Content uploaded by Ron E Durán
Author content
All content in this area was uploaded by Ron E Durán on Nov 09, 2019
Content may be subject to copyright.
Social Support, Positive States of Mind, and HIV Treatment Adherence in
Men and Women Living With HIV/AIDS
Jeffrey S. Gonzalez, Frank J. Penedo,
Michael H. Antoni, and Ron E. Dura´n
University of Miami
Maria Isabel Fernandez
University of Miami School of Medicine
Shvawn McPherson-Baker and Gail Ironson
University of Miami Nancy G. Klimas and Mary Ann Fletcher
University of Miami School of Medicine
Neil Schneiderman
University of Miami
Numerous studies have linked social support to better medication adherence among illness groups, but
few have examined potential mechanisms for this relationship. Relationships were examined between
social support, depression, positive states of mind (PSOM), and medication adherence among HIV
positive men who have sex with men (n⫽61) and women (n⫽29) on highly active antiretroviral
therapy. Depression and PSOM were evaluated as potential mediators of the relationship between support
and adherence. Cross-sectional data showed that greater social support and PSOM related to better
adherence whereas higher depression scores related to nonadherence. PSOM partially mediated the
relationship between social support and adherence. PSOM may be an important mechanism through
which social support is related to better medication adherence in this population.
Key words: adherence, HIV, antiretroviral therapy, HAART, social support, positive states of mind
HIV continues to be a major health problem in the United States
and worldwide. The development of successful antiretroviral ther-
apies has improved the health of HIV positive (HIV⫹) patients
although also demanding near-perfect medication adherence to
achieve maximum benefit and avoid treatment failure (Rabkin &
Chesney, 1999). Increasing patient adherence remains an impor-
tant challenge in the control of the virus and provides a major
opportunity for behavioral scientists to improve the health and
quality of life of HIV⫹patients. Highly active antiretroviral treat-
ment (HAART) is a very rigorous, demanding, and unforgiving
regimen commanding up to 22 pills a day (Friedland & Williams,
1999). Research shows that even modest or occasional nonadher-
ence to HAART greatly diminishes the benefits of treatment. In an
HIV drug trial, for example, the omission of even a single dose in
28 days was strongly associated with treatment failure (Montaner
et al., 1998). Perhaps the most sobering evidence for this comes
from Paterson et al. (2000), who reported that among patients who
adhered to between 80% and 90% of their HIV treatment doses—a
rate of adherence that would be considered acceptable in many
other treatment populations—only 50% of patients achieved viral
load levels below detectable limits. Approximately 81% of patients
who adhered to more than 95% of their medications had undetect-
able viral loads. Thus, even very minor nonadherence in HIV
treatment has unparalleled consequences for treatment success.
Social and psychological variables are among the most signifi-
cant factors that influence adherence to medical therapy. Social
support, often defined as the degree of one’s satisfaction with his
or her social relationships, has consistently been found to relate to
adherence behavior among various populations with chronic ill-
ness (e.g., Levy, 1983). Studies of HIV patients on combination
therapy have shown a positive association between perceived
social support quality and HAART adherence (e.g., Catz, Kelly,
Bogart, Benotsch, & McAuliffe, 2000) and medical appointment
attendance (Catz, McClure, Jones, & Brantley, 1999). Less satis-
faction with overall social support has been shown to relate to
poorer adherence as measured by pharmacy refill count (Singh et
al., 1999). Few studies have examined the mechanisms through
which social support is related to medication adherence (for ex-
ceptions, see Gonzalez et al., 2001; Simoni, Frick, Lockhart, &
Liebovitz, 2002). In the current study, we aimed to examine how
social support may be related to better medication adherence by
considering depression and positive states of mind (PSOM) as
potential mediators of this relationship.
Jeffrey S. Gonzalez, Frank J. Penedo, Michael H. Antoni, Ron E. Dura´n,
Shvawn McPherson-Baker, Gail Ironson, and Neil Schneiderman, Depart-
ment of Psychology, University of Miami; Maria Isabel Fernandez, De-
partment of Epidemiology and Public Health, University of Miami School
of Medicine; Nancy G. Klimas and Mary Ann Fletcher, Department of
Microbiology and Immunology, University of Miami School of Medicine.
