Clinical Endocrinology (2007) 66, 235–245 doi: 10.1111/j.1365-2265.2006.02715.

ORIGINAL ARTICLE

Effects of testosterone and nandrolone on cardiac function:

Blackwell Publishing Ltd

a randomized, placebo-controlled study

T. Chung*, S. Kelleher†, P. Y. Liu†, A. J. Conway†, L. Kritharides* and D. J. Handelsman†

Departments of *Cardiology and †Andrology, Concord Hospital & ANZAC Research Institute, University of Sydney, Sydney,
New South Wales, Australia

8·3 ± 1·8, P = 0·02) and shortened PR interval on the electrocardio-

Summary gram (167 ± 13 vs. 154 ± 12, P = 0·03).
Conclusion Four weeks of treatment with testosterone or nan-
Background Androgens have striking effects on skeletal muscle, drolone had no benefical or adverse effects compared with placebo
but the effects on human cardiac muscle function are not well on cardiac function in healthy young men.
defined, neither has the role of metabolic activation (aromatization,
5α reduction) of testosterone on cardiac muscle been directly studied. (Received 2 May 2006; returned for revision 6 June 2006; finally
Objective To assess the effects of testosterone and nandrolone, a revised 1 August 2006; accepted 31 August 2006)
non-amplifiable and non-aromatizable pure androgen, on cardiac
muscle function in healthy young men.
Design Double-blind, randomized, placebo-controlled, three-arm
parallel group clinical trial. Introduction
Setting Ambulatory care research centre. Testosterone has been used at either physiological doses as androgen
Participants Healthy young men randomized into three groups of replacement therapy for men with disorders of the hypothalamo-
10 men. pituitary-testicular axis1,2 or at pharmacological doses to exploit
Intervention Weekly intramuscular injections of testosterone selective effects on muscle, bone, haematopoeitic marrow or other
(200 mg mixed esters), nandrolone (200 mg nandrolone decanoate) salutary effects in men with a range of chronic diseases 2,3 since 1937.4
or matching (2 ml arachis oil vehicle) placebo for 4 weeks. Recent studies have demonstrated striking effects of testosterone on
Main outcome measures Comprehensive measures of cardiac 5–10
skeletal muscle mass and strength. Yet despite nearly seven decades
muscle function involving transthoracic cardiac echocardiography of clinical use, the effects of androgens on cardiac muscle function
measuring myocardial tissue velocity, peak systolic strain and strain has attracted little systematic study especially using modern non-
rates, and bioimpedance measurement of cardiac output and sys- invasive methods.11
tematic vascular resistance. Androgen abuse among elite power athletes and bodybuilders has
Results Left ventricular (LV) function (LV ejection fraction, LV become entrenched since its origin in international sporting com-
modified TEI index), right ventricular (RV) function (ejection area, petition as proxy conflicts of the Cold War era.12 Androgen abuse
tricuspid annular systolic planar motion, RV modified TEI index) characteristically involves high doses of nandrolone and testosterone
as well as cardiac afterload (mean arterial pressure, systemic vascular in particular, but also other synthetic androgens to a lesser extent.
resistance) and overall cardiac contractility (stroke volume, cardiac These are usually taken in intermittent, short-term cyclic regimens.13,14
output) were within age- and gender-specific reference ranges and In addition, the abuse of nonprescribed androgens for cosmetic and
were not significantly (P < 0·05) altered by either androgen or recreational motives has extended to the wider community.15,16 The
placebo over 4 weeks of treatment. Minor changes remaining within potential for cardiovascular harm arising from androgen abuse is
normal range were observed solely within the testosterone group primarily derived from anecdotal case reports of cardiomyopathy,
for: increased LV end-systolic diameter (30 ± 7 vs. 33 ± 5 mm, P = 0·04) myocardial infarction, arrhythmia, hypertension, cardiac failure,
and RV end-systolic area (12·8 ± 1·3 vs. 14·6 ± 3·3 cm2, P = 0·04), stroke, pulmonary embolism and sudden death.17–21 However, sudden
reduced LV diastolic septal velocity (Em, 9·5 ± 2·6 vs. 8·7 ± 2·0 cm/s, cardiac deaths among healthy young athletes not using androgens
P = 0·006), increased LV filling pressure (E/Em ratio, 7·1 ± 1·6 vs. are well known22 and there are few studies examining direct cardiac
effects of androgens to evaluate deleterious and/or beneficial effects.
In theory, if anabolic effects of androgens were similar between car-
diac and skeletal muscle, chronic heart failure might be improved
Correspondence: Professor D. J. Handelsman, ANZAC Research Institute, with androgen therapy.23,24 Indeed, randomized placebo-controlled
Sydney NSW 2139, Australia. Tel.: +61-2-9767 9100; Fax: +61-2-9767 9101; studies show that testosterone administered to patients with chronic
E-mail: [email protected] heart failure reduces systemic vascular resistance and increases both

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd 235
236 T. Chung et al.

cardiac output and overall exercise capacity.24–26 However, effects on At baseline and weeks 2 and 4, blood was obtained 7 days after
objective echocardiographic findings were minimal, raising the the last treatment to measure trough hormones (LH, testosterone),
possibility that improvements could have been indirect through haematological (full blood count) and biochemical [total, low-density
improved exercise endurance arising from skeletal muscle or lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol,
psychological effects. The effects of androgens on damaged and triglycerides, urea, creatinine, liver function tests, serum iron] variables.
undamaged myocardium might also differ. One recent randomized Participants completed a mood and sexual function questionnaire
placebo-controlled study has also shown no deleterious effects on that evaluated sexual function (six questions reflecting desire and
cardiac structure or function in 16 athletes following 8 weekly satisfaction) over the last week and mood (11 variables) over the last
injections of 200 mg nandrolone decanoate.11 However, the effects day. A 12-lead electrocardiogram, transthoracic echocardiogram
of testosterone have not been similarly evaluated to determine and non-invasive haemodynamic measurements were performed at
whether androgen activation via amplification by 5α reduction and/ baseline and week 4. At the end of study participants were asked to
or aromatization to oestradiol influences effects of androgens on guess whether they were on active drug or placebo.
cardiac function.
Conventional echocardiographic evaluation, as applied in the
Transthoracic echocardiogram
above studies, may not have sufficient sensitivity to detect subtle
alterations of myocardial function associated with drugs including Transthoracic echocardiograms were performed with the patients in
androgens. For example, preclinical cardiomyopathy unrecognized the left lateral position using a Vingmed Vivid 5 (GE Medical Sys-
by conventional echocardiography can be reliably detected using more tems, Milwaukee, WI, USA). All images were acquired from standard
sensitive evaluation of systolic and diastolic function incorporating views, stored digitally on magneto optical disks and analysed offline
measurement of tissue velocity, strain and strain rate imaging.27 using EchoPAC PC Version 3·1·3 (GE Vingmed Ultrasound). To reduce
Whether acute androgen usage affects myocardial contraction or variability of the measurements, a single experienced operator blinded
relaxation as assessed by these methods is unknown. to treatment performed and analysed all the echocardiograms.
We therefore undertook a comprehensive non-invasive evaluation The primary outcome was a change in left ventricular (LV) systolic
of cardiac function using transthoracic echocardiography (myocardial function, quantified by LV ejection fraction (LVEF). LV dimensions
tissue velocity, peak systolic strain and strain rate) and bioimpedance (LV end-diastolic and LV end-systolic diameters) were measured in
analysis (cardiac output, systematic vascular resistance) in healthy the parasternal long axis view.28 LVEF was measured by biplane
28
young men treated with supraphysiological doses of testosterone, Simpson’s rule. Apical views were used for colour tissue Doppler
its non-amplifiable and non-aromatizable analogue nandrolone imaging (TDI) of the LV with all measurements averaged over three
or a matching placebo for 4 weeks in a double-blind, randomized consecutive cardiac cycles. For myocardial tissue velocity measure-
placebo-controlled study. ments, sample areas with a 5 mm diameter were placed at the septal
and lateral mitral annulus and within the basal, mid and apical seg-
ments of the six LV walls (septal, lateral, inferior, anterior, posterior
Methods
and anteroseptal walls).29 Strain and strain rate imaging were derived
The study was approved by the Human Ethics Committee of the for the basal and mid LV segments with the same sample areas used
Central Sydney Area Health Service (Concord Zone). Healthy young for tissue velocity measurements.29,30 An offset distance of 12 mm
men aged 18– 45 years were eligible for the study. Exclusion criteria was used for both strain studies. Tissue velocities [systolic (Sm), early
included (i) contraindications to testosterone administration (breast diastolic (Em) and late diastolic (Am)], peak systolic strains and peak
or prostate cancer); (ii) previous or current use of androgens; (iii) use systolic strain rates were measured. Measurements for the six LV
of disallowed (including illicitly obtained) drugs that may interfere walls were averaged to derive mean values for the basal, mid and apical
with androgen absorption, distribution, metabolism, excretion or LV segments. Global LV function was measured using the modified
action (e.g. androgens, anti-androgens, finasteride); (iv) any chronic TEI index.31 LV filling velocities [early peak (E wave) and atrial peak
medical condition requiring regular medication or likely to interfere (A wave)] and the derived E/A ratio were measured from pulse wave
with safe participation; (v) and significant cardiac disease or abnormal Doppler signal of mitral inflow.28 Deceleration time, a marker of LV
baseline echocardiogram. All participants were warned about risks stiffness, is measured from the early filling of the mitral inflow. Using
of disqualification from elite sports and temporary lowering of male the transmitral E wave and the averaged Em from the septal and
fertility arising from participation in the study and were required to lateral mitral annulus, an E/Em ratio was derived for estimation of
provide written informed consent prior to entry. LV end-diastolic filling pressures.32,33
Immediately prior to first treatment, volunteers (n = 30) were Right ventricular (RV) function was assessed on apical four-
allocated to three equal groups according to a computer-generated chamber view with all measurements averaged over three consecutive
randomization list with a block size of 6. The randomization list was cardiac cycles. Markers of RV dilatation (RV end-diastolic area, RV
held in the pharmacy that drew up all injections to maintain blinding end-systolic area and RV:LV end-diastolic area ratios) and RV systolic
of study personnel. Each group received 4 weekly intramuscular function (RV ejection area and tricuspid annular motion) were
injections of either 200 mg testosterone esters (Sustanon), 200 mg measured.34–37 Colour TDI derived myocardial tissue velocities (Em,
nandrolone decanoate (Deca Durabolin), or placebo (arachis oil Am, Sm), peak systolic strains and peak systolic strain rates were
vehicle) from Organon (Australia) Pty Ltd. All injections were measured at the tricuspid annulus and the basal, mid and apical seg-
administered by study personnel. ments of the RV free wall.35,38 Global RV function was measured

