University of Southern California Dissertations and Theses

Social support, self-efficacy, and gender in treatment adherence of heart failure patients

(USC Thesis Other)

Social support, self-efficacy, and gender in treatment adherence of heart failure patients

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
SOCIAL SUPPORT, SELF-EFFICACY, AND GENDER
IN TREATMENT ADHERENCE OF HEART FAILURE PATIENTS

by

Uta Maeda

_____________________________________________________

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(PSYCHOLOGY)

August 2010

Copyright 2010 Uta Maeda

ii

Table of Contents

List of Tables iii
List of Figures iv
Abstract v
Chapter 1: Introduction 1
Chapter 2: Methods 8
Chapter 3: Results 16
Chapter 4: Discussion 27
Bibliography 34

iii

List of Tables

Table 1: Demographic and Clinical Characteristics of Participants 18
Table 2: Descriptive statistics for key independent and dependent 19
variables for the full sample
Table 3: Gender comparisons of key independent and dependent variables 19

iv

List of Figures

Figure 1: A mediation model 14
Figure 2: Mediation analyses using the Baron & Kenny (1986) approach 25

v

Abstract
Nonadherence to medical recommendations is a leading preventable cause of
rehospitalization and premature mortality in chronic heart failure (HF) patients. This
study examined the contribution of functional and structural social support to general and
specific adherence, and whether self-efficacy and gender mediated and moderated these
relationships, respectively. Participants consisted of 252 HF patients (31% women) with
a mean age of 54 years. All analyses controlled for age, gender, marital status, education,
ethnicity, New York Heart Association (NYHA) class, and comorbidity. Structural
support was significantly associated with both general (β=.156, p<.05) and specific
adherence (β=.165, p<.05). Similarly, functional support was significantly associated
with both general (β=.275, p<.001) and specific (β=.274, p<.001) adherence. The
tangible, emotional/informational, affectionate, and positive social interaction subtypes of
functional support were highly intercorrelated (Pearson’s r=.77 - .88, ps <.001) and not
differentially associated with adherence. Self-efficacy significantly mediated the effects
of structural support on general (Sobel Z = 2.71, p=.007) and specific adherence (Sobel Z
= 3.09, p=.002), as well as the effects of functional support on general (Sobel Z = 3.13,
p=.002) and specific adherence (Sobel Z = 4.18, p=.000). Gender did not appear to
moderate any of these relationships (ps>.05). These results suggest that both the quality
and quantity of social support are important predictors of both global and domain-specific
adherence, and that this positive influence occurs indirectly via self-efficacy. Implications
and future directions for HF interventions are discussed.

1

Chapter 1: Introduction
Nonadherence in Heart Failure
Chronic heart failure (HF) is a progressive condition in which the cardiac muscles
become unable to pump sufficient blood to meet the body’s metabolic needs. It often
occurs as a result of severe cardiovascular disease, and affects over 5 million Americans
today, the majority of whom are over the age of 65 (Jessup & Brozena, 2003, Lloyd-
Jones et al., 2010). The disease takes a devastating toll on patients and their families not
only due to the rapid deterioration of physical health, but by creating a tremendous
financial and emotional burden as well. The annual cost of HF continues to grow rapidly,
with direct and indirect costs estimated to reach $39.2 billion in 2010 (Lloyd-Jones et al.,
2010). Despite a large amount of progress in the field, HF remains a condition with no
cure and very poor prognosis. Thus the primary goal of current treatments is not full
recovery, but rather, improved health outcomes such as increased survival, reduced rates
of rehospitalizations, and higher quality of life.
Treatment adherence is a critical determinant of the beneficial effect of medical
care on a patient’s physical health, functional capacity, and quality of life. However,
nonadherence is a major problem that plagues patients: it is estimated that approximately
a quarter of patients with cardiovascular disease do not adequately follow their treatment
regimens (DiMatteo, 2004a). This behavior is associated with 15 - 42% of cases of HF
decompensation and exacerbation, and is a leading preventable cause of
rehospitalizations and premature mortality (Michalsen et al., 1998; Opasich et al., 1996;

2

Tsuyuki et al., 2001; Sokol et al., 2005). If studies do not take into account the level of
adherence demonstrated by each patient, the resulting conclusions regarding the effects of
psychological, sociodemographic, and clinical predictors on health outcomes could be
very misleading. The identification and modification of potential causes and predictors of
nonadherence is critical for interventions for HF patients. Two such predictors that have
received a lot of attention in the health literature are social support and self-efficacy.
Social Support
Social support is one of the most robust psychosocial factors associated with
health-related behaviors and outcomes in HF patients, predicting hospital readmissions,
mortality, adaptive coping, quality of life, and depression prevalence (Luttik et al., 2005;
Murberg, 2004; Bennett et al., 2001). Social support has also been associated with greater
adherence in patients with cardiovascular diseases including HF (Sayers et al., 2008;
DiMatteo, 2004b). These findings are consistent with Cohen’s (1988) model, which states
that social support can positively impact health through behavioral processes such as
boosting health maintenance behaviors in various domains via encouragement or social
control.
Although social support is a common predictor examined in the health
psychology literature, studies have differed widely in their conceptualization and
definition of this construct. Depending on the way and context in which social support is
operationalized, it could have differential effects on health outcomes of HF patients
(Friedmann et al., 2006). Social support can be broadly classified into structural support

3

and functional support. Structural support refers to the size and density of social networks
as well as the frequency of social interactions and may at times provide limited
information due to the lack of distinction between positive and negative social interaction,
thus confusing quantity with quality. Functional support, on the other hand, refers to
different types of social interactions that really are “supportive” in nature, such as
tangible, affectionate, emotional-informational, and positive social interaction support
(Sherbourne & Stewart, 1991). A major goal of this study was to examine these various
types of social support and their differential effects on medical adherence specifically in
HF patients.
Self-Efficacy
Self-efficacy, or the belief that one can carry out a behavior necessary to achieve
a certain desired outcome, has been indicated as one of the strongest and final
determinants of health-promoting behavior in chronically ill populations, independent of
disease severity or level of physical impairment (Bandura, 1986; O’Leary, 1985; Strecher
et al., 1986). In patients with cardiovascular disease, self-efficacy predicted adherence to
exercise and diet regimens (Schweitzer et al., 2007; Joekes et al., 2007; Ewart et al.,
1986; Robertson & Keller, 1992; Gilroy, 1990), and confidence ratings related to
maintaining normal function and controlling symptoms had particularly significant long-
term effects (Sullivan et al., 1998). A study of older women with HF showed that those
who displayed higher self-efficacy at baseline demonstrated higher levels of disease
management behavior (taking medication as prescribed, exercise, diet, stress
management) at 4 and 12 month follow-ups (Clark & Dodge, 1999). On the other hand,

4

poorer adherence was associated with lower perceived self-efficacy and lack of
knowledge about self-care (Ni et al., 1999).
Managing a chronic illness such as HF is rarely an individual matter, and external
sources can positively influence a patient’s confidence and motivation to adhere to
treatments, both by reinforcing beneficial disease management behaviors as well as
monitoring and helping to eliminate detrimental ones. The literature on marital support
and health has suggested that a spouse’s confidence in the patient’s self-care abilities can
boost the patient’s own efficacy, which can then lead to better treatment adherence and
improved cardiac health outcomes (Taylor et al., 1985; Rohrbaugh et al., 2004). Such
results indicate that social support and self-efficacy may not predict adherence through
entirely independent pathways. This study aimed to explore whether self-efficacy acts as
a mediator of the relationship between social support and adherence.
Gender differences
Overall, survival rates for patients are improving despite the increasing incidence
of HF. However, this effect is significantly greater in men than in women (Barker et al.,
2006; Roger et al., 2004), which suggests that the field may not be adequately addressing
gender-specific needs in caring for HF patients. A closer examination of the potential
moderating role of gender in the relationship between social support and treatment
adherence may ultimately help clinicians identify and enhance the types of support that
are most beneficial to each patient.

