A Study of Antihypertensive Drugs and Depressive Symptoms (SADD-Sx) in

Patients Treated With a Calcium Antagonist Versus an Atenolol Hypertension

Treatment Strategy in the International Verapamil SR-Trandolapril Study
(INVEST)
L. DOUGLAS RIED, PHD, MICHAEL J. TUETH, MD, EILEEN HANDBERG, PHD, STUART KUPFER, MD, CARL J. PEPINE, MD,
AND THE INVEST STUDY GROUP

Background: The International Verapamil/Trandolapril Study (INVEST) demonstrated comparable efficacy between verapamil SR
and atenolol antihypertensive treatment strategies for clinical outcomes and blood pressure (BP) control in hypertensive patients
with coronary artery disease (N ⫽ 22,576). Effects of these antihypertension strategies on mood-related issues are not well
understood. Objectives: The objectives of this study were 1) to compare depressive symptoms by strategy and 2) to identify
predictors of depressive symptoms in INVEST patients after 1 year of follow up. Design, Setting, and Patients: Depressive
symptoms were assessed in a subset (N ⫽ 2317) of consecutively randomized U.S. patients enrolled between April 1, 1999, and
October 31, 1999. Patients were mailed surveys after randomization and after 1 year of treatment. Intervention: Patients were
assigned to either a verapamil SR or atenolol strategy to achieve Joint National Committee on Prevention, Detection, Evaluation,
and Treatment of High Blood Pressure BP goals. Trandolapril and/or hydrochlorothiazide were recommended as add-on agents.
Main Outcome Measure: Depressive symptoms were measured by the Center for Epidemiologic Studies–Depression (CES-D)
scale. Results: CES-D scores improved 1.45 points (p ⬍ .001) after 1 year in patients assigned to the verapamil SR strategy,
whereas a nonsignificant improvement was observed in patients assigned to the atenolol strategy (0.27 points, p ⫽ .44). Predictors
of higher depressive symptoms were higher baseline CES-D score (p ⬍ .001), history of depression diagnosis (p ⫽ .03), history
of stroke (p ⬍ .001), and assignment to the atenolol strategy (p ⬍ .001). Conclusions: A verapamil SR strategy is a viable
alternative to beta-blocker therapy for hypertensive patients with coronary artery disease, especially those at risk of depression. Key
words: antihypertensive, beta-blocker, calcium channel blocker, coronary artery disease, depressive symptoms, quality of life.

CAD ⫽ coronary artery disease; MI ⫽ myocardial infarction;
SADD-Sx ⫽ Study of Antihypertensive Drugs and Depressive
Symptoms; INVEST ⫽ International Verapamil SR-Trandolapril
Study; CES-D ⫽ Center for Epidemiologic Studies–Depression
scale; JNC VI ⫽ Joint National Committee on Prevention, Detec-
tion, Evaluation, and Treatment of High Blood Pressure.
INTRODUCTION
A n important and often overlooked consequence of coro-
nary artery disease (CAD) and hypertension, and their
treatment, is mental depression. Depression is a risk factor for
death and myocardial infarction (MI) among patients with
CAD (1–3). Choice of antihypertensive treatment may influ-
ence risk for depression and consequently clinical outcomes.
Rates of “depression” for beta-blockers are reportedly higher
than with calcium antagonists (4,5). Some have called for a
halt of the use of beta-blockers to treat hypertension among
older persons because of depression-related side effects,
among other reasons (6). However, evidence regarding the
relationship between beta-blockers and depression is unclear
From the Rehabilitation Outcomes Research Center (L.D.R.) and the De-
partment of Psychiatry (M.J.T.), Malcom Randall Veterans Affairs Medical
Center, Department of Veterans Affairs, Gainesville, Florida; the College of
Pharmacy, University of Florida, Gainesville, Florida (L.D.R.); the Depart-
ment of Psychiatry (M.J.T.) and the Division of Cardiovascular Medicine,
Department of Medicine (E.H., C.J.P.), College of Medicine, University of
Florida, Gainesville, Florida; and Abbott Laboratories, Abbott Park, Illinois
(S.K.).
Address correspondence and reprint requests to L. Douglas Ried, PhD,
Pharmacy Health Care Administration, College of Pharmacy, PO Box
100496, University of Florida, Gainesville, FL 32610-0496. E-mail:
ried@cop.ufl.edu
Received for publication July 21, 2004; revision received November 30,
2004.
This study was supported by Abbott Laboratories and the University of
Florida.
DOI: 10.1097/01.psy.0000160468.69451.7f
(7,8), and clarification of this relationship may have signifi-
cant clinical implications (9).
There is a paucity of rigorous studies comparing risk dif-
ferences of depression and depressive symptoms among pa-
tients prescribed antihypertensive agents, including beta-
blockers and calcium antagonists. Studies that examined risk
of depression associated with antihypertensive treatment have
been relatively small and limited by study design (8). More-
over, the relationship between beta-blockers and depressive
symptoms may be confounded in patients with concomitant
diseases. Without large, randomized studies, risk differences
between hypertension treatments, including beta-blockers and
calcium antagonists, remains uncertain.
To the best of our knowledge, no large-scale, prospective,
randomized clinical trials have compared calcium antagonist-
based and beta-blocker-based hypertension treatment strate-
gies with respect to risk of depression or depressive symptoms
among patients with CAD. Accordingly, the objectives of this
study were 1) to compare depressive symptoms after 1 year of
hypertension treatment with either a verapamil SR or atenolol
strategy in a subset of patients with CAD from the Interna-
tional Verapamil SR-trandolapril Study (INVEST) and 2) to
determine predictors of depressive symptoms.
METHODS
Study Design
The Study of Antihypertensive Drugs and Depressive Symptoms (SADD-
Sx) is a substudy of INVEST. A description of INVEST and its patients,
randomization procedures, intervention, and implementation have previously
been reported (10,11). Briefly, INVEST was a randomized, open-label,
blinded end-point study of 22,576 hypertensive patients with CAD aged ⱖ50
years conducted from September 1997 to February 2003. Patients were
randomized to antihypertensive treatment with either a verapamil SR- or
atenolol-based strategy to achieve blood pressure (BP) control according to
the sixth report of the Joint National Committee on Prevention, Detection,

