Full text of DOI:10.1592/phco.21.7.676.34571

Drug Therapy and Prevalence of Erectile Dysfunction in
the Massachusetts Male Aging Study Cohort
Carol A. Derby, Ph.D., Marilyn M. Barbour, Pharm.D., Anne L. Hume, Pharm.D.,
and John B. McKinlay, Ph.D.
Study Objective. To examine the association of commonly used drugs with
erectile dysfunction (ED) at two time points.
Design. Population-based, cross-sectional, survey analysis.
Participants. Randomly selected cohort of men in the Massachusetts Male
Aging Study (MMAS) that included 1476 men for the baseline (1987–1989)
and 922 for the follow-up (1995–1997) analyses.
Intervention. Crude associations between specific drug categories were
2
examined with ␹ statistics. Logistic regression analysis was used to
separate the effect of drugs from the influence of heart disease,
hypertension, untreated diabetes, or depressive symptoms.
Measurements and Main Results. In the MMAS, medical history, current
drug use, and erectile function status were ascertained with in-home
interviews. In unadjusted analyses, thiazide and nonthiazide diuretics, ␤-
blockers, calcium channel blockers, angiotensin-converting enzyme
inhibitors, benzodiazepines, digitalis, nitrates, 3-hydroxy-3-methylglutaryl
coenzyme A reductase inhibitors, and histamine2 receptor antagonists were
associated with prevalent ED. Adjustment for comorbidities and health
behaviors attenuated these associations, with only nonthiazide diuretics
and benzodiazepines remaining statistically significant.
Conclusion. Several common drugs may increase prevalence of ED; however,
additional data from larger populations are needed to determine whether
these associations are independent of underlying health conditions and to
explore the effects of dosage and duration of use.
(Pharmacotherapy 2001:21(6):676–683)
affect 10–20 million American men.2 Although it
is not life-threatening, ED has a profound effect
on the well-being of aging men.3 Despite recent
advances in the understanding of erectile
physiology and the causes and treatments of ED,
the epidemiology of ED remains poorly
described. Drugs, including antihistamines,
antihypertensive agents, antidepressants,
tranquilizers, appetite suppressants, and
cimetidine, have been associated with increased
rates of ED,3–6 but studies on drug-induced ED
show conflicting results. The available data
regarding the prevalence and incidence of drug-
induced ED are based on clinical series of
Erectile dysfunction (ED), defined as the
consistent inability to sustain an erection
sufficient for sexual performance,1 is estimated to
From New England Research Institutes, Watertown,
Massachusetts (Drs. Derby and McKinlay); the Department
of Pharmacy Practice, University of Rhode Island, Kingston,
Rhode Island (Drs. Barbour and Hume); and the
Departments of Medicine (Dr. Barbour) and Family
Medicine (Dr. Hume), Memorial Hospital of Rhode Island,
Pawtucket, Rhode Island.
Supported by grants from the National Institute on Aging
and the National Institutes of Diabetes and Digestive and
Kidney Disorders, Bethesda, Maryland.
Address reprint requests to Carol A. Derby, Ph.D., New
England Research Institutes, 9 Galen Street, Watertown, MA
02472.

DRUG THERAPY AND ERECTILE DYSFUNCTION Derby et al
677
patients or on case reports.6 Population-based
data are rare, particularly prospective data from
community samples of aging men.
the Boston, Massachusetts, area. The selection of
the study population and composition of the
analysis samples are illustrated in Figure 1.