This work was supported by National Institute of Mental Health Grants
PO1 MH49548 and T32 MH18917.
Correspondence concerning this article should be addressed to Michael
H. Antoni, Department of Psychology, University of Miami, P.O. Box
248185, Coral Gables, FL 33134. E-mail: mantoni@miami.edu
Health Psychology Copyright 2004 by the American Psychological Association
2004, Vol. 23, No. 4, 413–418 0278-6133/04/$12.00 DOI: 10.1037/0278-6133.23.4.413
413
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
It is well established that increased social support is predictive
of lower levels of depression (S. Cohen & Wills, 1985). Although
numerous studies have reported a relationship between social
support and depression, on the one hand (e.g., Leserman et al.,
1994), and depression and adherence, on the other hand (in chronic
illness populations, DiMatteo, Lepper, & Croghan, 2000; in per-
sons living with HIV, Catz et al., 2000; Singh et al., 1996), few
have examined how social support may relate to positive psycho-
logical variables. Social support may influence adherence through
associated increases in positive psychological states. S. Cohen and
Syme (1985) suggested that social support results in increases in
positive affect and other positive psychological states through the
stability, predictability, and control that it provides. Indeed, the
ability to maintain PSOM such as focused attention, productivity,
responsible caretaking, relaxation, sensuous pleasure, and a sense
of sharing has been conceptualized as an important aspect of
adjustment to stressful situations (Horowitz, Adler, & Kegeles,
1988). Research on this positive pathway from social support to
positive psychological factors is limited despite the importance of
social relationships in current theoretical models of positive human
health (e.g., Ryff & Singer, 1998). Satisfaction with social support
related to HIV/AIDS has been linked to increased PSOM in a
sample of HIV⫹men and women (Turner-Cobb et al., 2002). The
question of whether PSOM is related to adherence or other health
behaviors has also received limited research attention. One recent
study suggests that related factors—feeling that one has a mean-
ingful life, feeling comfortable and well cared for, using time
wisely, and taking time for important things—result in better
medication adherence in older HIV patients (Holzemer et al.,
1999). These findings lend support for the positive pathway, but
research has been limited.
In this study, we aimed to contribute to the understanding of
how social support is related to medication adherence in HIV⫹
patients by considering two possible mediational paths: (a) social
support’s relationship to less depressive symptoms and (b) social
support’s relationship to more PSOM. Last, we hypothesized that
these paths would be independent of each other.
Method
Participants
Participants were 90 HIV⫹men who have sex with men (MSM) and
women (of any sexual orientation) who were participants in a larger
longitudinal psychosocial intervention study at the University of Miami.
Only preintervention data were included in our analyses. Participants were
recruited through community newspapers, HIV support service agencies,
doctors’offices and medical clinics, HIV conferences and community
events, and fliers placed throughout the South Florida area.
Inclusion criteria included being an HIV⫹female or MSM between the
ages of 18 and 65 and currently on a prescribed regimen of antiretroviral
combination therapy. Individuals identified with cognitive impairment or
active suicidal ideation, panic attacks, psychosis, or alcohol or drug de-
pendence within the past 3 months (First, Spitzer, Gibbon, & Williams,
1997) were also excluded. Additional temporary exclusionary criteria
included use of antibiotics within the previous 2 weeks, changes in any
HAART medications within the previous month, having had an infection
within the prior month, having had surgery within the prior 3 months,
initiation of formal psychotherapy or a formal aerobic fitness training
program within the past 3 months, and intravenous drug use within the
prior 6 months. Participants meeting inclusion criteria signed an informed-
consent form, completed a psychosocial battery with a trained interviewer,
and received monetary compensation ($50).
Measures
Demographic and health-related variables. The following variables
were assessed by a brief questionnaire as potential control variables:
demographic characteristics including age, income, education, race,
months since diagnosis of HIV, HIV symptoms (in the past 2 weeks),
alcohol and drug consumption (drinks/times used in the past 30 days), and
recent stressful life events (in the past 3 months).