398 Psychosomatic Medicine 67:398 – 406 (2005)

0033-3174/05/6703-0398
Copyright © 2005 by the American Psychosomatic Society
ANTIHYPERTENSIVE DRUGS AND DEPRESSIVE SYMPTOMS
Evaluation, and Treatment of High Blood Pressure (JNC VI) (12). Patients
were excluded from INVEST if they were taking a beta-blocker at random-
ization.
Consecutively randomized INVEST patients residing in the United States
were mailed SADD-Sx questionnaires between April 1, 1999, and October 31,
1999 (N ⫽ 2317). SADD-Sx patients were mailed a health-related quality-
of-life survey that contained a measure of depressive symptoms. Information
on educational level, living status, and psychiatric history was obtained by the
survey. SADD-Sx was conducted according to the principles of the Declara-
tion of Helsinki. The University of Florida Institutional Review Board ap-
proved the study protocol.
Patients were mailed baseline surveys the day after randomization and
follow-up surveys were sent at 1 year. If surveys were not returned within 10
working days, a second survey was mailed. Approximately 2 weeks before
follow-up surveys were mailed, a letter asking for continued support was
mailed to each patient to enhance response rate. If patients did not respond to
the second survey, no further attempts were made to contact them for the
purposes of SADD-Sx.
Self-rated Depressive Symptoms
The Center for Epidemiologic Studies-Depression (CES-D) scale was
used to assess depressive symptoms (13). The CES-D is a 20-item self-
reported rating scale designed to measure current levels of depressive symp-
toms. It is widely used and is a reliable (13,14) and valid instrument (13,15–
17). Scores range from 0 to 60; higher scores indicate more depressive
symptoms. As a screening tool, the CES-D has been used to estimate the
presence of clinically significant depression (15,17). Scores ⱖ16 are gener-
ally consistent with depressive symptoms of clinically depressed patients
(13,17). People with major chronic medical conditions are most likely to score
in the high depressive symptoms range (18), so a higher threshold of 23 is
recommended for studies of older persons with chronic illnesses (19). Study
conclusions were similar whether the ⱖ16 or ⱖ23 threshold was used, so only
results of the higher threshold analysis are reported. Patients were excluded if
they were missing more than four of the 20 CES-D items. If the patient was
missing between one and four CES-D items, person mean imputation was
used for those missing items (20,21).
Predictor Variables
Pharmacologic Hypertension Treatment Strategy
Assignment to pharmacologic hypertension treatment strategy was the
primary explanatory variable of this study (10). Study patients assigned to
the atenolol strategy were assigned a value of zero (0) and those assigned
to the verapamil SR strategy were assigned value of one (1).
Covariates
Sociodemographic variables commonly associated with depression and
depressive symptoms were included in the model as covariates. Gender, race
(white versus nonwhite), cohabitation (living alone or with someone), and
education (high school graduate vs nonhigh school graduate) were collected
from the mail survey. Patient age and medical history were obtained during
the baseline INVEST visit and abstracted from the baseline forms electroni-
cally submitted to the INVEST data coordinating center. Finally, patients
self-reported their overall feeling of well-being as excellent, good, fair, or
poor, which were dichotomized into “poor/fair” (0) and “good/excellent” (1).
History of cancer, stroke/transient ischemic attack, hypercholesterolemia,
abnormal coronary angiogram, angina, MI, coronary artery bypass graft, left
ventricular hypertrophy, congestive heart failure, and cardiac arrhythmias
were noted at baseline. Other medical conditions noted were Alzheimer’s
disease, diabetes, gastrointestinal bleeding, Parkinson’s disease, renal impair-
ment, and peripheral vascular disease. Patients were assigned a value of “1”
if the condition was present. In addition, patients’ smoking history (1) and
history of stopping and restarting smoking (1) were included in the model.
Finally, information about patients’ psychiatric history was obtained (22).
Patients were asked if “. . . a medical doctor or psychiatrist has ever told you
that you were depressed” (0 ⫽ no or I don’t know; 1 ⫽ yes). No further
corroboration of the history of treatment for depression was made, and no
Psychosomatic Medicine 67:398 – 406 (2005)
additional information was collected regarding onset, duration, severity, or
antidepressant or psychologic treatment (23).
Statistical Analyses
Differences in sociodemographic characteristics, medical and cardiac
conditions, and mental health risk factors between participants in the two
treatment strategies were compared using Pearson’s chi-square and inde-
pendent t tests for categorical and continuous data, respectively. Baseline
and 1-year mean scores within each treatment strategy were compared
using paired t tests. Unadjusted risk ratios were calculated for the explan-
atory variables and risk of reporting high levels of depressive symptoms
(ⱖ23).
Depressive symptoms were examined using multivariate models. The
independent effects of sociodemographic, medical and cardiac conditions, and
mental health risk factors on depressive symptoms were evaluated stepwise,
entering them into the equation in clinically and theoretically relevant groups.
Sociodemographic characteristics associated with depression and depressive
symptoms were first entered, followed by medical conditions and mental
health risk factors associated with depressive symptoms. Lastly, hypertension
treatment strategy was added to the regression equation. If the change in
explained variance (R2) in the final step was statistically significant, then
addition of the hypertension treatment strategy significantly improved pre-
diction of depressive symptoms. We tested for interactions in all regression
models.
The a priori level of statistical significance was alpha ⫽ 0.05. A sufficient
number of patients were enrolled to detect a small effect size (d ⫽ 0.20) with
a power of 80 at the a priori alpha level. Statistical analyses were conducted
with the Statistical Package for the Social Sciences (SPSS) version 10.0.05
(24).
RESULTS
The final sample consisted of 2317 INVEST patients as-
signed to either the verapamil SR-based (n ⫽ 1184) or ateno-
lol-based (n ⫽ 1133) treatment strategy. Useable survey re-
sponses were obtained for the baseline survey only, and both
the baseline and 1-year follow-up surveys from 68.1% (n ⫽
1578) and 51.4% (n ⫽ 1192) of study patients, respectively.
Gender (␹2 ⫽ 7.19, p ⫽ .007) and race (␹2 ⫽ 92.57, p ⬍
.001) were the only significant predictors of nonresponse to
the baseline survey. Within gender and race, baseline return
rates were similar for the two BP treatment strategies
(females: ␹2 ⫽ 1.92, p ⫽ .17; males: ␹2 ⫽ 0.08, p ⫽ .78 and
whites: ␹2 ⫽ 1.36, p ⫽ .24 and nonwhites: ␹2 ⫽ 0.15, p ⫽
.70). Figure 1 details reasons for not completing either the
baseline or follow-up survey (n ⫽ 1125). Only data for
those returning both the baseline and 1-year surveys are
presented.
Baseline Characteristics
Table 1 summarizes baseline characteristics, including fre-
quently occurring medical conditions that may have influ-
enced the patient’s mood. Nearly 44% (n ⫽ 518) of the
SADD-Sx patients were female and the majority was ⱖ65
years old (59%) and white (80%). The remaining patients
were black (16%), Hispanic or another ethnic origin (4.3%).
At baseline, approximately 86% of SADD-Sx patients were
taking antihypertensive medications. Among those who com-
pleted both surveys, nearly 68% were high school graduates or
had some college education. Finally, 75% of the study patients
lived with someone, usually a spouse (60%) or another rela-
tive (20%).
399