5

The literature has suggested that there are gender differences in the way patients
react to and cope with chronic cardiovascular illnesses. For example, women with heart
disease consistently report poorer quality of life and endorse higher levels of perceived
stress and emotional problems such as sadness, tearfulness, and other signs of depression
(Bennett et al., 2000, Shumaker, 1997, Loose & Fenhall, 1995, Czajkowski, 1993). They
also report lower scores in vitality and physical functioning, with reductions in social
activities and activities of daily living (Chin & Goldman, 1998, Riedinger et al., 2001).
The poorer prognosis of women overall may be partly explained by gender
differences in the nature of social support networks. The literature has consistently found
that women with heart disease receive less structural support than men (Shumaker, 1997;
Woloshin, 1997). Since cardiovascular disease usually has a later onset in women than
men, due in part to the protective effect of estrogen, female patients tend to be older and
more likely to be widowed and living alone (Grace et al., 2002; Barber et al., 2001;
Cooper et al., 2002). Structural support may actually have a weaker protective effect for
women than it does for men, as it has been suggested that women bear the stress of
additional social ties that men do not (Unger, 1999; Gore & Colten, 1991).
The literature also appears to suggest that women may feel negatively pressured
by tangible support, which was reported by DiMatteo (2004b) to be the functional
support subtype most strongly associated with adherence in the chronically ill population
in general. Women who are recovering from a cardiac event or condition tend to
minimize their own health situation and avoid burdening people in their social networks,
and report feeling guilty when they must rely on family members to help with household

6

tasks (Benson et al., 1997; Bjarnason-Wehrens et al., 2007). These results are equivocal,
however, as a study of older women with HF found that greater tangible support was
associated with lower levels of negative affect (Friedman & King, 1994). In contrast, low
levels of perceived general emotional support predicted subsequent cardiovascular events
at an odds ratio that was significantly higher in older female HF patients compared to
older male patients, while higher levels of emotional support were associated with greater
positive affect and life satisfaction (Krumholz et al., 1998; Friedman & King, 1994).
In addition to discrepancies in the influence of social support, previous studies
have also suggested that there may be gender differences in levels of self-efficacy among
HF patients. Female cardiac patients report receiving significantly less information from
their healthcare practitioners regarding their diagnosis and treatment needs, which lowers
their self-efficacy and subsequent health maintenance behaviors (Stewart et al., 2004).
Female HF patients, but not male patients, had increased self-efficacy, better treatment
adherence, and improved cardiac health outcomes when their spouse expressed as much
confidence in the patients’ disease management abilities as the patients themselves did
(Taylor et al., 1985; Rohrbaugh et al., 2004). Taken together, these and other results in
the literature suggest that adherence behaviors and health outcomes of women with HF
may be influenced by psychological variables in ways that are different from those of
men.
In summary, the main goals of this study were to examine (1) the association of
structural and functional social support with general and specific adherence, (2) the role

7

of self-efficacy as a potential mediator of these relationships, and (3) the role of gender as
a potential moderator of these relationships.

8

Chapter 2: Methods
Participants
The sample included 252 patients with a primary diagnosis of HF as determined
by their cardiologist, who were recruited from the University of Miami Jackson
Memorial Hospital (JMH) and the Cedars-Sinai Medical Center (CSMC) for a larger
study examining the effects of psychosocial factors on health functioning and outcomes
in HF patients. Participants were outpatients older than 18 years of age who demonstrated
competence in reading, writing, and speaking the English or Spanish language at the 6th
grade level. All participants had clinical evidence of HF, and were assigned a New York
Heart Association (NYHA) functional class of I, II, III, or IV. Individuals scoring 24 or
lower on the Mini Mental Status Exam (MMSE) were excluded from the study, in order
to keep the sample free of individuals with cognitive impairments, neurological disorders,
and psychiatric disorders that may interfere with the validity of the self-reported data.
Individuals with diagnosis and/or treatment for severe life-threatening comorbidities were
also excluded from the study. Written informed consent was obtained from all
participants before the study.
Measures
Social Support.
The Medical Outcomes Study (MOS) Social Support survey is a 20-item
questionnaire which assesses multiple dimensions of an individual’s perceived level of
support. A single item measures structural support by asking the participant to indicate
the number of close friends and relatives he or she has. The remaining 19 items assess

9

functional support by asking participants to indicate how often different kinds of support
are available to them, using a 5-point Likert scale that ranges from “None of the time (1)”
to “All of the time (5).” The overall index was calculated by averaging scores across all
items, with higher scores indicating greater levels of social support. Confirmatory factor
analysis supported a total social support index encompassing four subscales: tangible
support (e.g., “someone to help you if you were confined to bed”), affectionate support
(e.g., “someone who hugs you”), emotional-informational support (e.g., “someone you
can count on to listen to you when you need to talk”), and positive social interaction (e.g.,
“someone to get together with for relaxation”) (Sherbourne & Stewart, 1991). Prior use
of this scale with patients with chronic illnesses provided evidence of high internal
consistency for the measure as a whole (Cronbach’s α = 0.97), as well as the individual
subscales (Cronbach’s αs = 0.91-0.96). Although the subscales are highly correlated
(rs=0.69 to 0.82), multitrait and factor analyses revealed good discriminant validity
between them (Sherbourne & Stewart, 1991). An exception was item #14 (“someone to
do things with to help you get your mind off things”), which contributed significantly to
both the positive interaction and emotional-informational subscales; thus this item was
removed from further analyses.
General Adherence.
The MOS General Adherence survey is a 5-item questionnaire assessing a
patient’s general ability to follow their doctor’s suggestions, regardless of the specific
treatment plan. Examples of items are: “I followed my doctor’s suggestions exactly” and
“I was unable to do what was necessary to follow my doctor’s treatment plans.” Responses are