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
 Androgen effects on cardiac function 237

using the modified TEI index.39 The pulmonary artery systolic pressure study were measured together for each analyte within a single assay
was estimated from the RV-RA pressure difference derived from the having a within-assay CV of < 10%.
tricuspid regurgitant envelope obtained by continuous Doppler.34
Atrial areas and ejection fractions, respectively, were measured at
Statistical methods
end-systole by planimetry and by the area-length method on apical
four-chamber view.28 Pulse wave Doppler signal derived systolic, Baseline characteristics were expressed as mean ± standard deviations.
diastolic and atrial reversal velocities from the pulmonary venous All measurements were analysed in a two-way repeated measure
inflow28 and colour TDI derived atrial velocities (Am) from the mid- analysis of variance () with time (baseline, weeks 2 and 4) as
atrial walls were measured as markers of atrial function. three level within-subject factor and treatment groups (testosterone,
The coefficient of variation (CV, expressed as percentage of mean nandrolone, placebo) as a three level between-subject factor. The
value) for echocardiographic measurements was established in 10 main effects of time and treatment groups and their interaction effect
healthy adults in our hospital echocardiographic laboratory. The CV for each measurement were examined. As randomization would be
for echocardiographic measurements ranged between 5% and 15% expected to balance baseline variables, a statistical effect of treatment
for most parameters. Cardiac chamber dimensions and left ventricu- was defined as a time–treatment interaction with a P-value of < 0·05.
lar ejection fraction demonstrated CV of < 6%. Tissue Doppler Post hoc Tukey test was used to identify differences within groups
determinations of myocardial velocity, strain and strain rate had a when a significant interaction effect was present. Repeated measures
CV of between 4% and 16%.  and power analysis were performed using SPSS (SPSS Inc.,
Chicago, IL, USA) or NCSS 2001 (Statistical Systems, Kaysville, UT,
USA) software.
Non-invasive haemodynamic measurements

Haemodynamic measurements were obtained by thoracic electrical

Results
bioimpedance (Bioz, CardioDynamics International Corp., San
Diego, CA, USA). The technique has been validated by comparison The three groups were well matched for age and anthropometric
with other methods40 and is highly reproducible.41,42 Surface elec- measures (overall mean age, height, weight, body mass index (BMI),
trodes were applied at the base of the neck and thorax where a low- body surface area (BSA)) as well as baseline haematological and
voltage high-amplitude alternating current is introduced through biochemical variables (Table 1). All 30 men completed the study and
the outermost sensors and sensed through the innermost sensors. none reported any adverse effects. Men receiving active androgens
The difference between the voltages introduced and sensed indicate (12/20, 60%) were no more likely than those on placebo (5/10, 50%)
the thoracic impedance and is inversely proportional to the thoracic to identify correctly they were on an active drug rather than placebo
blood volume.43 Stroke volume was derived from change in imped- (OR = 1·5, 95% CI 0·25–9·0).
ance/time measured during electrical systole, calculated using the Z Both androgens had the expected supraphysiological hormonal
MARC (modulating Aortic Compliance) algorithm.42 Coupled with effects (Fig. 1). There was a single unexplained outlier in the testo-
heart rate and blood pressure measurements, real-time measure- sterone treatment group for basal LH, but the basal LH levels
ments of cardiac output, cardiac index, stroke index and systemic did not differ significantly between groups. Both testosterone and
vascular resistance were calculated.42 nandrolone treatment suppressed LH (P = 0·045). Nandrolone sup-
pressed and testosterone marginally increased blood testosterone
concentrations (P < 0·001).
Hormone assays
Consistent with known androgen effects, compared to baseline;
Blood LH and testosterone were measured by immunoflurometric the haemoglobin, haematocrit, red cell counts and platelet counts
immunoassays as described previously.44–47 Nandrolone did not were mildly increased at weeks 2 and 4 in the testosterone and nan-
cross-react in the testosterone immunoassay. All samples from the drolone groups. Cholesterol and triglyceride levels were unchanged

Fig. 1 Plot of serum testosterone (left panel)

and LH (right panel) for men treated with
nandrolone (filled diamond), testosterone
(filled square) or placebo (open circle). Data are
mean and standard error of mean before and
after 2 and 4 weeks of treatment with matching
injections.

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
 238 T. Chung et al.
Table 1. Haematological and biochemical markers

Controls (n = 10) Testosterone (n = 10) Nandrolone (n = 10)

Baseline Week 2 Week 4 Baseline Week 2 Week 4 Baseline Week 2 Week 4 F value P value

Haematological markers
Haemoglobin (g /l) 151 ± 9 151 ± 9 149 ± 8 144 ± 5 152 ± 7** 151 ± 6* 148 ± 6 152 ± 7 153 ± 10* 4·0 0·007**
White cell count (109/l) 7·9 ± 3·29 8·2 ± 3·3 7·4 ± 2·8 6·0 ± 1·6 7·1 ± 1·5 6·1 ± 1·2 13·4 ± 25·9 7·2 ± 1·9 6·7 ± 1·8 0·7 0·57
Platelets (109/l) 309 ± 74 306 ± 68 302 ± 73 239 ± 46 274 ± 51** 259 ± 55* 232 ± 48 268 ± 50** 261 ± 39** 5·4 0·001**
Red cell count (1012/l) 5·1 ± 0·5 5·1 ± 0·5 5·0 ± 0·5 4·8 ± 0·4 5·1 ± 0·4** 5·1 ± 0·4** 4·8 ± 0·2 5·0 ± 0·3** 5·0 ± 0·4** 6·2 0·001**
Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245