Figure 1. Flow diagram of patients’ progression through the SADD-Sx substudy of the International Verapamil SR-Trandolapril Study (n ⫽ 2317). SADD-Sx ⫽
Study of Antihypertensive Drugs and Depressive Symptoms.
At baseline, depressive symptoms, history of depression,
proportion of individuals at high risk of depression (CES-D
ⱖ23), and CES-D score were similar between treatment
groups, as were sociodemographic characteristics, systolic and
diastolic BP, medical and cardiovascular health, and history of
hypertension treatment (Table 1). Similar results were ob-
served when all randomized patients (N ⫽ 2317) were exam-
ined (data not shown).
Change in Depressive Symptoms
On average, depressive symptoms improved between base-
line and 1 year in this sample of SADD-Sx patients (CES-D
score ⫽ 14.12 versus 13.24, paired t test ⫽ 3.65, p ⬍ .001).
Depressive symptoms were lower after 1 year among patients
assigned to the verapamil SR strategy (CES-D score ⫽ 12.54,
standard deviation [SD] ⫽ 10.31) compared with patients
assigned to the atenolol strategy (CES-D score ⫽ 14.00, SD ⫽
11.60; independent t test ⫽ 2.30, p ⫽ .02, Fig. 2). Patients
assigned to the verapamil SR strategy improved an average of
1.45 points (14.00 versus 12.54; paired t test ⫽ 4.37, p ⬍
.001) on the CES-D scale, whereas improvement was negli-
gible among patients assigned to the atenolol strategy (14.27
versus 14.00; 0.27 points, paired t test ⫽ 0.77, p ⫽ .44).
400
L. D. RIED et al.
Predictors of Depressive Symptoms
After 1 year, a lower proportion of patients assigned to the
verapamil SR strategy reported high depressive symptoms
than patients assigned to the atenolol strategy (17.0% versus
21.6%; relative risk [RR] ⫽ 0.95, 95% confidence interval
[CI] ⫽ 0.89 – 0.99). When the unadjusted risk was examined
at 1 year (Fig. 3), women, nonwhite patients, patients with
lower educational level, patients with poorer perceptions of
well-being, and those living alone were at higher risk. On
average, patients reporting high depressive symptoms were
younger. High-risk patients averaged 64.1 year compared with
67.3 years (t ⫽ 4.69, p ⬍ .001). Patients at higher risk also
were more likely to have a history of smoking, stopping and
restarting smoking at baseline, a history of angina, as well as
stroke. All baseline depression indicators were significant risk
factors for higher depressive symptoms. High baseline risk of
depression increased the 1-year risk by nearly sevenfold, and
a diagnosis of depression before randomization was associated
with more than triple the risk.
Multivariate Model
When entered as a group in the first model, female gender,
living alone, lower educational status, poorer health status,
Psychosomatic Medicine 67:398 – 406 (2005)
ANTIHYPERTENSIVE DRUGS AND DEPRESSIVE SYMPTOMS

TABLE 1. Comparison of Patients at Baseline Assigned to the Verapamil SR Strategy and the Calcium and Atenolol Strategy Among SADD-Sx
Patients Returning Both Baseline and 1-Year Surveys (N ⴝ 1192)