10

scored on a 6-point Likert scale, ranging from “Completely Disagree (1)” or “None of the
Time (1)” to “Always Agree (6)” or “All of the Time (6).” The total score was calculated
by averaging responses across items after reverse-scoring the negatively worded items,
with higher scores indicating better adherence. The measure has shown high internal
consistency reliability (Cronbach’s α = 0.80-0.87) in patients with chronic illnesses
including cardiovascular disease (Sherbourne et al., 1992; Kravitz et al., 1993).
Specific Adherence.
Two items were added to the MOS Specific Adherence survey in order to create a
17-item questionnaire assessing the frequency that the patient followed specific
recommendations regarding their treatment, in areas such as medication, diet, exercise,
and stress management. Due to the behavior-specific nature of adherence (e.g., an
individual may have an easier time following dietary restrictions but have difficulty
adhering to exercise regimens), it was anticipated that this scale may be able to detect
individual differences in self-care abilities that the general adherence scale would not.
Responses are scored on a 6-point Likert scale, and can range from “None of the time
(1)” to “All of the time (6),” with an option to indicate “NA” for recommendations that
do not apply to the individual. The total score was calculated by averaging responses
across items, with higher scores indicating better adherence. The original measure has
demonstrated fair reliability (Cronbach’s α = 0.69) in patients with chronic illnesses
including cardiovascular disease, and an appropriately low two-year stability of r=0.24
which indicates substantial change over time (Sherbourne et al., 1992). The internal
consistency of the revised measure in the current sample was α = 0.84 when the

11

reliability analysis excluded two items (related to smoking and alcohol) that did not apply
to the majority of the participants.
Self-Efficacy.
The Self-Efficacy scale is a 17-item questionnaire that assesses the patient’s
degree of confidence in his/her abilities to follow specific treatment recommendations,
particularly in the domains of diet, exercise, stress management, and other lifestyle
changes. The format follows Bandura’s (1986) recommendation that self-efficacy should
be assessed as a patient’s confidence in tackling specific challenges rather a global
judgment of ability. The items are based largely on those of the Specific Adherence
survey described above, and represent the major goals for treatment in HF patients.
Responses can range from “Not at all (1)” to “All of the time (6).” The total score was
calculated by averaging responses across items, with higher scores indicating higher self-
efficacy. Internal consistency was α = 0.87 for this participant sample when reliability
analysis excluded three items (related to smoking, alcohol, and work) that did not apply
to a majority of the participants.
Procedures
Details of the study were introduced to patients while they were in the hospital for
an inpatient stay or routine outpatient visit. Participation was voluntary and no data was
collected from those who refused to partake in the study. Once the interested patients
gave consent and attained stable outpatient status, they were scheduled for a study session
at JMH or CSMC, where they completed a packet of questionnaires (including the

12

measures described above) assessing their level of functioning and quality of life as a
patient with HF. Semistructured interviews were conducted to obtain information on age,
marital status, number of medical comorbidities, race/ethnicity, and disease severity.
Medical chart reviews were conducted to obtain other relevant clinical information such
as history of cardiovascular or psychiatric diseases, history of major hospitalizations and
medical procedures, and current medications (including both cardiovascular and other
prescriptions).
Analyses
All analyses were performed using the SPSS (version 15.0) statistical package.
While participants were free to skip any questionnaire items they did not wish to answer,
packets that were less than 75% complete were deemed unusable, and their data were not
included in the analysis. For the remaining participants, missing data were addressed
using pair-wise deletions.
Relevant sociodemographic and clinical variables were tested for significant
correlations with the study variables in order to identify covariates. Age, gender, ethnicity,
marital status, education, comorbidity, and NYHA class were always included in the
models as core covariates regardless of level of significance. Comorbidy was calculated
using the weighted index system devised by Charleson et al. (1987).
Hierarchical multiple regressions were used to examine the cross-sectional
relationship between social support and treatment adherence after controlling for the core
demographic and clinical variables listed above. Separate sets of analyses were conducted

13

for the total sample using general and specific adherence as outcome variables. Study
variables were tested for the assumptions of independence, normality, linearity, and
homoscedasticity, and transformations were applied as necessary.
Several analysis methods were used in order to test whether self-efficacy
mediated the relationship between social support and both general and specific adherence.
According to the traditional method suggested by Baron and Kenny (1986), self-efficacy
can be considered a mediator between social support and adherence if: (1) social support
is significantly associated with self-efficacy (path a in Figure 1), (2) social support is
significantly associated with adherence (path c in Figure 1), (3) self-efficacy is
significantly associated with adherence after controlling for social support (path b in
Figure 1), and (4) the impact of social support on adherence is significantly reduced after
including self-efficacy as a predictor (path c’ in Figure 1). Hierarchical multiple
regression was conducted to test these conditions, and the statistical significance of each
regression coefficient was evaluated using the t test.

14

Figure 1: A mediation model
Above: Illustration of direct effect of social support on adherence (c = total effect of IV
on DV through mediator). Below: Illustration of mediation in which social support
affects adherence (a = IV to mediator, b = direct effect of mediator on DV, c’ = direct
effect of IV on DV)

The indirect effect of social support on adherence via self-efficacy was tested
using the SPSS (Chicago, IL) macro for simple mediation provided by Preacher and
Hayes (2004). The Sobel test was used to test whether the indirect effect of social support
on adherence, defined as the product between paths a and b in Figure 1, was significantly
different from zero. The Sobel test and other product of coefficients tests were found to
have high statistical power and accurate Type I error rates when compared with multiple
other methods of assessing mediation effects (MacKinnon et al., 2002). However,
because the assumptions of the Sobel test—particularly that of the normality of the
sampling distribution of ab—are often violated, the significance of indirect effects was

15

also examined by bootstrapping the sampling distribution of ab to obtain bias-corrected
95% confidence intervals. This nonparametric approach has been recommended as a
means of reducing the likelihood of underpowered analyses (Preacher & Hayes, 2004).
Specifically, 1000 bootstrap samples of size N=252 (the original sample size) were taken
from the data, sampling with replacement, and the indirect effect ab was calculated in
each sample. The estimate of ab is the mean of ab calculated over the bootstrap samples,
and the estimated standard error is the standard deviation of the ab estimates. To create
the 95% CI, the 1000 estimates of ab were sorted from low to high, and the 25
th
and 976
th

scores were taken as the lower and upper limits, respectively.
To test whether gender moderated the association between social support and
adherence, hierarchical multiple regressions were conducted in which the predictors were
(1) social support, (2) dummy-coded gender, and (3) a gender x social support cross
product, in addition to the covariates listed above. Social support was centered in order
to reduce collinearity.