Social Provisions Scale. The Social Provisions Scale (Cutrona & Rus-
sell, 1987) is a 24-item scale that assesses agreement with statements
concerning relationships with other people on a scale from 1 (strongly
disagree)to4(strongly agree). The Social Provisions Scale measures six
constructs of social relationship provisions: (a) attachment, (b) social
integration, (c) opportunity for nurturance, (d) reassurance of worth, (e)
reliable alliance, and (f) guidance. The coefficient alpha for the total score
was .88.
Beck Depression Inventory (BDI). The BDI (Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961) is a 21-item self-report questionnaire that mea-
sures depressive symptom severity. Subscale scores were computed to
examine the psychological (Items 1–14) and somatic (Items 15–21) aspects
separately. Coefficient alpha for the BDI was .87. Coefficient alphas for the
psychological and somatic composites were .85 and .67, respectively.
PSOM. The PSOM Scale (Horowitz et al., 1988) is a state measure that
assesses individuals’capacities to enter positive cognitive and interper-
sonal states over the past week. As a construct, PSOM is related to positive
mood but is conceptually broader as it also includes one’s ability to attain
and appreciate positive experiences related to general well-being. The
positive states measured are (a) focused attention, (b) productivity, (c)
responsible caretaking, (d) restful repose, (e) sensuous nonsexual pleasure,
(f) sharing, and (g) sensuous sexual pleasure. The total score is conceptu-
alized as a measure of overall positive mood and life satisfaction. The
PSOM Scale has demonstrated divergent validity from measures of social
desirability, stress, and negative mood and has shown convergent validity
with measures of positive mood (Adler, Horowitz, Garcia, & Moyer,
1998). Internal reliability was .82 in the current study.
Medication adherence. The Adherence to Combination Therapy Guide
(Chesney et al., 2000) is a validated instrument that assesses the proportion
of antiretroviral medications taken compared with the amount of medica-
tion prescribed for the participant over the previous 4-day period. A trained
interviewer administered this measure in an interview format. For the
current study, an overall measure of adherence was computed by summing
the number of pills taken during the previous 4 days and dividing by the
number of pills prescribed for the same period. The resulting ratio repre-
sented the percentage of HAART pills taken as prescribed.
Immune measures. Viral load and CD4⫹T-lymphocyte cell counts
were assessed from peripheral blood samples taken during the assessment
between 9 a.m. and 12 p.m. Serum HIV RNA viral load was measured by
standard laboratory analyses determining the number of HIV virions per
microliter of peripheral blood plasma by using an Amplicor assay (Model
83988, Roche Laboratories, Pleasanton, CA). The lower limit of quantifi-
cation for this assay was 50 copies/
l. CD4⫹T-lymphocyte cell counts
were assessed by flow cytometry methods using commercially prepared
antibodies as described by Fletcher, Baron, Ashman, Fischl, and Klimas
(1987). Viral load and CD4⫹count were assessed for descriptive purposes,
and viral load was additionally used to assess its relationship to self-report
of medication adherence.
Results
Table 1 reports demographic and HIV-related variables. To
establish the validity of the 4-day measure of self-reported medi-
414 BRIEF REPORTS
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
cation adherence, we tested the relationship between adherence
and HIV viral load on study entry. For all analyses, participants
who had missed any medication over the previous 4 days (n⫽31)
were compared with those who reported perfect medication adher-
ence (n⫽59). Analysis of variance controlling for number of days
between viral load measure and self-report adherence indicated
that nonadherent participants had a higher mean log10 viral load
(M⫽3.19, SD ⫽1.50) than adherent participants (M⫽2.50,
SD ⫽1.29), F(1, 82) ⫽5.51, p⫽.02. Examination of actual mean
differences in untransformed viral load values suggested that the
differences between adherent (M⫽12,091 copies/
l, SD ⫽
31,162) and nonadherent (M⫽43,438 copies/
l, SD ⫽102,160)
participants were clinically significant. After establishing that non-
adherence to HAART was associated with poorer health status, we
examined psychosocial factors in relation to nonadherence.
Of all potential demographic and health-related control vari-
ables, only age and alcohol use were related to medication adher-
ence at the p⬍.10 level. Nonadherent participants were younger
(M⫽36 years) than adherent participants (M⫽41 years); t(88) ⫽
2.91, p⬍.01, and reported more alcohol consumption in the
previous week (M⫽3.4 drinks, SD ⫽3.7) than adherent partic-
ipants (M⫽1.5 drinks, SD ⫽2.7), t(88) ⫽⫺2.55, p⫽.01.
