerability. A social network analysis of family caregiving in later life paid direct attention to
caregiving relationships in social contexts. It was through such relationships that social networks
were transformed to influence the health and well-being of older family caregivers. The results
of this study recognized that the functional dimension rather than structural dimension was the
most determinative factor in social networks analysis of health aging among older family
caregivers.

137

APPENDICES

138

APPENDIX A

INITIAL IRB APPLICATION APPROVAL

139

Michigan State University
Initial IRB Application Approval

July 7, 2010
To: DeBrenna Agbényiga
254 Baker Hall
School of Social Work
Re: IRB# 10-528 Category: EXPEDITED 2-7
Approval Date: July 6, 2010
Expiration Date: July 5, 2011
Title: Effects of social networks on health of family caregivers in later life
The Institutional Review Board has completed their review of your project. I am pleased to
advise you that your project has been approved.
The committee has found that your research project is appropriate in design, protects the rights
and welfare of human subjects, and meets the requirements of MSU's Federal Wide Assurance
and the Federal Guidelines (45 CFR 46 and 21 CFR Part 50). The protection of human subjects
in research is a partnership between the IRB and the investigators. We look forward to working
with you as we both fulfill our responsibilities.
Renewals: IRB approval is valid until the expiration date listed above. If you are continuing your
project, you must submit an Application for Renewal application at least one month before
expiration. If the project is completed, please submit an Application for Permanent Closure.

140

Revisions: The IRB must review any changes in the project, prior to initiation of the change.
Please submit an Application for Revision to have your changes reviewed. If changes are made
at the time of renewal, please include an Application for Revision with the renewal application.
Problems: If issues should arise during the conduct of the research, such as unanticipated
problems, adverse events, or any problem that may increase the risk to the human subjects, notify
the IRB office promptly. Forms are available to report these issues.
Please use the IRB number listed above on any forms submitted which relate to this project, or
on any correspondence with the IRB office.
Good luck in your research. If we can be of further assistance, please contact us at 517-355-2180
or via email at IRB@msu.edu. Thank you for your cooperation.
Sincerely,
Gail M. Dummer, Ph.D.
SIRB Chair
c: Lihua Huang

141

APPENDIX B

INVITATION LETTER

142

July 12, 2010

Dear Mr. Smith,
We are writing to ask for your help for a study “Family caregiving and social networks.” We are
asking people like you to reflect on your experience as caregivers for your family member.
Your responses to this survey are very important and will help in advancing human services and
research in family caregivers and social networks.
Your participation is entirely voluntary and all of your responses will be kept confidential. No
personal identifiable information will be associated with your responses in any reports of this
data.
The survey will involve approximately one hour of your time. You may choose not to
participate, or choose not to answer a particular question. You may withdraw your participation
at any time. You are encouraged to ask the researchers any questions you may have.
If you decide to participate, we will be asking you to fill out the survey. There is no right or
wrong answers to the questions to be asked. The important thing is for you to share your
experience and opinions.
If you have any questions about this study, please contact Dr. DeBrenna Agbényiga, the
Responsible Project Investigator, School of Social Work, MSU, 104 Baker Hall, East Lansing,
MI 48824, USA, at 1-517-432-4459, fax: 1-517-353-3038, or email: agbenyi1@msu.edu. You
may also contact the director of MSU’s Human Research Protection Programs, Judy McMillan,
at 1-517-432-4502, Fax: 1-517-432-4503, or email mcmill12@ora.msu.edu, or regular mail at
202 Olds Hall, East Lansing, MI 48824.

143

We appreciate your time and consideration in completing the survey. Thank you for participating
in this study! It is only through the help of grandparents like you that we can provide information
to help human services and policy making related to health and well-being of grandparent
caregivers in Michigan.
Many thanks.
Lihua Huang
Study Coordinator
School of Social Work, Michigan State University

144

APPENDIX C

INFORMED CONSENT

145

A Research Study of Family Caregiving and Social Networks Survey
Consent Form
Hi, we are researchers from Michigan State University. We are conducting a study that involves
research about family caregiving and social networks. The study involves completing survey and
interviewing. You are invited to complete in this survey because you belong to one of three
groups: primary and secondary family caregivers (mostly family caregivers), care recipients of
family caregiving, and non-family caregivers. 3–5 of you may be selected for a follow-up study
that involves interviewing. No matter whether you have direct experience with family caregiving
or not, your response to this survey will help in advance in human services for family caregivers
in Michigan.
The survey will take about 15 minutes to complete. Your participation is entirely voluntary. You
may choose complete or not complete this survey. Refusal to participate will involve no penalty
or loss of benefits to which the subjects is otherwise entitled. You may choose not to answer any
question(s), and you may end your participation at any time. There is no penalty if you decide
that you do not want to complete some or all of the survey.
How will the information you provide be protected?
All of your responses will be kept confidential and your confidentiality will be protected to the
maximum extent allowable by law. Do not write your name on this form or anywhere on your
survey. We will keep all of the surveys in a locked cabinet at 104 Baker Hall, Michigan State
University for at least 3 years after the project closed. Only the two Principal Investigators and
authorized organization, including the MSU Institutional Review Boards, will have access to the
surveys.

146

When the results of the research are published or discussed at conferences, no information will
be included that would reveal your identity.
What are the risk and benefits of participation?
There are no known physical, social, legal or economic risks to you if you choose to participate.
Some of the questions may cause you some emotional discomfort if they remind you of
situations or people. Remember, your participation is voluntary and you may choose not to
participate at all, or you may skip questions or end your participation at any time. We will ensure
that no one else knows what you have responded in your survey.
You will not directly benefit from your participation in this study. However, your participation in
this study may contribute to the understanding of family caregiving and social networks.
Who should you contact if you have questions or concerns?
If you have concerns or questions about this study, such as scientific issues, how to do any part
of it, or to report an injury, please contact the Responsible Project Investigator: Dr. DeBrenna
Agbényiga at Michigan State University. Phone: 517-432-4459. Address: 104 Baker Hall, East
Lansing, MI 48824.
If you questions regarding your role and rights as a study participant, or would like to register a
complaint about this research study, or report a research-related injury (i.e. physical,
psychological, social, financial, or otherwise), you may contact, anonymously, if you wish the
Human Research Protection Program, at 517-355-2180, Fax 517-432-4503, or email
irb@msu.edu, or regular mail at HRPP, 207 Olds Hall, MSU, East Lansing, MI 48824.
You indicate your voluntary agreement to participate by completing and returning this survey.
THANK YOU FOR YOUR TIME AND HELP!

147

This consent form was approved by the Social Science/Behavioral/Education Institutional
Review Board (SIRB) at Michigan State University. Approved 07/06/10 – valid through
07/05/11. This version supersedes all previous versions. IRB # 10-528.

148

APPENDIX D

QUESTIONNAIRE

149

Family Caregiving and Social Networks
A Research Study to Understand Family Caregiving and Social Network in Michigan

To be completed by the adult age 64 and over in your household
School of Social Work
Michigan State University
East Lansing, MI 48442

150

Instructions for completing the survey
On this survey, you are asked to answer questions to help us understand your social networks,
health, and family caregiving. There are five sections. Please take the time to read and answer
each question carefully by filling in the blank or checking the box that best represents your
response. Please remember, your answers will not be shared with anyone. Your name or
identification will not be on this form at all.
-------------------------------------------------------------------------------------------SECTION ONE: SOCIAL NETWORKS
1. How many relatives/neighbors/friends do you see or hear from at least once a month?
0

1

2

3–4

5–9

9 and more

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

2. How often do you see or hear from them with whom you have the most contact?
Never

Seldom

Sometimes Often

Very Often Always

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

3. How many do you feel at ease with that you can talk about private matters?
0

1

2

3–4

5–9

9 and more

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

151

1. How many relatives/neighbors/friends do you feel close to such that you could call
on them for help?
0

1

2

3–4

5–9

9 and more

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

5. When one of your relatives/neighbors/friends has an important decision to make, how
often do they talk to you about it?
Never Seldom Sometimes Often Very Often

Always

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

6. How often is one of your relatives/neighbors/friends available for you to talk to when
you have an important decision to make?
Never

Seldom Sometimes

Often

Very Often

Always

Relatives

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Neighbors

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

Friends

[ ]

[ ]

[ ]

[ ]

[ ]

[ ]

7. Do you attend religious services?
[ ] No. Go to Question 8.

[ ] Yes

If yes, about how often have you attended services in the last four weeks?
[ ] More than once a week [ ] Once a week [ ] Twice [ ] Once [ ] None
8. Not including attendance at religious services, do you belong to national or local
community organizations?

152

[ ] No. Go to Question 9.

[ ] Yes

If yes, how many times did you participate in activities during the past year?
[ ] More than once a month

[ ] Once a month

[ ] 2 or 3 times a year

[ ] About once a year

[ ] Less than once a year

9. Do you belong to groups or clubs?
[ ] No. Go to Question 10.

[ ] Yes

If yes, how many times did you participate in activities during the past year?
[ ] More than once a month

[ ] Once a month

[ ] 2 or 3 times a year

[ ] About once a year

[ ] Less than once a year

10. Have you spent any time in the past 12 months doing volunteering work for religious,
educational, health-related or other charitable organizations?
[ ] Yes

[ ] No

11. Please check the life events that have occurred during the past two years. 0= never, 1=
once, 2= twice, and so on and so for.
1) Marriage
2) Marriage separation/divorce
3) Major illness of spouse
4) Death of spouse
5) Death of immediate family member
6) Retirement
7) Death of a close friend
8) Gain a new family member
9) Change in financial state
10) Change in residence

153

11) Change in social activities
12) Change in number of family get-togethers
13) Diagnosis of a severe disease of a loved one
14) Legal issue of a social network member
15) Beginning school, training, or other learning
-------------------------------------------------------------------------------------------SECTION TWO: HEALTH
1. In general, would you say your health is:
[ ] Excellent

[ ] Good

[ ] Neutral

[ ] Poor

[ ] Very Poor

2. Are you able to drive?
[ ] Yes
[ ] No
[ ] Never drove
If yes, do you limit your driving to nearby places, or do you also drive on
longer trips?
[ ] Limit to nearby
3. About how tall are you?
4. About how much do you weigh?