16

Chapter 3: Results
Sample Characteristics and Study Variables
The final sample consisted of 252 participants (174 men) with ages ranging from
20 to 85 (M=54.25, SD=11.09). Over half (54.8%) of the participants were married or
partnered. There was considerable ethnic diversity, with nearly half of the participant
group identifying themselves as Hispanic (43.6%). Over 65% of the group had at least a
high school education. At the time of the study, the mean duration of diagnosed HF was
approximately 5.5 years (SD=66.96 months), and the majority (76.6%) of participants
had a NYHA class of II or III. A detailed description of demographic and clinical
characteristics is reported in Table 1.
The mean level of perceived structural support for the entire sample was 9.25
(SD=7.57). The mean level of overall perceived functional support was 4.04 (SD=1.02),
with means of 3.98 (SD=1.13), 4.02 (SD=1.08), 4.24 (SD=1.07), and 4.06 (SD=1.11) in
the tangible, emotional-informational, affectionate, and positive social interaction
subscales, respectively. The mean level of self-efficacy was 4.52 (SD=.90). The mean
score for general and specific adherence were 4.73 (SD=.93) and 4.47 (SD=.97),
respectively. A detailed description of the distributions of these independent and
dependent variables is reported in Table 2.
A comparison of study sites using independent samples t-tests revealed that
participants from CSMC were significantly younger [mean difference = 7.04,
t(43.24)=2.985, p=.005], had greater resting ejection fractions [mean difference = -8.87,
t(31.393)=-2.715, p=.011], had greater comorbidity [mean difference = .77, t(31.410)=-

17

2.058, p=.048], and had longer duration of HF at the time of study [t(222)=2.005, p=.046]
than those from JMH. In addition, the patient sample from CSMC was composed of a
significantly greater proportion of Caucasians [χ
2
=11.336, p=.001] and smaller
proportion of Hispanics [χ
2
=11.771, p=.001] than the sample recruited from JMH. The
participants at the two sites did not differ significantly in their self-reported levels of
social support, general adherence, specific adherence, or self-efficacy.
Independent sample t-tests were also used to compare profiles of the demographic,
clinical, and study variables between genders. Women had higher resting ejection
fractions, [mean difference = 4.37, t(243)=-2.499, p=.013], a smaller proportion of
Caucasians [χ
2
=4.687, p=.030], and a greater proportion of Blacks [χ
2
=5.081, p=.024]
than men. Structural support was not significantly different between genders (p=.480).
However, perceived overall functional social support was significantly higher in women
than men [mean difference = .30, t(175.6)=2.320, p=.022], particularly in the emotional
[mean difference = .37, t(184.1)=2.809, p=.006] and affectionate [mean difference = .30,
t(184.9)=2.232, p=.027] subscales. There were no significant gender differences in the
tangible and positive social interaction support subscales (p=.234 and p=.085,
respectively). There was a significant gender effect for self-efficacy scores, with men
reporting higher scores than women [mean difference = -.26, t(249)=-2.120, p=.035].
Self-reported general and specific adherence were not significantly different between
men and women (p=.553 and p=.194, respectively). Gender differences in self-reported
values of the key independent and dependent variables are outlined in Table 3.

18

Table 1: Demographic and Clinical Characteristics of Participants
M(SD) or N(%)
Total Sample
(N=252)
Men
(N=174)
Women
(N=78)
Gender 174 (69%) 78 (31%)
Age (years) 54.25 (11.09) 55.06 (10.42) 52.44 (12.34)
Race
White 71 (28.4%) 56 (32.6%) 15 (19.2%)
Black 55 (22.0%) 31 (18.0%) 24 (30.8%)
Hispanic 109 (43.6%) 76 (44.2%) 33 (42.3%)
Other 15 (6.0%) 9 (5.2%) 6 (7.7%)
Married or Partnered 138 (54.8%) 105 (60.3%) 33 (42.3%)
Income ($K/year) 43.58 (53.48) 44.67 (59.26) 40.97 (36.19)
Education
< HS 31 (12.4%) 15 (8.6%) 16 (20.5%)
HS 164 (65.3%) 118 (68.2%) 46 (59.0%)
College 35 (14.0%) 22 (12.7%) 13 (16.6%)
Advanced 21 (8.4%) 18 (10.4%) 3 (3.9%)
NYHA class
I 37 (15.2%) 25 (14.7%) 12 (16.3%)
II 105 (43.0%) 71 (41.8%) 34 (45.9%)
III 82 (33.6%) 58 (34.1%) 24 (32.5%)
IV 20 (8.2%) 16 (9.4%) 4 (5.4%)
Ejection Fraction (%) 26.93 (13.56) 25.57 (13.03) 29.94 (14.30)
Time Since HF Diagnosis (months) 65.47(66.96) 62.97 (65.80) 71.07 (69.66)
Weighted Comorbidity Index 1.42 (1.42) 1.51 (1.47) 1.23 (1.27)

19

Table 2: Descriptive statistics for key independent and dependent variables for the full
sample
Mean SD Range
Structural Social Support 9.25 7.57 0 - 35
Functional Social Support 4.04 1.02 1.00 – 5.00
Tangible 3.98 1.13 1.00 – 5.00
Emotional/Informational 4.02 1.08 1.00 – 5.00
Affectionate 4.24 1.07 1.00 – 5.00
Positive Social Interaction 4.06 1.11 1.00 – 5.00
Self-Efficacy 4.52 .90 2.13 – 6.00
General Adherence 4.73 .93 1.80 – 6.00
Specific Adherence 4.47 .97 2.00 – 6.00

Table 3: Gender comparisons of key independent and dependent variables
Men
(N=174)
M(SD)
Women
(N=78)
M(SD)
Test Statistic
t
(*p<.05)
Structural Social Support 9.01 (7.35) 9.76 (8.05) .707
Functional Social Support 3.95 (1.06) 4.25 (.89) 2.32*
Tangible 3.92 (1.19) 4.11 (.99) 1.19
Emotional/Informational 3.90 (1.13) 4.28 (.90) 2.81*
Affectionate 4.14 (1.13) 4.44 (.89) 2.23*
Positive Social Interaction 3.98 (1.14) 4.24 (1.02) 1.73
Self-Efficacy 4.60 (.86) 4.34 (.97) -2.12*
General Adherence 4.75 (.91) 4.67 (.98) -.59
Specific Adherence 4.52 (.94) 4.35 (1.01) -1.30

20

Social Support and Treatment Adherence
Model 1: Structural Social Support - General Adherence.
Log-transformed structural support was significantly associated with general
adherence [β=.156, t(207)=2.274, p=.024]. The model also explained a modest but
significant portion of the variation in general adherence [R
2
=.090, F(10, 207)=2.05,
p=.03]. Among the covariates in the model, only Black/African American ethnic status
was a significant predictor of general adherence, with those who were Black
demonstrating lower levels than those who were not [β=-.225, t(207)=-2.769, p=.006].
Model 2: Functional Social Support – General Adherence.
Perceived functional social support was significantly associated with general
adherence [β=.275, t(226)=4.264, p=.000]. The model accounted for a significant
proportion of the variance in general adherence [R
2
=.136, F(10,226)=3.566, p=.000].
Among the covariates in the model, only Black ethnic status had a significant effect, with
those who were Black demonstrating poorer general adherence those who were not [β=-
.193, t(226)=-2.534, p=.012].
Model 3: Structural Social Support – Specific Adherence.
Log-transformed structural support was also significantly associated with specific
adherence [β=.165, t(207)=2.436, p=.016]. The model explained a significant proportion
of variance in specific adherence [R
2
=.104, F(10, 207)=2.404, p=.01]. Among the
covariates in the model, only age had a significant effect, with those who were older

