Accepted Manuscript

Title: Relation of Testosterone Levels to Mortality in Men with Heart Failure

Author: Akiomi Yoshihisa, Satoshi Suzuki, Yu Sato, Yuki Kanno, Satoshi Abe,
Makiko Miyata, Takamasa Sato, Masayoshi Oikawa, Atsushi Kobayashi,
Takayoshi Yamaki, Hiroyuki Kunii, Kazuhiko Nakazato, Takafumi Ishida,
Yasuchika Takeishi

PII: S0002-9149(18)30247-9
DOI: https://doi.org/10.1016/j.amjcard.2018.01.052
Reference: AJC 23141

To appear in: The American Journal of Cardiology

Received date: 14-12-2017

Accepted date: 30-1-2018

Please cite this article as: Akiomi Yoshihisa, Satoshi Suzuki, Yu Sato, Yuki Kanno, Satoshi Abe,
Makiko Miyata, Takamasa Sato, Masayoshi Oikawa, Atsushi Kobayashi, Takayoshi Yamaki,
Hiroyuki Kunii, Kazuhiko Nakazato, Takafumi Ishida, Yasuchika Takeishi, Relation of
Testosterone Levels to Mortality in Men with Heart Failure, The American Journal of Cardiology
(2018), https://doi.org/10.1016/j.amjcard.2018.01.052.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will
undergo copyediting, typesetting, and review of the resulting proof before it is published in its
final form. Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.
Relation of Testosterone Levels to Mortality in Men with Heart Failure

Akiomi Yoshihisa, MD, PhD1*, Satoshi Suzuki, MD, PhD 1, Yu Sato, MD,1 Yuki Kanno,
MD, PhD 1, Satoshi Abe, MD, PhD 1, Makiko Miyata, MD, PhD 1, Takamasa Sato, MD,
PhD 1, Masayoshi Oikawa, MD, PhD 1, Atsushi Kobayashi, MD, PhD 1, Takayoshi
Yamaki, MD, PhD 1, Hiroyuki Kunii, MD, PhD 1, Kazuhiko Nakazato, MD, PhD 1,
Takafumi Ishida, MD, PhD 1 and Yasuchika Takeishi1
1
Department of Cardiovascular Medicine, Fukushima Medical University, Fukushima,
Japan

A short title: Serum testosterone in men with heart failure

*Address for correspondence. Tel: +81 24 547 1190, Fax: +81 24 548 1821, Email:

yoshihis@fmu.ac.jp; Akiomi Yoshihisa, MD, PhD. Department of Cardiovascular

Medicine, Fukushima Medical University. 1 Hikarigaoka, Fukushima 960-1295, Japan

Total word count: 1970 words.

6 tables and 2 figures

Page 1 of 24
Abstract

We aimed to investigate the impact of testosterone on prognosis of heart failure (HF), as

well as underlying cardiac function, cardiac damage and exercise capacity. We analyzed

consecutive 618 men with HF (age 65.9 years). These patients were divided into

quartiles based on their serum levels of total testosterone (TT): 1st (TT>631 ng/dl,

n=154), 2nd (462<TT≤631 ng/dl, n=155), 3rd (300<TT≤462 ng/dl, n=156) and 4th

(TT≤300 ng/dl, n=153) quartiles. In the Kaplan-Meier analysis (mean 1281 days),

all-cause mortality progressively increased throughout from the 1st to the 4th groups. In

the multivariable Cox proportional hazard analysis, TT was found to be an independent

predictor of all-cause mortality (hazard ratio 0.929, P=0.042). In addition, we compared

the parameters of echocardiography and cardiopulmonary exercise testing, as well as

levels of B-type natriuretic peptide and cardiac troponin I among the four groups. Left

ventricular ejection fraction and B-type natriuretic peptide did not differ among the

groups. In contrast, the 4th quartile, compared to 1st, 2nd and 3rd groups, had higher levels

of troponin I and lower peak VO2 (P<0.05, respectively). Decreased serum testosterone

is associated with myocardial damage, lower exercise capacity, and higher mortality in

men with HF.

Key words: heart failure, men, testosterone, myocardial damage, exercise capacity

Page 2 of 24
Introduction

Testosterone affects multiple cardiovascular systems, including cardiomyocytes (protein

synthesis, hypertrophy), cardiac electrophysiology (ion channel, arrhythmias), cardiac

contractility (adrenoreceptor regulation, heavy myosin chain modification), and vascular

function.1 It has been reported that deficiency of anabolic hormone is associated with

poor prognosis in men with heart failure (HF).2 In contrast, low testosterone levels are

associated with low mortality from ischemic heart disease.3 Thus, we aimed to clarify

associations between testosterone levels and prognosis in men with HF while taking

into consideration the underlying comprehensive clinical parameters (e.g. cardiac

function, cardiac damage, arterial stiffness and exercise capacity).