Haematocrit 0·44 ± 0·02 0·44 ± 0·02 0·45 ± 0·02 0·42 ± 0·02 0·45 ± 0·02** 0·45 ± 0·03** 0·42 ± 0·02 0·45 ± 0·02** 0·45 ± 0·03** 7·7 0·001**

Biochemical markers
Urea (mmol /l) 5·2 ± 1·0 5·5 ± 1·4 5·4 ± 1·3 6·3 ± 1·8 5·8 ± 1·4 6·4 ± 2·4 5·8 ± 0·8 4·3 ± 1·0** 5·4 ± 1·2 3·2 0·02*
Creatinine (µmol / l) 89 ± 10 80 ± 27 76 ± 25 89 ± 13 102 ± 15 99 ± 11 83 ± 13 89 ± 10 89 ± 10 1·9 0·13
Bilirubin (µmol /l) 14 ± 6 16 ± 9 15 ± 9 12 ± 10 12 ± 10 11 ± 6 11 ± 5 13 ± 10 12 ± 6 0·4 0·79
ALP (U/ l) 79 ± 18 80 ± 21 70 ± 15 65 ± 21 63 ± 16 57 ± 16 88 ± 55 86 ± 52 83 ± 49 0·9 0·45
GGT (U/ l) 27 ± 20 25 ± 13 25 ± 16 24 ± 17 26 ± 19 24 ± 17 22 ± 16 22 ± 19 20 ± 19 2·2 0·37
ALT (U/ l) 38 ± 25 26 ± 12 25 ± 11 28 ± 16 28 ± 20 29 ± 19 30 ± 20 31 ± 20 32 ± 25 2·2 0·09
AST (U/l) 37 ± 24 25 ± 7 25 ± 7 24 ± 5 22 ± 7 24 ± 6 26 ± 7 26 ± 6 27 ± 7 1·9 0·13
Iron (µmol/ l) 14 ± 6 14 ± 4 14 ± 4 17 ± 2 16 ± 7 15 ± 4 16 ± 5 13 ± 7 17 ± 8 1·1 0·36
Cholesterol (mmol / l) 4·4 ± 1·0 4·6 ± 1·0 4·3 ± 1·0 4·5 ± 0·9 4·8 ± 1·2 4·7 ± 1·2 4·4 ± 0·7 4·5 ± 0·7 4·4 ± 0·9 1·5 0·21
LDL (mmol /l) 1·3 ± 0·2 1·5 ± 0·4 1·4 ± 0·4 1·4 ± 0·3 1·3 ± 0·4 1·2 ± 0·2 1·3 ± 0·3 1·3 ± 0·3 1·2 ± 0·2 1·1 0·34
HDL (mmol /l) 2·5 ± 0·8 2·3 ± 0·8 2·1 ± 0·8 2·4 ± 0·8 2·5 ± 1·0 2·6 ± 1·1 2·6 ± 0·6 2·7 ± 0·7 2·3 ± 1·0 1·0 0·42
Triglycerides (mmol /l) 1·5 ± 0·7 1·7 ± 1·0 1·6 ± 1·4 1·4 ± 1·0 1·7 ± 1·0 1·8 ± 1·2 1·2 ± 0·9 1·2 ± 0·4 1·7 ± 1·7 0·6 0·63

Data presented are mean ± standard deviation. *P < 0·05, **P < 0·01. ALP, alkaline phosphatise; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamic transpeptidase;
HDL, high-density lipoprotein; LDL, low-density lipoprotein.
© 2007 The Authors
 Androgen effects on cardiac function 239

Table 2a. Standard echocardiographic parameters

Controls (n = 10) Testosterone (n = 10) Nandrolone (n = 10)

Baseline Week 4 Baseline Week 4 Baseline Week 4 F value P value

Left ventricular parameters

LVED (mm) 47 ± 5 47 ± 5 48 ± 7 49 ± 4 50 ± 5 50 ± 5 0·2 0·79
LVES (mm) 32 ± 5 31 ± 5 30 ± 7 33 ± 5* 34 ± 4 32 ± 5 3·5 0·04*
LVEF (%) 60 ± 6 62 ± 6 64 ± 5 63 ± 5 61 ± 7 61 ± 7 0·7 0·50
LV modified TEI Index 0·37 ± 0·10 0·40 ± 0·08 0·41 ± 0·10 0·43 ± 0·09 0·39 ± 0·12 0·42 ± 0·13 0·1 0·94

Standard Doppler parameters

Mitral valve inflow
E wave (m/s) 0·8 ± 0·2 0·7 ± 0·1 0·8 ± 0·2 0·8 ± 0·1 0·8 ± 0·1 0·8 ± 0·2 0·4 0·71
E/A ratio 1·6 ± 0·5 1·6 ± 0·4 1·9 ± 0·8 2·0 ± 0·7 1·7 ± 0·4 1·8 ± 0·6 0·1 0·92
DT (ms) 203 ± 62 221 ± 70 210 ± 55 206 ± 65 230 ± 64 212 ± 45 0·6 0·56

Pulmonary venous inflow

S wave (m/s) 0·6 ± 0·1 0·6 ± 0·1 0·5 ± 0·1 0·5 ± 0·1 0·6 ± 0·1 0·6 ± 0·2 0·3 0·78
D wave (m/s) 0·6 ± 0·2 0·5 ± 0·2 0·6 ± 0·2 0·5 ± 0·1 0·5 ± 0·1 0·6 ± 0·1 2·8 0·08
S/D ratio 1·0 ± 0·3 1·3 ± 0·5 1·0 ± 0·4 1·1 ± 0·3 1·2 ± 0·3 1·0 ± 0·3 3·5 0·05
A reversal (m/s) 0·4 ± 0·1 0·4 ± 0·1 0·3 ± 0·1 0·3 ± 0·04 0·3 ± 0·04 0·3 ± 0·02 0·9 0·41
A reversal duration (ms) 147 ± 14 120 ± 13 128 ± 23 138 ± 33 146 ± 14 143 ± 35 1·5 0·24

Right ventricular parameters

2
RVED area (cm ) 20·3 ± 4·9 18·7 ± 5·7 23·1 ± 4·3 23·3 ± 4·5 22·5 ± 5·0 21·1 ± 4·7 0·7 0·49
2
RVES area (cm ) 12·2 ± 4·5 10·9 ± 4·5 12·8 ± 1·3 14·6 ± 3·3* 13·8 ± 3·3 12·9 ± 3·3 3·5 0·04*
RV ejection area 0·4 ± 0·1 0·4 ± 0·1 0·4 ± 0·1 0·4 ± 0·1 0·4 ± 0·1 0·4 ± 0·1 2·7 0·09
Tricuspid annular motion (cm) 2·4 ± 0·3 2·5 ± 0·4 2·6 ± 0·4 2·5 ± 0·3 2·6 ± 0·5 2·5 ± 0·3 1·8 0·18
RV modified TEI index 0·36 ± 0·05 0·38 ± 0·11 0·37 ± 0·09 0·36 ± 0·09 0·36 ± 0·09 0·40 ± 0·09 0·7 0·51
Pulmonary artery systolic 30 ± 7 31 ± 6 30 ± 9 31 ± 7 31 ± 6 30 ± 4 0·2 0·84
Pressure (mmHg)

Atrial parameters
2
LA end-systolic area (cm ) 17 ± 5 17 ± 4 18 ± 2 18 ± 2 19 ± 3 19 ± 4 0·2 0·81
LA ejection fraction (%) 56 ± 11 56 ± 10 51 ± 14 54 ± 10 60 ± 14 58 ± 10 0·2 0·81
RA end-systolic area (cm2) 14 ± 4 14 ± 4 16 ± 4 16 ± 2 15 ± 4 15 ± 3 0·2 0·80
RA ejection fraction (%) 46 ± 10 39 ± 12 49 ± 18 50 ± 15 46 ± 13 46 ± 8 0·6 0·56