21

demonstrating better adherence overall to specific recommendations in the domains of
diet, exercise, stress management, and lifestyle changes [β=.206, t(207)=2.935, p=.004.]
Model 4: Functional Social Support – Specific Adherence.
Perceived functional support was also significantly associated with specific
adherence [β=.274, t(226)=4.224, p=.000]. The model accounted for a significant
proportion of variance in specific adherence [R
2
=.128, F(10, 226)=3.320, p=.000].
Among the covariates in the model, only age had a significant effect, with older patients
demonstrating better levels of specific adherence [β=.186, t(226)=2.853, p=.005].
Comparison of Functional Support Subscales.
The four social support subscales were highly intercorrelated, with Pearson
correlations ranging from .77 to .88 (all ps < .001). There did not appear to be any
significant differences in the regression coefficients of the four regression equations for
either the model with general (β=.228-.275) or specific (β=.214-.277) adherence as the
outcome. As a result, functional social support was treated as a single index for all
subsequent analyses.
Self-efficacy as a mediator between functional social support and treatment
adherence
Model 1: Structural Social Support - General Adherence.
The mediator model accounted for a significant amount of variance in general
adherence [total R
2
=.182, F(11, 206)=4.168, p=.000]. This represented an R
2
increase

22

of .092 compared to the model with structural support only. Log-transformed structural
support was significantly and positively associated with self-efficacy [β=.174,
t(207)=2.538, p=.012], and self-efficacy was significantly associated with general
adherence after controlling for structural support [β=.317, t(206)=4.812, p=.000]. When
self-efficacy was introduced into the model, the direct effect of structural support on
general adherence was significantly reduced from β=.156 to β=.100 and was no longer
significant (p=.024), which is consistent with full mediation (see Figure 2a). Evaluation
of the indirect effect also supported the finding that self-efficacy mediated the
relationship between structural support and general adherence (β=.075, Sobel test Z =
2.71, p=.007). The indirect effect was significant in that the 95% bias corrected
bootstrapped confidence interval (95% CI = .0205-.1446 with 1000 resamples) excluded
zero. These results indicate that greater perceived social network sizes contribute to
greater general adherence via higher self-efficacy.
Model 2: Functional Social Support - General Adherence.
The mediator model accounted for a significant amount of variance in general
adherence [total R
2
=.187, F(11, 225)=4.707, p=.000]. Higher levels of perceived
functional social support were significantly associated with greater self-efficacy [β=.327,
t(226)=5.133, p=.000], and self-efficacy was significantly associated with general
adherence after controlling for social support [β=.246, t(225)=3.749, p=.000]. When self-
efficacy was introduced into the model, the direct effect of functional support on general
adherence was significantly reduced from β=.275 to β=.195 but still significant (p=.004),
which is consistent with partial mediation (see Figure 2b). Evaluation of the indirect

23

effect also supported the finding that self-efficacy mediated the relationship between
functional social support and general adherence (β=.068, Sobel test Z = 3.13, p=.002).
The indirect effect was significant in that the 95% bias corrected bootstrapped confidence
interval (95% CI = .0301-.1238 with 1000 resamples) excluded zero. These results
indicate that higher levels of perceived functional social support contribute to greater
general adherence via higher self-efficacy.
Model 3: Structural Social Support - Specific Adherence.
The mediator model accounted for a significant amount of variance in specific
adherence [total R
2
=.463, F(11, 206)=16.148, p=.000]. This represented an R
2
increase
of .359 compared to the model with structural social support only. Greater self-efficacy
was significantly associated with specific adherence after controlling for structural social
support [β=.626, t(206)=11.735, p=.000]. The relationship between structural support and
specific adherence was fully mediated by self-efficacy, as the direct path between social
support and specific adherence was substantially reduced from β=.165 to β=.056 and no
longer significant (p=.294; see Figure 2c). Evaluation of the indirect effect also supported
the finding that self-efficacy mediated the relationship between structural social support
and specific adherence (β=.1471, Sobel test Z = 3.09, p=.002). The 95% bias corrected
bootstrap confidence interval excluded zero (95% CI = .0423-.2649), indicating that the
indirect effect was significantly different from zero. These results indicate that greater
sizes of social networks contribute to greater specific adherence via higher self-efficacy.

24

Model 4: Functional Social Support - Specific Adherence.
The mediator model accounted for a significant amount of variance in specific
adherence [total R
2
=.450, F(11, 225)=16.763, p=.000]. Greater self-efficacy was
significantly associated with specific adherence after controlling for social support
[β=.619, t(225)=11.487, p=.000]. The relationship between functional support and
specific adherence was fully mediated by self-efficacy, as the direct path was
substantially reduced from β=.274 to β=.072 and no longer significant (p=.191; see
Figure 2d). Evaluation of the indirect effect also supported the finding that self-efficacy
mediated the relationship between functional social support and specific adherence
(β=.160, Sobel test Z = 4.18, p=.000). The 95% bias corrected bootstrap confidence
interval excluded zero (95% CI = .0788-.2561), indicating that the indirect effect was
significantly different from zero. These results indicate that higher levels of perceived
social support contribute to greater specific adherence via higher self-efficacy.

25

Figure 2: Mediation analyses using the Baron & Kenny (1986) approach
Self-efficacy mediated the influence of social support on adherence in all models.
Mediation was partial in Model 2 (*p<0.05).
(a) Model 1: Structural Support – General Adherence

(b) Model 2: Functional Support – General Adherence

(c) Model 3: Structural Support – Specific Adherence

(d)Model 4: Functional Support – Specific Adherence

26

Gender did not moderate the relationship between social support and adherence
The gender x social support cross product was not statistically significant in any
of the four hierarchical regression models predicting adherence (p>.05), indicating that
structural or functional support do not differentially predict general or specific adherence
for men and women.

27

Chapter 4: Discussion
The current investigation provides a step toward gaining a comprehensive
understanding of the complex relationship between social support and adherence in heart
failure. By examining differential contributions of the various types of social support on
adherence, this study aimed to clarify the findings from past studies that have tackled the
topic from limited angles by using a sample of only women, comparing tangible versus
emotional support, looking only at structural support, looking at characteristics of social
support and adherence rather than the relationship between them, etc. (Sayers et al.,
2008; Bennett, 2001; Clark & Dodge, 1999). The results indicate that both the quantity
and quality of social support may play important roles in predicting both global and
domain-specific treatment adherence in HF patients. The fact that social support did not
influence adherence directly, but did so indirectly, indicates that self-efficacy may be an
important mediating variable that explains the effects of various facets of social
relationships on disease management behaviors.
These results suggest a potential cognitive mechanism for the effect of social
support on adherence behavior. While external sources can facilitate and influence a
patient’s coping response to HF, adaptive disease management behavior ultimately
depends on internal resources. Personal beliefs regarding perceived capabilities may
influence preference for specific tasks (e.g., walking around the neighborhood versus
going to the gym), the amount of effort one puts forth, and persistence despite obstacles
or frustrations. While past experiences of accomplishment and failure generally provide
the most influential sources of efficacy information, self-perception and knowledge of