Methods

This was a prospective observational study of 702 decompensated men with HF who

were discharged from Fukushima Medical University Hospital between 2009 and 2015.

The diagnosis of decompensated HF was made by several cardiologists based on the HF

guidelines.4 Patients who were prescribed androgenic steroids, glucocorticoid, thyroid

hormone, and/or anti-thyroid drugs during follow up were excluded. The patient flow

chart is shown in Figure 1. Of the total of 702 patients, 618 were finally enrolled. The

average total testosterone (TT) level of the study population was 479.2 ± 267.7 (range

Page 3 of 24
6–2538 ng/dl), and the patients were divided into quartiles based on their TT levels: 1st

(632 ng/dl ≤ TT, n = 154), 2nd (463 ≤ TT ≤ 631, n = 155), 3rd (462 ≤ TT ≤ 301, n = 156),

and 4th quartiles (TT ≤ 300, n = 153). TT of ≤ 300 ng/dl is generally considered as low

TT levels.5 We compared the clinical features and parameters of laboratory data,

echocardiography, pulse wave velocity (PWV) and cardiopulmonary exercise testing.

These assessments were performed within one week of hospital discharge.

The patients were followed up until 2017 for all-cause death. The status and/or

dates of death of all patients were obtained from the patients’ medical records or the

attending physicians at the patient’s referring hospital. We were able to follow up all

patients. Survival time was calculated from the date of hospitalization until the date of

death or last follow-up. Written informed consent was obtained from all study subjects

at discharge. The study protocol was approved by the Ethics Committee of Fukushima

Medical University, and was carried out in accordance with the principles outlined in

the Declaration of Helsinki. Reporting of the study conforms to STROBE along with

references to STROBE and the broader EQUATOR guidelines.6

Blood samples were obtained at hospital discharge each morning, with the

patients in a fasted state. Serum TT was measured by electro chemiluminescence

immunoassay (ARCHITECT 2nd Generation Testosterone, Abbott Japan, Tokyo,

Page 4 of 24
Japan) using analyzer (Architect i2000 analyzer; Abbott Diagnostics, Abbott Park, IL,

USA). High-sensitivity cardiac troponin I levels were measured in EDTA anticoagulated

plasma using the refined assay (Abbott-Architect, Abbott Laboratories, Abbott Park, IL,

USA). These immunoassays were blindly performed by clinical laboratory technologists

at Abott Japan Co. Ltd. B-type natriuretic peptide (BNP) levels were measured using a

specific immunoradiometric assay (Shionoria BNP kit, Shionogi, Osaka, Japan).

Echocardiography was performed blindly by experienced echocardiographers

using ultrasound systems (ACUSON Sequoia, Siemens Medical Solutions USA, Inc.,

Mountain View, CA, USA) with standard techniques.7 Left ventricular ejection fraction

(LVEF) was calculated using Simpson’s method in four chamber view.7 LVEF of less

than 40% was considered as reduced LVEF, LVEF of 40 to 49% was considered as

mid-range LVEF, and LVEF of more than 50% was considered as preserved LVEF.8

PWV was estimated using a Mobil-O-Graph PWA Monitor (I.E.M. GmbH,

Stolberg, Germany), which is the automated self-measurement blood pressure

monitoring device to estimate brachial and ankle PWV.9-11 This device uses a transfer

function-like method (ARCSolver algorithm) with brachial cuff-based waveform

recordings.

The patients underwent incremental symptom-limited exercise testing before

Page 5 of 24
discharge, using an upright cycle ergometer with a ramp protocol (Strength Ergo 8,

Fukuda Denshi Co. Ltd., Tokyo, Japan). Breath-by-breath oxygen consumption (VO2),

carbon dioxide production (VCO2), and minute ventilation (VE) were measured during

exercise using an AE-300S respiratory monitor (Minato Medical Science, Osaka,

Japan).12 Peak VO2 was measured as an average of the last 30 s of exercise. Ventilatory

response to exercise (slope of the relationship between ventilation and carbon dioxide

production, VE/VCO2 slope) was calculated as the regression slope relating VE to CO2

from the start of exercise until the respiratory compensation point.12

Parametric variables are presented as mean ± SD, non-parametric variables

(e.g. ferritin, vitamin B12, erythropoietin, C-reactive protein, BNP and troponin I) are

presented as median and interquartile range, and categorical variables are expressed as

numbers and percentages. The chi-square test was used for comparisons of categorical

variables. We used the analysis of variance for continuous variables, followed by

Tukey’s post-hoc test. We performed multiple regression analysis, allowing for

interaction between serum TT level and clinical confounding factors: age, body mass

index, New York Heart Association (NYHA) functional class, presence of ischemic

etiology, hypertension, diabetes, dyslipidemia, chronic kidney disease (CKD), anemia

and atrial fibrillation. Correlations between serum TT levels and other parameters (e.g.