Variable Atenolol Strategy (N ⫽ 575) Verapamil SR Strategy (N ⫽ 617) P Value

Demographic characteristics
Femalea 41.2% 45.5% 0.13
Whitea 79.1% 80.4% 0.59
Not a high school graduateb 33.9% 31.3% 0.33
Live aloneb 25.3% 25.3% 0.99
Poor or fair well-beinga 20.2% 22.9% 0.26
Patient average age at entry—years (SD) 66.8 (9.4) 66.5 (9.4) 0.63c
Medical history at baseline
Hypercholesteremiaa 66.6% 65.5% 0.68
History of smoking at baselinea 59.3% 53.6% 0.83
Within past 30 daysa 15.5% 14.6% 0.67
Diabetesa 28.5% 28.0% 0.85
Cancera 5.9% 4.2% 0.18
Alzheimer’sa 0.3% 0.5% 0.72
Cardiovascular history at baselinea
Mean systolic blood pressurea (SD) 148.2 (18.2) 147.5 (19.3) 0.47c
Mean diastolic blood pressurea (SD) 83.1 (11.3) 83.0 (11.6) 0.91c
Abnormal coronary angiograma 65.0% 64.7% 0.89
Angina pectorisa 46.6% 43.8% 0.33
Myocardial infarctiona 44.0% 45.9% 0.52
Coronary artery bypass graft ⱖ1 month agoa 30.1% 27.9% 0.40
Left ventricular hypertrophya 19.0% 17.8% 0.62
Peripheral vascular diseasea 12.9% 15.1% 0.27
Congestive heart failurea 7.1% 5.5% 0.25
Arrhythmiasa 7.8% 7.5% 0.81
Strokea 5.4% 6.5% 0.43
Medication history at baselined
Angiotensin-converting enzyme inhibitora 41.7% 40.2% 0.59
Calcium antagonista 43.5% 41.8% 0.56
Risk factors for depressive symptoms
High risk of depression at baselineb 39.1% 36.3% 0.32
History of depression diagnosisb 17.1% 17.8% 0.74
Family member with a history of depressionb 18.8% 21.8% 0.20
Mean score on CES-Db (SD) 14.27 (10.90) 13.99 (11.05) 0.67c

SD ⫽ standard deviation; SADD-Sx ⫽ Study of Antihypertensive Drugs and Depressive Symptoms.

a
Information regarding these variables was obtained from INVEST data.
b
Information regarding these variables was obtained from the SADD-Sx baseline mail survey from those patients who returned both the baseline and 1-year
follow-up surveys.
c
Independent t test.
d
Patients were excluded from INVEST if they were taking a beta-blocker at randomization.

and younger age were associated with more depressive symp-
toms (Table 2). Sociodemographic characteristics accounted
for nearly 10% of the explained variance.
When adjusted for baseline medical conditions, only stroke
and angina added significantly to the second model. As ex-
pected, the mental health variables added to the prediction of
higher depressive symptoms to the greatest degree in the third
model. Patients at high risk of depression at baseline or with
a history of depression before randomization reported more
depressive symptoms at 1 year. Stroke was the only medical
condition that remained statistically significant after adjusting
for the mood-related and psychiatric history variables. After
these adjustments, the influence of female gender, living
alone, lower educational status, poorer well-being, younger
age, and baseline angina on depressive symptoms was no
longer statistically significant.
Finally, patients assigned to the verapamil SR strategy had
lower depressive symptoms after adjusting for all covariates in
the fourth model. When the variables were entered simulta-
neously, significant predictors of higher depressive symptoms
after 1 year were baseline history of stroke, higher baseline
levels of depressive symptoms, self-reported history of a de-
pression diagnosis, and BP treatment strategy. After adjusting
for the covariates, patients assigned to the verapamil SR
strategy scored 1.49 points lower on the CES-D scale (95%
CI ⫽ ⫺0.60 to ⫺2.38) than patients assigned to the atenolol
strategy. Overall, the model explained 53% of the variance.
DISCUSSION
SADD-Sx compared change in depressive symptoms after
1 year of follow up in U.S. patients with CAD and hyperten-
sion who were assigned to a verapamil SR or atenolol strategy