Data presented are mean ± standard deviation. *P < 0·05. A, late transmitral flow velocity; A rev, atrial pulmonary venous velocity; Adur, atrial pulmonary
venous velocity duration; DT, deceleration time; E, early transmitral flow velocity; E/A, early to late transmitral flow velocity ratio; D, pulmonary venous
diastolic velocity; LA, left atrial; LV, left ventricular; LVED, left ventricular end-diastolic; LVEF, left ventricular ejection fraction; LVES, left ventricular end-
systolic; RA, right atrial; RV, right ventricular; RVED, right ventricular end-diastolic; RVES, right ventricular end-systolic; S, pulmonary venous systolic velocity;
S/D, systolic to diastolic pulmonary venous wave ratio.

throughout the study in all groups. The urea level was slightly lower of E/Em ratio (7·1 ± 1·6 vs. 8·3 ± 1·8, F = 4·4, P = 0·02). Although
at week 2 following nandrolone injections. There were no significant statistically significant, these echocardiographic measurements of
changes in sexual function or mood (data not shown). the testosterone group remained in the accepted normal ranges.29
Markers of overall LV function (LVEF and LV modified TEI index)
were unchanged in all groups. Em velocity of the LV apex increased
Echocardiography
mildly in the nandrolone group (4·3 ± 1·3 vs. 5·3 ± 1·1 cm/s, F = 6·1,
Echocardiographic parameters are summarized in Table 2 (a–c). P = 0·0006); however, this was comparable to changes observed in
There was no change in LV end-diastolic diameter with treatment. the control group (4·8 ± 1·5 vs. 5·5 ± 1·6 cm/s). Furthermore, it was
However, a significant increase of LV end-systolic diameter at week not associated with change in Em velocity in other segments. No
4 compared with baseline was observed in the testosterone group conventional Doppler variables measured from the mitral inflow or
(30 ± 7 vs. 33 ± 5 mm, F = 3·5, P = 0·04). The testosterone group also pulmonary venous inflow showed any significant changes. No changes
had a mild reduction of the LV diastolic parameter of mitral septal in peak systolic strain or peak systolic strain rates of the LV were seen.
annular Em velocity (9·5 ± 2·6 vs. 8·7 ± 2·0 cm/s, F = 6·2, P = 0·006) Similar to changes in LV dimensions, the RV end-diastolic area
and an associated small increase in the LV filling pressure parameter was unchanged but the RV end-systolic area was increased mildly

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
240 T. Chung et al.

Table 2b. Colour-tissue Doppler derived tissue velocity echocardiographic parameters

Controls (n = 10) Nandrolone (n = 10) Testosterone (n = 10)

Baseline Week 4 Baseline Week 4 Baseline Week 4 F value P value

LV mitral annular tissue velocities (cm/s)

Septal Em 8·7 ± 2·4 9·7 ± 2·9* 9·5 ± 2·6 8·7 ± 2·0* 9·5 ± 2·1 9·4 ± 2·4 6·2 0·006**
Septal Am 6·9 ± 1·3 6·8 ± 2·0 7·4 ± 1·7 8·0 ± 0·6 6·3 ± 1·4 6·6 ± 2·0 0·3 0·72
Septal Sm 6·2 ± 0·5 7·0 ± 0·9 7·1 ± 1·3 7·3 ± 0·9 6·8 ± 1·0 7·16 ± 1·3 0·9 0·42
Lateral Em 12·2 ± 3·2 12·1 ± 2·9 12·4 ± 2·8 11·4 ± 2·7 12·1 ± 2·2 12·4 ± 3·0 1·8 0·18
Lateral Am 5·4 ± 1·9 5·5 ± 1·8 5·6 ± 1·2 5·6 ± 1·4 6·7 ± 3·4 5·6 ± 1·4 0·6 0·57
Lateral Sm 7·2 ± 1·4 7·3 ± 1·5 7·0 ± 1·1 7·4 ± 1·2 7·5 ± 2·0 7·8 ± 2·9 0·1 0·95
E/Em ratio 7·9 ± 1·8 7·3 ± 1·9 7·1 ± 1·6 8·3 ± 1·8** 7·8 ± 1·3 7·7 ± 1·5 4·4 0·02*

LV myocardial tissue velocities (cm/s)

Averaged basal segments
Em 10·1 ± 2·4 10·7 ± 2·4 10·1 ± 2·0 9·6 ± 2·4 10·1 ± 2·0 10·4 ± 2·5 2·3 0·11
Am 4·4 ± 1·2 5·2 ± 2·2 5·3 ± 1·2 5·8 ± 1·4 4·3 ± 1·3 5·0 ± 1·6 0·2 0·85
Sm 6·5 ± 0·9 6·9 ± 0·6 6·6 ± 0·9 6·8 ± 0·7 6·8 ± 1·2 7·4 ± 1·2 0·2 0·80
Averaged mid segments
Em 7·9 ± 2·2 8·2 ± 2·6 7·6 ± 2·0 7·7 ± 2·2 7·4 ± 1·3 8·0 ± 2·0 0·9 0·41
Am 3·2 ± 0·9 3·7 ± 1·8 3·6 ± 0·8 4·1 ± 0·9 3·0 ± 0·9 3·6 ± 0·8 0·1 0·95
Sm 4·9 ± 0·5 5·3 ± 0·7 5·0 ± 1·2 5·0 ± 0·8 5·2 ± 1·1 5·7 ± 0·8 0·7 0·49
Averaged apical segments
Em 4·8 ± 1·5 5·5 ± 1·6* 4·9 ± 1·7 4·5 ± 1·4 4·3 ± 1·3 5·3 ± 1·1** 6·1 0·006**
Am 1·4 ± 0·4 2·0 ± 1·1 1·7 ± 0·4 2·4 ± 1·3 1·4 ± 0·5 1·8 ± 0·8 0·1 0·92
Sm 3·4 ± 0·5 3·8 ± 0·6 3·2 ± 0·9 3·1 ± 0·6 3·6 ± 0·9 4·2 ± 1·6 0·8 0·46

RV tricuspid annulus tissue velocity (cm/s)

Em 9·6 ± 2·5 9·1 ± 1·9 9·2 ± 1·7 9·0 ± 1·5 10·1 ± 1·2 10·0 ± 2·5 0·1 0·94
Am 7·9 ± 2·4 7·9 ± 2·9 8·8 ± 1·9 9·1 ± 2·6 7·4 ± 3·2 9·0 ± 4·8 1·3 0·30
Sm 9·5 ± 1·4 10·0 ± 1·2 10·6 ± 1·9 10·9 ± 2·4 10·2 ± 1·1 10·6 ± 2·6 0·02 0·89

RV free wall myocardial tissue velocities (cm/s)

Basal Em 10·9 ± 4·1 9·4 ± 1·6 9·4 ± 1·5 10·0 ± 1·4 10·0 ± 4·1 10·5 ± 2·2 0·06 0·22
Basal Am 8·8 ± 3·2 8·5 ± 3·2 8·8 ± 2·2 8·8 ± 1·9 7·6 ± 4·0 9·0 ± 5·1 1·3 0·21
Basal Sm 9·5 ± 1·8 9·7 ± 1·5 10·0 ± 1·7 10·2 ± 2·0 10·3 ± 1·5 10·5 ± 2·3 0·02 0·98
Mid Em 6·7 ± 2·2 7·2 ± 1·8 7·2 ± 1·7 8·0 ± 1·7 7·8 ± 2·2 9·0 ± 2·4 0·5 0·63
Mid Am 5·0 ± 3·2 6·9 ± 3·7 5·9 ± 2·4 5·8 ± 3·0 4·9 ± 2·6 6·9 ± 4·3 2·4 0·11
Mid Sm 6·0 ± 1·8 6·7 ± 1·3 6·8 ± 1·5 7·8 ± 2·6 7·5 ± 1·6 8·5 ± 3·0 0·07 0·93
Apical Em 4·3 ± 1·9 5·0 ± 2·2 4·7 ± 1·3 4·9 ± 2·2 5·2 ± 2·6 5·0 ± 2·5 0·7 0·50
Apical Am 2·6 ± 2·3 3·5 ± 3·4 1·9 ± 0·8 2·3 ± 1·8 1·9 ± 1·4 2·6 ± 2·2 0·9 0·92
Apical Sm 3·3 ± 1·5 3·8 ± 1·2 3·4 ± 1·0 4·5 ± 2·6 4·4 ± 2·0 4·7 ± 2·2 0·5 0·61