Baseline surveys were conducted from
1987–1989 in a random sample of men aged
40–70 years who were residing in 11 cities and
towns in the vicinity of Boston. The communities
were randomly selected by use of probabilities
proportional to the population within six strata
defined by size and median income. Men born
between 1917 and 1946 were drawn at random
from the annual state census list with sampling
fractions that were adjusted to produce a uniform
age distribution across the age groups 40–49,
50–59, and 60–69 years. Introductory letters
were sent to 5287 men and were followed by a
phone call encouraging participation. No
financial incentive was offered. Successful
telephone contacts were completed for 4014
The Massachusetts Male Aging Study (MMAS),
which includes a complete drug database,
provided the opportunity to examine drug-
induced ED in a randomly selected cohort of men
from the general population. The objective of
our analysis was to evaluate the cross-sectional
association of ED with various drugs at two time
points by using data from the baseline and
follow-up evaluations of the participants in the
MMAS.
Methods
Subjects
The MMAS is a longitudinal study of health
status and aging in a cohort of men residing in
Figure 1. Identification of the Massachusetts Male Aging Study (MMAS) cohort and selection of the samples for baseline and
follow-up analyses.

678
PHARMACOTHERAPY Volume 21, Number 6, 2001
men. Of these, 756 subjects were deceased,
inaccessible, ineligible, or unable to complete the
interview. A total of 1709 respondents (52% of
those eligible) completed a baseline evaluation in
their homes. Written informed consent was
obtained, and the protocol was approved by our
institutional review board.
The characteristics of the study population
were consistent with the composition of the
Massachusetts population in 1990.7 The cohort
was predominantly Caucasian (95%). The
majority were employed (78%), married (75%),
and had completed high school (89%). Many
participants had earned at least a bachelor’s
degree (42%).
Follow-up evaluations were completed
between 1995 and 1997. Of the 1709 baseline
participants, 180 (11%) were confirmed
deceased, and 33 (2%) were ineligible because
they had moved out of the country (5 men) or
denied participation due to serious illness (28
men). This left 1496 men who were alive and
eligible for follow-up, of whom 1156 (77%)
completed a follow-up questionnaire. Compared
with men in the baseline sample, men not
included in follow-up were slightly older, were
more likely to have a history of hypertension or
heart disease, had a higher body mass index, and
were less physically active, whereas the
prevalence of cigarette smoking was similar.
For the present analysis, men were excluded if
they reported a history of prostate cancer (82
men at baseline, 62 at follow-up) or current
treatment for diabetes (80 men at baseline, 70 at
follow-up), because of the strong correlation
between these conditions and ED, which could
mask the associations of interest in this analysis.
Men also were excluded if they were missing
information on ED status (71 men at baseline,
102 at follow-up). The analysis samples
consisted of 1476 men at baseline and 922 at
follow-up.
rate 99.2% at baseline, 95.5% at follow-up). The
questionnaire included items related to specific
aspects of erectile function (e.g., frequency and
quality of erections). Subsequent to the baseline
assessment, a National Institutes of Health
consensus panel identified global subjective self-
assessment of ED as the method of choice.1
Therefore, in addition to these items, the follow-
up questionnaire included a single-question
global self-assessment that classified men into
one of four categories of ED status: none,
minimal, moderate, or complete.9 The follow-up
data were used to perform discriminant analysis
of the relationship between the specific item
responses and the global self-assessment. This
yielded the probability that an individual with a
given set of responses would fall into a particular
ED category. These functions were applied to
each MMAS subject at baseline and follow-up,
with each respondent assigned to the ED
category for which they had the greatest
probability of membership.10 The classification
has been shown to be highly correlated with
widely used clinical assessments.9, 10 For the
present analysis, ED status was dichotomized as
none to mild, or moderate to complete. Men
with minimal ED were combined with no ED to
prevent the inclusion of men who did not truly
have the condition in the affected group.
Drug use was ascertained by interviewers’
visual inspection of prescription and over-the-
counter drugs, which were recorded along with
the respondent’s stated reason for use.
Alternative therapy use, such as natural products,
was not reported. Drugs later were coded by a
classification system similar to that of the
American Hospital Formulary Service.11, 12 For
this analysis, drugs were classified into broad
categories according to major drug classes. Only
one man reported use of a drug to treat ED
(alprostadil).
