Relationship between lower urinary tract symptoms and serum levels of sex hormones in men with symptomatic benign prostatic hyperplasia

Authors


Vincenzo Favilla, Department of Urology, University of Messina, Consolare Valeria Avenue Messina, Messina 98100, Italy.
e-mail: vincenzo.favilla@libero.it

Abstract

Study Type – Aetiology (case series)
Level of Evidence 4

OBJECTIVES

To investigate a possible association between the severity of lower urinary tract symptoms (LUTS) and the serum levels of sex hormones in men with symptomatic benign prostatic hyperplasia (BPH) that underwent surgery for severe benign prostatic obstruction.

PATIENTS AND METHODS

In all, 127 selected men with symptomatic BPH attending our urology clinic were recruited. The clinical conditions of BPH were assessed by digital rectal examination, serum prostate-specific antigen (PSA) determination, International Prostate Symptom Score (IPSS), transrectal ultrasonography and maximum urinary flow rate (Qmax) value at uroflussimetry. Before surgery, we measured the serum concentrations of total testosterone (TT) and free testosterone (FT), oestradiol, prolactin, luteinizing hormone and follicle-stimulating hormone. We excluded men with endocrine diseases, those with prostate disease who were receiving antiandrogen therapy and those with psychological diseases. The relationships between the IPSS score and serum sex hormone levels were determined.

RESULTS

The final study population consisted of 122 men (mean age of 70.66 years), as five were excluded (three due to incomplete evaluation and two who were diagnosed with prostate cancer). On statistical analysis, the total IPSS was significantly associated with age (r= 0.405, P < 0.001) and TT (r= 0.298, P= 0.020) but not with FT or the serum levels of the other sex hormones. The serum levels of testosterone and IPSS did not correlate with prostate volume and Qmax. PSA level and age correlated with prostate volume (r= 0.394, P < 0.001; r = 0.374, P < 0.001, respectively). We distinguished two subgroups of patients: the first group of 40 men with an IPSS of <19 and the second group of 82 with an IPSS of >19, and we evaluated the median levels of TT in each group. There was an increased risk of LUTS in men with a greater serum concentration of TT (P= 0.042), although the mean TT level was in the normal range.

CONCLUSIONS

In the present study, the severity of LUTS was associated with age and serum levels of TT but only age correlated with the measures of BPH, especially prostate volume. The potential effects of testosterone on LUTS may well be indirect. Additional large studies are needed to confirm these preliminary results.

Abbreviations
TT

total testosterone

FT

free testosterone

E2

oestradiol

DHT

dihydrotestosterone

IPSS

International Prostate Symptom Score

Qmax

maximum urinary flow rate.

INTRODUCTION

BPH is very common among ageing men and can cause LUTS, which can be bothersome and which tend to interfere with quality of life [1–3]. Two factors that are largely accepted to play a role in the aetiopathogenesis of the disease are ageing and androgens [3–6]. First, the prevalence of BPH increases from 8% in the fourth decade of life to >70% in the seventh decade [3]. Second, it has been established that androgens are important for prostate development and growth and the maintenance of structural and functional integrity [4–6]. Because sex steroid hormones are thought to be an important determinant of prostate growth [5,6], they are also thought to contribute to the development and maintenance of LUTS secondary to BPH in older men. Although, in ageing men, the impact of androgen status on the prostate’s condition, such as prostate volume and LUTS, have not been consistent [7–9]. Several studies have found that ethnic differences exist in the prevalence of LUTS and BPH [10–12] and in serum androgen levels [13]. Many studies have tried to establish a relationship between sex steroids and BPH but few studies have analysed the relationship between circulating testosterone and LUTS. One study reported a relationship between symptoms of LUTS and plasma total testosterone (TT) and bioavailable testosterone but this relationship disappeared after statistical adjustment for age [14]. There were no consistent correlations between TT and calculated free testosterone (FT) and symptoms of LUTS in another study, but there was a relationship with androstanediolglucuronide, a metabolite of dihydrotestosterone (DHT) and oestradiol (E2) [15]. Overall, it has been difficult to relate plasma testosterone to LUTS, but it is noteworthy that, within certain testosterone levels limits, the signs and symptoms of testosterone deficiency in men do not relate in a uniform pattern to testosterone concentrations, which may be in part explained by the properties of the androgen receptors [16]. The aim of the present study was to investigate a possible association between the severity of LUTS and the serum levels of sex hormones in men with symptomatic BPH who underwent surgery for severe benign prostatic obstruction.