Atrial myocardial tissue velocity (cm/s)

Left atrial aAm 5·4 ± 1·7 5·5 ± 1·8 6·1 ± 1·5 7·0 ± 1·7 4·8 ± 1·5 5·6 ± 1·4 0·9 0·42
Right atrial Am 6·5 ± 2·3 7·4 ± 2·9 7·0 ± 1·2 8·3 ± 2·9 6·3 ± 3·4 8·2 ± 4·3 0·6 0·55

Data presented are mean ± standard deviation. *P < 0·05, **P < 0·01. Am, late diastolic velocity; Em, early diastolic velocity; E/Em, early transmitral flow/
early diastolic velocity ratio; LV, left ventricular; RV, right ventricular; Sm, systolic velocity.

with testosterone treatment (12·8 ± 1·3 vs. 14·6 ± 3·3 cm2, F = 3·5,
Haemodynamics
P = 0·04). No parameters of RV function (RV ejection area, tricuspid
annular motion and RV modified TEI index) were altered following Non-invasive haemodynamic measurements were normal at baseline
androgen usage. Furthermore, colour TDI derived myocardial tissue in all groups (Table 3). There was no effect of 4 weeks treatment with
velocities, peak systolic strain and peak systolic strain rates of the RV either androgen or placebo on cardiac contractility (stroke volume,
free wall did not change from baseline at week 4. No change was cardiac output) or afterload (systolic, diastolic or mean blood pres-
observed in atrial areas (LA end-systolic area, RA end-systolic area sure; systemic vascular resistance) (Table 3).
and RA:LA end-systolic area ratio), ejection fractions or mean atrial Electrocardiograms were normal in all men before and after the
myocardial velocities. 4-week treatment period. Small reduction in the PR interval was

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
 Androgen effects on cardiac function 241

Table 2c. Colour-tissue Doppler derived strain and strain rate echocardiographic parameters

Controls (n = 10) Testosterone (n = 10) Nandrolone (n = 10)

Baseline Week 4 Baseline Week 4 Baseline Week 4 F value P value

LV peak systolic strain (%)

Averaged basal 19·8 ± 6·1 20·8 ± 3·7 20·0 ± 3·6 20·2 ± 3·5 19·9 ± 4·7 21·6 ± 5·5 0·2 0·82
Averaged mid 19·1 ± 5·3 18·4 ± 3·7 18·9 ± 3·7 20·7 ± 5·5 20·0 ± 3·0 20·0 ± 3·6 0·7 0·52

LV peak systolic strain rate (1/s)

Averaged basal 1·8 ± 0·6 1·6 ± 0·5 1·6 ± 0·4 1·7 ± 0·5 1·7 ± 0·5 1·8 ± 0·5 0·7 0·51
Averaged mid 1·6 ± 0·4 1·4 ± 0·3 1·3 ± 0·3 1·5 ± 0·3 1·6 ± 0·5 1·5 ± 0·4 1·5 0·24

RV free wall peak systolic strain (%)

Basal 25·1 ± 9·3 28·7 ± 13·4 24·2 ± 6·8 27·4 ± 11·2 23·9 ± 8·8 27·5 ± 6·2 0·2 0·82
Mid 29·9 ± 12·9 30·7 ± 3·4 33·1 ± 9·9 33·7 ± 9·8 31·6 ± 8·9 32·1 ± 7·4 0·2 0·86
Apical 25·4 ± 9·0 27·9 ± 9·2 29·5 ± 8·0 26·7 ± 9·8 32·9 ± 4·3 27·7 ± 5·1 0·8 0·47

RV free wall peak systolic strain rate (1/s)

Basal 2·1 ± 1·2 2·3 ± 1·3 1·8 ± 0·8 2·3 ± 0·9 2·0 ± 0·6 2·1 ± 0·7 1·5 0·24
Mid 2·1 ± 0·8 2·1 ± 0·4 2·1 ± 0·7 2·5 ± 1·1 2·7 ± 0·8 2·6 ± 0·9 0·7 0·50
Apical 2·0 ± 0·7 2·4 ± 1·0 2·2 ± 0·8 2·2 ± 1·0 2·9 ± 0·9 2·1 ± 0·8 2·4 0·12

Data presented are mean ± standard deviation. LV, left ventricular; RV, right ventricular.

Table 3. Haemodynamic measurements by thoracic electrical bioimpedence

Controls (n = 10) Testosterone (n = 10) Nandrolone (n = 10)

Baseline Week 4 Baseline Week 4 Baseline Week 4 F value P value

Electrocardiogram
Heart rate (beats/min) 74 ± 9 74 ± 16 67 ± 15 70 ± 13 67 ± 14 71 ± 15 0·8 0·45
PR interval (ms) 152 ± 12 152 ± 10 167 ± 13 154 ± 12* 162 ± 28 159 ± 29 4·3 0·03*
QRS duration (ms) 86 ± 10 85 ± 12 92 ± 4 88 ± 6 94 ± 8 93 ± 9 0·9 0·43
Corrected QT interval (ms) 381 ± 34 382 ± 17 391 ± 28 354 ± 25 384 ± 21 271 ± 31 2·9 0·07

Blood pressure (mmHg)

Systolic 124 ± 11 128 ± 5 124 ± 13 123 ± 13 131 ± 9 130 ± 13 0·3 0·77
Diastolic 80 ± 10 82 ± 10 78 ± 11 77 ± 11 85 ± 10 82 ± 12 0·6 0·59
Mean 97 ± 12 95 ± 5 92 ± 9 90 ± 11 100 ± 10 97 ± 13 0·7 0·51
Cardiac output (l/min) 6·6 ± 1·5 6·4 ± 1·6 5·8 ± 1·5 5·9 ± 1·4 6·5 ± 1·5 6·6 ± 1·7 0·7 0·51
Cardiac index (2·5–4·7) 3·5 ± 0·7 3·4 ± 0·6 2·9 ± 0·6 3·0 ± 0·6 3·1 ± 0·7 3·1 ± 0·7 0·5 0·59
Stroke volume (mL) 92 ± 28 78 ± 27 88 ± 26 84 ± 21 98 ± 24 100 ± 28 0·2 0·81
Stroke index (35–65) 49 ± 15 41 ± 12 44 ± 10 43 ± 10 47 ± 10 47 ± 11 0·2 0·80
Systemic vascular resistance (742 –1378) 1133 ± 293 1211 ± 464 1260 ± 334 1209 ± 417 1258 ± 610 1154 ± 393 1·2 0·34

Data presented are mean ± standard deviation. *P < 0·05.

observed in the testosterone group (167 ± 13 vs. 154 ± 12 ms,

Discussion
F = 4·3, P = 0·03) and in corrected QT interval in the nandrolone
group (384 ± 21 vs. 271 ± 31, F = 2·9, P = 0·07). Androgens have been widely used in medicine since the Nobel Prize-
We then performed a post hoc power analysis of LVEF assuming winning identification of testosterone as the major mammalian
the observed pooled mean and standard deviation (of 62% and 6%, androgen in 1935 48–50 and its first clinical use in 1937.4 Despite great
respectively, from Table 2a). Our sample size of 30 men (10 allocated advances in knowledge of the molecular and cellular biology of
to each group) therefore has 80% power to detect an absolute 6·5% androgen action 51–53 and the clinical pharmacology of androgens2, the
increase or decrease in LVEF with testosterone or nandrolone therapy. cardiovascular effects and safety of physiological or supraphysiological

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
242 T. Chung et al.