28

one’s capabilities to carry out certain behaviors do not rely exclusively on previous
mastery (Bandura, 1986). Social support, in the form of vicarious experiences, verbal
persuasion or encouragement, and other normative or informational influence may alter
one’s attitudes, motivation, and emotional reactions regarding particular behaviors. The
fact that social support was significantly and concurrently associated with self-efficacy in
this sample suggests that HF patients weigh and integrate efficacy information from
diverse sources in order to form their personal judgments of self-efficacy, which in turn
promote appropriate disease management behaviors.
The analyses of the present study did not find any evidence that gender moderated
the relationships between social support and adherence, indicating that there may not be
any differences in the extent to which disease management behaviors of men and women
with HF are influenced by social support. However, it is possible that due to the heavily
(nearly 70%) male sample used in this study, there may not have been sufficient power to
detect significant differences between genders. Curiously, while previous literature
suggested that women with cardiovascular disease have lower levels of structural support
than their male counterparts, no gender difference for this variable was observed in the
current sample. This may be explained by the fact that the current study sample was
younger than the typical HF patient population with no gender difference in age, while
much of the previous literature is based upon older HF patients among whom women
were more likely to be widowed than men (Grace et al., 2002; Barber et al., 2001; Cooper
et al., 2002). These differences in sample characteristics may provide some clues to

29

explain why gender differences in the relationship between social support and adherence,
if any, failed to emerge.
Implications
Given the alarming rates and implications of nonadherence in patients with HF,
interventions that incorporate elements that are likely to facilitate an increase in the
patient’s efficacy cognitions will potentially reduce medical costs and improve health
outcomes. The treatment regimen for HF often necessitates a major shift in the lifestyle
that a patient has carried out for decades. Simply providing an extensive list of
recommendations such as a new diet, new exercise regimen, and numerous medications
can be daunting and frustrating to the patient. Instead, behavioral management techniques
such as self-monitoring and recording, tracking progress with charts, keeping personal
exercise and nutrition diaries, and reasonable goal-setting strategies may provide
evidence of personal mastery for the patient and consequently increase the perception of
self-efficacy.
Managing HF is rarely an individual matter, and a patient’s attitudes toward his or
her own health management behavior can be influenced significantly by the nature and
availability of close relationships with one’s spouse, family, and friends. It then becomes
critical for clinicians to address this component of the complex biopsychosocial matrix of
HF by assessing each patient’s social support availability in order to design customized
interventions that meet his or her unique needs. Despite strong evidence linking social
support to positive health outcomes, current interventions for HF are still usually directed

30

at the patient as an individual. Although some recent efforts have been made to develop
programs targeted at enhancing social support in patients with chronic illnesses
(Daugherty et al., 2002; Riegel & Carlson, 2004), it is not yet a standard component of
care for most patients. Interventions may introduce resources for enhancing the quality
and quantity of a patient’s social network in the form of peer-support groups and
community organizations that can socially integrate a newly diagnosed patient into a
group of other patients living with chronic illnesses and allow for sharing of experiences
and resources to boost confidence. At the same time, programs can also educate
caregivers, family members, and friends about specific ways in which they can not only
provide tangible, emotional-informational, affectionate, and positive social interaction
types of support, but more importantly, communicate this support in a manner that will be
perceived by the patient as positive and meaningful. Consistent with social control theory,
how others attempt to influence self-care can lead to differing results. Positive tactics
such as acknowledging successful changes generally lead to more desirable outcomes
than negative attempts such as making the patient feel guilty, which may increase
psychological distress without improving adherence (Lewis & Rook, 1999). Those
individuals who are closest to the patient may be coached to become more cognizant of
ways to communicate verbal reminders (to take medication or exercise, for example) in a
way that will more likely be perceived by the patient as encouraging rather than nagging,
patronizing, or suggestive of the patient’s incompetence. Customized supportive
interventions may not only help the patient boost his or her self-efficacy and health-
promoting behaviors, but may also provide caregivers (and others in the support network)

31

with the confidence that they can make a critical contribution to maintaining the health of
the HF patient.
Limitations
It should be noted that the analyses conducted in this study were cross-sectional in
nature, thus limiting the causal inferences that can be drawn regarding the relationships
between the study variables. While this study examined the positive influence of social
support on adherence through heightened self-efficacy, the directionality of the links
between these variables may not be limited to the one observed here. Supplemental
exploratory analyses conducted to test some of these other mechanisms demonstrated that
social support does not mediate the influence of self-efficacy on adherence in this sample
(ps>.05). However, adherence may positively influence social support or self-efficacy.
For example, patients who adhere well to their regimens are likely healthy enough to
maintain social networks and subsequently may be more pleasant and approachable, and
receive more support for their efforts. As mentioned above, positive reinforcement may
also be at work, in which a history of positive experiences with high adherence boosts a
patient’s sense of mastery and increases confidence in his or her future self-care abilities.
Future analyses with longitudinal data are warranted in order to also clarify the effect of
changes (either increases or decreases) in social support or self-efficacy on adherence
over time, since initial levels of these variables may be time bound and diminish with
experience. For example, while a new exercise plan might seem easy at the outset, it is
conceivable that a struggle to fit the regimen into one’s schedule or overcome physical
aches and pains may dilute behavior in the long term.

32

Furthermore, the results from the current study may be limited in their
generalizability to the broader HF population at large. Since no data was obtained from
patients who declined participation in the study, it could not be investigated whether
participants and nonparticipants differed in terms of their demographic or clinical
characteristics. Furthermore, as mentioned above, the participants had a mean age of
approximately 54, which is significantly younger than the typical HF patient population
consisting primarily of individuals over the age of 65. This may be a reflection of the
greater severity of HF and wider variety of etiological factors observed at tertiary care
centers such as CSMC and JMH. As a function of the complex nature of cases treated at
these sites, participants also tended to have a high level of medical comorbidity (despite
excluding those with life-threatening diseases from analyses) and take many types of
medications, both of which may introduce confounders that may not be addressed
completely by the analyses of this study.
Future Directions
While the present study examines the linear effect of social support on adherence,
it is possible that this relationship may be curvilinear in nature. For example, it is possible
that too much social support (e.g., leading to overprotection or loss of patient’s
independence) may be equally or even more harmful as too little support. This is a topic
that has not received much attention in the literature and cannot be addressed by the
measures used in the present study. Future work in this area calls for the development of
new measures to capture finer discrepancies between inadequate, optimal, and excessive
support. The mismatch between desired and actual support can also occur in the aspect of

33

quality as well as quantity of social interactions. Although researchers usually focus on
the benefits of social support, it has been suggested that negative aspects of social
interactions may actually be more influential than positive ones. Consistent with this idea,
Carels et al. (2004) reported that social conflict but not social support was significantly
associated with physical symptoms expression in HF patients. Furthermore, there may be
a complex interplay between various sources and types of social support for each patient.
For example, an individual may desire and benefit from tangible support from a spouse
and informational support from a physician, but may experience stress or interpersonal
strain if they receive the “wrong” type of support from the “wrong” person. The
differential effects of these combinations on adherence have not been examined in detail
in the literature, and this is an area that may be clarified by future research.
Conclusion
Both structural and functional types of social support were positively associated
with general and specific treatment adherence in chronic heart failure patients. These
relationships were mediated by self-efficacy but not moderated by gender. These results
suggest that boosting self-efficacy may be a critical goal for HF interventions, which may
be facilitated by enhancing a patient’s quantity and quality of social support.