PATIENTS AND METHODS

In all, 127 randomly selected men with severe, symptomatic BPH attending our general urology clinic from January to September 2009 were recruited for the study. We excluded men with endocrine diseases, those with prostate diseases who were receiving antiandrogen therapy and those with psychological diseases. The clinical conditions of BPH were assessed by DRE, serum PSA level determination, International Prostate Symptom Score (IPSS), TRUS and maximum urinary flow rate (Qmax) at uroflussimetry. Patients with abnormal DRE findings or an elevated serum PSA level (i.e. >4.0 ng/mL) were referred for prostate biopsy to exclude the possibility of prostate cancer. Before surgery, we measured the serum concentrations of TT and FT, E2, prolactin, LH and FSH. Blood samples were collected before any prostate evaluation, including DRE and TRUS, and were taken between 0800 hours and 1100 hours to minimize the confounding effects of diurnal variation in the concentration levels of the hormones. The levels of PSA, TT, FT, LH, FSH, prolactin, and E2 were measured by radioimmunoassay. The correlations between the different variables were first estimated using Pearson’s correlation coefficients. Statistical significance was defined as P < 0.05. Pearson’s rank correlation test was first used to determine the relationship between the IPSS and serum sex hormone levels. The strongly significantly (P < 0.01) associated factors with IPSS on Pearson’s rank correlation test were estimated using multivariate linear regression models. All participants provided written informed consent for participation in this study, which was approved by the Ethical Committee of our hospital.

RESULTS

The final study population consisted of 122 men with a mean (range) age of 70.66 (39–86) years. In all, five men were excluded: three due to incomplete evaluation and two who were diagnosed with prostate cancer. The mean (sd) IPSS of the study population was 21.82 (4.08), the mean (sd) prostate volume was 62.60 (29.83) cm3 and the mean (sd) Qmax was 9.7 (7.9) mL/s. The mean (sd) values for PSA, TT, FT, LH, FSH, prolactin, and E2 were 3.70 (2.17) ng/mL, 405.17 (160.34) ng/mL, 7.50 (3.43) pg/mL, 7.01 (3.73) mIU/mL, 10.26 (8.51) mIU/mL, 202.64 (119.37) ng/mL, 27.76 (18.07) ng/mL, respectively. The respective endocrine variables and clinical characteristics of the study population are given in Table 1. On statistical analysis, the IPSS correlated positively with age (r= 0.405, P < 0.001) and TT (r= 0.298, P= 0.020) but did not correlate with serum FT and the other sex hormones levels, prostate volume and Qmax. Prostate volume correlated positively with age (r= 0.374, P < 0.001) and PSA level (r= 0.394, P < 0.001) but not with the serum levels of the sex hormones, IPSS and Qmax. Age correlated positively with prostate volume (r= 0.305, P < 0.001), IPSS (r= 0.165, P= 0.027) PSA level (r= 0.381, P < 0.001) and E2 (r= 0.183, P= 0.023) and negatively with FT (r=−0.404, P < 0.001). There was no correlation with age for TT and the other sex hormone levels (Table 2). The results from the multivariate linear regression models indicated that age and TT were significantly associated with the total IPSS (Table 3). We distinguished two subgroups of patients. The first group of 40 men with an IPSS of <19 and the second group of 82 with an IPSS of >19, and we evaluated the median levels of TT in each group (346.80 ng/mL and 425.69 ng/mL, respectively). There was an increased risk of LUTS in men with a greater serum concentration of TT, although the mean level of testosterone was in the normal range. In particular, the difference between the mean TT of 78.89 ng/mL that existed between these two groups was significantly associated with severity of LUTS (P= 0.042).

Table 1.  Principal clinical and endocrinological data (n= 122)
VariableMean (sd)
Clinical
Age, years 70.66 (9.08)
Total IPSS 21.82 (4.08)
Prostate volume, cm3 62.60 (29.83)
Qmax, mL/s  9.7 (7.9)
PSA level, ng/mL  3.70 (2.17)
Endocrinological
TT, ng/mL405.17 (160.34)
FT, pg/mL  7.50 (3.43)
LH, mIU/mL  7.01 (3.73)
FSH, mIU/mL 10.26 (8.51)
Prolactin, ng/mL202.64 (119.37)
E2, ng/mL 27.76 (18.07)
Table 2.  Pearson correlations of endocrinological variables with age, prostate volume, IPSS and Qmax
 Prostate volumePIPSSPAgeP
  • *

    P= 0.01 but <0.05.

