Archives of Sexual Behavior, Vol. 19, No. 3, 1990
Effects of Androgen Treatment in Impotent Men With Normal and Low Levels of Free Testosterone Cesare Carani, M.D., ~ D a n t e Zini, M.D., L4 A u g u s t o B a l d i n i , M.D., ~ L u c i a n o D e l l a Casa, M.D., 2 A n n a G h i z z a n i , M.D., 3 and P a o l o M a r r a m a , M.D. 1
The relation between sexual function and serum free testosterone (fT) levels, which represent the active fraction o f circulating testosterone, was evaluated. Two groups of impotent male subjects with mild hypogonadism were treated with oral testosterone undecanoate (TU); these men presented with tT/luteinizing hormone (LH) ratio and tT levels at the lower limits of normal. The first group had serum f T below 6.6 ng/ml, considered the lower normal value, according to our laboratory method, whereas the second group had normal f T limits. Administration o f TU improved sexual function only in impotent men with low f T levels, but not in subjects with normal f T levels, even though the tT levels and the t T / L H ratio o f the two groups were not significantly different. The results of our study suggest the presence o f a minimun serum f T threshold, lying near the lower normal range, which determines the male sexual function. Moreover, serum f T levels were a more sensitive index than tT for identifying impotent men who can be successfully treated with androgens. KEY WORDS: testosterone; free testosterone; impotence; testosterone undecanoate; luteinizing hormone.
~Department of Endocrinology, University of Modena, Policlinico, Via del Pozzo, 71, 41100, Modena, Italy. 2Department of Metabolic Diseases, University of Modena. ~Department of Obstetrics, University of Siena, Italy. ~To whom correspondence should be addressed. 223 0004-0002/90/0600-0223506.00/0© 1990PlenumPublishingCorporation
224
Carani et aL
INTRODUCTION Various studies have suggested that hypogonadism in males leads to a reduction in sexual interest and to an increase of erectile difficulties. A marked reduction in frequency of sexual activity is a consequence of low androgen levels, with frequency improved by androgen replacement therapy (Schiavi and White, 1976; Franchimont et aL, 1978; Davidson et aL, 1979; Luisi and Franchi, 1980; Maisey et al., 1981; Skakkebaeck et al., 1981; Bancroft and Wu, 1983; Della Casa et al., 1983; Zini et al., 1985), Androgen replacement therapy represents an effective therapeutic approach in hypogonadal impotent men, and Davidson et al. (1979) and Skakkebaeck et al. (1981) demonstrated that exogenous androgen administration is superior to placebo. Two questions are outstanding regarding the relation between testosterone (T) and male sexual behavior: (i) Does a relation exist between circulating levels of T and sexual behavior? (ii) Is there a minimum T level that allows the expression of male sexuality? In that case, above this threshold T would exert its effects of sexuality fully, whereas subthreshold levels would cause a deficit in sexuality, which would be improved by androgen replacement therapy. In this respect, Damassa et al. (1977), in a study on male rats, found that serum T levels below 1.0 ng/ml compromised sexual behavior, whereas T levels in the normal reference range were not correlated with behavior. The dose-response curve between T levels and copulatory behavior seems markedly below the threshold value which, in turn, is below normal values and appears to shift with age (Gray et al., 1981). The relative T threshold for the different components of sexuality has not yet been established. In man, the dose-response curve between serum total testosterone (tT) levels and sexual behavior does not coincide with the normal range of values of tT (about 3-10 ng/ml, according to Pirke and Doerr, 1973) but seems to hover around the lower values of the normal range. In this regard, studies have shown that oral administration of testosterone undecanoate (TU) may stimulate sexual behavior in hypogonadal patients without elevating tT values to normal (Skakkebaeck et al., 1981); whereas Klinefelter patients, with androgen levels within the lower normal range, may show a certain degree of sexual fuction (Wu et al., 1982). On the other hand, it is difficutt to find significant correlations between androgens and sexual behavior in eugonadal men (Brown and Monti, 1978; Rubin et al., 1979; Lange et aL, 1980; Tsitouras et al., 1982; O'Carrol and Bancroft, 1984). According to Salminies et al. (1982), a threshold for tT, below which sexual function is impaired, also exists in man. On the other hand, there is wide variation between individuals with regard to this threshold, which varies from 2.0 to 405 ng/ml. Then, below 2.0 ng/ml, patients respond to testosterone replacement therapy.