Because alcohol use was significantly skewed, a square-root-
transformed variable was used as the covariate in regression anal-
yses below. No significant gender differences were found in ad-
herence rates (p⬎.10). A set of moderated logistic regression
analyses tested gender as a moderator of the relationships between
social support, depression, PSOM, and medication adherence. In-
teraction terms were tested between gender and each independent
variable in each of the regressions presented below. Analyses
revealed no significant interaction effects (all ps⬎.40), suggest-
ing that the relationships did not vary as a function of gender.
Table 2 shows the results of three separate logistic regression
analyses controlling for age and alcohol use and testing social
support, depression, and PSOM as correlates of medication adher-
ence. Results of these regression analyses indicated that social
support and PSOM were significantly related to medication adher-
ence. An initial analysis showed that total BDI score was nega-
tively related to adherence (odds ratio [OR] ⫽0.47; 95% confi-
dence interval [CI] ⫽0.28, 0.89; p⬍.01). To rule out the
confound of health on somatic symptoms of depression, we ex-
amined the psychological symptom subscale of the BDI, which
was also related to adherence, as a mediator. The ORs in Table 2
are based on standardized variables and reflect the increase in the
odds of being a member of the adherent group associated with a 1
standard deviation unit increase in the independent variable.
Table 2
Independent Logistic Regression Models Predicting Medication
Adherence Group Membership
Step and variable Model
2a
OR 95% CI Wald’s
2a
Logistic Regression 1
Step 1 14.43***
Age 1.08 1.02, 1.15 6.65**
Alcohol 0.57 0.36, 0.90 5.73*
Step 2 20.78***
SPS total score 1.89 1.12, 3.18 5.66*
Logistic Regression 2
Step 1 14.43***
Controls
Step 2 19.61***
PSYBDI 0.58 0.35, 0.94 4.98**
Logistic Regression 3
Step 1 14.43***
Controls
Step 2 25.44***
PSOM 2.32 1.35, 4.00 9.20**
Note. Medication adherence was coded as 1 ⫽100% and 0 ⫽⬍100%.
Except for age, all odds ratios (ORs) are based on standardized variables to
facilitate comparison between variables. The OR for age reflects the
increase in odds associated with each 1-year increase in age. Alcohol ⫽
square-root-transformed alcohol use; Controls ⫽age and alcohol; CI ⫽
confidence interval; SPS ⫽Social Provisions Scale; PSYBDI ⫽psycho-
logical symptom score from Beck Depression Inventory; PSOM ⫽positive
states of mind.
a
For Step 1, df ⫽2; for Step 2, df ⫽3; N⫽90.
*p⬍.05. ** p⬍.01. *** p⬍.001.
Table 1
Sociodemographic and HIV-Related Variables
Variable n%MSD
Age (years) 39.4 9.5
Months since HIV diagnosis 67.6 46.1
Years on antiretrovirals 1.9 2.1
Total HIV medications (total
pills in last 4 days) 45.8 25.7
HAART adherence
Adherent (100%) 59 66
Nonadherent (⬍100%) 31 34
HIV RNA copies/
l 22,785.7 65,865.2
Total CD4⫹T cells/mm
3
392.2 244.0
Gender
Male 68
Female 32
Ethnicity
Hispanic 37
Non-Hispanic White 31
Black 28
Other 4
Education
Grades 7–12 19
High school 19
Some college 27
Graduated college 21
At least some graduate school 14
Annual income
⬍$10,001 47
$10,001–$30,000 30
$30,001–$50,000 12
⬎$50,000 9
Unknown 2
Employment status
Full time 31
Part time 10
Student 4
Unemployed 19
Disabled 32
Unknown 4
Note. HAART ⫽highly active antiretroviral treatment.