Health status and treatment for selected
conditions were determined by prompted self-
report. For the analysis, history of heart disease,
diabetes, and hypertension were defined as either
having been told by a health professional that the
condition was present or self-report of current
treatment for the condition.
Height and weight were measured by standard
protocol. Body mass index was computed as
weight in kilograms divided by height in meters
squared. Cigarette smoking status was obtained
from self-report. Physical activity level was
ascertained by recall of activities in the past 7
days, including frequency and duration.13 Total
Design
Details of the MMAS design were described
previously.2, 8 In brief, trained interviewer-
phlebotomists visited subjects in their homes and
completed a health questionnaire, psychologic
instruments, physiologic measurements, and
blood sampling for hormone assays within 2
hours of the subject’s awakening.
A self-administered sexual activity question-
naire was completed in private and returned to
the interviewer in a sealed envelope (response

DRUG THERAPY AND ERECTILE DYSFUNCTION Derby et al
679
Table 1. Characteristics of the Baseline (1987–1989) and
Follow-up (1995–1997) Analysis Samples from the MMAS
energy expenditure for the week was estimated
by rating moderate, vigorous, and heavy physical
activity at 17, 25, and 47 kcal/kg/hour,
respectively. Based on the scale described in a
recent recommendation from the Centers for
Disease Control and Prevention,14 total weekly
energy expenditure was categorized as 0, less
than 200, or 200 or more kcal/day of
moderate–intense physical activity. The presence
of depressive symptoms was assessed with the
Center for Epidemiologic Studies Depression
Scale (CES-D).15
Baseline
(n=1476)
Follow–up
(n=922)
Characteristic
Mean (SD) age, yrs
ED (%)
53.8 (8.6)
61.6 (8.3)
Mild
Moderate
23.6
7.6
21.9
10.2
Complete
11.0
27.3 (4.3)
23.9
16.8
27.6 (4.3)
12.4
Mean (SD) BMI, kg/m2
Current smoking (%)
Moderate to intense
physical activity (%)
0 kcal/day
6.5
27.7
65.8
10.7
28.7
3.2
8.7
23.3
68.0
14.7
31.0
2.5
< 200 kcal/day
≥ 200 kcal/day
History of heart disease
History of hypertension
History of diabetesa
Statistical Methods
Crude associations between specific drug
2
categories were examined by using ␹ statistics.
MMAS = Massachusetts Male Aging Study, BMI = body mass index.
aPositive response to “Have you ever been told by a doctor that you
have diabetes?” Men currently treated for diabetes were excluded
from the analysis.
To separate the effect of drugs from the influence
of underlying disease state (i.e., heart disease,
hypertension, untreated diabetes, or depressive
symptoms as defined by the CES-D), crude
associations between ED and drugs were
explored in logistic regression models in which
ED was the dependent variable, drug category
was the independent variable of interest, and
history of heart disease, hypertension, untreated
diabetes, and presence of depressive symptoms
were confounding variables, along with age,
smoking status, and physical activity level.
Testosterone level and body mass index were also
treated as potential confounders but were
dropped from the final models due to lack of
statistical significance (F test p>0.10) and
because they had no substantial effect on
parameter estimates.
of men using thiazide diuretics declined slightly.
The use of 3-hydroxy-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors and histamine
(H)2 receptor antagonists also increased
substantially during the study.
Unadjusted associations between drugs and ED
are shown in Table 2. Use of nonthiazide
diuretics, ␤-blockers, and ACE inhibitors was
reported more frequently by men with, than
those without, ED at both baseline and follow-
up. In the baseline sample, men with ED were
more likely to report use of thiazide diuretics,
whereas in the follow-up sample, those with ED
were more likely to report using calcium channel
blockers. Among the other agents studied,
benzodiazepines, digitalis, or nitrates were
associated with prevalent ED at baseline and
follow-up, whereas HMG-CoA reductase
inhibitors and H2 receptor antagonists were each
associated with ED in the follow-up sample.