[ ] Drive long trips
Feet______

Inches_______

(lb)

5. Do you have any difficulty with these activities?
No Yes
Shopping for groceries ………………………………………………………

[ ] [ ]

Preparing a hot meal ………………………………………………………….

[ ] [ ]

Making phone calls ………………………………………………………….

[ ] [ ]

Taking medications if you needed to do so …………………………………

[ ] [ ]

154

No Yes
Housework ……………………………………………………………………

[ ] [ ]

Handling your money–e.g. paying bills and keeping tracking of expenses

[ ] [ ]

6. Do you take prescription drugs daily?
[ ] No. Go to Question 7.

[ ] Yes

If yes, how many prescription drugs you take daily?
(Write the number of drugs in the box.)
7. Has a doctor ever told you that you have any of the following conditions?
No

Yes

High blood pressure or hypertension …………………………………..

[ ]

[ ]

Diabetes or high blood sugar …………………………………………….

[ ]

[ ]

Cancer ……………………………………………………………………

[ ]

[ ]

Chronic lung disease such as asthma, chronic bronchitis or emphysema

[ ]

[ ]

Congestive heart failure ………………………………………………..

[ ]

[ ]

Stroke …………………………………………………………………..

[ ]

[ ]

Arthritis or rheumatism ………………………………………………

[ ]

[ ]

[ ]

[ ]

Heart attack, coronary heart disease, congestive heart failure or other heart
problems …………………………………………………………...
8. Have you fallen down in past five years?
[ ] No.

[ ] Yes

If yes, how many times have you fallen in past five years?
9. Are you taking any medication or treatment to control my condition?
No Yes
High blood pressure or hypertension

…………………………………..

Diabetes or high blood sugar ……………………………………………...
155

[ ] [ ]
[ ] [ ]

No Yes
Cancer …………………………………………………………………….

[ ] [ ]

Chronic lung disease such as asthma, chronic bronchitis or emphysema

[ ] [ ]

Congestive heart failure …………………………………………………

[ ] [ ]

Stroke ……………………………………………………………………

[ ] [ ]

Arthritis or rheumatism ………………………………………………...

[ ] [ ]

Heart attack, coronary heart disease, congestive heart failure or other heart
problems ………………………………………………………………..

[ ] [ ]

10. Have you ever
No Yes
Fractured your hip ………………………………………………………….

[ ][ ]

Had any of persistent back pain or problems ……………………………..

[ ][ ]

Had persistent headaches ……………………………………………………

[ ][ ]

Had persistent and severe fatigue or exhaustion …………………………..

[ ][ ]

11. How have you felt over the past week? Choose the best answer.
Yes

No

Am I basically satisfied with my life? ………………………………………

[ ]

[ ]

Have I dropped many of my activities and interests? ………………………

[ ]

[ ]

Do I often get bored? ……………………………………………………….

[ ]

[ ]

Am I in good spirits most of the time? …………………………………….

[ ]

[ ]

Am I afraid that something bad is going to happen to me? ………………..

[ ]

[ ]

Do I feel happy most of the time? …………………………………………

[ ]

[ ]

Do I often feel helpless? …………………………………………………

[ ]

[ ]

Do I prefer to stay at home, rather than going out and doing new things?

[ ]

[ ]

156

Yes

No

[ ]

[ ]

Do I think it is wonderful to be alive now? …………………………………. [ ]

[ ]

Do I feel pretty worthless the way I am now? ………………………………

[ ]

[ ]

Do I feel full of energy?

[ ]

[ ]

Do I feel that my situation is hopeless? …………………………………..

[ ]

[ ]

Do I think that most people are better off than I am? ……………………

[ ]

[ ]

Do I feel I have more problems with memory than most? ……………

……………………………………………………

-------------------------------------------------------------------------------------------SECTION THREE: HEALTH CARE
1. Do you have health insurance?
[ ] No
[ ] Yes
If yes, what kind of health insurance coverage? Choose applicable
[ ] Private insurance
[ ] Employment-based private insurance
[ ] Direct purchased private insurance
[ ] Medicaid
[ ] Medicare
[ ] Other government insurance_______________
Are you currently covered by the following Medicare plans? Choose applicable.
[ ] Yes, Part A of Medicare
[ ] Yes, Part B of Medicare
[ ] Yes, Part D of Medicare
[ ] Don’t know
What is the Medicaid plan you are covered?
157

(Write the name of the plan)

2. Are you receiving benefits from the Social Security Disability program?
[ ] No
[ ] Yes
If yes, when did you start to receive?
(MM/YYYY)
3. Are you receiving benefits from the Supplemental Security Income program?
[ ] No
[ ] Yes
If yes, when did you start to receive?
(MM/YYYY)
4. How many full meals do you eat daily?
[ ] One

[ ] Two

[ ] Three

[ ] Other_______

5. Do you eat or drink the following:
No

Yes

At least one serving of dairy products (milk, cheese, yogurt) per day

[ ]

[ ]

Two or more servings of beans or eggs per week ………………………..

[ ]

[ ]

Meat, fish, or poultry per day ……………………………………………….

[ ]

[ ]

6. Do you ever drink alcoholic beverages such as beer, wine, or liquor?
[ ] No
[ ] Yes
If yes, in the last four weeks, on average, how many days per week have
you had any alcohol to drink?

158

(1–7)

In the last four weeks, on the days you drink, about how many drinks do
you have?
(Write the number in the box.)
7. Have you ever smoked cigarettes? (By smoking we mean more than 100 cigarettes in your
lifetime. Do not include pipes or cigars.)
[ ] No
[ ] Yes
If yes, do you smoke cigarettes now?
[ ] Yes

[ ] No

10. Do you exercise regularly?
[ ] No
[ ] Yes
If yes, how many times do you do aerobic exercise each week? Examples of aerobic
exercises are walking, running, and biking.
[ ] Hardly once a week

[ ] Once a week

[ ] More than three times a week

[ ] Two–three times a week

[ ] Other______________________

How many times do you do strengthening and toning exercise each week? Examples
of strengthening and toning exercises are lifting and weights.
[ ] Hardly once a week

[ ] Once a week

[ ] More than three times a week

[ ] Two–three times a week

[ ] Other______________________

How many times do you do stretching and flexibility exercise each week? Examples
for stretching and flexibility exercises are lifting and weights.
[ ] Hardly once a week

[ ] Once a week

[ ] More than three times a week

[ ] Two–three times a week

[ ] Other______________________
159

SECTION FOUR: FAMILY CAREGIVING
1. Do you consider yourself a caregiver for or a care receiver from a loved one (a family
member, relative, friend, neighbor or other)?
[ ] Both [ ] Caregiver

[ ] Care receiver

[ ] Neither. Go to Page 11.

2. Over the past 30 days, have you regularly helped a loved one or been helped by a loved
one with following activities? Check all that apply.
No

Yes

Getting in and out of beds and chairs ……………………..

[ ]

[ ]

Getting dressed …………………………………………..

[ ]

[ ]

Getting to and from the toilet …………………………….

[ ]

[ ]

Bathing or showering ……………………………………

[ ]

[ ]

Dealing with incontinence or diapers …………………..

[ ]

[ ]

Feeding …………………………………………………

[ ]

[ ]

Managing medicines, pills, injections …………………..

[ ]

[ ]

3. Over the past 30 days, have you regularly helped a loved one or been helped by a loved
one with following activities? Check all that apply.
No

Yes

Managing finances ………………………………………….

[ ]

[ ]

Grocery shopping ……………………………………………

[ ]

[ ]

Housework …………………………………………………

[ ]

[ ]

Preparing meals ………………………………………………

[ ]

[ ]

Transportation …………………………………………………

[ ]

[ ]

Arranging or supervising services ……………………………

[ ]

[ ]

160

4. Who are you currently caring for or cared by? Choose all applicable. Writing down 2 after
“sibling” if you care for two siblings, for example.
Grandparent/Parent/step parent

Spouse/ Partner

Sibling/step sibling

Adult child/step child

Grandchild/Great grandchild

In-law

Friend

Neighbor

5. How long have you been caring for or cared by the loved one? Choose the longest time if
you care for more than one.
[ ] Less than 6 month
[ ] 3 or 4 years

[ ] 6–11 months
[ ] 5–9 years

[ ] 1 or 2 years
[ ] 10 years or more]

[ ] Other, specify _______________________________________
6. What brought him/her or you into the care? Choose all applicable.
[ ] Absence or unavailability of birth parents of my grandchild
[ ] Chronic ill, specify____________ [ ] Disability, specify_______________
7. How many hours did you care for him/her or were you cared last week?
[ ] Less than 9 hours

[ ] 9–20 hours

[ ] 21–40 hours

[ ] Over 40 hours

8. Except professionals, are you the only person caring for him/her or are you cared by
only one family member/relative/friend/neighbor?
[ ] Yes. Go to Question 9.
[ ] No.
If no, who are the other people providing cares? Choose all applicable
[ ] Parent/step parent

[ ] Spouse/ Partner [ ] Sibling/step sibling

[ ] Adult child/step child [ ] In-law

[ ] Friend

[ ] Grandchild/Great grandchild

[ ] Neighbor
161

9. Where does the loved one live currently? Choose all applicable.
[ ] Live in the same household with me
[ ] Live less than 20 minutes away [ ] Live between 20 to 60 minutes away
[ ] Live between 1 to 2 hours away [ ] Live more than 2 hours away
10. Have you used human and health services for family caregivers in the past year? If your
are not a caregiver, skip Question 10-14, and go to Page 11.
[ ] No. Go to Question 11.
[ ] Yes
If yes, what services have you used? How many times have you used them in the last year?
Are you satisfied with them?
Service Use

Times Used

Satisfaction

No=0 Yes=1

Last Year

No=0 Yes=1

Personal care …………….

[ ]

[

]

[ ]

Home making...……………

[ ]

[

]

[ ]

Home delivered meals ….

[ ]

[

]

[ ]

Adult day care ……………

[ ]

[

]

[ ]

Case management ………….

[ ]

[

]

[ ]

Congregate meal ………….

[ ]

[

]

[ ]

Nutrition counseling …….

[ ]

[

]

[ ]

Psychosocial services……..

[ ]

[

]

[ ]

Transportation ……………..

[ ]

[

]

[ ]

Legal assistance .....................

[ ]

[

]

[ ]

Nutrition education ……..

[ ]

[

]

[ ]

Information & assistance .