415
BRIEF REPORTS
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Mediation Analyses
Psychological symptoms of depression. Following the proce-
dures outlined by Baron and Kenny (1986), we conducted a series
of regression analyses to test depressive symptoms as a mediator
of the relationship between emotional social support and medica-
tion adherence. Age and alcohol use were controlled in all logistic
regressions for which medication adherence was the dependent
variable. First, medication adherence (nonadherent vs. adherent)
was regressed on social support (B⫽.07, p⬍.02). Then, in a
separate regression model, medication adherence was regressed on
the psychological symptom subscale of the BDI (B⫽⫺.09, p⬍
.03). Subscale scores were then regressed on social support (B⫽
⫺.21, p⬍.01). A final analysis was conducted in which medica-
tion adherence was regressed on the potential mediator and on
social support. In this final hierarchical logistic regression model,
neither social support (

⫽.05, p⫽.08) nor the BDI subscale
(B⫽⫺.07, p⫽.15) was significantly related to medication
adherence. However, as a block, they were significant,
2
(2, N⫽
90) ⫽8.41, p⬍.02. The failure of these variables to indepen-
dently relate with medication adherence suggests that the unique
variance of each independent variable was not associated with
medication adherence. Because of their failure to uniquely relate to
medication adherence, psychological symptoms of depression
could not be considered a mediator.
PSOM. A similar series of four regressions were used to test
PSOM as a mediator of social support. Figure 1 shows the un-
standardized regression coefficients for the analyses testing PSOM
as a mediator. In the final model including both social support and
PSOM, the relationship of support to medication adherence was
attenuated (B⫽.04, p⫽.13), whereas PSOM retained its rela-
tionship to adherence (B⫽.16, p⫽.01). Sobel’s variance estimate
of the mediated effect (Baron & Kenny, 1986; MacKinnon &
Dwyer, 1993) revealed that the indirect path of social support
to medication adherence, mediated by PSOM, was significant
(z
indirect
⫽1.99, p⬍.05). The equation
␣
/(
␣
⫹
⬘) showed that
approximately 37.5% of the relationship between social support
and medication adherence was mediated by PSOM (MacKinnon &
Dwyer, 1993). Thus, PSOM was found to be a substantial partial
mediator of the relationship between perceived support and adher-
ence. This model accounted for 37% of the variability in medica-
tion adherence (Nagelkerke R
2
⫽.367).
Independence of PSOM From Depression
We hypothesized that the relationship of PSOM to medication
adherence would be independent of the relationship of depression
and adherence. A hierarchical logistic regression model revealed
that PSOM remained significantly associated (OR ⫽2.07, CI ⫽
1.17, 3.66; p⫽.01) with medication adherence when the BDI
subscale score was entered into the model. Psychological symp-
toms of depression were not significantly related to medication
adherence in this model. Furthermore, when psychological symp-
toms of depression were added to the PSOM mediation model, the
pattern of relationships presented in Figure 1 remained unchanged.
These results indicate that the relationship of PSOM to medication
adherence and its partial mediation of social support’s relationship
to medication adherence are independent of depression.
Discussion
In this study, we examined the relationship between social
support and medication adherence among HIV⫹MSM and
women who were prescribed a regimen of HAART. We hypoth-
esized that social support would relate to adherence through its
relationship with lower depression and greater PSOM. Our hy-
pothesis was based on research showing consistent relationships
between social support and depression (e.g., S. Cohen & Wills,
1985) and theoretical models predicting associations between so-
cial support and positive psychological factors (S. Cohen & Syme,
1985). Perceived quality of social support was significantly asso-
ciated with medication adherence even after we controlled for age
and alcohol consumption. These findings are consistent with pre-
vious research that has shown relationships between aspects of
support and medication adherence among HIV⫹patients on com-
bination therapy (Catz et al., 2000; Singh et al., 1999). Perceived
quality of social support was also associated with less depressive
symptomatology and higher levels of PSOM, consistent with pre-
vious research (S. Cohen & Wills, 1985; Turner-Cobb et al., 2002).
We also found that depression and PSOM were each individu-
ally related to medication adherence. Depressive symptomatology
was inversely related to medication adherence. Higher levels of
PSOM were related to successful medication adherence. The cur-
rent study’s finding that higher levels of depressive symptomatol-
ogy were related to medication nonadherence is consistent with
previous reports (e.g., Catz et al., 2000; Singh et al., 1996). It is
important to note that this relationship was maintained even when
we examined only psychological symptoms of depression, thus
eliminating possible somatic confounds.