Psychosomatic Medicine 67:398 – 406 (2005) 401


Figure 2. Mean difference in Center for Epidemiologic Studies–Depression
scale depressive symptom scores by assignment group at baseline and 1year
(n ⫽ 1192).
as part of INVEST. We found that depressive symptoms
improved for the overall SADD-Sx population. For those
assigned to the verapamil SR strategy, depressive symptoms
improved significantly, whereas there was little change for
those assigned to the atenolol strategy. On average, patients
assigned to the verapamil SR strategy had lower depressive
symptoms than those assigned to the atenolol strategy at 1
year when adjusted for baseline conditions.
Few studies have directly compared strategies containing a
calcium antagonist versus a beta-blocker with regard to de-
pression or depressive symptoms. Our results appear to differ
from previous studies. Two nonrandomized, retrospective
studies found no differences between treatments containing a
beta-blocker versus a calcium antagonist for measures of
depression (25,26). In a prospective, observational study, cal-
cium antagonists and beta-blockers were both associated with
depressive symptoms, but only calcium antagonists were as-
sociated with a depression diagnosis (27). In each of these
studies, beta-blockers and calcium antagonists were examined
as drug classes.
Our study’s design has other important distinctions from
previous studies (8). First, to reduce confounding, we focused
on patients with medical conditions that already placed them
at high risk of depressive symptoms. It also used a prospective
study design, random allocation to treatment, and adequate
statistical power. To our knowledge, ours is the first trial to
compare risk of depression between atenolol and verapamil
strategies with JNC VI BP target goals (12). More impor-
tantly, this study is applicable to everyday practice. The strat-
egy approach of BP treatment for patients with cardiovascular
disease reflects everyday practice and treatment guidelines
(12,28). Our most important finding is that the improvement
in depressive symptoms with the verapamil SR strategy was
significant, even after considering patients’ mental health his-
tory and baseline depressive symptoms.
Several explanations for our findings are possible. First, the
402
L. D. RIED et al.
observed difference in depressive symptoms could be ex-
plained if there was a corresponding difference in clinical end
points. However, risk for the INVEST primary outcome (first
occurrence of death, nonfatal MI, or nonfatal stroke) was
similar for the verapamil SR- and atenolol-strategy groups at
1 year in SADD-Sx (RR ⫽ 1.02, 95% CI ⫽ 0.73–1.42) and
was similar to all US INVEST patients (RR ⫽ 0.93, 95% CI ⫽
0.80 –1.07). Although no difference in adverse outcome was
seen in both cases, higher mortality has been observed at
levels of depressive symptoms not generally considered clin-
ically significant and below levels usually considered predic-
tive of increased postacute MI mortality (3). In those studies,
longer time periods than we were able to observe in INVEST
transpired before a relation between depression and cardio-
vascular outcomes became apparent (29 –31).
Next, physiological mechanisms that account for the im-
provement in BP could account for the decline in depressive
scores at 1 year. Vascular disease burden (i.e., hypertension,
diabetes, heart disease, metabolic syndrome) contributes to
small-vessel disease and has been hypothesized to be associ-
ated with depressive symptoms (32–34). INVEST patients in
both strategies achieved similar levels of BP control (11). So,
the finding that changes in depressive symptoms at 1 year was
not associated with follow-up systolic or diastolic BP is not
completely explained by the vascular burden hypothesis and
may indicate differential effects from the verapamil SR strat-
egy. One plausible explanation for the observed difference
between the two INVEST BP treatment strategies is verapam-
il’s mood-related effects. Clinically, although verapamil is not
indicated as first-line monotherapy for major depressive dis-
order, it does have a role as an add-on to a mood stabilizer in
bipolar affective disorder or antidepressant treatment for
mood enhancement among severely depressed or refractory
patients (35). Although verapamil was prescribed for its BP-
lowering effects in this study, a latent benefit might be its
effect on patients’ mood, albeit small. On the other hand,
although BP reduction may have improved depressive symp-
toms of patients assigned to the atenolol strategy, its adverse
somatic side effects such as decreased energy and increased
fatigue may have negated its beneficial effects, resulting in no
observable change in patients’ depressive symptoms.
Our findings add to the evidence regarding atenolol and
depressive symptoms. On average, patients’ depressive symp-
toms did not change over the 1-year period when assigned to
the atenolol strategy. With the exception of the Beta-Blocker
Heart Attack Trial (BHAT) (36,37), ours is the only prospec-
tive, randomized study designed to evaluate the association
between depression and an atenolol-containing antihyperten-
sive strategy. This finding has important clinical implications
given the value of beta-blockers in the treatment of specific
conditions (i.e., acute MI). Beta-blockers may be underused in
elderly MI survivors because of concerns about their depres-
sogenic effects, leading to measurable adverse outcomes (9).
SADD-Sx provides new evidence that a BP treatment strategy
containing atenolol can be given to patients with hypertension
and CAD without significantly increasing their risk for de-
Psychosomatic Medicine 67:398 – 406 (2005)
ANTIHYPERTENSIVE DRUGS AND DEPRESSIVE SYMPTOMS

Figure 3. Comparison of patients’ unadjusted risk of high depressive symptoms after 1 year (n ⫽ 1192). *Unadjusted risk of being classified as “at risk of
depressive symptoms” Center for Epidemiologic Studies–Depression scale score ⱖ23. ACE ⫽ angiotensin-converting enzyme; BP ⫽ blood pressure; CABG ⫽
coronary artery bypass graft; RR ⫽ relative risk; SD ⫽ standard deviation; SADD-Sx ⫽ Study of Antihypertensive Drugs and Depressive Symptoms.

pression. All the same, when both are equally clinically indi-
cated, it seems prudent to recommend a verapamil SR strategy
rather than an atenolol strategy for patients with CAD who
already are despondent, depressed, dysphoric, or have a per-
sonal or family history of depression.
The patients’ initial mental health status, history of stroke,
and hypertension treatment strategy were important predictors
of subsequent depressive symptoms. Of these predictors, the
most potent one was the initial level of depressive symptoms.
It is important for physicians making decisions about hyper-
tension treatment to do a brief screening to ascertain whether
a specific patient may be at greater risk. Brief screening
instruments are available that require only 2 to 5 minutes to
fully complete and may indicate the need for further assess-
ment of the patient’s mood before making a final decision
regarding therapy.
The magnitude of change in CES-D scores for patients in
the verapamil-SR strategy was small but statistically signifi-
cant. Previous research has shown that an increase of five
points or more in the CES-D score may be associated with
increased risk of stroke, MI, and mortality (38). Potent inter-
ventions such as exercise have resulted in five-point improve-
ments in CES-D score (39). Our finding of an average 1.5-
point reduction for patients assigned to the verapamil-SR
strategy was a little less than one third of the five-point
difference deemed clinically significant in these studies. Nev-
ertheless, this was not an intervention study specifically in-
tended to modify mood. The optimization of BP control using
two equally effective strategies, however, did result in differ-
ences in CES-D scores between the two strategies. In SADD-
Sx, 23% of patients assigned to the atenolol strategy worsened
by five points or more over the year, whereas 19% assigned to
the verapamil SR strategy worsened by five points or more.
Conversely, 29% of patients assigned to the verapamil-SR