[ ]

[

]

[ ]

162

Service Use

Times Used

Satisfaction

No=0 Yes=1

Last Year

No=0 Yes=1

Home environmental skill
building …………………

[ ]

[

]

[ ]

Outreach ……………………

[ ]

[

]

[ ]

Support group ………………

[ ]

[

]

[ ]

Recreation/social center

[ ]

[

]

[ ]

Medical education program

[ ]

[

]

[ ]

Other________________

[ ]

[

]

[ ]

11. Have you ever received services through the internet?
[ ] No

[ ] Yes

If Yes, have you received help from relatives, friends, neighbors or professionals to use
internet services? Choose applicable.
[ ] Relatives

[ ] Friends

[ ] Neighbors

[ ] Professionals

12. How did you know of these services?
[ ] From friends

[ ] From neighbors

[ ] From relatives

[ ] From physicians/nurses [ ] From social workers [ ] From media
[ ] From other professionals

[ ] I never know them available

13. Rank these changes you have experienced since serving in the family caregiving role,
using a scale of -3 to 3 with -3 = changed for the worst and 3 = changed for the best.
-3

-2

Changed for the worst

-1

0

1

No change

2

3

Changed for the best

[ ] My self-esteem

163

[ ] My physical health
[ ] My time for family activities
[ ] My ability to cope with stress
[ ] My relationship with friends
[ ] My future outlook
[ ] My emotional well-being
[ ] My time for social activities friends
[ ] My relationships with family
[ ] My relationship with the loved one I care for
14. Your answers to the following two questions are very important for understanding what
family caregiving brings to you and your family.
1) How does caregiving make you feel?

2) In general, how has your life changed as a result of family caregiving?

164

SECTION FIVE: DEMOGRAPHICS
1. What is your gender?

[ ] Male

[ ] Female

2. What is your race?
[ ] Caucasian/White, non-Hispanic
[ ] Black/African American
[ ] American Indian and Alaska Native
[ ] Asian (Asian Indian, Chinese, Filipina, Japanese, Korean, Vietnamese, etc.)
[ ] Native Hawaiian and other Pacific Islander
[ ] Multiracial
3. What is your date of birth? ____ / ____ / _______(DD / MM / YYYY)
4. Were you born in the United States?

[ ] No

[ ] Yes

5. Are you Hispanic or Latino?

[ ] Yes

[ ] No

6. What is your current marital status?
[ ] Single, Never married

[ ] Married/live with a partner

[ ] Separated

[ ] Divorced

[ ] Widowed
How many years and months have you been widow/widower?
(Year & month)

8. What is the highest education you have completed?
[ ] None

[ ] Less than High School Diploma

[ ] A High School Diploma/GED

[ ] Some College, no Degree

[ ] Associate degree

[ ] Bachelor’s Degree

[ ] Master’s Degree

[ ] Post Master’s

9. What is your employment status? Choose all applicable.
165

[ ] Retired

[ ] Part-time employed

[ ] Full-time employed

[ ] Unemployed and looking for work

[ ] On disability

[ ] Temporarily laid off

[ ] On sick leave

[ ] Homemaker

[ ] Student

[ ] Other_______________

Did you retire because of an ongoing health problem, impairment or disability?
[ ] No

[ ] Yes

10. Where do you live?
[ ] House

[ ] Mobile home

[ ] Condo

[ ] Apartment

[ ] Farm/Ranch

11. What is your total household income (Including all earners and benefits in your
household)?
[ ] Less than $25,000

[ ] $25,000–$49,999

[ ] $50,000–$74,999

[ ] $75,000–$99,999

[ ] $100,000 and more
12. Not including you, how many people currently live in the residence?
(Write down the total number of people living in the household with you.)
If one or more, who are they? Check all applicable.
Spouse/partner

Parent

Child/step child

Grandchild

Sibling

In-law

Other___________________
13. Do you and your family speak a language other than English at home?
[ ] Yes
[ ] No. The end of the Section.
166

If yes, what language is it?

Thank you for taking the time to complete our survey. Your contribution is greatly appreciated.
Please use the pre-stamped envelope to return your survey.
Thank You!

167

APPENDIX E

DEMOGRAPHIC COMPARISON OF RESEARCH SITES

168

Table A1
Demographic Comparison of Research Sites

48197
YT

49506
GR

49874
PV

Metropolitan

Suburban

Rural

Region

Southeast

Southwest

Upper Peninsula

County

Washtenaw

Kent

Menominee

Zip Code Tabulation Areas
Residence

Total 64 and over
Aging (%)
Race (Non-white, %)

3,684

2,715

6.1

115

13.3

.347

.254

Source: U.S. Census 2000.

169

27.5
.020

REFERENCES

170

REFERENCES

Addams, J. (1910). Twenty years at Hull-House with Autobiographical notes. New York: The
Macmillan Company.
Agency for Health care Research and Quality (AHRQ). (2001). The Characteristics of Longterm Care Users. AHRQ Publication No. 00-0049. Rockville, MD: Agency for Health care
Research and Quality. Retrieved from
http://www.ahrq.gov/RESEARCH/ltcusers/ltcuse1.htm
Albert, S. M., & Levine, C. L. (2005). Family caregiver research and the HIPAA factor. The
Gerontologist, 45(4), 432–437.
Alwin, D. F., Converse, P. E., Martin, S. S. (1985). Living arrangements and social integration.
Journal of Marriage and Family, 47(2), 319–334.
Arno, P. S., Levine, C., & Memmott, M. M. (1999). The economic value of informal caregiving.
Health Affairs, 18(2), 182–188.
Aschbrenner, K. A., Greenberg, J. A., Allen, S. M., & Seltzer, M. M. (2010). Subjective burden
and personal gains among older parents of adults with serious mental health. Psychiatric
Services, 61(6), 605–611.
Bakas, T., & Champion, V. (1999). Development and psychometric testing of the Bakas
Caregiving Outcomes Scale. Nursing Research, 48(5), 250–259.
Baker, K. L., Robertson, N., & Connelly, D. (2010). Men caring for wives or partners with
dementia: masculinity, strain and gain. Aging and Mental Health, 14(3), 319–327.
Balducci, C., Minich, E., McKee, K. J., Lamura, G., Beckmann, A., Krevers, B., et al. (2008).
Negative impact and positive value in caregiving: Validation of the COPE Index in a sixcountry sample of careers. The Gerontologist, 48(3), 276–286.
Barker, J. C. (2002). Neighbors, friends, and other non-kin caregivers of community-living
dependent elders. Journal of Gerontology, 57B(3), S158–S167.
Barrett, A. E., & Lynch, S. M. (1999). Caregiving networks of elderly persons: Variation by
marital status. The Gerontologist, 39(6), 695–704.
Beach, S. R., Schulz, R., Williamson, G. M., Miller, S., Weiner, M. F., & Lance, C. E. (2005).
Risk factors for potentially harmful informal caregiver behavior. Journal of American
Geriatrics Society, 53(2), 255–261.
Beeber, A. S., Thorpe, J. M., & Clipp, E. C. (2008). Community-based service use by elders with
dementia and their Caregivers: A latent class analysis. Nursing Research, 57(5), 312–321.

171

Beeson, R. A. (2003). Loneliness and depression in spousal caregivers of these with Alzheimer’s
disease versus non-caregiving spouses. Achieves of Psychiatric Nursing, 17, 135–143.
Berkman, L. F. (1984). Assessing the physical health effects of social networks and social
support. Annual Review of Public Health, 5, 413–432.
Berkman, L. F., Glass, T., Brissette, I., & Seeman, T. E. (2000). From social integration to
health: Durkheim in the new millennium. Social Science & Medicine, 51(6), 843–857.
Bergman, E. J., & Haley, W. E. (2009). Depressive symptoms, social network, bereavement
service utilization and preferences among spouses of former hospice patients. Journal of
Palliative Medicine, 12(2), 170–176.
Berkman, L. F., & Syme, S. L. (1979). Social networks, host resistance, and mortality: a nineyear follow-up study of Alameda County residents. American Journal of Epidemiology,
109(2), 186–204.
Berks, J., & McCormich, R. (2008). Screening for alcohol misuse in elderly primary care
patients: A systematic literature review. International Psychogericatics, 20(6), 1090–1103.
Berrick, J. D. (2008). From private to public: Paying grandparents as caregivers. In J. D. Berrick
& N. Gilbert (Eds.) Raising children: Emerging needs, modern risks, and social responses
(pp. 27–43). New York, NY: Oxford University Press.
Blazer, D. G. (1982). Social support and mortality in an elderly community population.
American Journal of Epidemiology, 115 (5), 684–694.
Boden-Albala, B., Litwak, E., Elkind, M. S. V., Rundek, T., & Sacco, R. L. (2005). Social
isolation and outcomes post stroke. Neurology, 64, 1888–1892.
Bonsang, E. (2009). Does informal care from children to their elderly parents substitute for
formal care in Europe? Journal of Health Economics, 28(1), 143–154.
Bott, E. (1957). Family and social network. London: Tavistock.
Braddock, D. L. (2002). Disability at the dawn of the 21st century and the state of the states.
Washington, DC: American Association on Mental Retardation.
Braddock, D., Hemp, R., & Rizzolo, M. (2008). The state of the states in developmental
disabilities. Boulder: Department of Psychiatry and Coleman Institute for Cognitive
Disabilities, University of Colorado.
Brodaty, H., Thomson, C., Thompson, C., & Fine, M. (2005). Why caregivers of people with
dementia and memory loss don’t use services. International Journal of Geriatric Psychiatry,
20, 537–546.