The finding that PSOM was a significant correlate of medication
adherence is, to the best of our knowledge, novel. Although other
research has reported relationships between similar positive as-
pects of quality of life and medication adherence among HIV⫹
patients (Holzemer et al., 1999), we are unaware of research
showing a relationship between PSOM, or other similar positive
psychological state variables, and medication adherence in any
illness group. This finding deserves further replication and could
have important implications for understanding medication adher-
ence behavior and other important health outcomes among HIV-
Figure 1. Path diagram representing mediation analysis for Positive
States of Mind (PSOM) Scale and path diagram for model testing direct
and indirect effects of social support on medication adherence. Unstand-
ardized regression coefficients for factors entered in model individually are
in parentheses, and unstandardized regression coefficients with all factors
in the model entered simultaneously are outside of parentheses. N.S. ⫽
nonsignificant. *p⬍.05. **p⬍.01. ***p⬍.001.
416 BRIEF REPORTS
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
infected persons. Our findings complement those of Simoni et al.
(2002), which showed that depression scores and self-efficacy
partially mediated the relationship between need for social support
and nonadherence. Although their analyses did not compare the
independence of the effects, self-efficacy was found to be more
strongly related to nonadherence than depressive symptoms. Inter-
estingly, the Simoni et al. measure of depressive mood, the Center
for Epidemiological Studies–Depression Scale (Radloff, 1977),
like many other measures, includes items that reflect negative
affect (e.g., “I felt sad”) as well as items that reflect positive affect
(e.g., “I enjoyed life”). Positive affect items are reverse coded and
included in the overall score of negative affect. However, it seems
plausible that positive affect may have independent relationships
to adherence and other health outcomes. In fact, a recent study
suggests that positive affect is the “active ingredient”in the asso-
ciation of Center for Epidemiological Studies–Depression Scale
scores and mortality in a large sample of HIV⫹men (Moskowitz,
in press). Our findings, together with those of Moskowitz, suggest
that there is justification for expanding our focus on negative
psychosocial predictors of health outcomes to include positive
factors as well and to consider the overlap between these con-
structs in our analyses.
The current study showed that PSOM was also a significant
mediator of the relationship between social support and medication
adherence. Moreover, the relationship of PSOM to medication
adherence was independent of depression, suggesting that PSOM
represents more than the absence of depression, as hypothesized.
These findings are partially supportive of S. Cohen and Syme’s
(1985) direct effect theory regarding the relationship between
social support and health behaviors. Their model suggests that
social support affects health behaviors by decreasing negative
psychological factors and also by increasing positive factors. It
appears, in the case of this HIV⫹population, that social support is
related to medication adherence more through positive psycholog-
ical processes than through negative ones.
Further research is needed to identify how positive psycholog-
ical states (i.e., PSOM) are related to medication adherence be-
havior. One study has shown that positive affect predicted better
health practices such as exercise and good nutrition in a sample of
healthy college students (Griffin, Friend, Eitel, & Lobel, 1993).
Are HIV⫹individuals high in PSOM more likely to adhere to their
medications because of positive moods associated with PSOM?
The broaden-and-build theory of positive emotions proposes that
the experience of positive emotion broadens one’s focus of atten-
tion, thinking, and behavioral repertoires and may facilitate prob-
lem solving (Fredrickson, 2001). This flexibility in problem solv-
ing may be adaptive to managing medication adherence in the
context of other life stressors that are commonly experienced by
HIV patients. More research is needed to examine the applicability
of this theory to the context of medication adherence.
Causal inferences cannot be drawn from the current study be-
cause of its cross-sectional design. Generalizing findings to other
populations should be cautioned because of our use of a conve-
nience sample that was paid for participation. The use of a number
of exclusionary criteria (e.g., alcohol dependence in past 3 months,
intravenous drug use in past 6 months) may also limit the gener-
alizability of our findings. The current study adds to most HIV
studies of medication adherence in that it included both MSM and
primarily ethnic minority women of lower socioeconomic status.
