The clinical relevance of sex hormone levels and sexual activity in the ageing male
Objectives To assess the changes in sex hormone levels with age and the relationship of sexual functioning to testosterone levels, evaluating serum testosterone levels and erectile function in men with lower urinary tract symptoms (LUTS).
Patients and methods The study included 213 men with LUTS (age range 31–78 years) who had no confirmed erectile dysfunction. Their serum total and free testosterone, and sex-hormone binding globulin (SHBG) levels were measured, and they completed the International Index of Erectile Function (IIEF) questionnaire.
Results The total and free testosterone levels decreased and SHBG increased with age, but only the change in free testosterone and SHBG were statistically significant. The correlation with age was closer for free testosterone (r = − 0.356, P < 0.001) than for SHBG (r = 0.177, P = 0.010). Regression analysis of the five domain scores of the IIEF and three hormonal levels, after correcting for age, showed that free testosterone level was significantly correlated with erectile function (r = 0.2136, P = 0.005) and orgasmic function (r = 0.179, P = 0.020), but SHBG levels were significantly correlated only with orgasmic function (r = − 0.154, P = 0.046). Total testosterone levels showed no significant correlation with any of the five domains of the IIEF.
Conclusions Of the sex hormone levels, the change in free testosterone correlated most closely with ageing and had the closest correlation with sexual activity. Contrary to previous reports, free testosterone and SHBG levels were significantly correlated with orgasmic function and/or erectile function rather than sexual desire. A complete study of sex hormone levels is needed to evaluate patients with erectile dysfunction.
Androgens play an important role in the development of male secondary sexual characteristics and androgen deficiency may result in structural abnormalities of the penis [1–3]. Androgen depletion via surgical or medical castration generally results in loss of libido and a decline in erectile function [3–5]. Reportedly, men with high testosterone levels are more likely to have higher levels of sexual activity, and men who desire sexual intercourse more than once a week have higher testosterone levels than men satisfied with a lower frequency . Nevertheless, the correlation between testosterone levels and libido is poor [7–9], and the testosterone levels required to sustain normal sexual interest are apparently low . However, Anderson et al. reported that increasing testosterone to supraphysiological concentrations can increase sexual activity, although the effect may be marginal. Most clinical studies of testosterone and sexual function are related to libido, ejaculation and nocturnal erections rather than reflexogenic erection. In animal models erection is androgen-dependent; in castrated rats the intracavernosal pressure was reduced in response to electrical-field stimulation of the cavernosal nerve. Testosterone treatment of castrated rats restored this response .
The role of sex hormones in human sexuality, in the mechanism of erection and in the pathogenesis of impotence has not yet been established, and testosterone supplementation for erectile dysfunction (ED) is under debate. Several studies have shown that ageing in men is associated with a decrease in sexual interest and behaviour, and with an increased prevalence of erectile disorders [12–16]. There is information showing a decline in gonadal function with increasing age [17–21]. These behavioural and hormonal data have led to the hypothesis that androgen deficiency may contribute to the decrease in sexual desire and activity in older men . The hypothesis rests on evidence, primarily derived from androgen replacement in hypogonadal men, that testosterone, and more specifically the bioavailable fraction of this hormone, is necessary to sustain sexual drive and behaviour [22–24]. However, few studies have assessed the role of gonadal hormones on age-related changes in male sexuality. Because ED increases markedly with age, from 8% at 55 years to > 50% at 75 years , androgen deficiency may play a subsidiary role in many cases of ED in the elderly, in addition to a non-hormonal cause, e.g. atherosclerosis or polyneuropathy.
The aim of the present study was to assess the changes in testosterone levels with age and the relationship of sexual function, including penile erection, with sex hormone levels, by evaluating serum testosterone levels and erectile function in men with LUTS.
Patients and methods
The study comprised 213 men with LUTS (aged 31–78 years) who were married or living in a stable relationship with a female sexual partner for at least 1 year. Sexual function was assessed using the International Index of Erectile Function (IIEF) at the first visit. Blood samples were collected between 08.00 and 09.00 hours after a 12-h overnight fast. Total and free testosterone and sex-hormone binding globulin (SHBG) were measured using RIAs (Coat-A-Count total Testosterone® and free Testosterone® kits, Diagnostic Products Corp., USA; and the DSL-7400® kit, ActiveTM, USA, respectively).