172

Brothers, D., & de Jong Gierveld, J. (2011). Challenges and opportunities for relationships in the
era of the third age. In D. C. Carr & K. Komp (Eds.) Gerontology in era of the third age:
Implications and next steps (pp. 189–205). New York: Springer Publishing Company.
Brown, J., & Chen, S. (2008). Help-seeking patterns of older spousal caregivers of older adults
with dementia. Issues in Mental Health Nursing, 29(8), 839–852.
Brown, J. W., Chen, S., Mitchell, C., & Province, A. (2007). Help-seeking by older husbands
caring for wives with dementia. Journal of Advanced Nursing, 59(4), 352–360.
Brown, S. L. (2007). Health effects of caregiving: Studies of helping behavior needed!
Alzheimer’s Care Today, 8(3):235–46.
Brown, S. L, Nesse, R. M., Vinokur, A. D., & Smith, D. M. (2003). Providing social support
may be more beneficial than receiving it: results from a prospective study of mortality.
Psychological Science, 14(4):320–7.
Brown, S. L., Smith, D. M., Schulz, R., Kabeto, M. U., Ubel, P. A., Poulin, M., et al. (2009).
Caregiving behavior is associated with decreased mortality risk. Psychological Science,
20(4), 488–494.
Burns, C. M., Abernethy, A. P., Leblanc, T. W., & Currow, D. C. (2011). What is the role of
friends when contributing care at the end of life? Findings from an Australian population
study. Psycho-Oncology, 20(2), 203–212.
Calasanti, T. M., & Slevin, K. (2001). Gender, care work, and family in old age. In T. M.
Clasanti & K. Slevin Gender, social inequalities, and aging (pp. 143–178). NY: Altamira
Press.
Caldwell, J. (2008). Health and access to health care of female family caregivers of adults with
developmental disabilities. Journal of Disability Policy Studies, 9(2), 68–79.
Campbell, L. D., Carroll, M. P. (2007). The incomplete revolution: Theorizing gender when
studying men who provide care to aging parents Men and Masculinities 9(4), 491–508.
Campbell, L. D., & Martin-Matthew, A. (2003). The gendered nature of men's filial care.
Journal of Gerontologist, 58B(6), S350–S358.
Cantor, M. H. (2001). Family and community: Changing roles in an aging society.
Gerontologist, 31(3), 337–346.
Carpentier, N., & Ducharme, F. (2003). Care-giver network transformations: The need for an
integrated perspective. Ageing & Society, 23, 507–525.
Carpentier, N., & Ducharme, F. (2005). Support network transformations in the first stages of the
caregiver’s career. Qualitative Health Research, 15(3), 289–311.

173

Carpentier, N., & Ducharme, F. (2007). Social network data validity: The example of the social
network of caregivers of older persons with Alzheimer-type dementia. Canadian Journal on
Aging, 26(Suppl. 1), 103–116.
Carpiano, R. M. (2008). Actual or potential neighborhood resources and access to them: Testing
hypotheses of social capital for the health of female caregivers. Social Science & Medicine
67, 568–582.
Carr, D. C., & Hendricks, J. (2011). Relevance of social capital and lifestyle for the third age. In
D. C. Carr & K. Komp (Eds.) Gerontology in era of the third age: Implications and next
steps (pp. 207–223). New York: Springer Publishing Company.
Carter, P. A., & Clark, A. P. (2005). Assessing and treating sleep problems in family caregivers
of intensive care unit patients. Critical Care Nurse, 25(1), 16–23.
Castro, C. M., King, A. C., Housemann, R., Bacak, S. J., McMullen, K. M., & Brownson, R. C.
(2007). Rural family caregivers and health behaviors: Results from an epidemiologic survey.
Journal of Aging and Health, 19(1), 87–105.
Chappell, N., & Blandford, A. (1991). Informal and formal care: Exploring the complementarity.
Ageing and Society, 11(3), 299–317.
Chatters, L. M., Taylor, R. J., & Jackson, J. S. (1985). Size and Composition of the Informal
Helper Networks of Elderly Blacks. Journal of Gerontology, 40 (5), 605–614.
Cheng, S., Lee, C. K. L., Chan, A. C. M., Leung, E. M. F., & Lee, J. (2009). Social network
types and subjective well-being in Chinese older adults. Journal of Gerontology:
Psychological Sciences, 64B(6), 713–722.
Choi, N. G., Burr, J. A., Mutchler, J. E., & Caro, F. G. (2007). Formal and informal volunteer
activity and spousal caregiving among older adults. Research on Aging, 29(2), 99–124.
Chou, K. , Yeung, S., & Chi, I. (2001). Does physical distance make a difference in caregiving?
Journal of Gerontological Social Work, 35(1), 21–37.
Christakis, N. A., & Fowler, J. H. (2007). The spread of obesity in a large social network over 32
years. The New England Journal of Medicine, 357(4), 370–379.
Clay, O. J., Roth, D. L., Wadley, V. G., & Haley, W. E. (2008). Changes in social support and
their impact on psychosocial outcome over a 5-year period for African American and White
dementia caregivers. International Journal of Geriatric Psychiatry, 23(8), 857–862.
Cohen, G. D. (2006). Research on creativity and aging: The positive impact of the arts on health
and illness. Generations, 30(1), 7–15.
Cohen, J. (2nd Ed.) (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ:
Lawrence Erlbaum Associates.

174

Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin, 98(2), 310–357.
Connidis, I.A. (2001). Family ties and aging. Thousand Oaks, CA: Sage.
Cooper, C., Selwood, A., Blanchard, M., Walker, Z., Blizard, R., & Livingston, G. (2009).
Abuse of people with dementia by family carers: Representative cross sectional survey.
British Medical Journal, 338(b115), 1–3.
Copen, C. E. (2006). Welfare reform: Challenges for grandparents raising grandchildren. Journal
of Aging and Social Policy, 18(3/4), 193–208.
Cornwell, B. (2009). Network bridging potential in later life: Life-course experiences and social
network position. Journal of Aging and Health, 21(1), 129–154.
Cornwell, B., Schumm, L. P. , Laumann, E. O. , & Graber, J. (2009). Social networks in the
NSHAP study: rationale, measurement, and preliminary findings. Journal of Gerontology,
64B(SI), I47–I55.
Cornwell, E. Y., & Waite, L. J. (2009). Social disconnectedness, perceived isolation, and health
among older adults. Journal of Health and Social Behavior, 50(1), 31–48.
Corsentino, E. A., Molinari, V., Gum, A. M., Roscoe, L. A., & Mills, W. L. (2008). Family
caregivers' future planning for younger and older adults with serious mental illness (SMI).
Journal of Applied Gerontology 27(4), 466–485.
Costa, D. L. (1999). A house of her own: Old age assistance and the living arrangements of older
nonmarriaged women. Journal of Public Economics, 72(1), 39–59.
Coward, R. T., & Dwyer, J. W. (1990). The association of gender, sibling network composition,
and patterns of parent care by adult children. Research on Aging, 12(2), 158–181.
Cox, C. (1999). Race and caregiving. Journal of Gerontological Social Work, 32(2), 5–19.
Cox, E. O. (1988). Social services and intergenerational caregivers: issues for social work. Social
Work, 33 (5), 430–434.
Crooks, V. C., Lubben, J., Petitti, D. B., Little, D., & Chiu, V. (2008). Social network, cognitive
function, and dementia incidence among elderly women. American Journal of Public Health,
98(7), 1221–1227.
Czaja, S. J., & Rubert, M. P. (2002). Telecommunications Technology as an Aid to Family
Caregivers of Persons with Dementia. Psychosomatic Medicine, 64, 469–476.
Daly, M. (Ed.). (2001). Care work: The quest for security. Geneva: International Labour Office.

175

Dillenburger, K., & McKerr, L. (2011). How long are we able to go on?’ Issues faced by older
family caregivers of adults with disabilities. British Journal of Learning Disabilities, 39(1),
29–38.
Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail, and mixed mode survey:
The tailored design method. New Jersey, Hoboken, John Wiley & Sons.
Dilworth-Anderson, P., Williams, I. C., & Gibson, B. E. (2002). Issues of race, ethnicity, and
culture in caregiving research: A 20-year review (1980–2000). Gerontologist, 42(2), 237–
272.
Donovan-Kicken, E., & Bute, J. J. (2008). Uncertainty of social network members in the case of
communication-debilitating illness or injury. Qualitative Health Research, 18(1), 5–18.
Dooley, W. K., Shaffer, D. R., Lance, C. E., & Williamson, G. M. (2007). Informal care can be
better than adequate: Development and evaluation of the Exemplary Care Scale.
Rehabilitation Psychology, 52(4), 359–369.
Dunne, E., Wrosch, C., & Miller, G. E. (2011). Goal disengagement, functional disability, and
depressive symptoms in old age. Health Psychology, 30(May 23), 1–8.
Eaves, Y. D. (2006). Caregiving in rural African American families for older adults stroke
survivors. Journal of Neuroscience Nursing, 38(4), 270–282.
Edwards, E., Higgins, D. J., & Zmijewski, N. (2007). The families caring for a person with a
disability study and the social lives of carers. Family Matters, 76, 8–17.
Edwards, J. R., & Cooper, C. L. (1988). The impacts of positive psychological states on physical
health: a review and theoretical framework. Social Science and Medicine, 27(12), 1447–59.
Edwards, O. W. (2003). Living with grandma: A grandfamily study. School Psychology
International 24(2), 204–218.
Edwards, P., Roberts, I., Clarke, M., DiGuiseppi, C., Pratrap, S., Wentz, R., et al. (2002).
Increasing response rates to postal questionnaires: systematic review. British Medical
Journal, 324 (7347), 1183–1185.
Ekwall, A., Sivberg, B., & Hallberg, I. R. (2004). Dimensions of informal care and quality of life
among elderly family caregivers. Scandinavian Journal of Caring Sciences, 18(3), 239–248.
Ekwall, A., Sivberg, B., & Hallberg, I. R. (2005). Loneliness as a predictor of quality of life
among older caregivers. Journal of Advanced Nursing, 49(1), 23–32.
Elmore, D. L., & Talley, R. C. (2009). Family caregiving and U.S. federal policy. In S. H. Qualls
& S. H. Zarit (Eds) Aging family and caregiving (pp. 209–231). Hoboken, NJ: John Wiley
& Sons.