However, generalizing findings to other HIV⫹groups should be
cautioned. We measured self-reported medication adherence,
which is known to produce higher estimates of medication adher-
ence than more objective measures such as electronic bottle-cap
monitoring (C. Cohen, 2000). However, even objective measures
of adherence such as electronic bottle-cap monitoring have been
criticized and may provide limited and/or distorted information
about adherence behavior (Samet, Sullivan, Traphagen, & Ickov-
ics, 2001). Additionally, dichotomization of the adherence variable
limited our analyses to comparisons between those participants
who were adherent to all of their medications and those who were
not. Because even very low levels of nonadherence have been
associated with treatment failure in HAART (e.g., Montaner et al.,
1998; Paterson et al., 2000), we do not believe that this limits the
validity of our findings. Future longitudinal research should eval-
uate these relationships by using multiple measures of medication
adherence to increase validity and reliability.
Findings from mediational analyses are consistent with our
hypothesis that social support, through the stability, predictability,
and control that it provides, may facilitate increases in PSOM in
the support receiver. These positive factors may provide psycho-
logical resources to help HIV⫹individuals cope successfully with
the stressful aspects of taking HIV medication and may increase
motivation to take medication as prescribed. These findings may
be useful in the development of intervention strategies aimed at
improving medication adherence among HIV⫹individuals.
References
Adler, N. E., Horowitz, M., Garcia, A., & Moyer, A. (1998). Additional
validation of a scale to assess positive states of mind. Psychosomatic
Medicine, 60, 26–32.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator distinction
in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51, 1173–
1182.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961).
An inventory for measuring depression. Archives of General Psychiatry,
4, 561–571.
Catz, S. L., Kelly, J. A., Bogart, L. M., Benotsch, E. G., & McAuliffe, T. L.
(2000). Patterns, correlates, and barriers to medication adherence among
persons prescribed new treatments for HIV disease. Health Psychology,
19, 124–133.
Catz, S. L., McClure, J. B., Jones, G. N., & Brantley, P. J. (1999).
Predictors of outpatient medical appointment attendance among persons
with HIV. AIDS Care, 11, 361–373.
Chesney, M. A., Ickovics, J., Chambers, D. B., Gifford, A. L., Neidig, J.,
Zwickl, B., & Wu, A. W. (2000). Self-reported adherence to antiretro-
viral medications among participants in HIV clinical trials: The ACTG
Adherence Instruments. AIDS Care, 12, 255–266.
Cohen, C. (2000, January–February). Gauging and improving adherence.
Paper presented at the Seventh Conference on Retroviruses and Oppor-
tunistic Infections, San Francisco.
Cohen, S., & Syme, S. L. (1985). Issues in the study and application of
social support. In S. Cohen & S. L. Syme (Eds.), Social support and
health (pp. 3–22). Orlando, FL: Academic Press.
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering
hypothesis. Psychological Bulletin, 98, 310–357.
Cutrona, C. E., & Russell, D. (1987). The provisions of social relationships
and adaptation to stress. In W. H. Jones & D. Perlman (Eds.), Advances
in personal relationships (Vol. 1, pp. 37–68). Greenwich, CT: JAI Press.
417
BRIEF REPORTS
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
DiMatteo, M. R., Lepper, H. S., & Croghan, T. W. (2000). Depression is
a risk factor for noncompliance with medical treatment: Meta-analysis of
the effects of anxiety and depression on patient adherence. Archives of
Internal Medicine, 160, 2101–2107.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1997).
Structured Clinical Interview for the DSM–IV Axis I disorders: Non-
patient edition (SCID-IV/NP Version 2.0-4/97 rev.). New York: Biomet-
rics Research Department, New York State Psychiatric Institute.
Fletcher, M. A., Baron, G., Ashman, M., Fischl, M., & Klimas, N. (1987).
Use of whole blood methods in assessment of immune parameters in
immunodeficiency states. Diagnostic Clinical Immunology, 5, 69–81.
Fredrickson, B. L. (2001). The role of positive emotions in positive
psychology: The broaden-and-build theory of positive emotions. Amer-
ican Psychologist, 56, 218–226.
Friedland, G. H., & Williams, A. (1999). Attaining higher goals in HIV
treatment: The central importance of adherence. AIDS, 13(Suppl. 1),
S61–S72.