doses of androgens remain poorly understood.17–21 Cardiac effects risks of myocardial infarction and stroke remain to be fully evaluated
of androgens might involve either indirect mechanisms involving by further studies with longer periods of treatment and follow-up.63
blood vessel disorders, notably atherosclerosis, as well as increased The direct effects of androgens on cardiac muscle have been less
erythropoiesis, haematocrit, hyperviscosity and hypertension, but in studied64 based mostly on anecdotal case reports of cardiomyopathy
addition androgens may have direct effects on cardiac muscle and at autopsy in some androgen abusers.19 Some but not all studies of
its function, since the androgen receptor is expressed in cardiac muscle anabolic steroid abuse in athletes suggest that androgen abuse may
and in vitro studies of nonhuman cardiac myocytes show that tes- be associated with left ventricular hypertrophy, which may persist.
tosterone can decrease action potential duration (thereby altering However, the lack of randomization, self-selection and the con-
repolarization) and peak shortening times.54 founding role of strength training are important limitations. The
Nandrolone is a naturally occurring steroid produced as an inter- present randomized placebo-controlled study of testosterone and
mediate during the aromatization of testosterone to oestradiol but nandrolone administered over 4 weeks is consistent with a previous
is minimally secreted into the circulation of humans.55 In humans, study using nandrolone for 8 weeks11 in showing no adverse effects
it can be considered a prototype pure androgen acting solely upon of androgens on sensitive non-invasive measures of cardiac function.
the androgen receptor as exogenous nandrolone is minimally con- While these negative findings are reassuring for short-term safety,
verted to oestradiol and its 5α reduced metabolite has negligible longer studies would be required to verify the maintenance of such
androgenic activity.56 This contrasts with testosterone that is subject a safety profile over longer usage. Retrospective data from the East
to both aromatization to oestradiol (capable of acting on oestrogen German national sports doping programme involving over 2000 elite
receptors) as well as amplification by 5α reductase to the more potent athletes regularly monitored medically during treatment with high
androgen dihydrotestosterone. Hence, comparing effects of nan- dose synthetic androgens reported no cardiovascular complications
drolone with testosterone can identify androgen effects that depend whereas liver disease and other complications were frequently
on androgen activation. For example, testosterone has more potent recorded.65 Given the numbers of men abusing androgens in the
effects on bone compared with nandrolone whereas both androgens community66 together with the rarity of cardiomyopathic deaths in
had similar effects on muscle in glucocorticoid-dependent men.57 androgen abusers suggests that, if it exists, androgen-induced cardio-
This observation highlighted the importance of aromatization to tes- myopathy is rare.
tosterone action on bone but not muscle and confirmed the utility Measurement of cardiac output and stroke volume provides global
of the design utilized for this study. In the present study we found assessment of cardiac function. The non-invasive measurement of
that both androgens had similar effects on haematological parameters these by thoracic electrical bioimpedance is particularly suitable for
and overall cardiovascular and haemodynamic function. This is con- repeated within-subject studies of healthy individuals. The healthy
sistent with observations that heart muscle expresses only low levels young men in this study had normal haemodynamic measurements
of androgen-activating enzymes, aromatase58 and 5α reductase.59 including cardiac output, stroke volume and systemic vascular resist-
However, they did demonstrate minor differences in their effects on ance before and after 4 weeks of treatment with either androgen or
ventricular chamber size, LV diastolic function and filling pressures. placebo. Androgen therapy may improve exercise capacity in men
In the present study, among the potential mechanisms for indirect with heart failure.26 Our results indicate potential differences with the
effects of androgens on cardiac function, we observed a modest expected cardiac response between those with cardiac dysfunction
increase in haemoglobin, haematocrit, erythrocyte and platelet and healthy young men with normal resting cardiac function.
counts for both androgens equally but no change in blood pressure Echocardiographic measurements identified several small differ-
or lipids when compared with placebo. The modest increase in ences between the testosterone and nandrolone groups. The traditional
erythrocytes and platelets confirms the efficacy of androgenic stimu- marker of chamber size, end-diastolic diameter, was unchanged by
lation by the doses used in this study and is consistent with the treatment in all three groups. However, a mild increase in the end-
known stimulatory effects of androgens on bone marrow60–62 regard- systolic diameter of the LV and RV was seen in the testosterone
less of mode of administration. The similar supraphysiological group. The observed small increases in chamber sizes were within
effects of nandrolone and testosterone at equal doses suggest that accepted normal ranges for healthy young men and do not have
these androgen-induced erythropoietic effects are mediated directly immediate clinical significance. Furthermore, the primary outcome
on the androgen receptor without the need for metabolic activation marker of LV systolic function (LVEF) remained unchanged follow-
of testosterone either by aromatization or 5α reduction/amplifica- ing androgen administration during the study. Preservation of LVEF
tion. However, these findings provide little support for older studies has particular significance in excluding deleterious effects on cardiac
suggesting that other testosterone metabolites, notably 5β andro- function as this marker has proven prognostic significance whereby
gens, may be involved in the erythropoietic response to androgens.62 reductions of LVEF < 40% predict increased cardiovascular morbidity
While increased circulating erythrocyte numbers could in theory and mortality.67 This study excludes a major short-term adverse
increase blood viscosity, the quantitatively minor changes observed effect on LVEF by either androgen.
are unlikely to have significant clinical effects. Nevertheless, at the Recent refinements of echocardiographic colour TDI derived
extremes, development of significant polycythaemia could increase parameters have allowed detection of early subclinical changes on
cardiac workload and therefore, although rare, the risk of poly- cardiac filling and myocardial velocities.68,69 A reduction in mitral
cythaemia warrants age-appropriate monitoring of haemoglobin annular Sm velocity is a sensitive marker of mildly impaired LV
responses especially during pharmacological androgen therapy. systolic function even in patients with apparently ‘normal’ LVEF and
Whether such mild haematological changes influence the long-term is reduced in LV diastolic dysfunction.69 When combined with the

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
 Androgen effects on cardiac function 243