Table 3 displays the relation of specific
covariates to ED in the analysis samples. Age
and physical activity level were associated with
ED in both the baseline and follow-up samples.
Current smoking was positively associated with
ED at baseline, whereas history of heart disease
and hypertension were associated with ED at
follow-up.
Results
Descriptive characteristics of the analysis
samples at baseline and follow-up are displayed
in Table 1. On average, the follow-up sample
was 8 years older than the baseline sample and
had approximately half the prevalence of
smoking, a higher proportion with history of
heart disease, and slightly higher prevalence of
hypertension. The prevalence rates of moderate
and complete ED in the baseline and follow-up
analysis samples were 18.6% and 27.0%,
respectively, with the higher prevalence at follow-
up reflecting the aging of the cohort.
For several drugs, the prevalence of use
increased substantially between the baseline and
follow-up surveys. The proportion of men using
␤-blockers, calcium channel blockers,
angiotensin-converting enzyme (ACE) inhibitors,
or ␣1-blockers increased, whereas the proportion
Results of adjusted analyses are displayed in
Figure 2. The majority of the odds ratios (OR)
exceeded one; however, adjustment for
comorbidities and other risk factors attenuated
the associations such that they were no longer

680
PHARMACOTHERAPY Volume 21, Number 6, 2001
Table 2. Prevalence of Use for Selected Drugs by Presence of ED at Baseline (1985–1987) and Follow-up (1995–1997)
Evaluations of the MMAS
Baseline ED Status
Moderate or
Follow-up ED Status
Moderate or
None or
Mild
None or
Mild
Agent
Total
Complete
Total
Complete
Total
100 (922)
81.4 (1201)
18.6 (275)
100 (1476)
73 (673)
27 (249)
Antihypertensive drugs
Thiazides
Other diuretics
␤-Blockers
Calcium channel blockers
ACE inhibitors
Sympatholytics
␣1-Blockers
6.4 (77)
0.9 (11)
7.7 (93)
3.2 (38)
2.3 (27)
1.3 (15)
0.6 (7)
10.6 (29)a
3.6 (10)a
13.1 (36)a
4.7 (13)
5.1 (14)a
1.5 (4)
7.2 (106)
1.4 (21)
8.7 (129)
3.5 (51)
2.8 (41)
1.3 (19)
0.7 (10)
4.8 (32)
1.3 (9)
8.6 (58)
5.9 (40)
8.8 (59)
0.3 (2)
7.6 (19)
6.0 (15)a
18.9 (47)a
16.5 (41)a
18.5 (46)a
0.4 (1)
5.5 (51)
2.6 (24)
11.4 (105)
8.8 (81)
11.4 (105)
0.3 (3)
1.1 (3)
4.9 (33)
4.8 (12)
4.9 (45)
Psychotropic drugs
Selective serotonin reuptake
inhibitors
Other antidepressants
Benzodiazepines
0.08 (1)
1.5 (18)
3.7 (44)
0 (0)
2.9 (8)
8.0 (22)
0.07 (1)
1.8 (26)
4.5 (66)a
3.9 (26)
3.0 (20)
3.6 (24)
4.4 (11)
3.2 (8)
7.6 (19)
4.0 (37)
3.0 (28)
4.7 (43)a
Other drugs
HMG-CoA reductase inhibitors
Digitalis
H2 receptor antagonists
Nitrates
0.2 (2)
0 (0)
3.6 (10)
3.3 (9)
4.4 (12)
0.1 (2)
11.3 (76)
1.6 (11)
8.0 (54)
1.3 (9)
17.7 (44)
6.8 (17)
12.5 (31)
5.6 (14)
13.0 (120)a
3.0 (28)a
9.2 (85)a
2.5 (23)a
1.5 (18)
2.3 (28)
2.0 (24)
1.9 (28)a
2.5 (37)
2.4 (36)a
ED = erectile dysfunction, MMAS = Massachusetts Male Aging Study; ACE = angiotensin-converting enzyme, HMG-CoA = 3-hydroxy-
methylglutaryl coenzyme A. Data are percentages. Numbers in parentheses are number of men.