Changes in the hormonal variables and the five domains of the IIEF (erectile function, orgasmic function, sexual desire, intercourse satisfaction, and overall satisfaction) with age were analysed using the Kruskal–Wallis anova using age groups. If there were statistically significant differences in the age groups the least significant difference test and the multiple analysis test were applied. Relationships between hormonal variables and sexual behaviour were analysed using Spearman partial correlation after controlling for age. To assess the associations with age, the hormonal variables were analysed using linear regression. To better approximate a normal distribution the free and total testosterone values were root-transformed and SHBG values log-transformed.
The mean serum total and free testosterone levels decreased by ≈17% and ≈36%, respectively, from age < 40 years to > 70 years, whereas the mean serum SHBG levels increased by ≈21% over the same range (Table 1). However, only the change in free testosterone and SHBG were statistically significant and the correlation with age was closer for free testosterone (r = − 0.356, P < 0.001) than for SHBG (r = 0.177, P = 0.010).
Table 1. The serum levels of total and free testosterone and SHBG, and the IIEF scores in the five age groups of Korean men
|Free testosterone, ng/mL*||4.45 (1.35)||4.44 (1.18)||4.61 (1.63)||4.39 (1.41)||3.63 (1.29)|
|Total testosterone, pg/mL||14.60 (3.44)||13.41 (3.55)||12.36 (3.55)||11.92 (4.01)||9.49 (3.17)|
|SHBG, nmol/L*||29.17 (10.60)||31.83 (11.61)||34.14 (12.39)||36.84 (17.11)||43.08 (12.76)|
|Erectile function*||19.3 (7.6)||17.3 (6.3)||14.2 (7.6)||11.3 (7.7)||7.1 (7.2)|
|Sexual desire||6.1 (1.8)||5.0 (1.8)||4.9 (2.0)||4.6 (2.1)||4.5 (2.3)|
|Orgasm function*||7.4 (2.8)||5.7 (3.1)||4.7 (3.2)||4.1 (3.7)||2.4 (3.0)|
|Satisfaction with intercourse||7.4 (3.6)||6.1 (2.8)||5.5 (3.6)||4.9 (3.8)||3.3 (3.7)|
|Overall satisfaction||5.0 (1.8)||4.6 (1.9)||4.0 (1.9)||4.8 (2.1)||4.2 (2.6)|
The total score on the IIEF decreased significantly with ageing; the score for ‘erectile function’ and ‘orgasm’ decreased consistently for every decade, but ‘desire’ differed only between < 40 and > 40 years, with no differences in scores in those > 40 years old. Although ‘intercourse satisfaction’ showed differences with ageing, overall ‘satisfaction with sex’ did not (Table 1). Among the five domains of the IIEF, ‘erectile function’ showed the most significant change, with large differences in scores between those < 50 and > 50 years old.
All the hormonal and IIEF differences among groups were compared after adjusting, by analysis of covariance, for the effect of age (Table 2). Free testosterone correlated significantly with ‘erectile function’ and ‘orgasmic function’, but SHBG correlated significantly with ‘orgasmic function’ only. Free testosterone and SHBG did not correlate significantly with ‘sexual desire’. Total testosterone did not correlate significantly with any of the five domains (Table 2). Thus free testosterone levels had the closest correlation with sexual activity.
Table 2. The Spearman partial correlation coefficients (with P in parentheses), controlling for age, among hormone levels and the five domains of the IIEF
|Erectile function||0.057 (0.460)||0.212 (0.005)||− 0.137 (0.217)|
|Sexual desire||0.066 (0.274)||0.047 (0.546)||− 0.009 (0.206)|
|Orgasm function||0.086 (0.274)||0.179 (0.020)||− 0.154 (0.046)|
|Satisfaction with intercourse||0.104 (0.177)||0.009 (0.956)||− 0.144 (0.112)|
|Overall satisfaction||0.031 (0.679)||0.079 (0.295)||0.012 (0.842)|
The first evidence for an age-associated decrease in testosterone secretion was reported by Hollander and Hollander in 1958, where testosterone (determined in spermatic vein blood) was lower in elderly than in young men . Many studies have assessed the changes in testosterone and associated reproductive hormones during adult life in men, showing that in healthy men the mean serum testosterone levels decrease by ≈30% from 25 to 75 years of age, whereas mean serum free testosterone levels decrease by as much a half over the same period. In the present study, the respective mean total and free testosterone levels decreased by ≈17% and 36% from < 40 to > 70 years, whereas the mean serum SHBG levels increased by up to 22%. The greater decrease in free testosterone levels is a consequence of the age-associated increase in binding capacity of SHBG.