Psychosomatic Medicine 67:398 – 406 (2005) 403

 TABLE 2. Final Model Regressing 1-Year Level of Depressive Symptoms on Sociodemographic, Medical History, Cardiovascular History at Baseline, Past Depressive Symptoms, and

404
Treatment Assignment (n ⴝ 1106)a

Variableb Model 1 ␤ Coeffc t Value P Value Model 2 ␤ Coeffc t Value P Value Model 3 ␤ Coeffc t Value P Value Model 4 ␤ Coeffc t Value P Value

Demographic
Female 0.09 2.89 0.09 2.82 0.00 ⫺0.11 0.00 0.38
0.004 0.005 0.91 0.97
White ⫺0.03 ⫺0.97 ⫺0.02 ⫺0.63 0.00 0.17 0.00 0.19
0.33 0.53 0.86 0.84
Not a high school graduate 0.19 6.52 0.18 6.41 0.04 1.69 0.04 1.63
p⬍.001 p⬍.001 0.09 0.10
Live alone ⫺0.09 ⫺3.19 ⫺0.09 ⫺3.21 ⫺0.04 ⫺1.68 ⫺0.04 ⫺1.64
0.001 0.001 0.09 0.10
Poor/fair self-reported well-being ⫺0.14 ⫺4.88 ⫺0.13 ⫺4.39 ⫺0.03 1.23 ⫺0.03 1.34
p⬍.001 p⬍.001 0.22 0.18
Patient age at randomization ⫺0.16 ⫺5.50 ⫺0.16 ⫺5.16 ⫺0.03 ⫺1.18 ⫺0.03 ⫺1.29
p⬍.001 p⬍.001 0.24 0.20
Medical history
Hypercholesteremia ⫺0.01 ⫺0.35 ⫺0.01 ⫺0.62 ⫺0.01 ⫺0.66
0.72 0.54 0.51
History of smoking at baseline 0.03 0.82 0.01 0.30 0.01 0.29
0.41 0.77 0.77
Within past 30 days 0.02 0.50 ⫺0.03 ⫺1.27 ⫺0.03 ⫺1.37
0.62 0.21 0.17
Cancer 0.02 0.61 0.01 0.38 0.01 0.22
0.54 0.70 0.86
Cardiovascular history at baseline
Abnormal coronary angiogram ⫺0.02 ⫺0.53 ⫺0.01 ⫺0.28 0.01 ⫺0.27
0.60 0.78 0.79
Angina pectoris 0.06 2.13 0.00 0.18 0.00 0.11
0.03 0.86 0.92
Myocardial infarction 0.02 0.78 0.03 1.47 0.03 1.56
0.44 0.14 0.12
Coronary artery bypass graft ⱖ1 month ago 0.05 1.46 0.03 1.45 0.03 1.37
0.15 0.15 0.17
Left ventricular hypertrophy 0.04 1.51 0.03 1.29 0.03 1.23
0.13 0.20 0.22
Congestive heart failure 0.02 0.61 0.03 1.35 0.03 1.34
0.54 0.18 0.18
Peripheral vascular disease 0.01 0.24 ⫺0.01 ⫺0.23 ⫺0.00 ⫺0.10
0.81 0.82 0.92
Stroke 0.09 3.13 0.07 3.44 0.07 3.54
0.002 0.001 p⬍.001
Arrhythmias ⫺0.03 ⫺1.00 ⫺0.01 ⫺0.58 ⫺0.01 ⫺0.56
0.32 0.60 0.57
Depressive symptoms
Risk of depression at baseline 0.68 28.32 0.68 28.40
p⬍.001 p⬍.001
History of a depression diagnosis 0.05 2.31 0.05 2.34
0.02 0.02
Antihypertensive treatment
Assignment to verapamil SR strategy ⫺0.07 ⫺3.29
0.001
Model-adjusted R2 0.10 0.11 0.52 0.53
F-ratio of model 21.37 8.11 59.73 58.0
Significance of model p⬍.001 p⬍.001 p⬍.001 p⬍.001
F-ratio of R2 change 1.89 483.50 10.80
Significance of R2 change 0.03 p⬍.001 0.001

a
The number of patients differs from the original sample because of incomplete data among the sociodemographic variables.
b
The variables indicating that the study patient was already taking an angiotensin-converting enzyme inhibitor or calcium antagonist at baseline were not included in the model because of their lack of
significance in any of the previous models.
c
L. D. RIED et al.

Psychosomatic Medicine 67:398 – 406 (2005)