a
␹2 p<0.05 for comparison of ED none or mild with ED moderate or complete.
effects are clinically relevant.6, 17 Some studies
suggest that thiazides in particular are most likely
to cause ED,18–21 but others have failed to support
this conclusion.22, 23 Authors of one study24 argue
that thiazides lack the necessary central and
autonomic nervous system effects and hormonal
effects to have a significant impact on sexual
dysfunction.
In the present analysis, among men using
nonthiazide diuretics, 76% (16 of 21 men) at
baseline and 75% (18 of 24 men) at follow-up
were using loop diuretics. The remainder were
using potassium-sparing diuretics. In the
analysis sample, among men treated with an
antihypertensive agent, those using a nonthiazide
diuretic were less likely to report history of
hypertension and more likely to report history of
heart disease. Loop diuretics and potassium-
sparing diuretics often are prescribed in the
presence of chronic heart failure or for men with
complicating conditions such as renal failure or
liver disease.25 Similarly, it is possible that men
with symptoms of ED are more likely to be
prescribed ACE inhibitors because they have a
lower risk of causing ED. Thus, the observed
association of these agents with ED may be
confounded by the presence of comorbidities.
Benzodiazepines were associated with ED at
statistically significant. The exceptions were
nonthiazide diuretics and benzodiazepines,
which were associated with ED in the baseline
sample (relative OR 2.81 and 1.98, 95%
confidence interval [CI] 1.11–7.14 and
1.11–3.53, respectively). These two drug
categories showed a similar pattern at follow-up,
although the 95% CIs around the ORs included
one.
Discussion
The study of drug-induced ED is confounded
by the fact that often the underlying disease state
or indication for therapy is itself related to
increased frequency of ED. Results from the
MMAS suggest cross-sectional associations
between ED and many commonly used drugs.
However, associations were attenuated after
adjustment for underlying chronic diseases, age,
and health behaviors. The exceptions were
nonthiazide diuretics and benzodiazepines,
which were independently associated with ED in
the baseline sample.
Diuretics are believed to cause ED by
decreasing vascular resistance4 or inducing low
serum zinc levels.16 However, there is a lack of
consensus regarding the extent to which these

DRUG THERAPY AND ERECTILE DYSFUNCTION Derby et al
681
Table 3. Association of Covariates with ED: Odds Ratios and 95% Confidence Intervals at
Baseline (1985–1987) and Follow-up (1995–1997) Evaluations of the MMAS
Baseline
Age-adjusted OR
Follow-up
Age-adjusted OR
Covariate
95% CI
95% CI
Age (per yr)
1.10
1.02
1.57
1.08–1.11
0.98–1.05
1.15–2.14
1.11
1.03
1.08
1.08–1.13
1.00–1.07
0.66–1.75
BMI (kg/m2)
Cigarette smoking status
Physical activity
Sedentary
Moderate
Active
History of heart disease
History of hypertension
History of diabetesa
2.22
1.35
1.00
1.33
1.20
1.37
1.36–3.65
1.0–1.84
1.83
1.44
1.00
2.04
2.00
1.81
1.06–3.15
0.99–2.09
0.90–1.98
0.89–1.62
0.70–2.70
1.36–3.06
1.44–2.79
0.72–4.54
ED = erectile dysfunction; MMAS = Massachusetts Male Aging Study; OR = odds ration; CI = confidence interval;
BMI = body mass index.
aPositive response to “Have you ever been told by a doctor that you have diabetes?” Men currently treated for
diabetes were excluded from the analysis.
both baseline and follow-up, although only the
baseline result reached statistical significance.