Free Testosterone and Sexuality
225
Another question relating to the importance of serum T for male sexual behavior is the relation between serum free testosterone (fT) levels, the active fraction of circulation T, and sexual behavior. Davidson et al. (1983) found that male sexual behavior was more strongly correlated to serum fT levels and to the fT/luteinizing hormone (LH) ratio than to tT levels. This is not surprising, since it is well known that the biological effects of T are exerted via its free fraction, either directly or after aromatization to estradiol or 5-alpha reduction to dihydrotestosterone (Griffin and Wilson, 1985). Our preliminary data on the endocrine and behavioral effects of TU administration indicated that male sexual response seems to be influenced more by fT levels than by tT levels during androgen therapy (Zini et al., 1985; Carani et al., 1987). On the basis of this work, we performed a double-blind cross-over controlled study in important male patients to verify the sexual effects of administration of T in relation to basal serum fT levels.
SUBJECTS AND METHODS Subjects
The study was performed on a group of 14 mildly hypogonadal patients, 37 _ 6.5 years old (mean _ SD), who presented tT levels at the lower range of normal and a low t T / L H ratio, which is a more sensitive index of Leydig cell function than the absolute levels of either of these hormones (Della Casa et al., 1983). Of the 14 patients, a group of 6 presented serum fT levels below 6.6 ng/ml, which represents the lower limit of the range of fT, by our method. The second group had fT levels lying within the normal range (6.6-19.8 ng/ml). Table I shows the clinical data of the patients studied. The clinical characteristics of impotence in the two groups were homogeneous and suggested progressively deteriorating organic impotence, in agreement with Cooper (1972) (impotence generally persistent and progressively worsening, absence of, or reduced, morning and/or spontaneous erections and absence of or diminished erotic desire in sexual situations). The control group for hormone level data consisted of 57 male subjects, matched for age (mean +_ SD = 38 _+6.2 years). The sexual behavior profile was assessed in 18 of the control subjects (35 +_8.2 years). Thirteen of 18 controls had a regular sexual partner; of these 10 were married. None complained of sexual dysfunctions. No subject had endocrine, liver, or kidney disease, nor were they taking drugs that interfered with their endocrine profile.
~
~
g
~
0 0 0 0 0 0 0 0
Z Z Z Z Z Z Z Z o
0 0 0 0 0 0
0
0
0
0
0
0
0
0
Z Z Z Z Z Z Z Z
ZZZZZZ
0 0 0 0 0 0
0
0
0
0
0
0
0
Z Z Z Z Z Z Z Z
ZZZZZZ
0
~ ft., gen
r~
g
6
N~N.~.~
I
~
a~ a~ a~
:~ ~ ' ~ g o~ ~ ~ ~ .~ ~ o~ -~ .~ ~ ~o o~
I "Un
i:~.o
11 "~ 1t
v~
v~
Free Testosterone and Sexuality
227 Design
Subjects were seen as outpatients in our Department of Endocrinology. All gave informed consent before beginning the study. The same double-blind cross-over design was used with both impotent patients having low fT levels and with the group having normal fT levels. Three successive hormonal assays and behavioral assessments were performed: (i) at the initial phase of treatment, (ii) in two successive 6-week periods in which one-half of the two groups was orally administered first placebo and than TU (80 mg at 7-8 AM, 80 mg at 7-8 PM), and the other half of the two groups first TU and then placebo. Hormones were assayed at the end of each period. Behavioral assessment was carried out through daily diaries and weekly self-ratings recorded by the patients from the 15th day after the beginning of replacement therapy or placebo and for 4 consecutive weeks. This was necessary to avoid the possible influence of the previous treatment procedure. A sexological assessment was made by the same physician at the end of the third period.