176

Emlet, C. A. (2006). An examination of the social networks and social isolation in older and
younger adults living with HIV/AIDS. Health and Social Work, 31(4), 299–308.
Eng, P. M., Rimm, E. B., Fitzmaurice, G., & Kawachi, I. (2002). Social ties and change in social
ties in relation to subsequent total and cause-specific mortality and Coronary heart disease
incidence in men. American Journal of Epidemiology, 155(8), 700–709.
Estes, C. L. (1999). Critical gerontology and the new political economy of aging. In M. Minkler
& C. L. Estes (Eds.) Critical gerontology: Perspectives from political and moral economy
(pp. 17–35). Amityville, NY: Baywood Publishing Company.
Evercare & National Alliance for Caregivers. (2009). The Evercare Survey of the economic
downturn and its impact on family caregiving: Report findings. Retrieved from
http://www.caregiving.org/data/EVC_Caregivers_Economy_Report%20FINAL_4-28-09.pdf
Fast, J., Keating, N., Otfinowski, P., & Derksen, L. (2004). Characteristics of family/friend care
networks of frail seniors. Canadian Journal on Aging, 23(1), 5–19.
Feinberg, L. F., & Newman, S. L. (2006). Preliminary experiences of the States in implementing
the National Family Caregiver Support Program: A 50-State Study. Journal of Aging and
Social Policy, 18(3/4), 95–113.
Fengler,A. P., & Goodrich, N. (1979). Wives of elderly disabled men. The hidden patients. The
Gerontologist, 19, 175–183.
Fingerman, K. L., Miller, L. M., & Seidel, A. J. (2009). Functions families serve in old age. In S.
H. Qualls & S. H. Zarit (Eds.) Aging families and caregiving (pp. 19–43). Hoboken, NJ: John
Wiley & Sons.
Fiori, K. L., Smith, J., & Antonucci, T. C. (2007). Social network types among older adults: A
multidimensional approach. Journal of Gerontology, 62B(6), 322–330.
Fortinsky, R. H., Tennen, H., Frank, N., & Affleck, G. (2007). Health and psychological
consequences of caregiving. In C. M. Aldwin, C. L. Park, & A. Spino III (Eds.) Handbook of
health psychology and aging (pp. 227–249). New York: The Guilford Press.
The Fostering Connections to Success and Increasing Adoption Act 0f 2008. P.L. 110-351.
Fredrickson, B. L. (2004). The broaden-and-build theory of positive emotions. Philosophical
Transactions of the Royal Society B: Biological Sciences, 359(1449), 1367–1378.
Fujiura, G. T. (2010). Aging families and the demographics of family financial support of adults
with disabilities. Journal of Disability Policy Studies, 20(4), 241–250.
Fuller-Thomson, E. & Minkler, M. (2003). Housing issues and realities facing grandparent
caregivers who are renters. The Gerontologist, 43(1), 92–98.

177

Fung, H. H., Yeung, D. Y., Li, K., & Lang, F. R. (2009). Benefits of Negative Social Exchanges
for Emotional Closeness. Journal of Gerontology, 64B(5), 612–621.
Furstenberg, F. F. (2005). Banking on families: How families generate and distribute social
capital. Journal of Marriage and Family, 67, 809–821.
Gable, S. L., & Haidt, J. (2005). What ( and why) is positive psychology? Review of General
Psychology, 9(2), 103–110.
Gerard, J. M. Landrey-Meyer, L. & Guzell, J. (2006). Grandparents raising grandchildren: The
role of social support in coping with caregiving challenges. The International Journal of
Aging and Human Development, 62(4), 359–383.
Gergen, M., Gergen, K. J. (2003). Positive aging. In J. F. Gubrium & J. A. Holstein Ways of
aging (pp. 205–224). Malden, MA: Niley-Bleckwell.
Ghosh, S., & Greenberg, J. (2009). Aging fathers of adult children with schizophrenia: The toll
of caregiving on their mental and physical Health. Psychiatric Services, 60(7), 982–984.
Gideon, C. A. (2007). Social environments of dementia caregivers: Relationships between social
support, negative social interactions, and caregiver emotional distress. Unpublished
dissertatin at Case Western Reserve University.
Gilley, D. W., McCann, J. J., Bienias, J. L., & Evans, D. A. (2005). Caregiver psychological
adjustment and institutionalization of persons with Alzheimer’s Disease. Journal of Aging
and Health, 17(2), 172–189.
Given, B., Sherwood, P. R., & Given, C. W. (2008). What knowledge and skills do caregivers
need? Journal of Social Work Education, 44(3), 115–123.
Glasgow, N. (2000). Rural/urban patters of aging and caregiving in the United States. Journal of
Family Issues, 21(5), 611–631.
Glueckauf, R. L., Ketterson, T. U., Loomis, J. S., & Dages, P. (2004). Online support and
education for dementia caregivers: Overview, utilization, and initial program evaluation.
Telemedicine Journal of e-Health, 10(2), 223–232.
Goins, R. T., Spencer, S. M., & Byrd, J. C. (2009). Research on Rural Caregiving: A Literature
Review. Journal of Applied Gerontology, 28(2), 139–170.
Gopalan, N., & Brannon, L. A. (2006). Increasing family members' appreciation of family
caregiving stress. The Journal of Psychology, 140(2), 85–94.
Goodman, C. C., Potts, M. K. & Pasztor, E. M. (2007). Caregiving grandmothers with vs.
without child welfare system involvement: Effects of expressed need, formal services, and
informal social support on caregiver burden. Children and Youth Services Review, 29(4),
428–441.

178

Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6),
1360–1380.
Granovetter, Mark S. (1985). Economic Action and Social Structure: The Problem of
Embeddedness. The American Journal of Sociology, 91(3), 481–510.
Grant, J., Ramcharan, P., McGrath, M., Nolan, M., & Keady, J. (1998). Rewards and
gratifications among family caregivers: towards a refined model of caring and coping.
Journal of Intellectual Disability Research, 42(1), 58–71.
Greenberg, J. A., Seltzer, M. M., Krauss, M. W., & Kim, H. (1997). The differential effects of
social support on the psychological well-being of aging mothers of adults with mental illness
or mental retardation. Family Relations, 46(4), 383–394.
Greenberg, J. S., Seltzer, M. M., & Greenley, J. R. (1993). Aging parents of adults with
disabilities: The gratifications and frustrations of later-life caregiving. The Gerontologist,
33(4), 542–550.
Haber, C., & Gratton, B. (1994). Old Age and the Search for Security: An American Social
History. Bloomington: Indiana University Press.
Haley, W. E., Roth, D. L., George Howard, G., & Safford, M. M. (2010). Caregiving Strain and
Estimated Risk for Stroke and Coronary Heart Disease among Spouse Caregivers:
Differential Effects by Race and Sex. Stroke. 41(2), 331.
Haley, W. E., West, C. A. C., Wadley, V. G., Ford, G. R., White, F. A., Barrett, J. J., et al.
(1995). Psychological, Social, and Health Impact of Caregiving: A Comparison of Black and
White Dementia Family Caregivers and Noncaregivers. Psychology and Aging, 10(4), 540–
552.
Harrits, G. S. (2011). More than method? A discussion of paradigm differences within mixed
methods research. Journal of Mixed Methods Research, 5(2), 150–166.
Harrow, B. S., Mahoney, D. F., Mendelsohn, A. B., Ory, M. G., Coon, P. W., Belle, S. H. et al.
(2004). Variation in cost of informal caregiving and formal-service use for people with
Alzheimer’s disease. American Journal of Alzheimer’s Disease and other Dementia, 19(5),
299–308.
Harwood, D. G., Ownby, R. L., Burnett, K., Barker, W. W., & Duara, R. (2000). Predictors of
appraisal and psychological well-being in Alzheimer’s disease family caregivers. Journal of
Clinical Geropsychology, 6, 279–297.
Hayslip, B., Baird, A. H., Toledo, R., Toledo, C., & Emick, M. A. (2006). Cross-cultural
differences in traditional and custodial grandparenting: A qualitative approach. In B. Hayslip
& J. Hicks-Patrick (Eds.), Custodial grandparenting: Individual, cultural, and ethnic
diversity (pp. 169–181). New York: Springer.

179

Hayslip, B., & Kaminski, P. L. (2005). Grandparents Raising Their Grandchildren: A Review of
the Literature and Suggestions for Practice. The Gerontologist, 45(2), 262–269.
Hayward, L., Davies, S., Robb, R., Denton, M., & Auton, G. (2004). Publicly Funded and
Family-Friend Care in the Case of Long-Term Illness: The Role of the Spouse. Canadian
Journal on Aging, 23(Suppl. 1), S39–S47.
Heller, T., Caldwell, J., & Factor, A. (2007). Aging family caregivers: Policies and practices.
Mental Retardation and Developmental Disabilities Research Reviews, 13(2), 136–142.
Heller, T., & Kramer, J. (2009). Involvement of Adult Siblings of Persons With Developmental
Disabilities in Future Planning. Intellectual and Developmental Disabilities, 47(3), 208–219.
Hibbard, J., Neufeld, A., & Harrison, M. J. (1996). Gender differences in the support networks of
caregivers. Journal of Gerontological Nursing, 22(9), 15–23.
Hilgeman, M., M., Durkin, D. W., Fei, S., DeCoster, J., Allen, R. S., Gallagher-Thompson, D. et
al. (2009). Testing a theoretical model of the Stress Process in Alzheimer’s caregivers with
race as a moderator. Gerontologist, 49(2), 248–261.
Hinrichsen, G. A. (2010). Public policy and the provision of psychological services to older
adults. Professional Psychology: Research and Practice, 41(2), 97–103.
Hinterlong, J. E., Morrow-Howell, N., & Rozario, P. A. (2007). Productive engagement and late
life physical and mental health: Findings from a nationally representative panel study.
Research on Aging, 29, (4), 348–370.
Ho, C. J., Weitzman, P. F., Cui, X. & Levkoff, S. E. (2000). Stress and service use among
minority caregivers to elders with dementia. Journal of Gerontological Social Work, 33(1),
67–88.
Holmes, T. H., & Rahe, R. H. (1967). The social readjustment rating scale. Journal of
Psychosomatic Research, 11(2), 213–218.
Holstein, M. B., & Minkler, M. (2007). Critical gerontology: Reflections for the 21st century. In
M. Bernard & T. Scharf (Eds.) Critical perspectives on ageing societies (pp. 13–26). Bristol,
PA: The Policy Press.
Holzman, D. (1998). The place of filial piety in ancient China. Journal of the American Oriental
Society. 118(2), 185–199.
Houser, A. N., & Gibson, M. J. (2008). Valuing the invaluable: The economic value of family
caregiving, 2008 update. Insight on the Issues 13. AARP Public Policy Institute.
Hughes, M. E., Wait, L. J., LaPierre, T. A., & Luo, Y et al. (2007). All in the family: The impact
of caring for grandchildren on grandparents’ health. Journal of Gerontology, 62B(2), S108–
S119.

180

Hunt, C. K. (2003). Concepts in caregiver research. Journal of Nursing Scholarship, 35(1), 27–
32.
Iglehart, J. (2010). Long-term care legislation at long last? Health Affairs, 29(1), 8–9.
Jansen, A. P., Hein PJ van Hout, H. P. J., van Marwijk, H. W. J., Nijpels, G., Gundy, C.,
Vernooij-Dassen, M. J. F. J., et al. (2007). Sense of competence questionnaire among
informal caregivers of older adults with dementia symptoms: A psychometric evaluation.
Clinical Practice and Epidemiology in Mental Health, 3, 11.
Jankauskiene, L., Lesauskaite, V., & Naumaviciene, R. (2007). Empowering family members
for supporting a person with heart failure. Journal of the Indian Academy of Geriatrics, 3(1),
5–7.
Jewell, T. C., & Stein, C. H. (2002). Parental influence on sibling caregiving for people with
severe mental illness. Community Mental Health Journal, 38(1), 17–33.
Jick, T. D. (1979). Mixing qualitative and quantitative methods: Triangulation in action.
Administrative Science Quarterly, 24(4), 602–611.
Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a definition of mixed
methods research. Journal of Mixed Methods Research, 1(2), 112–133.
Johnson-Dalzine, P. (2007). Preparing social work students to work with grandparents in kinship
care: an approach to infusion of content materials into selected core social work courses.
Journal of Gerontological Social Work, 48(3/4), 405–20.
Kaplowitz, M. D., Hadlock, T. D., & Levine, R. (2004). A comparison of web and mail survey
response rates. Public Opinion Quarterly, 68 (1), 94–101.
Katz, S., Down, T.D., Cash, H.R., & Grotz, R.C. (1970). Progress in the development of the
index of ADL. The Gerontologist, 10(1), 20–30.
Katz, S., Ford, A. B., Moskonitz, R. W., Jackson, B. A., & Jaffe, M. W. (1963). Studies of illness
in the aged: The Index of ADL: A standardized measure of biological and psychosocial
function. Journal of American Medical Association, 185(12), 914–919.
Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health,
78(3), 458–467.
Kaye, L. W. (1985). Home care for the aged: a fragile partnership. Social Work, 30 (4), 312–317.
Kearns, R. A., & Andrews, G. J. (2005). Placing ageing: Positionings in the study of older
people. In Gavin J. Andrews & David R. Phillips (Eds.) Ageing and place: Perspective,
policy, practice (pp. 13–23). New York: Routledge.
Keating, N., & Dosman, D. (2009). Social capital and the care networks of frail seniors.
Canadian Review of Sociology. 46(4), 301–318.
181

Keating, N., Otfinowski, P., Wenger, C., Fast, J., & Derksen, L. (2003). Understanding the
caring capacity of informal networks of frail seniors: A case for care networks. Ageing and
Society, 23, 115–127.
Keith, P. M., Wachwe, R., & Collins, S. M. (2009). Family influence on caregiver resistance,
efficacy, and use of services in family elder care. Journal of Gerontological Social Work,
52(4), 377–400.
Kernisan, L. P., Sudore, R. L., & Knight, S. J. (2010). Information-seeking at a caregiving
website: A qualitative analysis. Journal of Medical Internet Research, 12(3), e31.
Kiecolt-Glaser, J. K., Preacher, K. J., MacCallum, R. C., Atkinson, C., Malarkey, W. B., &
Glaser, R. (2003). Chronic stress and age-related increases in the proinflammatory cytokine
IL-6. PNAS, 100(15), 9090–9095.
Kim, J., & Knight, B. G. (2008). Effects of caregiver status, coping styles, and social Support on
the physical health of Korean American caregivers. The Gerontologist, 48(3), 287–299.
Kim, Y., & Schulz, R. (2008). Family caregivers’ strains: Comparative analysis of cancer
caregiving with dementia, diabetes, and frail elderly caregiving. Journal of Aging and Health,
20(5), 483–503.
Kimuna, S. P., & Makiwane, M. (2007). Older people as resources in South Africa: Mpumulanga
households. Journal of Aging and Social Policy, 19(1), 97–114.
Knight, B. G., & Sayegh, P. (2010). Cultural values and caregiving: The updated sociocultural
stress and coping model. Journal of Gerontology: Psychological Sciences, 65B(1), 5–13.
Kolomer, S. R. & Lynch, K. Y. (2007). Challenges for grandparent housing programs. Journal of
Gerontological Social Work, 49(1/2), 65–79.
Kolomer, S. R., McCallion, P., & Overeynder, J. (2005). Why support groups help: Successful
interventions for grandparent caregivers of children with developmental disabilities. In B.
Hayslip Jr., & J. H. Patrick (Eds.) Working with custodial grandparents (pp. 111–125).
New York: Springer Publishing Company.
Kohn, S. J., & Smith, G. C. (2006). Social support among custodial grandparents within a
diversity of contexts. In B. Hayslip & J. Hicks-Patrick (Eds.), Custodial grandparenting:
Individual, cultural, and ethnic diversity (pp. 199–223). New York: Springer.
Kosberg, J. I., Kaufman, A. V., Burgio, L. D., Leeper, J. D., & Sun, F. (2007). Family
caregiving to those with dementia in rural Alabama: Racial similarities and differences.
Journal of Aging and Health, 19(1), 3–21.
Krause, N. (2009). Church-based volunteering, providing informal support at church, and selfrated health in late life. Journal of Aging and Health, 21(1), 63–84.

182

Lai, D. W. L. (2010). Filial piety, caregiving appraisal, and caregiving burden. Research on
Aging, 32(2), 200–223.
Lawton, M.P., & Brody, E.M. (1969). Assessment of older people: Self-maintaining and
instrumental activities of daily living. Gerontologist, 9, 179–186.
Lee, C. C., Czaja, S. J., & Schulz, R. (2010). The moderating influence of demographic
characteristics, social support, and religious coping on the effectiveness of a multicomponent
psychosocial caregiver intervention in three racial ethnic groups. Journal of Gerontology,
65B (2), 185–194.
Lee, Y., Back, J. H., Kim, J., Kim, S., Na, D. L., Cheong, H., et al. (2010). Systematic review of
health behavior at risks and cognitive health in older adults. International Psychogeriatrics,
22(2), 174–187.
Leos-Urbel, J., Bess, R., & Geen, R. (2002). The evolution of federal and state policies for
assessing and supporting kinship caregivers. Children and Youth Services Review, 24(1/2),
37–52.
Letiecq, B. L., Bailey, S. J., & Kurtz, M. A. (2008). Depression among rural Native American
and European American grandparents rearing their grandchildren. Journal of Family Issues,
29 (3), 334–356.
The Lifespan Respite Care Act of 2003. S. 538.
Litwin, H., & Attias-Donfut, C. (2009). The inter-relationship between formal and informal care:
A study in France and Israel. Ageing and Society, 29(1), 71–91.
Lohrer, S. P., Lukens, E. P., & Thorning, H. (2007). Economic expenditures associated with
instrumental caregiving roles of adult siblings of persons with severe mental illness.
Community Mental Health Journal, 43(2), 129–151.
Lubben, J. E. (1988). Assessing social networks among elderly populations. Family and
Community Health, 11, 42–52.
Lubben, J., & Gironda, M. (2003). Centrality and social ties to the health and well-being of older
adults. In B. Berkman & L. Harootyan (Eds.) Social work and health care in an aging society
(pp.319–345). New York: Springer.
Lyyra, T., & Heikkinen, R. (2006). Perceived Social Support and Mortality in Older People. The
Journals of Gerontology, 61B(3), S147–S152.
Magliano, L., Fiorillo, A., Malangone, C., De Rosa, C., Maj, M., & National Mental Health
Project Working Group. (2006). Social network in long-term diseases: A comparative study
in relatives of persons with schizophrenia and physical illnesses versus a sample from the
general population. Social Science & Medicine, 62(6), 1392–1402.

183

Martinson, M., & Minkler, M. (2006). Civic engagement and older adults: A critical perspective.
The Gerontologist, 46 (3), 318–324.
Marziali, E., & Donahue, P. (2006). Caring for others: Internet video-conferencing group
intervention for family caregivers of older adults with Neurodegenerative disease.
Gerontologist, 46(3), 398–403.
Maulik, P. K., Eaton, W. W., & Bradshaw, C. P. (2009). The role of social network and support
in mental health service use: Findings from the Baltimore ECA Study. Psychiatric Service,
60(9), 1222–1229.
McDaid, D. (2001). Estimating the cost of informal care for people with Alzheimer’s disease:
Methodological and practical challenges. International Journal of Geriatric Psychiatry,
16(4), 400–405.
McGarry, J., & Arthur, A. (2001). Informal caring in late life: a qualitative study of the
experiences of older carers. Journal of Advanced Nursing, 33(2), 182–189.
Mertler, C. A., & Vannatta, R. A. (3rd Ed.). (2005). Advanced and multivariate statistical
methods: Practical application and interpretation. Glendale, CA: Pyrczak Publishing.
Minkler, M., Driver, D., Roe, K. M., & Bedeian, K. (1993). Community interventions to support
grandparent caregivers. Gerontologist, 33(6), 807–811.
Minkler, M., & Estes, C. L. (Eds.) (1999). Critical gerontology: Perspectives from political and
moral economy. Amityville, NY: Baywood Publishing Company.
Moen, P. (2011). A life-course approach to the third age. In D. C. Carr & K. Komp (Eds.)
Gerontology in era of the third age: Implications and next steps (pp. 13–31). New York:
Springer Publishing Company.
Montoro-Rodríguez, J., Kosloski, K., Kercher, K., & Montgomery, R. J. V. (2009). The Impact
of Social Embarrassment on Caregiving Distress in a Multicultural Sample of Caregivers.
Journal of Applied Gerontology, 28(2), 195–217.
Moor, C., Zimprich, D., Schmitt, M., & Kliegel, M. (2006). Personality, aging self-perceptions,
and subjective health: A mediation model. International Journal of Aging and Human
Development, 63(3), 241–257.
Moren-Cross, J. L., & Lin, N. (2006). Social networks and health. In R. H. Binstock & L. K.
George (Eds.), Handbook of Aging and the Social Sciences (6th Ed.) (pp. 111–128). San
Diego, CA: Academic Press.
Moreno, J. L. (1953). Who shall survive? New York: Beacon House.
Murray, M. (2008). The networked family: Reframing the legal understanding of caregiving and
caregivers. Virginia Law Review, 94(2), 385–455.