Benzodiazepines interfered with sexual function
in some studies, although few studies examined
this issue directly. The available literature
suggests that the effects of benzodiazepines on
sexual function may be minimal.22, 26 It was
hypothesized that the sedative effect of
benzodiazepines results in decreased libido
among users, or that decreased libido stems from
Figure 2. Adjusted odds ratios (95% confidence intervals) for erectile dysfunction by drug use at the baseline (B) and follow-up
(F) evaluations of the MMAS cohort. Baseline values were adjusted for age, cigarette smoking status, and physical activity level.
Follow-up values were estimated from logistic models adjusted for age, physical activity level, heart disease, and hypertension.
At each time, models for benzodiazepines were adjusted for depressive symptoms as defined by the Center for Epidemiologic
Studies Depression Scale.

682
PHARMACOTHERAPY Volume 21, Number 6, 2001
the anxiety states for which treatment is
prescribed.26 An anxiety scale was not included
in our study; thus, we were not able to evaluate
whether anxiety confounded the observed
association. An alternative explanation for the
observed trend may be that both ED and
benzodiazepines are associated with ischemic
vascular disease. The results of a population-
based study27 suggest that benzodiazepine use
may be associated with increased incidence of
ischemic heart disease, whereas other reports28, 29
show an association between cardiovascular
disease and ED. However, a plausible biologic
mechanism for an association between
benzodiazepines and ischemic vascular disease is
lacking,27 and as noted by the authors, the
observed association may be confounded by the
inability to adequately adjust for the underlying
psychiatric condition.
ED at follow-up in relation to drug use among
men initially free of ED. However, the obser-
vation of a crude association between several
drugs and ED suggests that selection bias does
not entirely explain our findings.
A strength of the study design is that we were
able to confirm the baseline observations in the
follow-up sample. However, although the
prevalence of drug use increased during follow-
up, the power to detect statistically significant
associations was reduced due to attrition and the
smaller overall sample. Nevertheless, the follow-
up analyses yielded results that were in the same
direction and of a similar magnitude to those at
baseline.
Summary
Consistent with previous studies, these cross-
sectional results from the MMAS suggest that
several commonly used drugs are associated with
ED, but that the associations may be due in part
to confounding by the conditions for which the
drugs are prescribed. Although the MMAS is the
largest random-sample, population-based study
with data on the epidemiology of ED, a limitation
of these analyses is that for many of the drugs
studied, the sample size was limited. Thus, the
power to detect associations between drugs and
ED was low, and we were not able to explore the
relationship according to dosage or duration of
drug use. Whether the lack of statistically
significant results after adjustment is due to lack
of power or lack of an independent association
cannot be determined in this sample. It is also
possible that our analysis is limited by the
definition of ED used. Because the single-
question assessment was added after the baseline
survey, in response to the National Institutes of
Health consensus conference,1 we developed a
surrogate measure based on our sexual activity
questionnaire. However, this definition appears
to be highly correlated with widely used clinical
assessment instruments.9, 10
In these cross-sectional analyses, it is not
possible to establish whether drug use is causing
ED or whether the observed associations are due
to the underlying indications for treatment.
However, the associations persisted after
statistical adjustment for presence of underlying
conditions.
Additional data from even larger samples are
needed to further explore the association of ED
with dosage and duration of various drugs and to
separate the effects of various drugs taken in
␤-Blockers were associated with ED in several
studies.21, 30–32 The hypothesized mechanisms
include suppression of central nervous system
sympathetic outflow and inhibition of ␤2-
mediated peripheral vasodilation.4 Noncardio-
selective ␤-blockers in particular were associated
with increased frequency of ED, with
cardioselective agents showing smaller effects.4, 22,
26, 33
In the present study, the majority of ␤-
blockers were ␤1-selective agents, which
composed 72% and 80% of all ␤-blockers at
baseline and follow-up, respectively. By contrast,
noncardioselective agents (e.g., propranolol)
composed only 26% of all ␤-blocker use at
baseline, and 10% of ␤-blockers at follow-up.