Page 6 of 24
laboratory data, echocardiography, PWV and cardiopulmonary exercise test) were

assessed using Spearman’s correlation analysis. Kaplan-Meier analysis was used for

presenting the all-cause mortality, and the log-rank test was used for initial comparisons.

The prognostic value was tested by univariate and multivariate Cox proportional hazard

analyses. In the multivariate Cox proportional hazard analysis, to prepare for potential

confounding, we considered the following clinical factors, which are generally known

to affect the prognosis in HF patients: age, body mass index, presence of NYHA

functional class III or IV, LVEF, ischemic etiology, hypertension, diabetes,

dyslipidemia, CKD, anemia, atrial fibrillation, and TT levels. Univariate parameters

with P-values of <0.10 were included in the multivariate analysis. In addition, to assess

the potential heterogeneity of associations between TT levels and all-cause mortality,

we conducted subgroup analyses. Interactions between TT levels and clinically relevant

variables which were known to be important prognostic factors in HF patients and/or

were different clinical characteristics between the groups (age, body mass index, NYHA

functional class, LVEF, ischemic etiology, hypertension, diabetes, dyslipidemia, CKD,

anemia, atrial fibrillation, levels of total protein, albumin, C-reactive protein, BNP and

troponin I) were estimated by a Cox proportional hazard model. A P value of <0.05 was

Page 7 of 24
considered statistically significant for all comparisons. All analyses were performed

using a statistical software package (SPSS ver. 24.0, IBM, Armonk, NY, USA).

Results

The comparisons of the clinical features are shown in Table 1. Regarding factors which

affect serum TT levels, multiple regression analysis (Table 2) showed that age and the

presence of dyslipidemia and/or anemia are independent predictors of serum TT levels.

In the laboratory data (Table 3), hemoglobin, iron, total protein and albumin were

lowest, and ferritin, C-reactive protein and troponin I were highest in 4th quartile. The

PWV was highest, and the peak VO2 was lowest, in the 4th quartile (Table 3).

Additionally, correlation analyses with the TT levels and other parameters are presented

in Table 4. Although there was no significant correlation between serum TT levels and

echocardiographic parameters, there were many significant correlations between TT

levels and hemoglobin, iron, ferritin, unsaturated iron binding capacity, total protein,

albumin, log C-reactive protein, log troponin I, PWV, and peak VO2.

In the follow-up period (mean 1281 days), there were 180 deaths (94 cardiac

deaths and 86 non-cardiac deaths). In the Kaplan-Meier analysis (Figure 2), all-cause

mortality progressively increased from the 1st to 4th quartiles (log-rank, P=0.010). In the

Cox proportional hazard analysis (Table 5), after adjusting for other confounding

Page 8 of 24
factors, TT level was an independent predictor of all-cause mortality in HF patients. To

assess the potential heterogeneity of TT’s impact on all-cause mortality, we also

conducted subgroup analyses and examined interaction terms (Table 6). There were no

interactions between TT levels and other important variables, including age, NYHA

class, LVEF, co-morbidities, other laboratory markers including total protein, albumin,

C-reactive protein, BNP and troponin I. Thus, the impact of TT levels was consistent in

all subgroups.

Discussion

The present study is the first to report that men with HF with decreased TT levels

(especially, TT ≤300 ng/dl) experience a higher mortality accompanied by anemia,

inflammation, myocardial damage, increased arterial stiffness, and impaired exercise

capacity, without association with impaired cardiac function.

Testosterone strongly stimulates erythropoiesis through mechanisms such as

intestinal iron absorption, erythrocyte iron incorporation, hemoglobin synthesis, and

bone marrow hematopoietic stem cells through the induction of insulin growth factor-I

via androgen receptor-mediated mechanisms.1,13 Testosterone deficiency causes

resistance to erythropoiesis-stimulating agents in men with CKD.14 Recent study

Page 9 of 24
presented that testosterone supplementation significantly increased the hemoglobin

levels in older men with low testosterone levels with any cause of anemia.15

By modulating the activity of several ion channels and endothelial nitric oxide

synthase, testosterone influences the tonus of vascular smooth muscle cells, and leads to

vasodilation of systemic,16 coronary17 and pulmonary 18

vessels.19 Low testosterone

level is associated with increased levels of cholesterol, production of inflammatory

factors (high-sensitivity C-reactive protein, interleukin-6 and tumor necrosis factor-α)20

21
and thickness of the arterial wall, and contributes to endothelial dysfunction and

arterial stiffness.20,22,23

Testosterone inhibits cardiac fibroblast migration and proliferation in addition

to myofibroblast differentiation induced by transforming growth factor-β1.24

Testosterone affects cardiac efferent vagal activity,25 and induces hypertrophy of both

type I and type II muscle fibers.25 Myocardial damage is reduced by testosterone by

upregulating cardiac α1 adrenoceptor and possibly by activating cardiac mitochondrial

ATP-sensitive potassium channels.1,26

Skeletal muscle strength is also influenced by testosterone, by stimulating

change in muscle composition toward type I muscle fibers that are, compared to type II

fibers, associated with enhanced physical capability, strength and reduced exercise

10

Page 10 of 24
 27
capacity and cachexia. Although previous report suggested that only

dehydroepiandrosterone is correlated with peak VO2, whereas testosterone and

insulin-like growth factor-1 were not significantly correlated, 28 another report presented

that lowered circulating testosterone and its changes relate to peak VO2.29 Meta-analysis

of testosterone supplementation in moderate-severe HF patients revealed that

testosterone supplementation improves exercise capacity without any improvement of

myocardial structure or function.30 Thus, lower testosterone is associated with anemia,

inflammation, myocardial damage, arterial stiffness, exercise capacity, represents

chronic illness, and these mechanisms partly related to poor prognosis in men with HF

in the present study.

The present study has several limitations. First, as a prospective cohort study of

a single center with a relatively small number of patients, the present results may not be

representative of a general population. Although we performed both multivariate Cox

proportional hazard analysis and subgroup analysis under consideration with several

confounding factors, we cannot rule out residual confounding variables, and the effects

of differences in the backgrounds among the groups might not be completely adjusted.

Second, since the present study included variables during hospitalization for

decompensated heart failure, without taking into consideration changes in medical

11

Page 11 of 24
parameters and post-discharge treatment, we should pay attention to extrapolate our

findings to all men with HF. Third, although we encouraged cardiopulmonary exercise

testing and PWV as much as possible, we were not able to perform these measurements

in all patients for various reasons (e.g. patient refusal, medical reasons, etc). Thus, there

may be potential selection bias in these measurements. Fourth, since this was a

cross-sectional and prospective observational study without intervention for decreased

testosterone, the causal relationships and mechanisms of decreased testosterone on

hypoalbuminemia, anemia, inflammation, myocardial damage, increased arterial

stiffness, impaired exercise capacity and worse prognosis could not be fully explained.

Sixth, we did not examine other anabolic hormones (dehydroepiandrosterone sulfate

and insulin-like growth factor-1excluded) in this study. Therefore, the present results

should be viewed as preliminary, and further studies with a larger population are

needed.

In conclusion, decreased testosterone is associated with adverse prognosis,

accompanied by hypoalbuminemia, anemia, inflammation, myocardial damage,

increased arterial stiffness, and impaired exercise capacity, in men with HF.

12

Page 12 of 24
References

1. Busic Z, Culic V. Central and peripheral testosterone effects in men with heart failure: An

approach for cardiovascular research. World J Cardiol 2015;7:504-510.

2. Jankowska EA, Biel B, Majda J, Szklarska A, Lopuszanska M, Medras M, Anker SD,

Banasiak W, Poole-Wilson PA, Ponikowski P. Anabolic deficiency in men with chronic heart

failure: prevalence and detrimental impact on survival. Circulation 2006;114:1829-1837.

3. Araujo AB, Kupelian V, Page ST, Handelsman DJ, Bremner WJ, McKinlay JB. Sex

steroids and all-cause and cause-specific mortality in men. Arch Intern Med

2007;167:1252-1260.

4. Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE, Jr., Drazner MH, Fonarow GC,

Geraci SA, Horwich T, Januzzi JL, Johnson MR, Kasper EK, Levy WC, Masoudi FA,

McBride PE, McMurray JJ, Mitchell JE, Peterson PN, Riegel B, Sam F, Stevenson LW, Tang

WH, Tsai EJ, Wilkoff BL, American College of Cardiology F, American Heart Association

Task Force on Practice G. 2013 ACCF/AHA guideline for the management of heart failure: a

report of the American College of Cardiology Foundation/American Heart Association Task

Force on Practice Guidelines. J Am Coll Cardiol 2013;62:e147-239.

5. Kloner RA, Carson C, 3rd, Dobs A, Kopecky S, Mohler ER, 3rd. Testosterone and

Cardiovascular Disease. J Am Coll Cardiol 2016;67:545-557.

6. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, Initiative S.

13

Page 13 of 24
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)

statement: guidelines for reporting observational studies. BMJ 2007;335:806-808.

7. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, Picard MH,

Roman MJ, Seward J, Shanewise J, Solomon S, Spencer KT, St John Sutton M, Stewart W,

American Society of Echocardiography's N, Standards C, Task Force on Chamber Q,

American College of Cardiology Echocardiography C, American Heart A, European

Association of Echocardiography ESoC. Recommendations for chamber quantification. Eur J

Echocardiogr 2006;7:79-108.

8. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JG, Coats AJ, Falk V,

Gonzalez-Juanatey JR, Harjola VP, Jankowska EA, Jessup M, Linde C, Nihoyannopoulos P,

Parissis JT, Pieske B, Riley JP, Rosano GM, Ruilope LM, Ruschitzka F, Rutten FH, van der

Meer P, Authors/Task Force M. 2016 ESC Guidelines for the diagnosis and treatment of

acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and

chronic heart failure of the European Society of Cardiology (ESC)Developed with the special

contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J

2016;37:2129-2200.

9. Wei W, Tolle M, Zidek W, van der Giet M. Validation of the mobil-O-Graph: 24 h-blood

pressure measurement device. Blood Press Monit 2010;15:225-228.

14

Page 14 of 24
10. Wassertheurer S, Kropf J, Weber T, van der Giet M, Baulmann J, Ammer M, Hametner

B, Mayer CC, Eber B, Magometschnigg D. A new oscillometric method for pulse wave

analysis: comparison with a common tonometric method. J Hum Hypertens

2010;24:498-504.

11. Weber T, Wassertheurer S, Rammer M, Maurer E, Hametner B, Mayer CC, Kropf J, Eber

B. Validation of a brachial cuff-based method for estimating central systolic blood pressure.

Hypertension 2011;58:825-832.

12. Kanno Y, Yoshihisa A, Watanabe S, Takiguchi M, Yokokawa T, Sato A, Miura S, Shimizu

T, Nakamura Y, Abe S, Sato T, Suzuki S, Oikawa M, Saitoh S, Takeishi Y. Prognostic

Significance of Insomnia in Heart Failure. Circ J 2016;80:1571-1577.

13. Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS,

Montori VM. Testosterone therapy in adult men with androgen deficiency syndromes: an

endocrine society clinical practice guideline. J Clin Endocrinol Metab 2006;91:1995-2010.

14. Stenvinkel P, Barany P. Hypogonadism in males with chronic kidney disease: another

cause of resistance to erythropoiesis-stimulating agents? Contributions to nephrology

2012;178:35-39.

15. Roy CN, Snyder PJ, Stephens-Shields AJ, Artz AS, Bhasin S, Cohen HJ, Farrar JT, Gill

TM, Zeldow B, Cella D, Barrett-Connor E, Cauley JA, Crandall JP, Cunningham GR,

15

Page 15 of 24
Ensrud KE, Lewis CE, Matsumoto AM, Molitch ME, Pahor M, Swerdloff RS, Cifelli D, Hou

X, Resnick SM, Walston JD, Anton S, Basaria S, Diem SJ, Wang C, Schrier SL, Ellenberg SS.

Association of Testosterone Levels With Anemia in Older Men: A Controlled Clinical Trial.

JAMA Intern Med 2017;177:480-490.

16. Kang SM, Jang Y, Kim J, Chung N, Cho SY, Chae JS, Lee JH. Effect of oral

administration of testosterone on brachial arterial vasoreactivity in men with coronary

artery disease. Am J Cardiol 2002;89:862-864.

17. Webb CM, McNeill JG, Hayward CS, de Zeigler D, Collins P. Effects of testosterone on

coronary vasomotor regulation in men with coronary heart disease. Circulation

1999;100:1690-1696.

18. Jones RD, English KM, Pugh PJ, Morice AH, Jones TH, Channer KS. Pulmonary

vasodilatory action of testosterone: evidence of a calcium antagonistic action. J Cardiovasc

Pharmacol 2002;39:814-823.

19. Culic V. Androgens in cardiac fibrosis and other cardiovascular mechanisms. Int J

Cardiol 2015;179:190-192.

20. Oskui PM, French WJ, Herring MJ, Mayeda GS, Burstein S, Kloner RA. Testosterone

and the cardiovascular system: a comprehensive review of the clinical literature. Journal of

the American Heart Association 2013;2:e000272.

16

Page 16 of 24
21. Traish AM, Saad F, Feeley RJ, Guay A. The dark side of testosterone deficiency: III.

Cardiovascular disease. J Androl 2009;30:477-494.

22. Kyriazis J, Tzanakis I, Stylianou K, Katsipi I, Moisiadis D, Papadaki A, Mavroeidi V,

Kagia S, Karkavitsas N, Daphnis E. Low serum testosterone, arterial stiffness and

mortality in male haemodialysis patients. Nephrol Dial Transplant 2011;26:2971-2977.

23. Kelly DM, Jones TH. Testosterone: a vascular hormone in health and disease. J

Endocrinol 2013;217:R47-71.

24. Chung CC, Hsu RC, Kao YH, Liou JP, Lu YY, Chen YJ. Androgen attenuates cardiac

fibroblasts activations through modulations of transforming growth factor-beta and

angiotensin II signaling. Int J Cardiol 2014;176:386-393.

25. El-Mas MM, Afify EA, Mohy El-Din MM, Omar AG, Sharabi FM. Testosterone facilitates

the baroreceptor control of reflex bradycardia: role of cardiac sympathetic and

parasympathetic components. J Cardiovasc Pharmacol 2001;38:754-763.

26. Tsang S, Wu S, Liu J, Wong TM. Testosterone protects rat hearts against ischaemic

insults by enhancing the effects of alpha(1)-adrenoceptor stimulation. Br J Pharmacol

2008;153:693-709.

27. Volterrani M, Rosano G, Iellamo F. Testosterone and heart failure. Endocrine

2012;42:272-277.

17

Page 17 of 24
28. Pastor-Perez FJ, Manzano-Fernandez S, Garrido Bravo IP, Nicolas F, Tornel PL, Lax A,

de la Morena G, Valdes M, Pascual-Figal DA. Anabolic status and functional impairment in

men with mild chronic heart failure. Am J Cardiol 2011;108:862-866.

29. Jankowska EA, Filippatos G, Ponikowska B, Borodulin-Nadzieja L, Anker SD, Banasiak

W, Poole-Wilson PA, Ponikowski P. Reduction in circulating testosterone relates to exercise

capacity in men with chronic heart failure. J Card Fail 2009;15:442-450.

30. Toma M, McAlister FA, Coglianese EE, Vidi V, Vasaiwala S, Bakal JA, Armstrong PW,

Ezekowitz JA. Testosterone supplementation in heart failure: a meta-analysis. Circ Heart

Fail 2012;5:315-321.

18

Page 18 of 24
FIGURE LEGENDS

Figure 1

Patient flowchart.

Figure 2

Kaplan-Meier analysis for all-cause mortality stratified by serum testosterone.

Table 1. Comparisons of clinical features among testosterone quartiles (N=618)

Variable 1st N=154 2nd N=155 3rd N=156 4th N=153 P-
value
Testosterone (ng/dl) 827.2±20.0 546.6±48.2 376.6±47.6 165.1±91.4 <0.00
(632-2538 ng/dl) (463-631 ng/dl) (301-462 ng/dl) (6-300 ng/dl) 1
Age (years) 63.8±14.6 65.4±12.9 68.1±13.2 66.2±15.4 0.059
2
Body mass index (kg/m ) 23.2±3.6 24.0±4.1 24.1±4.0 23.5±4.4 0.614
NYHA class III or IV 4 (2.6%) 4 (2.6%) 0 (0%) 7 (4.6%) 0.075
Left ventricular ejection fraction 86(55.8%)/18(11 75(48.4%)/23(14 69(44.2%)/10(6. 76(49.7%)/16(10 0.037
reduced/ mid-range/ preserved .7%)/50(32.5%) .8%)/57(36.8%) 4%)/77(49.4%) .5%)/61(39.9%)
Ischemic etiology 41 (26.6) 50 (32.3) 47 (30.1) 61 (39.9) 0.085
Hypertension 118 (76.6%) 122 (78.7%) 137 (87.8%) 120 (78.4%) 0.057
Diabetes mellitus 58 (37.7%) 78 (50.3%) 63 (40.4%) 84 (54.9%) 0.006
Dyslipidemia 118 (76.6%) 120 (77.4%) 124 (79.5%) 134 (87.6%) 0.063
Chronic kidney disease 91 (59.1%) 95 (61.3%) 104 (66.7%) 101 (66.0%) 0.442
Anemia 67 (43.5%) 74 (47.7%) 86 (55.1%) 94 (61.4%) 0.009
Atrial fibrillation 71 (46.1%) 71 (45.8%) 66 (42.3%) 57 (37.3%) 0.363
Medications
Renin-angiotensin 135 (87.7%) 130 (83.9%) 133 (85.3%) 114 (74.5%) 0.013
aldosterone system inhibitors
Aldosterone antagonists 79 (51.3%) 58 (37.4%) 69 (44.2%) 77 (50.3%) 0.053
β-blockers 137 (89.0%) 138 (89.0%) 129 (82.7%) 123 (80.4%) 0.067
Diuretics 107 (30.5%) 98 (63.2%) 104 (66.7%) 114 (74.5%) 0.182

19

Page 19 of 24
Inotropic agents 27 (17.5%) 17 (11.0%) 26 (16.7%) 23 (15.0%) 0.379
Hypertension was defined as the recent use of antihypertensive drugs, a systolic blood
pressure of ≥140 mmHg, and/or a diastolic blood pressure of ≥90 mmHg. Diabetes
mellitus was defined as the recent use of antidiabetic dugs, a fasting glucose value of
≥126 mg/dl, a casual glucose value of ≥200 mg/dl, and/or a HbA1c percentage of
≥6.5% (National Glycohemoglobin Standardization Program). Dyslipidemia was
defined as the recent use of cholesterol-lowering drugs, a triglyceride value of ≥150
mg/dL, a low-density lipoprotein cholesterol value of ≥140 mg/dL, and/or a
high-density lipoprotein cholesterol value of <40 mg/dL. Chronic kidney disease was
defined as an estimated glomerular filtration rate of <60 ml/min/1.73 cm2 according to
the Modification of Diet in Renal Disease formula. Anemia was defined as hemoglobin
levels of <13.0 g/dL. Atrial fibrillation was identified by an electrocardiogram
performed during hospitalization and/or from medical records.

Table 2. Multiple regression analysis to determine serum testosterone levels

Univariate Multivariate
Factors β coefficient P-value β P-value
coefficient
Age -0.152 <0.001 -0.121 0.003
Body mass index -0.035 0.391
NYHA functional class -0.090 0.025 -0.026 0.522
Ischemic etiology -0.090 0.025 -0.035 0.399
Hypertension -0.031 0.439
Diabetes -0.105 0.009
Dyslipidemia -0.102 0.011 -0.098 0.017
Chronic kidney disease -0.076 0.059
Anemia -0.147 <0.001 -0.109 0.009
Atrial fibrillation 0.040 0.320

Table 3. Comparisons of laboratory data, echocardiography, cardiopulmonary

exercise testing and pulse wave velocity among quartile (N=618)
1st 2nd 3rd 4th P-value

Laboratory data
Testosterone (ng/dl) 827.2±20.0 546.6±48.2 376.6±47.6 165.1±91.4 <0.001
Hemoglobin (g/dl) 13.6±2.0 13.5±2.0 12.8±2.3 * 12.4±2.7 **†† <0.001

20

Page 20 of 24
Iron (µg/dl) 91.3±32.8 89.9±40.1 79.5±39.6 76.8±45.5 0.030
Ferritin (ng/ml) § 104.0 124.0 150.0 * 165.0 **† 0.003
(53.3-220.3) (59.0-268.0) (62.5-245.0) (84.5-297.8)
Unsaturated iron binding 222.7±69.7 224.4±63.9 215.3±62.3 205.6±75.0 0.250
capacity (µg/dl)
Vitamin B12 (pg/ml) § 405.0 408.0 429.0 516.0 0.615
(271.0-562.0) (286.0-599.5) (348.0-558.0) (321.3-742.8)
Folic acid (ng/ml) 6.1±2.7 7.6±3.5 6.7±4.4 7.1±3.6 0.115
Erythropoietin (mIU/ml) § 22.7 (13.1-39.8) 15.4 (9.7-26.1) 20.5 (11.8-33.3) 17.8 (12.5-46.6) 0.460
Glomerular filtration rate 56.1±24.1 57.9±22.0 52.9±23.3 54.1±25.9 0.287
2
(ml/min./1.73cm )
Sodium (mmol/l) 139.0±3.1 138.7±3.1 139.1±3.6 138.3±4.8 0.273
Total protein (g/dl) 7.1±0.7 7.2±0.7 7.0±0.7 6.7±0.8 <0.001
**††‡‡
Albumin (g/dl) 3.9±0.5 3.9±0.5 3.7±0.5 3.4±0.6 <0.001
**††‡‡
C-reactive protein (mg/dl) § 0.16 (0.07-0.52) 0.12 (0.06-0.74) 0.23 (0.08-0.76) 0.56 (0.14-1.84) <0.001
**††‡‡
B-type natriuretic paptide 98.5 75.2 114.8 103.3 0.342
(pg/ml) § (44.3-179.0) (41.2-164.3) (40.4-241.6) (47.1-232.9)
Troponin I (pg/ml) § 21.4 (10.4-49.8) 21.7 (10.2-97.3) 33.4 (12.2-92.5) 57.4 (22.3-258.9) 0.018
**††
Echocardiography
Left ventricular ejection 43.7±16.0 47.2±14.6 47.7±14.9 45.9±16.1 0.145
fraction (%)
Right ventricular 41.3±15.9 40.8±14.5 40.8±15.7 41.6±14.0 0.986
fractional area change
(%)
Tricuspid valve 28.7±13.4 26.7±15.8 28.6±12.7 28.8±13.5 0.676
regurgitation pressure
gradient (mmHg)
Cardio-pulmonary
exercise testing (n=299)
Peak VO2 (ml/kg/min) 16.4±5.1 16.5±4.5 15.8±4.7 14.5±4.9* 0.034
VE-VCO2 slope 34.3±7.0 34.0±7.7 34.8±8.2 35.9±10.2 0.593
Pulse wave velocity (n=156) 7.9±2.7 8.6±2.0 9.5±1.8 9.9±2.1**††‡ <0.001

21

Page 21 of 24
*P<0.05 and **P<0.01 vs. 1st quartile; †P<0.05 and ††P<0.01 vs. 2nd quartile; ‡ P<0.05
and ‡‡ P<0.01 vs. 3rd quartile. § Data are presented as median (interquartile range).

Table 4. Correlation analyses with serum testosterone levels and other parameters
R P value
Laboratory data
Hemoglobin 0.229 <0.001
Iron 0.139 0.008
Ferritin -0.174 0.002
Unsaturated iron binding capacity 0.143 0.007
Vitamin B12 -0.009 0.879
Folic acid -0.066 0.270
Erythropoietin -0.065 0.277
Glomerular filtration rate 0.060 0.154
Sodium 0.050 0.237
Total protein 0.209 <0.001
Albumin 0.315 <0.001
Log C-reactive protein -0.238 <0.001
Log B-type natriuretic peptide -0.019 0.649
Log troponin I -0.241 <0.001
Echocardiography
Left ventricular ejection fraction -0.102 0.128
Right ventricular fractional area change 0.003 0.962
Tricuspid valve regurgitation pressure gradient -0.002 0.966
Cardiopulmonary exercise test
Peak VO2 0.153 0.041
VE/VCO2 slope -0.086 0.136
Pulse wave velocity -0.228 0.004

Table 5. Cox proportional hazard model for All-cause mortality (event 180, N=618)
Univariate analysis Multivariate analysis
HR 95% CI P value HR 95% CI P value
Age 1.049 1.034-1.063 <0.001 1.040 1.024-1.057 <0.001

22

Page 22 of 24
Body mass index 0.884 0.849-0.921 <0.001 0.925 0.879-0.973 0.003
NYHA class III or IV 3.316 1.751-6.279 <0.001 2.195 1.046-4.609 0.038
Left ventricular ejection 0.997 0.986-1.008 0.581
fraction
Ischemic etiology 1.263 0.934-1.709 0.130
Hypertension 0.959 0.650-1.414 0.832
Diabetes 1.136 0.848-1.522 0.393
Dyslipidemia 1.201 0.811-1.777 0.361
Chronic kidney disease 1.831 1.306-2.566 <0.001 1.393 0.934-2.076 0.104
Anemia 2.288 1.666-3.143 <0.001 1.359 0.924-1.997 0.119
Atrial fibrillation 1.303 0.973-1.746 0.076 0.988 0.700-1.394 0.943
Testosterone 0.889 0.836-0.944 <0.001 0.929 0.865-0.997 0.042

Table 6. Subgroup analysis for all-cause mortality: impact of testosterone levels

(per 1 ng/dl increase)
Factor Subgroup n HR 95% Cl P value Interaction
P value
Total Total 618 0.889 0.836-0.944 <0.001 -
Age ≥70 303 0.902 0.835-0.975 0.009 0.941
<70 315 0.912 0.826-1.008 0.071
Body mass index ≥23.2 309 0.917 0.855-0.983 0.014 0.165
<23.2 309 0.837 0.749-0.936 0.002
NYHA class III or IV 15 0.868 0.708-1.064 0.174 0.594
III or IV I or II 603 0.900 0.845-0.959 0.001
Left ventricular ejection Reduced 306 0.866 0.799-0.939 <0.001 0.225
fraction Mid-range 67 0.867 0.732-1.027 0.099
Preserved 245 0.943 0.847-1.049 0.280
Ischemic etiology Present 199 0.839 0.752-0.936 0.056 0.213
Absent 419 0.914 0.850-0.982 0.015
Hypertension Present 497 0.884 0.826-0.947 <0.001 0.785
Absent 121 0.905 0.793-1.032 0.135
Diabetes Present 283 0.898 0.822-0.980 0.016 0.679
Absent 335 0.879 0.807-0.958 0.003
Dyslipidemia Present 496 0.868 0.812-0.929 <0.001 0.079
Absent 122 0.994 0.870-1.136 0.933

23

Page 23 of 24
 Chronic kidney disease Present 391 0.893 0.830-0.960 0.002 0.824
Absent 227 0.876 0.781-0.983 0.025
Anemia Present 321 0.918 0.854-0.987 0.021 0.666
Absent 297 0.898 0.804-1.003 0.057
Atrial fibrillation Present 265 0.861 0.790-0.939 0.001 0.508
Absent 353 0.905 0.831-0.986 0.022
Total protein ≥7.0 315 0.858 0.791-0.930 <0.001 0.090
<7.0 303 0.956 0.869-1.051 0.354
Albumin ≥3.6 290 0.869 0.803-0.941 0.001 0.070
<3.6 328 0.976 0.884-1.076 0.123
C-reactive protein ≥0.17 305 0.883 0.811-0.961 0.004 0.575
<0.17 313 0.907 0.831-0.991 0.031
B-type natriuretic peptide ≥92.2 309 0.887 0.829-0.950 0.001 0.974
<92.2 309 0.809 0.786-1.008 0.067
Troponin I ≥29.6 309 0.952 0.880-1.030 0.220 0.085
<29.6 309 0.852 0.771-0.942 0.002

24

Page 24 of 24