Age0.374<0.0010.405<0.0011 
Prostate volume1 −0.1700.1910.305<0.001
Qmax0.1890.145−0.1550.2340.0270.742
IPSS−0.1700.1911 0.1650.027*
PSA0.394<0.0010.0180.8360.381<0.001
TT0.1030.4030.2980.020*−0.1110.166
FT−0.1740.1790.0390.763−0.404<0.001
LH−0.0550.6750.2270.0790.0960.234
FSH−0.0020.9890.0320.3430.0760.349
PRL−0.0350.7900.0690.5980.0160.846
E20.1780.1710.0410.7510.1830.023*
Table 3.  Multiple linear regression-derived β coefficients and P values for factors significantly associated with IPSS
Factorβ coefficientP
  • *

    P= 0.01 but <0.05.

Age0.3200.042*
Prostate volume0.0950.575
Qmax0.2350.157
TT0.3330.048*

DISCUSSION

LUTS are common in older men, although intercultural and ethnic differences in the prevalence of LUTS and BPH have been reported in several studies [10–12]. An analysis of the data from the Third National Health and Nutrition Examination Survey (NHANES III) reported that 22% of USA men aged 70–79 years had undergone noncancer prostate surgery, presumably for LUTS and that the prevalence of specific urinary symptoms increased with age [17]. Furthermore, several large-scale epidemiological studies have identified age as the most important risk factor for BPH and LUTS [18–20]. These data were also confirmed in the present study. Androgens have been well accepted as the second risk factor for BPH. Sex steroid hormones are an important determinant of prostate growth [21,22] and they are also to contribute to the development and maintenance of LUTS secondary to BPH in older men. In the prostate, testosterone is converted into the more potent androgen, DHT by 5α-reductase type II. It is thought that DHT has a central role in BPH development and maintenance because the inhibition of 5α-reductase activity is associated with decreased serum DHT concentrations and decreased prostate size [23,24]. E2 has been hypothesized to potentiate the effects of androgens in inducing BPH by inducing the androgen receptor [4,25]. Despite the co-localization of androgen receptors with oestrogen receptors in the LUT, few studies have been done on the relationship between LUTS and sex hormones status [26].

Several studies have focused on the relationship between serum androgen concentrations and clinical BPH in elderly men, but the results have not been consistent. Joseph et al. [9] reported that a large prostate volume was marginally associated with increased TT level in African-American men, but Meikle et al. [7] found an inverse correlation between prostate volume and TT level in 214 male twins based on white populations. Others have not found a significant association between TT level and prostate volume [4,8,19]. There were no consistent correlations between TT and calculated FT and LUTS in another study, but there was a relationship with androstenediolglucuronide, a metabolite of DHT and with E2[15] Additionally, some studies found that prostate volume was not associated with FT level or bioavailable testosterone level after adjustment for age [2,4]

In the present study, it appeared that the dominant predictor of LUTS is the patient’s age. However, in older men with symptomatic BPH, there was a significant association between severity of LUTS and serum TT level, but there was no significant association between IPSS and other serum sex hormones levels. The prostate volume did not correlate with the severity of LUTS, serum testosterone levels or the levels of other sex hormones. Only age was a significant predisposing factor for prostate volume on statistical analysis. From the above-mentioned facts, the age-related growth of the prostate cannot be explained by a mere increase or decrease in serum androgens. Schultheiss et al. [27] reviewed previous studies and concluded that the link between androgens and age-related growth of the prostate might be explained by a shift of the hormonal ratio (e.g. the androgen/oestrogen ratio), the changing intraprostatic hormonal level, or a modified action of hormones and their respective receptors, as well as of intraprostatic enzymes (e.g. 5α-reductase). Additional large studies are needed to evaluate the above mechanisms.

In conclusion, the present results have shown that the severity of LUTS is associated with age and the serum levels of TT but only age was a significant predisposing factor for prostate volume on statistical analysis. Testosterone itself might not be the ‘prime mover’ of the effects of testosterone on those structures of the urinary tract anatomically and functionally related to LUTS. However, androgen receptors have been found to be present largely in the epithelial cells of the urethra and the bladder [28], and in another study, the role of testosterone and its metabolites on maintaining the reflex activity in the pelvic part of the autonomic nervous system could be shown in rats [29]. Others have postulated the influence of testosterone on postsynaptic nongenomic receptors, which can modulate detrusor activity [30,31]. Finally, not only in the penis but also in other parts of the urogenital tract nitric oxide acts as a nonadrenergic noncholinergic neurotransmitter and the action of testosterone on the urogenital tract may be mediated by this system [32]. Additional large studies are needed to confirm these preliminary results.

CONFLICT OF INTEREST

None declared.

Ancillary