In their review of male ageing and sexuality, Schiavi and Rehman  noted that in addition to a decline in sexual interest with age there is less sexual activity and an increased prevalence of sexual dysfunction, although there are substantial individual differences. Age-associated changes in male sexual response and function include: (i) decline in libido; (ii) decrease in the number and frequency of morning erections; (iii) reduced penile sensitivity; (iv) reduced sexual arousal, both mentally (visual, psychological) and physically, i.e. erectile capacity (delayed, less rigid, more difficult to sustain), reduced scrotal vasal congestion, impaired testicular elevation, and a poorly defined sense of impending orgasm; (v) a prolonged duration of the plateau phase with decreased or absent pre-ejaculatory secretions and emission; (vi) reduced duration and intensity of orgasm with a shorter period of ejaculatory inevitability; (vii) a prolonged refractory period; and (viii) reduced swelling and erection of the nipples, absence of maculopapular rash, and extragenital muscle spasm at climax.
In the present study the total IIEF score decreased significantly with age; among the five domains, the scores for erectile function, orgasmic function and intercourse satisfaction decreased significantly (Table 1). In contrast to other reports, sexual desire showed no change with age after 40 years. This difference is possibly attributable to cultural differences between Korean and western societies; in Korea it is considered a virtue not to openly express a desire for something, especially so for sex with increasing age. The erectile function domain of the IIEF showed the most significant change with age, agreeing with previous reports that age is the most significant risk factor for ED.
The relationship between testosterone levels and erectile function, especially in older men, is complex, and only a small percentage of ED in this age group appears to be secondary to ‘testosterone deficiency’. However, there is a common understanding that normal penile erection depends on androgen steroids, especially testosterone, and it is generally true that castration of adult men impairs erection. However, audiovisual sexual stimulation induced functional erections in four of 16 men who had been castrated to treat prostatic cancer ; all these patients were reportedly sexually active before castration. The men with a normal response to such stimulation had abnormally low (but significantly higher) testosterone levels than those who did not respond (1.125 vs 0.628 pg/mL). In the present study, of the three hormones assessed, free testosterone had the closest correlation with the IIEF for erectile and orgasmic function, after correcting for age, but there was no significant correlation with the remaining domains. These results differ from those in previous reports. Most clinical studies of testosterone and sexual function show that testosterone significantly increases sexual desire but has a minor effect on erectile function. However, there are several reports of a correlation between testosterone and erectile function. Schiavi and Rehman  noted that in addition to a decline in sexual interest, there is a decreased frequency of sexual activity and an increased prevalence of ED with age. Davidson et al., studying hypogonadal men given testosterone replacement or placebo, found significant dose-related increases in the frequency of erections, including nocturnal erections, and of intercourse, in the men given replacement therapy. Animal studies also provide evidence of the role of testosterone in erectile function [1,11,30,31]. The primary action of androgens in erectile function in the rat is postulated to be via stimulation of nitric oxide (NO) synthesis [30–35]. Neural NO synthase mRNA and protein expression were reduced in castrated rats [32,36], although other studies show that neural NO synthase protein expression was not reduced but NO synthase activity was diminished [37,38]. Castration also reduced NO synthase-containing nerve fibres, innervating the corpus cavernosum tissue, and enhanced nonadrenergic noncholinergic nerve-mediated relaxation in isolated corpus cavernosum strips, and increased the reactivity to α-adrenergic stimulation . In addition, castration induced programmed cell death of smooth muscle cells in the rat penis, suggesting that androgens may have an important role in maintaining smooth muscle growth and functional integrity .
In conclusion, in Korean men, changes in free testosterone were closely correlated with ageing and sexual activity. In contrast to previous reports, free testosterone and SHBG levels were also correlated with orgasmic function and/or erectile function rather than sexual desire. A detailed study of sex hormone levels is needed to evaluate patients with ED.
Sung Won Lee, Department of Urology, Samsung Medical Center, Sungkyunkwan University School of Medicine, #50 Ilwon-dong Kangman-ku, Seoul, Korea, 135–710.