Gonzalez, J. S., Antoni, M. H., Duran, R., McPherson-Baker, S., Fletcher,
M. A., & Schneiderman, N. (2001, March). Social support, positive
states of mind, self-efficacy and medication adherence among HIV⫹
men and women. Paper presented at the 20th Conference of the Society
of Behavioral Medicine, Seattle, WA.
Griffin, K. W., Friend, R., Eitel, P., & Lobel, M. (1993). Effects of
environmental demands, stress, and mood on health practices. Journal of
Behavioral Medicine, 16, 643–661.
Holzemer, W. L., Corless, I. B., Nokes, K. M., Turner, J. G., Brown, M. A.,
Powell-Cope, G. M., et al. (1999). Predictors of self-reported adherence
in persons living with HIV disease. AIDS Patient Care and STDs, 13,
185–197.
Horowitz, M., Adler, N., & Kegeles, S. (1988). A scale for measuring the
occurrence of positive states of mind: A preliminary report. Psychoso-
matic Medicine, 50, 477–483.
Leserman, J., DiSantostefano, R., Perkins, D. O., Murphy, C., Golden,
R. N., & Evans, D. L. (1994). Longitudinal study of social support and
social conflict as predictors of depression and dysphoria among HIV-
positive and HIV-negative men. Depression, 2, 189–199.
Levy, R. L. (1983). Social support and compliance: A selective review and
critique of treatment integrity and outcome measurement. Social Science
and Medicine, 17, 1329–1338.
MacKinnon, D. P., & Dwyer, J. H. (1993). Estimating mediated effects in
prevention studies. Evaluation Review, 17, 144–158.
Montaner, J. S., Reiss, P., Cooper, D., Vella, S., Harris, M., Conway, B.,
et al. (1998). A randomized, double-blind trial comparing combinations
of nevirapine, didanosine, and zidovudine for HIV-infected patients: The
INCAS Trial. Italy, the Netherlands, Canada and Australia Study. Jour-
nal of the American Medical Association, 279, 930–937.
Moskowitz, J. T. (in press). Positive affect predicts lower risk of AIDS
mortality. Psychosomatic Medicine.
Paterson, D. L., Swindells, S., Mohr, J., Brester, M., Vergis, E. N., Squier,
C., et al. (2000). Adherence to protease inhibitor therapy and outcomes
in patients with HIV infection. Annals of Internal Medicine, 133, 21–30.
Rabkin, J. G., & Chesney, M. A. (1999). Treatment adherence to HIV
medications. In D. G. Ostrow & S. C. Kalichman (Eds.), Psychosocial
and public health impacts of new HIV therapies (pp. 61–82). New York:
Kluwer Academic/Plenum Publishers.
Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for
research in the general population. Journal of Applied Psychological
Measurement, 1, 385–401.
Ryff, C. D., & Singer, B. (1998). The contours of positive human health.
Psychological Inquiry, 9, 1–28.
Samet, J. H., Sullivan, L. M., Traphagen, E. T., & Ickovics, J. R. (2001).
Measuring adherence among HIV-infected persons: Is MEMS consum-
mate technology? AIDS and Behavior, 5, 21–30.
Simoni, J. M., Frick, P. A., Lockhart, D., & Liebovitz, D. (2002). Medi-
ators of social support and antiretroviral adherence among an indigent
population in New York City. AIDS Patient Care and STDs, 16, 431–
439.
Singh, N., Berman, S. M., Swindells, S., Justis, J. C., Mohr, J. A., Squier,
C., & Wagener, M. M. (1999). Adherence of human immunodeficiency
virus-infected patients to antiretroviral therapy. Clinical and Infectious
Disease, 29, 824–830.
Singh, N., Squier, C., Sivek, C., Wagener, M., Nguyen, M. H., & Yu, V. L.
(1996). Determinants of compliance with antiretroviral therapy in pa-
tients with human immunodeficiency virus: Prospective assessment with
implications for enhancing compliance. AIDS Care, 8, 261–269.
Turner-Cobb, J. M., Gore-Felton, C., Marouf, F., Koopman, C., Kim, P.,
Israelski, D., & Spiegel, D. (2002). Coping, social support, and attach-
ment style as psychosocial correlates of adjustment in men and women
with HIV/AIDS. Journal of Behavioral Medicine, 25, 337–353.
418 BRIEF REPORTS
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