34

Bibliography
Bandura A. (1986). Social foundations of thought and action: a social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Barber K, Stommel M, Kroll J, et al. (2001). Cardiac rehabilitation for community-based
patients with myocardial infarction: factors predicting discharge recommendation and
participation. J Clin Epidemiol, 54, 1025-1030.
Barker WH, Mullooly JP, Getchell W. (2006). Challenging incidence and survival for HF
in a well-defined older population, 1970-1974 and 1990-1994. Circulation, 113, 799-805.
Baron & Kenny (1986). The moderator-mediator variable distinction in social
psychological research: conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
Bennett SJ, Cordes DK, Westmoreland G, Castro R, Donnelly E. (2000). Self-care
strategies for symptom management in patients with chronic heart failure. Nurs Res, 49,
139-145.
Bennett SJ, Perkins SM, Lane KA, Deer M, Brater DC, Murrary MD. (2001). Social
support and health-related quality of life in chronic heart failure patients. Quality of Life
Research, 10, 671-682.
Benson G, Arthur HM, Rideout E. (1997). Women and heart attack: a study of women’s
experiences. Can J Cardiovasc Nurs, 8(3), 16-23.
Bjarnason-Wehrens B, Grande G, Loewel H, Voller H, Mittag O. (2007). Gender-specific
issues in cardiac rehabilitation: do women with ischaemic heart disease need specially
tailored programmes? European Journal of Cardiovascular Prevention and
Rehabilitation, 14(2), 163-171.
Carels R, Musher-Eizenman, Cacciapaglia H, Perez-Benitez C, Christie S, O’Brien W.
(2004). Psychosocial functioning and physical symptoms in heart failure patients: a
within-individual approach. Journal of Psychosomatic Research, 56(1), 95-101.
Charleson M, Pompei P, Ales K, MacKenzie C. (1987). A new method of classifying
prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis,
40(5), 373-383.
Chin MH, Goldman L. (1998). Gender differences in 1-year survival and quality of life
among patients admitted with congestive heart failure. Med Care, 36, 1033-1046.

35

Clark NM, Dodge JA. (1999). Exploring self-efficacy as a predictor of disease
management. Health Education & Behavior, 26(1) 72-89.
Cohen S. (1988). Psychosocial models of the role of social support in the etiology of
physical disease. Health Psychology, 7(3), 269-297.
Cooper AF, Jackson G, Weinman J, et al. (2002). Factors associated with cardiac
rehabilitation attendance: a systematic review of the literature. Clin Rehabil, 16, 541-542.
Czajkowski SM. (1998). Psychosocial aspects of women's recovery from heart disease. In
K. Orth-Gomer, M. Chesney, & N.K. Wenger (Eds.), Women, stress, and heart disease
(pp. 151-164). Mahwah, NJ: Lawrence Erlbaum Associates.
Daugherty J, Saarmann L, Riegel B, Sornborger K, Moser D. (2002). Can we talk?
Developing a social support nursing intervention for couples. Clin Nurse Spec, 16(4),
211-218.
DiMatteo R. (2004a). Variations in patients’ adherence to medical recommendations: a
quantitative review of 50 years of research. Medical Care, 42(3), 200-209.
DiMatteo R. (2004b). Social support and patient adherence to medical treatment: a meta-
Analysis. Health Psychology, 23(2), 207-218.
Ewart CK, Stewart KJ, Gillilan RE, et al. (1986). Usefulness of self-efficacy in predicting
over-exertion during programmed exercise in coronary artery disease. Am J Cardiol, 57,
557-561.
Friedman M, King KB. (1994). The relationship of emotional and tangible support to
psychological well-being among older women with heart failure. Research in Nursing
and Health, 17(6), 433-440.
Friedmann E, Thomoas SA, Liu F, Morton PG, Chapa D, Gottlieb SS. (2006).
Relationship of depression, anxiety, and social isolation to chronic heart failure outpatient
mortality. Am Heart J, 152, 940e1-940e8.
Gilroy MBR. (1990). Multiple factors affect dietitians counseling practices for high blood
cholesterol. J Am Diet Assoc, 94, 1278-1283.
Gore S, Colten ME. (1991). Gender, stress, and distress: social-relational influences. In J.
Eckenrode (Ed.), The social context of coping (p. 139-163). New York: Plenum Press.

Grace SL, Abbey SE, Shnek ZM. (2002). Cardiac rehabilitation II: referral and
participation. Gen Hosp Psychiatry, 24, 127-134.

36

Jessup M, Brozena S. (2003). Heart Failure. The New England Journal of Medicine,
348(20), 2007-2018.
Joekes K, van Elderen T. (2007). Self-efficacy and overprotection are related to quality of
life, psychological well-being, and self-management in cardiac patients. Journal of
Health Psychology, 12(1), 4-16.
Kravitz RL, Hays RD, Sherbourne CD, et al. (1993). Recall of recommendations and
adherence to advice among patients with chronic medical conditions. Arch Intern Med,
153, 1869-1878.
Krumholz HM, Butler J, Miller J, Vaccarino V, Villiams CS, Mendes de Leon CF. (1998).
Prognostic importance of emotional support for elderly patients hospitalized with heart
failure. Circulation, 97, 958-964.
Lewis M, Rook K. (1999). Social control in personal relationships: impact on health
behaviors and psychological distress. Health Psychology, 18, 63-71.
Lloyd-Jones D, Adams R, Brown T, et al. (2010). Heart disease and stroke statistics—
2010 update: report from the American Heart Association. Circulation, 121(6), e46-e215.
Loose MS, Fenhall B. (1995). Differences in quality of life between male and female
cardiac rehabilitation participants. Journal of Cardiopulmonary Rehabilitation, 15, 225-
231.
Luttik ML, Jaarsma T, Moser D, Sanderman R, van Veldhuisen DJ. (2005). The
importance and impact on outcomes of social support on patients with heart failure.
Journal of Cardiovascular Nursing, 20(3), 162-169.
MacKinnon D, Lockwood C, Hoffman J, West S, Sheets V. (2002). A comparison of
methods to test mediation and other intervening variable effects. Psychological Methods,
7(1), 83-104.
Michalsen A, Konig G, Thimme W. (1998). Preventable causative factors leading to
hospital admissions with decompensated heart failure. Heart, 80, 437-441.
Murberg TA. (2004). Long-term effect of social relationships on mortality in patients
with congestive heart failure. Int J Psychiatry Med, 34, 207-217.
Ni H, Nauman D, Burgess D, Wise K, Crispell K, Hershberger R. (1999). Factors
influencing knowledge of and adherence to self-care among patients with heart failure.
Arch Intern Med, 159, 1613-1619.