mitral inflow E wave, the derived E/Em ratio is useful to estimate 9 Storer, T.W., Magliano, L., Woodhouse, L., Lee, M.L., Dzekov, C.,
the LV filling pressures.32,33 While the LVEF was within normal limits Dzekov, J., Casaburi, R. & Bhasin, S. (2003) Testosterone dose-
in both androgen groups, minor reduction of mitral annular Em dependently increases maximal voluntary strength and leg power,
velocity and elevation of E/Em ratio were observed in the testosterone but does not affect fatigability or specific tension. Journal of Clinical
Endocrinology and Metabolism, 88, 1478 –1485.
group. These changes suggest a mildly adverse effect of testosterone
10 Bhasin, S., Woodhouse, L., Casaburi, R., Singh, A.B., Mac, R.P.,
on LV diastolic function. Whether these effects would translate to
Lee, M., Yarasheski, K.E., Sinha-Hikim, I., Dzekov, C., Dzekov, J.,
changes in overall cardiac structure and LVEF after prolonged exposure
Magliano, L. & Storer, T.W. (2005) Older men are as responsive as
to androgens remain unclear. Longer-term studies incorporating young men to the anabolic effects of graded doses of testosterone on
these colour TDI derived parameters will be required to assess the skeletal muscle. Journal of Clinical Endocrinology and Metabolism,
whether these changes persists, increase, resolve or become clinically 90, 678 – 688.
significant. 11 Hartgens, F., Cheriex, E.C. & Kuipers, H. (2003) Prospective echo-
Strengths of the present study include its well-controlled design cardiographic assessment of androgenic-anabolic steroids effects on
featuring randomization and placebo controls, the detailed non- cardiac structure and function in strength athletes. International
invasive evaluation of cardiac function, the contrast of testosterone Journal of Sports Medicine, 24, 344 – 351.
with its non-aromatizable and non-amplifiable analogue, nan- 12 Hoberman, J.M. & Yesalis, C.E. (1995) The history of synthetic testo-
sterone. Scientific American, 272, 76 – 81.
drolone and the controlled comparison of the two most widely used
13 Wilson, J.D. (1988) Androgen abuse by athletes. Endocrine Reviews,
and abused androgens. Weaknesses include its limited sample size,
9, 181–199.
4 weeks’ duration of treatment and follow-up. The short duration
14 Handelsman, D.J. (2002) Androgen misuse and abuse. In: S.M. Shalet,
of this study made physiological and functional changes more likely J.A. Wass eds. Oxford Textbook of Endocrinology. Oxford University
to be detected than structural change. Our study reassuringly indicates Press, Oxford, 1374 –1377.
the safety of short-term usage of androgens in young men. Using non- 15 Yesalis, C.E., Anderson, W.A., Buckley, W.E. & Wright, J.E. (1990)
invasive determination of myocardial function and cardiovascular Incidence of the nonmedical use of anabolic-androgenic steroids.
haemodynamics, we found no evidence of clinically significant In: G.C. Lin, L. Erinoff eds. Anabolic Steroid Abuse. US Department
adverse, or beneficial, cardiac effects following short-term testosterone of Health and Human Services, Rockville, MD, 97 –111.
or nandrolone usage in healthy young men. 16 Handelsman, D.J. & Gupta, L. (1997) Prevalence and risk factors for
anabolic-androgenic steroid abuse in Australian secondary school
students. International Journal of Andrology, 20, 159 –164.
References 17 Rockhold, R.W. (1993) Cardiovascular toxicity of anabolic steroids.
Annual Review of Pharmacology and Toxicology, 33, 497– 520.
1 Zitzmann, M. & Nieschlag, E. (2000) Hormone substitution in male 18 Melchert, R.B. & Welder, A.A. (1995) Cardiovascular effects of
hypogonadism. Molecular and Cellular Endocrinology, 161, 73 – 88. androgenic-anabolic steroids. Medicine and Science in Sports and
2 Handelsman, D.J. (2005) Androgen action and pharmacologic uses. Exercise, 27, 1252 – 1262.
In: L.J. Degroot, J.L. Jameson eds. Endocrinology. Elsevier Saunders, 19 Sullivan, M.L., Martinez, C.M., Gennis, P. & Gallagher, E.J. (1998)
Philadelphia, PA, 3121– 3138. The cardiac toxicity of anabolic steroids. Progress in Cardiovascular
3 Liu, P.Y. & Handelsman, D.J. (2004) Androgen therapy in non- Diseases, 41, 1–15.
gonadal disease. In: E. Nieschlag, H.M. Behre eds. Testosterone: 20 Weidemann, W. & Hanke, H. (2002) Cardiovascular effects of androgens.
Action, Deficiency and Substitution. Springer-Verlag, Berlin, 445 – Cardiovascular Drug Reviews, 20, 175 –198.
495. 21 Dhar, R., Stout, C.W., Link, M.S., Homoud, M.K., Weinstock, J., Estes,
4 Hamilton, J.B. (1937) Treatment of sexual underdevelopment with N.A. III (2005) Cardiovascular toxicities of performance-enhancing
synthetic male hormone substance. Endocrinology, 21, 649 – 654. substances in sports. Mayo Clinic Proceedings, 80, 1307 –1315.
5 Bhasin, S., Storer, T.W., Berman, N., Callegari, C., Clevenger, B., 22 Maron, B.J. (2003) Sudden death in young athletes. New England
Phillips, J., Bunnell, T.J., Tricker, R., Shirazi, A. & Casaburi, R. (1996) Journal of Medicine, 349, 1064 –1075.
The effects of supraphysiologic doses of testosterone on muscle size 23 Jones, R.D., Nettleship, J.E., Kapoor, D., Jones, H.T. & Channer, K.S.
and strength in normal men. New England Journal of Medicine, 335, (2005) Testosterone and atherosclerosis in aging men: purported
1– 7. association and clinical implications. American Journal of Cardio-
6 Bhasin, S., Storer, T.W., Berman, N., Yarasheski, K.E., Clevenger, B., vascular Drugs: Drugs, Devices, and Other Interventions, 5, 141–154.
Phillips, J., Lee, W.P., Bunnell, T.J. & Casaburi, R. (1997) Testosterone 24 Pugh, P.J., Jones, R.D., West, J.N., Jones, T.H. & Channer, K.S. (2004)
replacement increases fat-free mass and muscle size in hypogonadal Testosterone treatment for men with chronic heart failure. Heart, 90,
men. Journal of Clinical Endocrinology and Metabolism, 82, 407 – 413. 446 – 447.
7 Bross, R., Casaburi, R., Storer, T.W. & Bhasin, S. (1998) Androgen 25 Pugh, P.J., Jones, T.H. & Channer, K.S. (2003) Acute haemodynamic
effects on body composition and muscle function: implications for effects of testosterone in men with chronic heart failure. European
the use of androgens as anabolic agents in sarcopenic states. Baillieres Heart Journal, 24, 909–915.
Clinical Endocrinology and Metabolism, 12, 365 – 378. 26 Malkin, C.J., Pugh, P.J., West, J.N., van Beek, E.J., Jones, T.H. & Chan-
8 Bhasin, S., Woodhouse, L., Casaburi, R., Singh, A.B., Bhasin, D., ner, K.S. (2006) Testosterone therapy in men with moderate severity
Berman, N., Chen, X., Yarasheski, K.E., Magliano, L., Dzekov, C., heart failure: a double-blind randomized placebo controlled trial.
Dzekov, J., Bross, R., Phillips, J., Sinha-Hikim, I., Shen, R. & Storer, T.W. European Heart Journal, 27, 57 – 64.
(2001) Testosterone dose–response relationships in healthy young 27 Fung, K.C., Corbett, A. & Kritharides, L. (2003) Myocardial tissue
men. American Journal of Physiology. Endocrinology and Metabolism, velocity reduction is correlated with clinical neurologic severity in
281, E1172 –E1181. myotonic dystrophy. American Journal of Cardiology, 92, 177 –181.

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
244 T. Chung et al.