184

Nakanishi, M., Hoshishiba, Y., Iwama, N. Okada, T., Kato, E., & Takahashi, H. (2009).
Impact of the elder abuse prevention and caregiver support law on system development
among municipal governments in Japan. Health Policy, 90(2), 254–261.
Navaie-Wailser, M., Feldman, P. H., Gould, D. A., Levine, C. L., Kuerbis, A. N., & Donelan, K.
(2002). When the caregiver needs care: The plight of vulnerable caregivers. American
Journal of Public Health, 92(3), 409–413.
Netto, N. R., Jenny, G. Y. N., & Philip, Y. L. K. (2009). Growing and gaining through caring for
a loved one with dementia. Dementia, 8(2), 245–262.
Neufeld, A., & Kushnner. K. E. (2009). Men family caregivers’ experience of nonsupportive
interactions: Context and expectations. Journal of Family Nursing, 15(2), 171–197.
Neufeld, A., Harrison, M. J., Stewart, M., & Hughes, K. (2008). Advocacy of women family
caregivers: Response to nonsupportive interactions with professionals. Qualitative Health
Research, 18(3), 301–310.
Neundorfer, M. M., McClendon, M. J., Smyth, K. A., Strauss, M. E., & McCallum, T. J. (2006).
Does depression prior to caregiving increase vulnerability to depressive symptoms among
caregivers of persons with Alzheimer’s disease? Aging & Mental Health, 10(6), 606–615.
Newsom, J. T., Mahan, T. L., Rook, K. S., & Krause, N. (2008). Stable negative social
exchanges and health. Health Psychology, 27(1), 78–86.
Nijboer, C., Tempelaar, R., Triemstra, M., van den Bos, G. A. M., & Sanderman, R. (2001).The
role of social and psychological resources in caregiving of cancer patients. Cancer, 91(5),
1029–1039.
Nijboer, C., Triemstra, M., Tempelaar, R., Sanderman, R., & van den Bos, G. A. M. (1999).
Measuring both negative and positive reactions to giving care to cancer patients:
Psychometric qualities of Caregiver Reaction Assessment (CRA). Social Science &
Medicine, 48, 1259–1269.
Nolan, M., Keady, J., & Grant, G. (1996). Understanding family care: A multidimensional model
of caring and coping. Bristol, PA: Open University Press.
Noyes, B. B., Hill, R. D., Hicken, B. L., Luptak, M., Rupper, R., Dailey, N. K., et al. (2010). The
Role of Grief in Dementia Caregiving. American Journal of Alzheimer’s Disease & Other
Dementia, 25(1), 9–17.

185

Lee, C. C., Czaja, S. J., & Schulz, R. (2010). The moderating influence of demographic
characteristics, social support, and religious coping on the effectiveness of a multicomponent
psychosocial caregiver intervention in three racial ethnic groups. Journal of Gerontology,
65B (2), 185–194.
Leos-Urbel, J., Bess, R., & Geen, R. (2002). The evolution of federal and state policies for
assessing and supporting kinship caregivers. Children and Youth Services Review, 24(1/2),
37–52.
Lévesque, L., Ducharme, F., Zarit, S. H., Lachance, L., & Giroux, F. (2008). Predicting
longitudinal patterns of psychological distress in older husband caregivers: Further analysis
of existing data. Aging & Mental Health, 12(3), 333–342.
Lim, J., & Zebrack, B. (2004). Caring for family members with chronic physical illness: A
critical review of caregiver literature. Health and Quality of Life Outcomes, 2(1), 50–59.
Obama, B. (2011). Presidential Proclamation National Family Caregivers Month. Retrieved
from http://nfca.typepad.com/nfc_month_2011/presidential-proclamation-national-familycaregivers-month-1.html
Oktay, J. S. (2006). Family caregivers on the job: Moving beyond ADLs and IADLs. Social
Work, 51(3), 281–282.
The Older American Act Amendments of 2000. P.L. 106-501.
The Older American Act Amendments of 2006. P.L. 109-365.
O’Mara, A. (2005). Who’s talking care of the caregiver? Journal of Clinical Oncology, 23(28),
6820–6821.
Ong, A. D., Bergeman, C. S., Bisconti, T. L., & Wallace, K. A. (2006). Psychological resilience,
positive emotions, and successful adaptation to stress in later life. Journal of Personality and
Social Psychology, 91(4), 730–749.
Ornstein, K., Smith, K. L., & Boal, J. (2009). Understanding and Improving the Burden and
Unmet Needs of Informal Caregivers of Homebound Patients Enrolled in a Home-Based
Primary Care Program. Journal of Applied Gerontology, 28(4), 482–503.
Park, H. H. (2006). The economic well-being of households headed by a grandmother as
caregiver. Social Service Review, 80(3), 264–295.

186

Peek, C. W., Zsembik, B. A., Coward, R. T. (1997). The changing caregiving networks of older
adults. Research on aging, 19(3), 333–361.
Penrod, J. D., Kane, R. A., Kane, R. L., & Finch, M. D. (1995). Who Cares? The Size, Scope,
and Composition of the Caregiver Support System. The Gerontologist, 35(4), 489–497.
Phillips, L. R., & Crist, J. (2008). Social relationships among family caregivers: A CrossCultural Comparison Between Mexican Americans and Non-Hispanic White Caregivers.
Journal of Transcultural Nursing, 19(4), 326–337.
Pinquart, M. (2001). Correlates of subjective health in older adults: A meta-analysis. Psychology
and Aging, 16(3), 414–426.
Pinquart, M., & Sörensen, S. (2003). Differences between caregivers and noncaregivers in
psychological health and physical health: A meta-analysis. Psychology and Aging, 18(2),
250–267.
Pinquart, M., & Sörensen, S. (2006). Gender Differences in Caregiver Stressors, Social
Resources, and Health: An Updated Meta-Analysis. Journal of Gerontology, 61B (1), S33–
S45.
Pinquart, M., & Sörensen, S. (2007). Correlates of physical health of informal caregivers: A
meta-analysis. Journal of Gerontology, 62B(2), 126–137.
Putnam, R. (2000). Bowling alone: The collapse and revival of American community. New York:
Simon and Schuster.
Quentin, W., Riedel-Heller, S. G., Luppa, H., Rudolph, A., & König, H. H. (2010). Cost of
illness studies of dementia: A systematic review focusing on stage dependency of costs. Acta
Psychiatrica Scandinavica, 121(4), 243–259.
Raehurn, J., & Rootman, I. (2007). A new appraisal of the concept of health. In M. O’Neill, A.
Pederson, S. Dup÷r÷, & I Rootman (Eds.) Health promotion in Canada: Critical perspectives
(pp. 19–31). Toronto, Ontario: Canadian Scholars’ Press.
Raphael, C., & Cornwell, J. L. (2008). Influencing support for caregivers. Journal of Social
Work Education, 44(3), 97–103.
Ray, R. A., & Street, A. F. (2005). Who's there and who cares: age as an indicator of social
support networks for caregivers among people living with motor neurone disease. Health &
Social Care in the Community, 13(6), 542–552.
Ray, R. A., & Street, A. F. (2011). The dynamics of socio-connective trust within support
networks accessed by informal caregivers. Health, 15(2), 137–152.
Robison, J., Fortinsky, R., Kleppinger, A., Shugrue, N., & Porter, M. (2009). A broader view of
family caregiving: Effects of caregiving and caregiver conditions on depressive symptoms,
health, work, and social isolation. Journal of Gerontology, 64B(6), 788–798.
187

Rochat, S., Cumming, R. G., Blyth, F., Creasey, H., Handelsman, D., Le Couteur, D. G., et al.
(2010). Frailty and use of health and community services by community-dwelling older men:
the Concord Health and Ageing in Men Project. Age and Ageing, 39 (2), 228–233.
Roepke, S. K., Mausbach, B. T., von Känel, R., Ancoli-Israel, S., Harmell, A. L., Dimsdale, J.
E., et al. (2009). The moderating role of personal mastery on the relationship between
caregiving status and multiple dimensions of fatigue. International Journal of Geriatric
Psychiatry, 24(12), 1453–1462.
Rozario, P. A., Chadiha, L. A., Proctor, E. K.& Morrow-Howell, N. (2008). Predicting the
influence of social resources on African American wife and daughter caregivers' depressive
symptoms. Journal of Family Issues, 29(3), 317–333.
Ruggles, S. (1994). The transformation of American family structure. American Historical
Review, 99(1), 103–128.
Ruggles, S. (2003). Multigenerational families in nineteenth-century America. Continuity and
Change, 18(1) 139–165.
Ruggles, S., & Browner, S. (2003). Measurement of household and family composition in the
United States, 1850–2000. Population and Development Review, 29(1) 73–101.
Ruiz, D. S. (2004). Amazing grace: African American grandmothers as caregivers and
conveyors of traditional values. Westport, CT: Praeger.
Russell, R. (2004). Social Networks Among Elderly Men Caregivers. The Journal of Men's
Studies, 13(1), 121–142.
Russell, R. (2008). Their story, my story: Health of older men as caregivers. Generations, 32(1),
62–67.
Salovey, P., Rothman, A. J., Detweiler, J. B., & Steward, W. T. (2000). Emotional states and
physical health. American Psychologist, 55(1), 110–121.
Sandberg, J., & Eriksson, H. (2009). From alert commander to passive spectator: older male
carers’ experience of receiving formal support. International Journal of Older People
Nursing, 4(1), 33–40.
Sanders, S. (2007). Experiences of rural male caregivers of older adults with their informal
support networks. Journal of Gerontological Social Work, 49(4), 97–115.
Sanders, S., & Power, J. (2009). Roles, responsibilities, and relationships among older husbands
caring for wives with progressive dementia and other chronic conditions. Health and Social
Work, 34(1), 41–51.
Scott, J. P., & Roberto, K. A. (1985). Use of informal and formal support networks by rural
elderly poor. The Gerontologist, 37(1), 624–630.