37

O’Leary (1985). Self-efficacy and health. Behavior Research and Therapy, 23(4), 437-
451.
Opasich C, Febo O, Riccardi G, et al. (1996). Concomitant factors of decompensation in
chronic heart failure. Am J Cardiol, 78, 354-357.
Preacher KJ, Hayes AF. (2004). SPSS and SAD procedures for estimating indirect effects
in simple mediation mode. Behavior Research Methods, Instruments, and Computers,
36(4), 717-731.
Riedinger MS, Dracup KA, Brecht M, Padilla G, Sarna L, Ganz PA. (2001). Quality of
life in patients with heart failure: do gender differences exist? Heart Lung, 30, 105-116.
Riegel B, Carlson B. (2002). Is individual peer support a promising intervention for
persons with heart failure? Journal of Cardiovascular Nursing, 19(3), 174-183.
Robertson D, Keller C. (1992). Relationships among health beliefs, self-efficacy, and
exercise adherence in patients with coronary artery disease. Heart Lung, 21, 6-63.
Roger VL, Weston SA, Redfield MM, Hellermann-Homan JP, Killian J, Yawn BP,
Jacobsen SJ. (2004). Trends in heart failure incidence and survival in a community-based
population. JAMA, 292, 344-350.
Rohrbaugh M, Shoham V, Coyne J, Cranford J, Sonnega J, Nicklas J. (2004). Beyond the
“self” in self-efficacy: spouse confidence predicts patient survival following heart failure.
Journal of Family Psychology, 18(1), 184-193.
Sayers SL, Riegel B, Pawlowski S, Coyne J, Samaha F. (2008). Social support and self-
care of patients with heart failure. Annals of Behavioral Medicine, 35, 70-79.
Schweitzer R, Head KB, Dwyer J. (2007). Psychological factors and treatment adherence
behavior in patients with chronic heart failure. Journal of Cardiovascular Nursing, 22(1),
76-83.
Sherbourne CD, Stewart AL. (1991). The MOS Social Support Survey. Soc Sci Med,
32(6), 705-714.
Sherbourne CD, Hays RD, Ordway L, DiMatteo MR, Kravitz RL. (1992). Antecedents of
adherence to medical recommendations: results from the medical outcomes study. J
Behav Med, 15, 447-468.
Shumaker SA, Brooks MM, Schron EB, Hale C, Kellen JC, Inkster M, Winbush FV,
Wiklund I, Morris M, Investigators C. (1997). Gender differences in health-related
quality of life among postmyocardial infarction patients: brief report. Women’s Health:
Research on Gender, Behavior, and Policy, 3, 53-60.

38

Shumaker SA, Hill DR. (1991). Gender differences in social support and physical health.
Health Psychology, 10, 102-111.
Sokol M, McGuigan K, Verbrugge R, Epstein R. (2005). Impact of medication adherence
on hospitalization risk and healthcare cost. Medical Care, 43(6), 521-530.
Stewart D, Abbey S, Shnek Z, Irvine J, Grace S (2004). Gender differences in health
information needs and decisional preferences in patients recovering from an acute
ischemic coronary event. Psychosomatic Medicine, 66, 42-48.
Strecher V, Devillis B, Becker M, Rosenstock I. (1986). The role of self-efficacy in
achieving health behavior change. Health Education and Behavior, 13(1), 73-92.
Sullivan MD, LaCroix AZ, Russo J, Katon WJ. (1998). Self-efficacy and self-reported
functional status in coronary artery disease: a six-month prospective study.
Psychosomatic Medicine, 60, 473-478.
Taylor CB, Bandura A, Ewart CK, Miller NH, DeBusk RF. (1985). Exercise testing to
enhance wives’ confidence in their husband’s capability soon after clinically
uncomplicated myocardial infarction. Am J Cardiol, 55, 635-638.

Tsuyuki RT, McKelvie RS, Arnold JM, et al. (2001). Acute precipitants of congestive
heart failure exacerbations. Arch Intern Med, 161, 2337-2342.
Unger JB, McAvay G, Bruce ML, Berkman L, Seeman T. (1999). Variation in the impact
of social network characteristics on physical functioning in elderly persons: MacArthur
Studies of Successful Aging. Journal of Gerontology: Social Sciences, 54B, S245-S251.
Woloshin S, Schwartz LM, Tosteson ANA, Chang CH, Wright B, Plohman J, Fisher ES.
(1997). Perceived adequacy of tangible social support and health outcomes in patients
with coronary artery disease. Journal of General Internal Medicine, 12, 613-618.

Abstract (if available)

Abstract Nonadherence to medical recommendations is a leading preventable cause of rehospitalization and premature mortality in chronic heart failure (HF) patients. This study examined the contribution of functional and structural social support to general and specific adherence, and whether self-efficacy and gender mediated and moderated these relationships, respectively. Participants consisted of 252 HF patients (31% women) with a mean age of 54 years. All analyses controlled for age, gender, marital status, education, ethnicity, New York Heart Association (NYHA) class, and comorbidity. Structural support was significantly associated with both general (β=.156, p<.05) and specific adherence (β=.165, p<.05). Similarly, functional support was significantly associated with both general (β=.275, p<.001) and specific (β=.274, p<.001) adherence. The tangible, emotional/informational, affectionate, and positive social interaction subtypes of functional support were highly intercorrelated (Pearson’s r=.77 - .88, ps <.001) and not differentially associated with adherence. Self-efficacy significantly mediated the effects of structural support on general (Sobel Z = 2.71, p=.007) and specific adherence (Sobel Z = 3.09, p=.002), as well as the effects of functional support on general (Sobel Z = 3.13, p=.002) and specific adherence (Sobel Z = 4.18, p=.000). Gender did not appear to moderate any of these relationships (ps>.05). These results suggest that both the quality and quantity of social support are important predictors of both global and domain-specific adherence, and that this positive influence occurs indirectly via self-efficacy. Implications and future directions for HF interventions are discussed.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The influences of anxiety, coping, and social support on physical functioning among heart failure patients

PDF

Longitudinal relationships between bereavement and physical health‐related quality of life in middle‐ to older‐aged women

PDF

Social support, constraints, and protective buffering in prostate cancer patients and their partners

PDF

The formation and influence of online health social networks on social support, self-tracking behavior and weight loss outcomes

PDF

Self-reported and physiological responses to hate speech and criticisms of systemic social inequality: an investigation of response patterns and their mediation…

Asset Metadata

Creator Maeda, Uta (author)

Core Title Social support, self-efficacy, and gender in treatment adherence of heart failure patients

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 06/14/2010

Defense Date 05/05/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag adherence,gender,heart failure,OAI-PMH Harvest,self-efficacy,social support

Place Name California (states), Florida (states), Los Angeles (city or populated place), Miami (city or populated place)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Shen, Biing-Jiun (committee chair), Baker, Laura A. (committee member), Meyerowitz, Beth (committee member)

Creator Email umaeda@usc.edu,utamaeda@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3131

Unique identifier UC1460715

Identifier etd-Maeda-3781 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-348292 (legacy record id),usctheses-m3131 (legacy record id)

Legacy Identifier etd-Maeda-3781.pdf

Dmrecord 348292

Document Type Thesis

Rights Maeda, Uta

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu