Association of serum androgen concentrations with cardiovascular risk factors in elderly male patients with chronic systolic heart failure in China

Abstract

Objective: To survey the serum androgen concentrations and investigate the relationship between androgen levels and cardiovascular risk factors in elderly male patients with chronic systolic heart failure (HF) in China.

Methods: 106 consecutive male patients hospitalized for chronic systolic HF aged from 60 to 87 were enrolled. About 400 healthy age-matched men were compared as a control group. Total testosterone (TT), free testosterone (FT), dehydroepiandrosterone sulfate (DHEAS) and sex hormone binding globulin (SHBG) were measured. Differences of androgen levels between HF patients and healthy men were determined by t-test and associations of androgen with cardiovascular risk factors were evaluated by partial correlations analyses.

Results: Compared with healthy men, TT, FT and DHEAS levels in patients with HF decreased, whereas SHBG level increased significantly (both p < 0.01). TT was negatively correlated with TC, TG and DBP (p < 0.05), FT was negatively correlated with TC, LDL-C and DBP (p < 0.05). SHBG correlated with BMI and smoking history positively (p < 0.05).

Conclusions: Level of bio-available testosterone decreased with advancing age, especially in men with HF. Men with low levels of bio-available testosterone had worse profiles of cardiovascular risk factors. Treatment of HF is still challenging and testosterone supplementation therapy may be an effective therapeutic option.

• Androgen
• cardiovascular risk factors
• elderly
• heart failure
• male

Introduction

Heart failure (HF) is a progressive disease affecting millions of people worldwide. It is reported that there are more than 20 million patients with HF in the world now, with the majority of these patients being elderly [1]. The disease carries a significantly high morbidity and mortality risk. The natural history of untreated HF is very poor, with worsening cardiac function, symptoms and death. Apart from the poor prognosis, HF causes functional disability. There is therefore a need to explore other therapies that may prolong life and improve symptoms and that could be widely available. In recent years, testosterone has emerged as a possible therapeutic option for the treatment of men with HF [2].

Coronary heart disease is the major underlying cause of HF. Male sex is a proven independent risk factor for cardiovascular disease. There is a widespread belief that the higher cardiovascular risk in men than in women may due to the higher serum testosterone levels in men [2]. While the relationship between male gender and the higher prevalence of coronary heart disease does suggest a role for male sex hormones in the etiology of cardiovascular disease, the situation may be more complex than simply higher testosterone levels driving higher cardiovascular disease rates in men. Testosterone has a direct physiological role in maintaining cardiovascular health, and androgen deficiency contributes to the onset and progression of cardiovascular disease [3]. Multiple cross-sectional studies have demonstrated that androgen levels decrease with advancing age [4,5]. Testosterone deficiency is a generalized phenomenon seen in the course of chronic HF in men. Growing evidence indicates that a relative hypotestosteronaemia could be involved in the pathophysiology of HF in men, contributing to some features of this syndrome, such as the reduced muscle mass, abnormal energy handling, fatigue, dyspnea and, finally, cachexia [6].

Testosterone and anabolic hormone deficiency are common and have been shown to be independent risk markers for worse outcomes in patients with HF [2]. Low serum androgens do not simply mirror the catabolism that accompanies cardiac disease, but may aggravate HF symptoms and accelerate disease progression [7]. It has been hypothesized that hypotestosteronemia contributes to the poor outcome of HF in men by worsening cardiovascular risk factors [8]. Hypertension, dyslipidemia, diabetes mellitus, obesity and smoking are generally acknowledged as traditional cardiovascular risk factors.

Some studies have investigated the relation between testosterone and cardiovascular risk factors among middle-aged or younger men [9,10]. However, knowledge about the relationship between serum androgen levels and cardiovascular risk factors in elderly male patients with HF is limited. The aim of this study was to investigate the association of serum androgen concentrations with cardiovascular risk factors in Chinese elderly male patients with chronic systolic HF.

Methods

Study population

From January 1, 2002 to December 31, 2004, consecutive elderly male patients with chronic systolic HF hospitalized in five hospitals (located in Beijing, Shanghai and ShanDong Province, respectively) were enrolled in this study. The criteria for study inclusion were (1) male sex; (2) age 60 years or over; (3) a documented history of HF for at least 6 months; (4) LVEF of 45% or less as assessed by echocardiography; and (5) clinical stability and unchanged medications for at least 1 month preceding the study. Exclusion criteria included (1) patients with carcinoma of prostate and other carcinoma; (2) patients with alcoholism and cirrhosis; (3) any acute/chronic illness that might influence hormonal metabolism, and (4) any hormonal treatment, either at the time of the study or in the past. The study protocol was approved by the ethics committees of the participating hospitals, and informed consent was obtained from all participants.

After applying the inclusion and exclusion criteria, 106 patients with chronic systolic HF who agreed to participate were identified. The age of these patients ranged from 60 to 87 years. The mean age was 71.07 ± 7.09 years. Among the causes of HF, 79 patients were diagnosed as having coronary heart disease, 13 patients had a history of hypertension, 11 patients were diagnosed with dilated cardiomyopathy, and 3 patients were diagnosed as having alcoholic cardiomyopathy, pulmonary heart disease and hyperthyroid heart disease, respectively. All patients underwent a detailed clinical examination and demographic characteristics were obtained from the electronic medical record of participating hospitals.

At the same period (from May 2002 to May 2003), Li [4] in our research group surveyed the sex hormone levels in 1080 healthy Chinese men from Beijing, Shanghai, Xian and Chongqing. Four hundred of these men were aged more than 60 years. We included these men as a control group.

Serum concentrations of androgen measurements

Venous blood samples were taken in the morning between 07:00 and 08:30 following an overnight fast and after a supine rest of at least 15 min. After centrifugation, serum blood was collected and frozen at −70 °C until it was measured. Serum concentrations of total testosterone (TT), free testosterone (FT), dehydroepiandrosterone sulfate (DHEAS) and sex hormone binding globulin (SHBG) were measured by enzyme linked immunosorbent assay (DSL, USA). The inter-assay variability coefficients for TT, FT, DHEAS and SHBG were 4.8–6.8%, 5.4–7.5%, 3.7–6.3% and 4.5–6.3%, respectively. The intra-assay variability coefficients were 2.8–4.9%, 0.22–6.18%, 5.0–8.8% and 2.6–4.2%, respectively. All measurements were performed at the Laboratory of Endocrinology, Department of Chinese PLA General Hospital by staff who were blind to the clinical data. In our laboratory, the tenth percentiles of TT were 13.88 nmol/l for healthy men aged 60–69 years and 12.35 nmol/l for men aged 70 years or over, and the 10th percentiles of FT were 0.22 nmol/l for men aged 60–69 years and 0.19 nmol/l for men aged 70 years or over [4].

The serum TT, FT and SHBG concentrations of 400 Chinese healthy elderly men were measured as described previously [4]. It was reported [11] that the average serum DHEA concentration for healthy men was 1200 ng/ml, which we used as a reference value.

Statistical analysis

Continuous variables were expressed as mean ± standard deviation or median, whereas categorical variables were expressed as number (percentage). The statistical significance of differences of androgen concentrations between HF patients and healthy men were determined using the t-test. We used partial correlations adjusted for age to evaluate associations of TT, FT, DHEA and SHBG with various cardiovascular disease risk factors. Statistical data were analyzed using SPSS for Windows version 18 (SPSS Inc., Chicago, IL). A p value <0.05 was considered significant.

Results

The baseline clinical characteristics of 106 elderly male patients with chronic systolic HF are given in Table 1. Mean age was 71.07 ± 7.09 years (25th–75th percentile: 66–75.5 years). Among the 106 male patients with HF, 75% of them were diagnosed as having coronary heart disease. History of diabetes mellitus, smoking and hypertension were common in the patients with HF (52, 43 and 40%, respectively). The mean level of TT was 10.79 ± 2.88 nmol/l and that of FT was 0.049 ± 0.032 nmol/l. Using the age-specific tenth percentiles of TT and FT in healthy men in our laboratory as cutoff points, the prevalence of TT and FT deficiency in men with HF were 78.3% (83/106) and 91.5% (97/106), respectively. Compared with the age-matched healthy men, the serum concentrations of TT, FT and DHEAS in patients with chronic systolic HF were significantly reduced (p < 0.01), whereas the serum level of SHBG was significantly increased (p < 0.01; (Table 2).

Table 1. Baseline characteristics.

	Baseline value (n = 106)
Age (years)	71.07 ± 7.09
BMI (kg/m^2)	23.6 ± 2.98
SBP (mmHg)	125.8 ± 18.7
DBP (mmHg)	72.3 ± 11.4
LVEF (%)	0.37 ± 0.08
TC (mmol/l)	4.45 ± 1.10
TG (mmol/l)	1.75 ± 1.96
HDL-C (mmol/l)	1.09 ± 0.38
LDL-C (mmol/l)	2.49 ± 0.75
Fasting glucose (mmol/l)	5.96 ± 1.91
History of smoking (n, %)	46 (43%)
CHD	79 (75%)
DM	55 (52%)
HT	42 (40%)

Data are shown as mean ± standard deviation for continuous variables and numbers (%) for categorical variables.

BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; LVEF, left ventricular ejection fraction; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CHD, coronary heart disease; DM, Diabetes mellitus; HT, hypertension.

Table 2. Mean serum androgen concentrations in elderly male patients with chronic heart failure and healthy elderly men.

Parameters	Patients with HF (n = 106)	Healthy men (n = 400)	p Value
Age (years)	71.07 ± 7.09	70.35 ± 8.63	>0.05
TT (nmol/l)	10.79 ± 2.88	21.32 ± 7.54	<0.0001
FT (nmol/l)	0.049 ± 0.032	0.34 ± 0.11	<0.0001
DHEAS (ng/ml)	921.76 ± 688.59	1200	0.038
SHBG (nmol/l)	248.94 ± 107.42	49.83 ± 15.50	<0.0001

Data shown as mean ± SD. TT, total testosterone; FT, free testosterone; DHEAS, dehydroepiandrosterone sulfate; SHBG, sex hormone binding globulin.

Pearson correlation coefficients were computed between serum androgen levels and patients’ clinical variables (Table 3). Partial correlations adjusted for age were carried out to eliminate the effect of age on androgen levels. The serum level of TT was negatively correlated with TC and TG (r = −0.313, p = 0.021 and r = −0.378, p = 0.004, respectively), and FT was negatively correlated with TC and LDL-C (r = −0.272, p = 0.046 and r = −0.332, p = 0.023, respectively). Serum SHBG levels in elderly HF men correlated with BMI and smoking history positively (r = 0.418, p = 0.012 and r = 0.376, p = 0.031, respectively). Both TT and FT had a significant negative association with DBP (r = −0.345, p = 0.011 and r = −0.359, p = 0.008, respectively).

Table 3. Correlations of cardiovascular risk factors with serum concentrations of androgen in elderly male patients with chronic heart failure.

Variables	TT		FT		DHEA		SHBG
	r	p	r	p	r	p	r	p
BMI	−0.192	0.080	−0.184	0.072	0.132	0.091	0.418	0.012*
SBP	0.091	0.130	0.012	0.121	0.042	0.280	−0.075	0.450
DBP	−0.345	0.011*	−0.359	0.008*	0.174	0.054	−0.176	0.142
Glucose	−0.163	0.061	−0.160	0.090	−0.012	0.320	−0.200	0.267
TC	−0.313	0.021*	−0.272	0.046*	0.171	0.078	−0.160	0.112
TG	−0.378	0.004*	0.114	0.054	0.068	0.182	−0.027	0.530
HDL-C	0.132	0.075	0.050	0.110	−0.039	0.178	0.011	0.587
LDL-C	−0.249	0.062	−0.332	0.023*	0.087	0.162	0.012	0.579
CHD	0.085	0.452	0.128	0.258	−0.137	0.224	−0.135	0.242
DM	−0.078	0.490	−0.193	0.087	0.029	0.797	0.103	0.361
HT	−0.037	0.743	−0.126	0.264	−0.046	0.683	0.0002	0.981
History of smoking	−0.106	0.062	−0.161	0.058	0.195	0.067	0.376	0.031*

*p < 0.05. Partial correlations analyses adjusted for age were used to determine the associations of cardiovascular risk factors with serum concentrations of androgen. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CHD, coronary heart disease; DM, Diabetes mellitus; HT, hypertension.

Discussion

Testosterone is an anabolic steroid hormone secreted by the Leydig cells in men. Within the circulation, testosterone is present in three major fractions: SHBG-bound testosterone (65–80%), albumin-bound testosterone (20–40%) and free testosterone (1–3%). The free plus the albumin-bound testosterone constitutes the bio-available testosterone fraction [12]. Ageing in males is accompanied by a progressive decline of gonadal function manifested by a decrease in serum TT and FT levels. FT levels decline more rapidly than TT levels because of the age-associated increase in SHBG level [4]. It is known that DHEAS serves as a precursor for estrogens and androgens, and it is believed that DHEA may be an inactive precursor pool for the formation of bioactive steroid hormones. Testosterone deficiency was previously reported to occur in approximately one-third of men aged >66 years, an age group in which HF is more common [5].

There is accumulating evidence showing a relationship between HF symptoms and testosterone deficiency. Wehr and colleagues [13] found an inverse association of FT levels with impaired LV function and higher NYHA classes. The study by Jankowska et al. [5] reported that TT, FT and DHEAS were inversely related to NYHA class, irrespective of cause. Likewise our previous study showed that the serum FT level was inversely related to EF in elderly male patients with chronic systolic HF [14]. In a double-blind, placebo-controlled, randomized trial in 70 elderly patients with chronic HF, Caminiti et al. [15] showed that administration of intramuscular testosterone supplementation with optimal therapy improved functional exercise capacity and muscle strength. DHEA is produced in the healthy heart and there is evidence suggesting that it has beneficial cardiovascular effects [16]. Cardiac production of DHEA was shown to be suppressed in the failing heart, it may be possible that reduced cardiac DHEA production causes a worsening of status during HF [17].

In this study, we investigated the serum androgen concentrations in elderly male patients with chronic systolic HF in China. Reference ranges for testosterone in men are essential for making the diagnosis of androgen deficiency. The Endocrine Society’s guidelines recommend the lower limit of the normal range in healthy young men as an appropriate threshold, citing values of 9.8–10.4 nmol/l [18]. Bhasin et al. established reference ranges for TT and FT in a community-based sample of healthy non-obese young men (19–40 years old), using liquid chromatography tandem mass spectrometry. The 2.5th percentile value in the reference sample for TT was 348.3 ng/dl (12.1 nmol/l), and for FT it was 70.0 pg/ml (243 pmol/l) [19]. Yeap et al. studied healthy older men aged 70–89 years and reported that the 2.5th percentile value for TT was 184 ng/dl (6.4 nmol/l) [20]. In our laboratory reference ranges for TT and FT were established by Li [4]. Compared with age-matched healthy men, the levels of TT, FT and DHEAS were significantly lower and the level of SHBG was significantly higher in men with chronic systolic HF. Our study demonstrated that compared with the gradual decrease in androgen levels associated with aging in healthy men, the level of serum bio-available testosterone decreased more significantly in the elderly male patients with chronic systolic HF.

Low testosterone levels have been observed in men with a variety of chronic diseases and have been found to be associated with a range of cardiovascular risk factors [21]. It was previously believed that the higher prevalence of coronary disease in men may be explained by differences in risk factor profiles between genders. A few systematic reviews have been published that assessed the effect of testosterone on cardiovascular risk, but the results were inconsistent.

Serum testosterone level has been demonstrated to be correlated positively with the cardio-protective high-density lipoprotein cholesterol (HDL-C) and negatively with atherogenic low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG) [9,21,22]. In Page’s study, 70 men with low serum testosterone, age 65 year and older, were assigned to receive testosterone replacement therapy for 36 months. The results showed that testosterone therapy significantly decreased total cholesterol (TC), LDL-C, without affecting HDL-C [23]. Another study of testosterone supplement therapy in elderly men with hypoandrogenism showed that HDL-C and TC levels decreased significantly after treatment with exogenous testosterone [24]. In contrast, a meta-analysis by Isidori et al. [25] suggested that only at supraphysiological doses of testosterone was a decrease of HDL-C noted. Recently, Traish et al. [26] reported that long-term testosterone therapy decreased TC, LDL-C, TG and increased HDL-C levels. Findings from animal model studies indicate that physiological testosterone replacement therapy in animals fed a high cholesterol diet inhibits fatty streak formation and improves HDL-C [27]. Our results showed that the serum level of TT was negatively correlated with TC and TG (p = 0.021 and p = 0.004, respectively) and FT was negatively correlated with TC and LDL-C (p = 0.046 and p = 0.023, respectively), indicating that hypotestosteronemia was related to worse lipid profiles in elderly male patients with chronic HF.

Hypertension is a significant contributory factor to the development of HF. Testosterone and SHBG were reported to a significant inverse correlation with SBP [9,28]. In contrast, in Guder’s study both TT and FT were positively related to SBP, even after adjustment for age and NYHA class [29]. An inverse relationship between testosterone and DBP was observed in patients with dysfunction and HF [30]. However, a meta-analysis showed that testosterone preparations had no significant effects on SBP and DBP [31]. Our study showed that both TT and FT had a significant negative association with DBP, suggesting that male HF patients with higher DBP had lower serum testosterone levels.

Diabetes mellitus is acknowledged as a CAD risk equivalent. Low testosterone levels are common in individuals with diabetes, and clinical hypogonadism is also associated with type 2 diabetes [32]. Some studies showed that low TT and/or FT concentrations and low levels of SHBG were associated with insulin resistance and increased risk of type 2 diabetes [32,33]. A meta-analysis examining testosterone supplementation in patients with systolic HF showed that fasting glucose, fasting insulin and insulin resistance were all significantly improved [6]. However, another study showed the fasting insulin levels were not affected by testosterone supplement [34]. No significant associations were observed between androgen levels and fasting blood glucose or the history of diabetes in the present study. This was due, at least in part, to the relative small sample size and the fact that fasting blood glucose of these patients in our study was well controlled.

A number of studies have suggested that a significant inverse relationship exists between serum levels of testosterone and the degree of obesity in men. Specifically, waist/hip ratio and abdominal or central obesity have been shown to be inversely related to TT and FT [34]. It has been suggested that increased aromatase-mediated androgen inactivation occurs in the abdominal adipose tissue of men, which could be involved in central obesity [35]. Recently, two open-label, single-center, cumulative, prospective registry studies in hypogonadal men evaluated the long-term effects of normalizing testosterone levels on anthropometric parameters [36,37]. Both of these studies demonstrated that normalizing serum testosterone produced consistent reductions in body weight, waist circumference and BMI. These improvements were progressive over the full 5 years of these studies. In this study, we found that SHBG was positively related to BMI, suggesting that patients with relative obesity have lower serum bio-available testosterone.

In males, the effect of smoking on androgen levels is important. Various studies have examined the effects of smoking on serum testosterone levels and have reported conflicting findings. Various studies have reported increased, decreased or unchanged levels of TT in male smokers [38]. SHBG levels have been reported to be higher amongst smokers [39]. In this study, FT showed that a borderline significant negative correlation with smoking in elderly men with HF. This trend was consistent with the significant positive correlation between SHBG and smoking, indicating that the decreasing effects of smoking on testosterone levels is due to the increased SHBG levels.

This study provides additional data concerning the serum concentrations of androgens and their associations with cardiovascular risk factors in elderly men with chronic systolic HF. Further studies are needed to examine how androgens affect the cardiovascular risk profile in elderly HF patients and whether testosterone supplementation therapy can prevent or delay the progression of HF.

The findings of this study, which examined data from male patients aged more than 60 years with chronic systolic HF, may not be generalizable to women and young men. In addition, the modest sample size may have limited the power of our analyses. However, the association of low bio-available testosterone level with a worsened cardiovascular risk profiles was consistent with previous reports. The investigated patient cohort was from several cities and represented a typical real-world elderly HF population in China.

Testosterone is an anabolic hormone with a wide range of beneficial effects on men’s health. It is well known that the level of testosterone decreases with aging in men. In our study, we found that, compared with healthy elderly men, the level of bio-available testosterone in elderly male patients with chronic systolic HF decreased significantly, indicating that bio-available testosterone may play an important role in the pathophysiology of HF. There have been some small studies with testosterone supplementation therapy in men with chronic HF, the results of which have been encouraging. Further large prospective studies should be done to examine the effect of testosterone supplementation therapy on HF. The treatment of HF remains challenging and the testosterone supplement maybe is an effective therapeutic option.

Declaration of interest

The authors report no declarations of interest.

Acknowledgements

We would like to thank Guanghui Liu for help with data analysis.