188

Schneider, E. L., & Guralnik, J. M. (1990). The aging of America. Impact on health care costs.
Journal of the American Medical Association, 263(17), 2335–40.
Sheran, M., & Swann, C. A. (2007). The take-up of cash assistance among private kinship care
families. Children and Youth Services Review, 29(8), 973–987.
Shiovitz-Ezra, S., Leitsch, S., Graber, J., & Karraker, A. (2009). Quality of life and
psychological health indicators in the national social life, health, and aging project. Journal
of Gerontology: Social Sciences, 64B(S1), I30–I37.
Schulz, R., & Beach, S. R. (1999). Caregiving as a risk factor for mortality: The Caregiver health
Effects Study. JAMA, 282(23), 2215–2219.
Schulz, R., Herbert, R. S., Dew, M. A., Brown, S. L., Scheier, M. F., Beach, S. R., et al., (2007).
Patient suffering and caregiver Compassion: New opportunities for research, practice, and
policy. The Gerontologist, 47 (1), 4–13.
Schulz, R., & Sherwood, P. R. (2008). Physical and Mental Health Effects of Family Caregiving:
Adverse—and even positive—outcomes in a chronic stress experience. American Journal of
Nursing, 108(9 Suppl.), 23–27.
Seeman, T. E., Kaplan, G. A., Knudsen, L., Cohen, R., & Guralnik, J. (1987). Social network ties
and mortality among the elderly in the Alameda County Study. American Journal of
Epidemiology, 126(4), 714–723.
Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Positive psychology: An introduction.
American Psychologist, 55(1), 5–14.
Silverstein, M., Conroy, S. J., & Gans, D. (2008). Commitment to caring: Filial responsibility
and the allocation of support by adult children to older mothers. In M. E. Szinovaz & A.
Davey (Eds.) Caregiving contexts: Cultural, familial, and society implications (pp. 71–91).
New York: Springer Publishing Company.
Simmons, T., & Dye, J. (2003). Grandparents living with grandchildren: 2000, Census 2000
Brief. Washington, DC: U.S. Census Bureau.
Sims-Gould, J., & Martin-Matthews, A. (2007). Family caregiving or caregiving along: Who
helps the helper? Canadian Journal of Aging, 26(Suppl. 1), 27–46.
Sit, J. W. H., Wong, T. K. S., Clinton, M., Li, L. S. W., & Fong, Y-M. (2004). Stroke care in the
home: the impact of social support on the general health of family caregivers. Journal of
Clinical Nursing 13(7), 816–824.
Smith, K. P., & Christakis, N. A. (2008). Social networks and health. Annual Review of
Sociology, 34, 405–429.
Spillman, B. C., & Black, K. J. (2005). Staying the course: Trends in family caregiving.
Washington, DC: AARP Public Policy Institute.
189

Stack, C. (1974). All our kin: Strategies for survival in a black community. New York: Harper
and Row.
Standing, T. S., Musil, S. M., & Warner, C. B. (2007). Grandmothers’ transitions in caregiving
to grandchildren. Western Journal of Nursing Research, 29(5), 613–631.
Stoller, E. P. (1985). Exchange patterns in the informal support networks of the elderly: The
impact of reciprocity on morale. Journal of Gerontology, 47B (2), 335–342.
Sun, F., Hilgeman,M. D., Durkin, D. W., Allen, R. S., & Burgio, L. D. (2009). Perceived
income inadequacy as a predictor of psychological distress in Alzheimer’s caregivers.
Psychology and Aging. 24(1), 177–189.
Sundström, G., Johansson, L., & Hassing, L. B. (2002). The shifting balance of long-term care
in Sweden. The Gerontologist, 42 (3), 350–355.
Szinovacz, M. E., & Davey, A. (2006). Effect of retirement and grandchild care on depressive
symptoms. The International Journal of Aging and Human Development, 62(1), 1–20.
Szinovacz, M. E., & Davey, A. (2007). Changes in adult child caregiver networks. The
Gerontologis, 47(3), 280–295.
Szinovaz, M. E. & Davey, A. (2008.) Caregiving contexts: Cultural, familial, and society
implications. New York: Springer Publishing Company.
Talley, R. C., & Crews, J. E. (2007). Framing the public health of caregiving. American Journal
of Public Health, 97(2), 224–228.
Tarlow B. J, Wisniewski, S. R., Belle, S. H., Rubert, M., Ory, M. G., & Gallagher-Thompson, D.
(2004). Positive aspects of caregiving: contributions of the REACH project to the
development of new measures for Alzheimer’s caregiving. Research on Aging; 26(4), 429–
480.
Taylor, S. E., Kemeny, M. E., Reed, G. M., Bower, J. E. & Gruenewald, T. L. (2000).
Psychological resources, positive illusions, and health. American Psychologist, 55(1), 99–
109.
Tebb, S., & Jivanjee, P.(2000). Caregiver isolation. Journal of Gerontological Social Work,
34(2), 51–72.
Tennstedt, S. L., John B. McKinlay, J. B., & Sullivan, L. M. (1989). Informal Care for Frail
Elders: The Role of Secondary Caregivers. The Gerontologist, 29 (5), 677–683.
Teschendorf, B., Schwartz, C., Ferrans, C. E., O’Mara, A., Novotny, P., & Sloan, J. (2007).
Caregiver Role Stress: When Families Become Providers. Cancer Control, 14(2), 183–189.
The United States Census. (2010). 2008 American Community Survey. Retrieved from
http://www.census.gov/compendia/statab/2011/tables/11s0035.pdf
190

The United States Census. (2011). 2010 Census shows nation's population is aging. Retrieved
from http://2010.census.gov/news/releases/operations/cb11-cn147.html
Thoits, P. (1995). Stress, coping and social support processes: Where are we? What next?
Journal of Health and Social Behavior, (extra issue), 53–79.
Thompson, L. (2004). Long-term care: Support for family caregivers. Washington, DC:
Georgetown University, Long-Term Care Financing Project.
Tolkacheva, N., van Groenou, M. B., de Boer, Al, & van Tilburg, T. (2011). The impact of
informal care-giving networks on adult children's care-giver burden. Ageing & Society,
31(1), 34–51.
Traustadóttir, R., & Sigurjónsdóttir, H. B. (2008). The ‘mother’ behind the mother: Three
generations of mothers with intellectual disabilities and their family support networks.
Journal of Applied Research in Intellectual Disabilities, 21(4), 331–340.
Uhlenberg, P., & Cheuk, M. (2008). Demographic change and the future of informal caregiving.
In M. E. Szinovaz & A. Davey (Eds.) Caregiving contexts: Cultural, familial, and society
implications (pp. 9–33). New York: Springer Publishing Company.
Umberson, D., & Montez, J. K. (2010). Social relationships and health: A flashpoint for health
policy. Journal of Health and Social Behavior, 51(1Suppl.), S54–S66.
Valle, R. (2004). Ethnic differences in social network help-seeking strategies among Latino and
Euro-American dementia caregivers. Aging & Mental Health, 8(6), 535–543.
Vaillant, G. E. (2000). Adaptive mental mechanisms: Their role in a positive psychology.
American Psychologist, 55, 89–98.
Van Houtven, C. H., & Norton, E. C. (2008). Informal care and Medicare expenditures: Testing
for hereroneneous treatment effects. Journal of Health Economics, 27, 134–156.
Vitaliano, P. P., Zhang, J., & Scanlan, J. M. (2003). Is caregiving hazardous to one's physical
health? A meta-analysis. Psychological Bulletin, 129(6), 946–97.
Wasserman, S., & Faust, K. (1995). Social network analysis: Methods and applications. New
York: Cambridge University Press.
Weeks, L. E., Nilsson, T., Bryanton, O., & Kozma, A. (2009). Current and future concerns of
older parents of sons and daughters with intellectual disabilities. Journal of Policy and
Practice in Intellectual Disabilities, 6(3), 180–188.
Weinland, J. A. (2009). The lived experience of informal African American male caregivers.
American Journal Men’s Health, 3(1), 16–24.
Weiss, R. (1973). Loneliness: The experience of social and emotional isolation. Cambridge, MA:
The MIT Press.
191

Whittington, F. J. (2010). The art and science of caregiving. The Gerontologist, 50(1), 133–140.
Whyte, S. R., Alber, E., & Geissler, P. W. (2004). Lifetimes intertwined: African grandparents
and grandchildren. Africa, 24(1), 1–5.
Williams, S. W., & Dilworth-Anderson, P. (2002). Systems of social support in families who
care for dependent African American elders. The Gerontologist, 42 (2), 224–236.
Williams, S. W., Williams, C. S., Zimmerman, S. Munn, S., Dobbs, D., & Sloane, P. D. (2008).
Emotional and physical health of informal caregivers of residents at the end of life: The role
of social support The Journals of Gerontology, 63B(3), S171–S183.
Williamson, D. L., & Carr, J. (2009). Health as a resource for everyday life: Advancing the
conceptualization. Critical Public Health, 19(1), 107–122.
World Health Organization. (1946). Constitution of the World Health Organization. New York:
Author.
Yang, Y. (2008). Social inequalities in happiness in the United States, 1972 to 2004: An AgePeriod-Cohort analysis. American Sociological Review, 73 (2), 204–226.
Yap, P., Luo, N., Ng, W. Y., Chionb, H. L., & Lim, J. (2010). Gain in Alzheimer Care
Instrument—A new scale to measure caregiving gains in dementia. American Journal of
Geriatric Psychiatry, 18(1), 68–76.
Yeh, P., Wierenga, M. E., & Yuan, S. (2009). Influences of Psychological Well-being, Quality of
Caregiver-patient Relationship, and Family Support on the Health of Family Caregivers for
Cancer Patients in Taiwan. Asian Nursing Research, 3(4), 154–166.
Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M. et al. (1983).
Development and validation of a geriatric depression screening scale: a preliminary report.
Journal of Psychiatric Research, 17(1), 37–49.
Zacharakis, J., & Flora, J. (2005). Riverside: A case study of social capital and cultural
reproduction and their relationship to leadership development. Adult Education Quarterly,
55, 288–307.
Zarit, S., & Femia, E. (2008). Behavioral and psychosocial interventions for family caregivers.
Journal of Social Work Education, 44(3), 49–57.
Zarit, S. H., Pearlin, L. I., & Schaie, K. W. (1992). Caregiving systems: informal & formal
helpers. Hillsdale, New Jersey: Lawrence Erlbaum Associates.

192