Standardized regression coefficient.
ANTIHYPERTENSIVE DRUGS AND DEPRESSIVE SYMPTOMS
strategy improved by five points or more, whereas 26% of
patients assigned to the atenolol strategy improved by a sim-
ilar amount. In both cases, the direction of change favored the
verapamil-SR strategy. Even more importantly, although the
clinical significance of the choice of BP medication may be
small by itself, it may become significant in combination with
other depression risk factors. For example, the risk of depres-
sion is typically greater for females than males. In SADD-Sx,
females assigned to the atenolol strategy were 63% more
likely to worsen by five points or more than females assigned
to the verapamil SR group (p ⬍ .04). In contrast, males’ risk
for worsening by five points or more was similar for both BP
treatment strategies. Finally, an issue to consider when placing
these findings into clinical perspective is the influence of BP
treatment choice relative to other clinical factors known to
foreshadow depression. In SADD-Sx, BP treatment strategy
choice was similar in magnitude to stroke and a history of
depression in the prediction of depressive symptoms. More-
over, it was even greater than other signs and symptoms of
serious cardiovascular disease that foreshadow depression,
including congestive heart failure, arrhythmias, angina, and
MI, to name a few. In summary, although the impact of BP
treatment choice on clinical outcomes may be small by itself,
it may be significant in combination with patients’ other risk
factors. One goal of preventive care is to modify risk factors
to reduce the likelihood of adverse outcomes. Why add an-
other risk factor without a clinically compelling indication if
given an equivalent choice? The next step in this line of
inquiry is an in-depth study to ascertain the clinical signifi-
cance of the role of medication choice in the relationship
between worsening mood and the occurrence of death, non-
fatal MI, and nonfatal stroke.
Limitations of this study should be noted. First, although
study noncompletion and survey nonresponse are potential
confounders, patients assigned to the treatment strategies were
similar with respect to noncompletion and survey nonresponse
rates. In addition, none of the other sociodemographic factors
associated with nonresponse were significant in the final
model, although potential limitations resulting from the sam-
ple being largely older and approximately one third without a
high school degree is acknowledged and should be investi-
gated. Second, the nonsignificant finding between BP and
depressive symptoms may be partially explained by a floor
effect. The baseline mean diastolic blood pressure was within
normal range and the mean systolic BP only was about 148
mm Hg in both BP strategies. The implications of these
findings merits further investigation with respect to the de-
pressogenic effects of higher vascular burden found in persons
with higher BP or for persons with combinations of cardio-
vascular risk factors (i.e., high BP, diabetes, and hypercholes-
terolemia). Finally, whether the findings from this study are
generalizable to the whole U.S. INVEST population is un-
clear. For example, the proportion of females and Hispanic
patients was lower in SADD-Sx compared with the entire U.S.
INVEST sample (data available on request). This finding
needs further investigation.
Psychosomatic Medicine 67:398 – 406 (2005)
In conclusion, the primary implication of our findings is
that a practitioner’s choice between two equally effective
hypertension treatment strategies may affect a patient’s mood
to the same extent as other important depression risk factors,
including stroke or history of depression. On average, the
mood of patients assigned to the verapamil SR strategy im-
proved, whereas it did not improve among those assigned to
the atenolol strategy. A verapamil SR strategy is a viable
alternative to a beta-blocker strategy for hypertensive patients,
especially those already at risk of depression.
We thank Wolfgang Wittenberg, Heather Bristol, Rhonda Cooper-
DeHoff, and Summana Moolasarn for their support during the con-
duct of the study and writing of the manuscript.
REFERENCES
1. Ariyo AA, Haan M, Tangen CM, Rutledge JC, Cushman M, Dobs A,
Furberg CD. Depressive symptoms and risks of coronary heart disease
and mortality in elderly Americans. Cardiovascular Health Study Collab-
orative Research Group. Circulation 2000;102:1773–9.
2. Blumenthal JA, Lett HS, Babyak MA, White W, Smith PK, Mark DB,
Jones R, Mathew JP, Newman MF. Depression as a risk factor for
mortality after coronary artery bypass surgery. Lancet 2003;362:604 –9.
3. Bush DE, Ziegelstein RC, Tayback M, Richter D, Stevens S, Zahalsky H,
Fauerbach JA. Even minimal symptoms of depression increase mortality
risk after acute myocardial infarction. Am J Cardiol 2001;88:337– 41.
4. Physician’s Desk Reference. Oradell, NJ: Medical Economics; 2003.
5. Lacy CF, Armstrong LL, Goldman MP, Lance LL. Drug Information
Handbook 2000 –2001, 8th ed. Hudson, OH: Lexi-Comp, Inc; 2000.
6. Beevers DG. Beta-blockers for hypertension: time to call a halt. J Hum
Hypertens 1998;12:807–10.
7. Ko DT, Hebert PR, Coffey CS, Sedrakyan A, Curtis JP, Krumholz HM.
Beta-blocker therapy and symptoms of depression, fatigue, and sexual
dysfunction. JAMA 2002;288:351–7.
8. Ried LD, McFarland BH, Johnson RE, Brody KK. Beta-blockers and
depression: the more the murkier? Ann Pharmacother 1998;32:699 –708.
9. Soumerai SB, McLaughlin TJ, Spiegelman D, Hertzmark E, Thibault G,
Goldman L. Adverse outcomes of underuse of beta-blockers in elderly
survivors of acute myocardial infarction. JAMA 1997;277:115–21.
10. Pepine CJ, Handberg-Thurmond E, Marks RG, Conlon M, Cooper-
DeHoff R, Volkers P, Zellig P, for the INVEST Investigators. Rationale
and design of the International Verapamil SR/Trandolapril Study
(INVEST): an Internet-based randomized trial in coronary artery disease
patients with hypertension. J Am Coll Cardiol 1998;32:1228 –37.
11. Pepine CJ, Handberg EM, Cooper-DeHoff RM, Marks RG, Kowey P,
Messerli FH, Mancia G, Cangiano JL, Garcia-Barreto D, Keltai M,
Erdine S, Bristol HA, Kolb HR, Bakris GL, Cohen JD, Parmley WW,
INVEST Investigators. A calcium antagonist vs a non-calcium antagonist
hypertension treatment strategy for patients with coronary artery disease.
The International Verapamil-Trandolapril Study (INVEST): a random-
ized controlled trial. JAMA 2003;290:2805–16.
12. Joint National Committee on Prevention Detection Evaluation and Treat-
ment of High Blood Pressure. The sixth report of the Joint National
Committee on prevention, detection, evaluation, and treatment of high
blood pressure. Arch Intern Med 1997;157:2413– 46.
13. Radloff LS. The CES-D scale: a self-report depression scale for research
in the general population. J Appl Psychol Meas 1977;1:385– 401.
14. Turner RJ, Noh S. Physical disability and depression: a longitudinal
analysis. J Health Soc Behav 1988;29:23–37.
15. Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ.
Assessing depressive symptoms in five psychiatric populations: a vali-
dation study. Am J Epidemiol 1977;106:203–14.
16. Husaini BA, Neff JA, Harrington JB, Hughes MD, Stone RH. Depression
in rural communities: validating the CES-D scale. J Comm Psychol
1980;8:20 –7.
17. Roberts RE, Vernon SW. The Center for Epidemiologic Studies Depres-
sion Scale: its use in a community sample. Am J Psychiatry 1983;140:
41– 6.
18. Berkman LF, Berkman CS, Kasl S, Freeman DH, Leo L, Ostfeld AM,
Cornoni-Huntley J, Brody JA. Depressive symptoms in relation to phys-
405