28 Gottdiener, J.S., Bednarz, J., Devereux, R., Gardin, J., Klein, A., hemodynamic measures in clinically stable heart failure patients.
Manning, W.J., Morehead, A., Kitzman, D., Oh, J., Quinones, M., Congestive Heart Failure, 6, 74 – 80.
Schiller, N.B., Stein, J.H. & Weissman, N.J. (2004) American Society 42 Taler, S.J., Textor, S.C. & Augustine, J.E. (2002) Resistant hypertension:
of Echocardiography recommendations for use of echocardiography comparing hemodynamic management to specialist care. Hypertension,
in clinical trials. Journal of the American Society of Echocardiography, 39, 982 – 988.
17, 1086 –1119. 43 Strobeck, J.E., Silver, M.A. & Ventura, H. (2000) Impedance
29 Sun, J.P., Popovic, Z.B., Greenberg, N.L., Xu, X.F., Asher, C.R., cardiography: noninvasive measurement of cardiac stroke volume
Stewart, W.J. & Thomas, J.D. (2004) Noninvasive quantification of and thoracic fluid content. Congestive Heart Failure, 6, 56 – 59.
regional myocardial function using Doppler-derived velocity, 44 Turner, L., Conway, A.J., Jimenez, M., Liu, P.Y., Forbes, E.,
displacement, strain rate, and strain in healthy volunteers: effects of McLachlan, R.I. & Handelsman, D.J. (2003) Contraceptive efficacy
aging. Journal of the American Society of Echocardiography, 17, 132 – of a depot progestin and androgen combination in men. Journal of
138. Clinical Endocrinology and Metabolism, 88, 4659 – 4667.
30 Gilman, G., Khandheria, B.K., Hagen, M.E., Abraham, T.P., Seward, 45 Kelleher, S., Conway, A.J. & Handelsman, D.J. (2004) Blood testoster-
J.B. & Belohlavek, M. (2004) Strain rate and strain: a step-by-step one threshold for androgen deficiency symptoms. Journal of Clinical
approach to image and data acquisition. Journal of the American Society Endocrinology and Metabolism, 89, 3813 – 3817.
of Echocardiography, 17, 1011–1020. 46 Kelleher, S., Howe, C., Conway, A.J. & Handelsman, D.J. (2004)
31 Harada, K., Tamura, M., Toyono, M., Oyama, K. & Takada, G. (2001) Testosterone release rate and duration of action of testosterone pellet
Assessment of global left ventricular function by tissue Doppler implants. Clinical Endocrinology, 60, 420 – 428.
imaging. American Journal of Cardiology, 88, 927 – 932, A929. 47 Aminorroaya, A., Kelleher, S., Conway, A.J., Ly, L.P. & Handelsman,
32 Ommen, S.R., Nishimura, R.A., Appleton, C.P., Miller, F.A., Oh, J.K., D.J. (2005) Adequacy of androgen replacement influences bone den-
Redfield, M.M. & Tajik, A.J. (2000) Clinical utility of Doppler sity response to testosterone in androgen-deficient men. European
echocardiography and tissue Doppler imaging in the estimation Journal of Endocrinology, 152, 881– 886.
of left ventricular filling pressures: a comparative simultaneous 48 David, K., Dingmanse, E., Freud, J. & Lacqueur, E. (1935) Uber
Doppler-catheterization study. Circulation, 102, 1788 –1794. krystallinisches mannliches Hormon aus Hoden (Testosteron),
33 Nagueh, S.F., Middleton, K.J., Kopelen, H.A., Zoghbi, W.A. & wirksamer als aus Harn oder aus Cholestrin bereites Androsteron.
Quinones, M.A. (1997) Doppler tissue imaging: a noninvasive tech- Zeischrift Physiologische Chemie, 233, 281– 282.
nique for evaluation of left ventricular relaxation and estimation of 49 Butenandt, A. & Hanisch, G. (1935) Uber die Umwandlung des
filling pressures. Journal of the American College of Cardiology, 30, Dehydroandrosterons in Androstenol-(17)-one-(3) (Testosterone);
1527 –1533. Umweg zur darstellung des testosterons aus Cholsterin (vorlauf
34 Miller, R.L., Das, S., Anandarangam, T., Leibowitz, D.W., Alderson, mitteilung). Zeischrift Physiologische Chemie, 237, 89 – 97.
P.O., Thomashow, B. & Homma, S. (1998) Association between right 50 Ruzicka, L. & Wettstein, A. (1935) Uber die krystallische Herstellung
ventricular function and perfusion abnormalities in hemodynami- des Testikelhormons, Testosteron (androsten-3-on-17-ol). Helvetica
cally stable patients with acute pulmonary embolism. Chest, 113, Chimica Acta, 18, 1264 –1275.
665 – 670. 51 Mooradian, A.D., Morley, J.E. & Korenman, S.G. (1987) Biological
35 Kjaergaard, J., Sogaard, P. & Hassager, C. (2004) Right ventricular actions of androgens. Endocrine Reviews, 8, 1– 28.
strain in pulmonary embolism by Doppler tissue echocardiography. 52 Quigley, C.A., DeBellis, A., Marschke, K.B., El-Awady, M.K.,
Journal of the American Society of Echocardiography, 17, 1210 –1212. Wilson, E.M. & French, F.F. (1995) Androgen receptor defects:
36 Mansencal, N., Joseph, T., Vieillard-Baron, A., Qanadli, S.D., historical, clinical and molecular perspectives. Endocrine Reviews, 16,
Jondeau, G., Lacombe, P., Jardin, F. & Dubourg, O. (2003) Com- 271– 321.
parison of different echocardiographic indexes secondary to right 53 Gao, W., Bohl, C.E. & Dalton, J.T. (2005) Chemistry and structural
ventricular obstruction in acute pulmonary embolism. American biology of androgen receptor. Chemical Reviews, 105, 3352 – 3370.
Journal of Cardiology, 92, 116 –119. 54 Kimura, N., Mizokami, A., Oonuma, T., Sasano, H. & Nagura, H.
37 Hammarstrom, E., Wranne, B., Pinto, F.J., Puryear, J. & Popp, R.L. (1993) Immunocytochemical localization of androgen receptor with
(1991) Tricuspid annular motion. Journal of the American Society of polyclonal antibody in paraffin-embedded human tissues. Journal of
Echocardiography, 4, 131–139. Histochemistry and Cytochemistry, 41, 671– 678.
38 Kukulski, T., Hubbert, L., Arnold, M., Wranne, B., Hatle, L. & 55 Reznik, Y., Dehennin, L., Coffin, C., Mahoudeau, J. & Leymarie, P.
Sutherland, G.R. (2000) Normal regional right ventricular function (2001) Urinary nandrolone metabolites of endogenous origin in
and its change with age: a Doppler myocardial imaging study. Journal man: a confirmation by output regulation under human chorionic
of the American Society of Echocardiography, 13, 194 – 204. gonadotropin stimulation. Journal of Clinical Endocrinology and
39 Harada, K., Tamura, M., Toyono, M. & Yasuoka, K. (2002) Com- Metabolism, 86, 146 –150.
parison of the right ventricular Tei index by tissue Doppler imaging 56 Toth, M. & Zakar, T. (1982) Relative binding affinities of testosterone,
to that obtained by pulsed Doppler in children without heart disease. 19-nortestosterone and their 5α-reduced derivatives to the androgen
American Journal of Cardiology, 90, 566 – 569. receptor and to other androgen-binding proteins: a suggested role
40 Belardinelli, R., Ciampani, N., Costantini, C., Blandini, A. & of 5α-reductive steroid metabolism in the dissociation of ‘myotropic’
Purcaro, A. (1996) Comparison of impedance cardiography with and ‘androgenic’ activities of 19-nortestosterone. Journal of Steroid
thermodilution and direct Fick methods for noninvasive measure- Biochemistry, 17, 653 – 660.
ment of stroke Volume and cardiac output during incremental 57 Crawford, B.A., Liu, P.Y., Kean, M., Bleasel, J. & Handelsman, D.J.
exercise in patients with ischemic cardiomyopathy. American Journal (2003) Randomised, placebo-controlled trial of androgen effects on
of Cardiology, 77, 1293 –1301. bone and muscle in men requiring long-term systemic glucocorti-
41 Greenberg, B.H., Hermann, D.D., Pranulis, M.F., Lazio, L. & coid therapy. Journal of Clinical Endocrinology and Metabolism, 88,
Cloutier, D. (2000) Reproducibility of impedance cardiography 3167 – 3176.

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245
 Androgen effects on cardiac function 245

58 Simpson, E.R. (2004) Aromatase: biologic relevance of tissue-specific structure and function in male bodybuilders. Journal of the American
expression. Seminars in Reproductive Medicine, 22, 11– 23. College of Cardiology, 37, 224 – 230.
59 Russell, D.W. & Wilson, J.D. (1994) Steroid 5 alpha-reductase: two 65 Franke, W.W. & Berendonk, B. (1997) Hormonal doping and andro-
genes/two enzymes. Annual Review of Biochemistry, 63, 25 – 61. genization of athletes: a secret program of the German Democratic
60 Kennedy, B.J. & Gilbertsen, A.S. (1957) Increased erythropoiesis Republic government. Clinical Chemistry, 43, 1262 –1279.
induced by androgenic hormone therapy. New England Journal of 66 Bahrke, M.S. & Yesalis, C.E. (2004) Abuse of anabolic androgenic
Medicine, 256, 719 – 726. steroids and related substances in sport and exercise. Current Opinion
61 Shahidi, N.T. (1973) Androgens and erythropoiesis. New England in Pharmacology, 4, 614 – 620.
Journal of Medicine, 289, 72 – 80. 67 Thohan, V. (2004) Prognostic implications of echocardiography
62 Gardner, F.H. & Besa, E.C. (1983) Physiologic mechanisms and the in advanced heart failure. Current Opinion in Cardiology, 19, 238 –
hematopoietic effects of the androstanes and their derivatives. 249.
Current Topics in Hematology, 4, 123 –195. 68 Smiseth, O.A., Stoylen, A. & Ihlen, H. (2004) Tissue Doppler imaging
63 Liverman, C.T. & Blazer, D.G., eds. (2004) Testosterone and Aging: for the diagnosis of coronary artery disease. Current Opinion in
Clinical Research Directions. Institute of Medicine: The National Cardiology, 19, 421– 429.
Academies Press, Washington, DC. 69 Sanderson, J.E., Wang, M. & Yu, C.M. (2004) Tissue Doppler imag-
64 Sader, M.A., Griffiths, K.A., McCredie, R.J., Handelsman, D.J. & ing for predicting outcome in patients with cardiovascular disease.
Celermajer, D.S. (2001) Androgenic anabolic steroids and arterial Current Opinion in Cardiology, 19, 458 – 463.

© 2007 The Authors

Journal compilation © 2007 Blackwell Publishing Ltd, Clinical Endocrinology, 66, 235–245