The lack of association between ACE inhibitors
or calcium channel blockers with ED is
consistent with results of previous studies.4, 33–35
Because subclasses of calcium channel blockers
have differential primary sites of action, they may
be related to ED to different degrees, although to
our knowledge no studies have provided direct
comparisons.4 Because of the limited sample, we
were unable to divide the category of calcium
channel blockers into dihydropyridine and
nondihydropyridine groups.
In this cross-sectional analysis, we were not
able to determine whether symptoms or
perceived risk of ED led to the choice of
treatment or whether the treatment itself led to
the development of ED. For example, the failure
to find an association between thiazide use and
ED may be due to selection for treatment such
that men with ED symptoms are less likely to be
prescribed this class of drug. Unfortunately, the
study sample was too small to examine incident

DRUG THERAPY AND ERECTILE DYSFUNCTION Derby et al
683
research in the general population. Appl Psychol Meas
1977;1:385–401.
16. Morely JE. Impotence. Am J Med 1986;80:897–905.
17. Jaffe A, Chen Y, Kisch ES, Fischel B, Alon M, Stern N. Erectile
dysfunction in hypertensive subjects: assessment of potential
determinants. Hypertension 1996;28:859–62.
18. Chang SW, Fine R, Siegel D, Chesney M, Black D, Hulley SB.
The impact of diuretic therapy on reported sexual function.
Arch Intern Med 1991;151:2402–8.
combination. Also needed are data from
populations that include diverse ethnic groups.
This is particularly important given the aging of
the population and the age-associated increase in
prevalence of ED, comorbidities, and use of
prescription drugs.
19. Croog SH, Levine S, Sudilovsky A, Baume RM, Clive J. Sexual
symptoms in hypertensive patients: a clinical trial of
antihypertensive medications. Arch Intern Med
1988;148:788–94.
20. Muller SC, El-Damanhoury H, Ruth J, Lue TF. Hypertension
and impotence. Eur Urol 1991;19:29–34.
Acknowledgment
We wish to thank Dr. John Morley for sharing his
expertise by reviewing this manuscript.
21. Medical Research Council Working Party on Mild to
Moderate Hypertension. Adverse reactions to bendrofluazide
and propranolol for the treatment of mild hypertension. Lancet
1981;2:539–43.
22. Weiss RJ. Effects of antihypertensive agents on sexual function.
Am Fam Physician 1991;44:2075–82.
23. Grimm RH, Cohen JD, Smith W, Falvo-Gerard L, Neaton JD,
and the Multiple Risk Factor Inter vention Group .
Hypertension management in the multiple risk factor
intervention trial (MRFIT). Six-year intervention results for
men in special intervention and usual care groups. Arch Intern
Med 1985;145: 1191–9.
24. Stevenson JG, Ulmstead GS. Sexual dysfunction due to
antihypertensive agents. Drug Intell Clin Pharm 1984;18:
113–21.
25. Johnson JA, Parker RB, Geraci SA. Heart failure. In: DiPiro JT,
Talbert RL, Yee GC, Matzke GR, Wells BG, Posey LM.
Pharmacotherapy: a pathophysiologic approach, 4th ed.
Stamford, CT: Appleton & Lange, 1999:153–181.
26. Brock GB, Lue TF. Drug-induced sexual dysfunction: an
update. Drug Safety 1993;8:414–26.
27. Lapane KL, Zierler S, Lasater TM, Barbour MM, Carleton RA,
Hume AL. Is the use of psychotropic drugs associated with
increased risk of ischemic heart disease? Epidemiology
1995;6:376–81.
References
1. NIH Consensus Development Panel on Impotence.
Impotence. JAMA 1993;270:83–90.
2. Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ,
McKinlay JB. Impotence and its medical and psychosocial
correlates: results of the Massachusetts Male Aging Study. J Urol
1994;151:54–61.
3. Krane RJ, Goldstein I, de Tejada IS. Impotence. N Engl J Med
1989;321:1648–59.
4. Barksdale JD, Gardner SF. The impact of first-line
antihypertensive drugs on erectile dysfunction. Pharmaco-
therapy 1999;19:573–81.
5. Goldstein I, Krane RJ. Drug-induced sexual dysfunction.
World J Urol 1983;1:239–43.
6. Wein AJ, Van Arsdalen KN. Drug-induced male sexual
dysfunction. Urol Clin North Am 1988;15:23–31.
7. U.S. Census Bureau. Census of population and housing.
Summary tape file 3C, part 1, 1990.
8. McKinlay JB, Feldman HA. Age-related variation in sexual
activity and interest in normal men: results from the
Massachusetts male aging study. In: Rossi AS, ed. Sexuality
across the life course. Chicago: University of Chicago Press,
1994.
9. Derby CA, Araujo AB, Johannes CB, Feldman HA, McKinlay
JB. Measurement of erectile dysfunction in population–based
studies: the use of a single question self-assessment in the
Massachusetts Male Aging Study. Int J Impotence Res
2000;12:197–204.
10. Kleinman KP, Feldman HA, Johannes CB, Derby CA,
McKinlay JB. A new surrogate variable for erectile dysfunction
status in the Massachusetts male aging study. J Clin Epidemiol
2000;53:71–7.
28. Feldman HA, Johannes CB, Derby CA, et al. Erectile
dysfunction and cardiovascular risk: prospective results from
the Massachusetts male aging study. Prev Med 2000;4:328–38.
29. Wabrek AJ, Burchell RC. Male sexual dysfunction associated
with coronary heart disease. Arch Sex Behav 1980;9:69–75.
30. Burnett WC, Chahine RA. Sexual dysfunction as complication
of propranolol therapy in men. Cardiovasc Med 1979;4:811–15.
31. Hollifield JW, Sherman K, Zeagg RU, Shand DG. Proposed
mechanisms of propranolol’s antihypertensive effect on
essential hypertension. N Engl J Med 1976;295:68–73.
32. Wassertheil-Smoller S, Blaufox MD, Oberman A, et al. Effect
of antihypertensives on sexual function and quality of life: the
TAIM study. Ann Intern Med 1991;114:613–20.
33. Grimm RH, Grandits GA, Prineas RJ, et al. Long-term effects
on sexual function of five antihypertensive drugs and
nutritional hygienic treatment in hypertensive men and
women: Treatment of mild hypertension study (TOMHS).
Hypertension 1997;29:8–14.
34. Suzuki H, Tominaga T, Kumagai H, Saruta T. Effects of first-
line antihypertensive agents on sexual functioning and sex
hormones. J Hypertens 1988;6:S649–51.
11. McEvoy GK, ed. AHFS drug information. Bethesda, MD:
American Society of Hospital Pharmacists, 1989.
12. Barbour MM, Hume AL, Assaf AR, Carleton RA. The drug
database of the Pawtucket heart health program: description
and development. Pharmacotherapy 1993;13:235–43.
13. U.S. Department of Health and Human Services. Physical
activity and health: a report of the surgeon general. Atlanta,
GA: U.S. department of health and human services, centers for
disease control and prevention, national center for chronic
disease prevention and health promotion, 1996.
14. Pate RR, Pratt M, Blair SN, et al. Physical activity and public
health: a recommendation from the centers for disease control
and prevention and the American college of sports medicine.
JAMA 1995;273:402–7.
35. Croog SH, Levine S, Testa MA, et al. The effects of
antihypertensive therapy on quality of life. N Engl J Med
1986;314:1657–64.
15. Radloff LS. The CES-D scale: a self-report depression scale for