ical health and functioning in the elderly. Am J Epidemiol 1986;124:
372– 88.
19. Himmelfarb S, Murrell SA. Reliability and validity of five mental health
scales in older persons. J Gerontol 1983;38:333–9.
20. Ying YW. Nonresponse on the Center for Epidemiological
Studies–Depression scale in Chinese Americans. Int J Soc Psychiatry
1989;35:156 –93.
21. Callahan CM, Wolinsky FD. The effect of gender and race on the
measurement properties of the CES-D in older adults. Med Care 1994;
32:341–56.
22. Robins LN, Helzer JE, Croughan J, Ratcliff KS. National Institute of
Mental Health Diagnostic Interview Schedule: its history, characteristics,
and validity. Arch Gen Psychiatry 1981;38:381–9.
23. Cohen HW, Madhavan S, Alderman MH. History of treatment for
depression: risk factor for myocardial infarction in hypertensive patients.
Psychosom Med 2001;63:203–9.
24. Statistical Package for the Social Sciences, version 10.0.5R. Chicago:
SPSS Inc; 1999.
25. Gerstmann BB, Jolson HM, Bauer M, Cho P, Livingston JM, Platt R. The
incidence of depression in new users of beta-blockers and selected
antihypertensives. J Clin Epidemiol 1996;49:809 –15.
26. Carney RM, Rich MW, teVelde A, Saini J, Clark K, Freedland KE.
Prevalence of major depressive disorder in patients receiving beta-
blocker therapy versus other medications. Am J Med 1987;83:223– 6.
27. Ried LD, Johnson RE, McFarland BH, Brody KK. Antihypertensive drug
use and the risk of depression among older hypertensives in an HMO.
J Pharmacoepidemiol 2000;8:1–28.
28. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL
Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ, Joint
National Committee on Prevention, Detection, Evaluation, and Treatment
of High Blood Pressure, National Heart, Lung, and Blood Institute,
National High Blood Pressure Education Program Coordinating Commit-
tee. The Seventh Report of the Joint National Committee on Prevention,
406
L. D. RIED et al.
Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7
report. JAMA 2003;289:2560 –72.
29. Jorge RE, Robinson RG, Arndt S, Starkstein S. Mortality and poststroke
depression: a placebo-controlled trial of antidepressants. Am J Psychiatry
2003;160:1823–9.
30. Schulz R, Beach SR, Ives DG, Martire LM, Ariyo AA, Kop WJ. Asso-
ciation between depression and mortality in older adults: the Cardiovas-
cular Health Study. Arch Intern Med 2000;160:1761– 8.
31. Wulsin LR, Vaillant GE, Wells VE. A systematic review of the mortality
of depression. Psychosom Med 1999;61:6 –17.
32. Lyness JM, Caine ED, King DA, Conwell Y, Cox C, Duberstein PR.
Cerebrovascular risk factors and depression in older primary care
patients: testing a vascular brain disease model of depression. Am J
Geriatr Psychiatry 1999;7:252– 8.
33. Mast BT, Neufeld S, MacNeill SE, Lichtenberg PA. Longitudinal support
for the relationship between vascular risk factors and late-life depressive
symptoms. Am J Geriatr Psychiatry 2004;12:93–101.
34. Alexopoulos GS, Meyers BS, Young RC, Campbell S, Silbersweig D,
Charlson M. ‘Vascular depression’ hypothesis. Arch Gen Psychiatry
1997;54:915–22.
35. Hollister LE, Trevino ES. Calcium channel blockers in psychiatric
disorders: a review of the literature. Can J Psychiatry 1999;44:658 – 64.
36. Hjalmarson AC. Use of beta blockers in postinfarct prophylaxis: aspects
on quality of life. Am Heart J 1987;114:245–50.
37. Davis BR, Furberg CD, Williams CB. Survival analysis of adverse effects
data in the Beta-Blocker Heart Attack Trial. Clin Pharmacol Ther 1987;
41:611–5.
38. Wassertheil-Smoller S, Applegate WB, Berge K, Chang CJ, Davis BR,
Grimm R, Kostis J, Pressel S, Schron E. Change in depression as a
precursor of cardiovascular events. Arch Intern Med 1996;156:553– 61.
39. Gowans SE, deHueck A, Abbey SE. Measuring exercise-induced mood
changes in fibromyalgia: a comparison of several measures. Arthritis
Rheum 2002;47:603–9.
Psychosomatic Medicine 67:398 – 406 (2005)