Assessment
Behavioral
(a) Sexual activity (with partner or masturbation) was recorded daily on a diary form. A daily diary was kept by each subject at home over a 4-week period of study. A record was made of the occurrence of (i) walking erections, (ii) sexual acts (masturbation or sexual intercourse with partner), and (iii) whether ejaculation occurred. (b) Sexual interest was measured by weekly self-ratings of the following parameters: (i) frequency of sexual thoughts (0-4 scale; 0 = no sexual thoughts, 4 = sexual thoughts several times daily) and (ii) the extent to which those thoughts were associated with sexual excitement (0-2 scale), according to Skakkebaeck et al. (1981) and Bagcroft and Wu (1983). (c) Interview data: At the end of each study period (no treatment, treatment with TU and then placebo, or treatment with placebo and then TU), evaluations for sexual desire and erectile difficulty were made by the same investigator. Ratings were assessed as follows: (i) sexual desire (0-4 scale: 0 = never achieves sexual excitement, 4 = achieves sexual excitement in the majority of occasions), (ii) erectile function (0-4 scale: 0 = never ejaculates when awake, 4 = can ejaculate during vaginal intercourse on more that 50% of occasions).
228
Carani et aL
Hormonal Morning intravenous blood samples were collected at 7:45, 8:00, and 8:15 AM; each determined hormone level represents the mean of the three values. The determinations of serum levels of tT, fT, and LH were performed by previously described RIA methods (Zini et al., 1985). RIA of fT was performed after an ultrafiltration step, according to Vlahos et al. (1982). RIA of both fT and tT was preceded by ether extraction and celite column chromatography and used a charcoal-dextran bound/free separation. Serum LH concentrations were determined by a double-antibody RIA technique. The standard employed was a pituitary preparation calibrated by RIA against a second IRP HMG. Intra- and interassay coefficients of variation were below 10070 for all hormonal assays, and sensitivity for all hormones was acceptable.
Statistical Analysis Levels of tT, fT, and LH and the t T / L H ratio and f T / L H ratio were compared by means of analysis of variance. Behavioral data were evaluated with Mann Whitney's U test for nonparametric data.
RESULTS Table II illustrates the hormonal effects of TU and placebo administration in mildly hypogonadal males with different levels of fT. Of interest is that tT was not different in the two groups of patients with different levels of fT. There was a significant correlation of fT levels with tT levels in the group having iT levels within the normal range (r = 0.640, p < 0.01), whereas in the group with low fT levels there was no correlation (r = -0. 06, ns). Treatment with TU resulted in a marked increase of serum fT and tT levels and of the t T / L H and f T / L H ratios which reached upper normal limits in both study groups. A significant improvement in all the studied behavioral parameters of sexual function paralleled the improved hormone profiles in subjects with low fT (Figs. 1-3). On the other hand, treatment with TU in the subjects with normal fT levels led to a hormone profile comparable to that in the first study group, but it did not ameliorate sexual function. No relationship between fT levels within the normal range and sexological parameters was found.
Free Testosterone and Sexuality
229
0 r~
+1
+1
+i
+1+1+1+1 ~ O ~ O
0
0 0 0 0 4-1 +1 +1
+l+l+l+l
o
o
o
~
"~" ~
~: ~
r~ o
o
~
~t 0 ~ 1 0 ~
~1"~
~
~
o=
.!
0
~
~
+1
+1 +1
+1
o ~
+1
+1
+1
+,~
~ o o o ~
.~.~ ~ ~ .~0
~
v~
.~
~
v~
°
o
~VVVvV
230
Carani et aL
Dno
therapy
p
Effects of Androgen Treatment in Impotent Men With Normal and Low Levels of Free Testosterone Cesare Carani, M.D., ~ D a n t e Zini, M.D., L4 A u g u s t o B a l d i n i , M.D., ~ L u c i a n o D e l l a Casa, M.D., 2 A n n a G h i z z a n i , M.D., 3 and P a o l o M a r r a m a , M.D. 1
The relation between sexual function and serum free testosterone (fT) levels, which represent the active fraction o f circulating testosterone, was evaluated. Two groups of impotent male subjects with mild hypogonadism were treated with oral testosterone undecanoate (TU); these men presented with tT/luteinizing hormone (LH) ratio and tT levels at the lower limits of normal. The first group had serum f T below 6.6 ng/ml, considered the lower normal value, according to our laboratory method, whereas the second group had normal f T limits. Administration o f TU improved sexual function only in impotent men with low f T levels, but not in subjects with normal f T levels, even though the tT levels and the t T / L H ratio o f the two groups were not significantly different. The results of our study suggest the presence o f a minimun serum f T threshold, lying near the lower normal range, which determines the male sexual function. Moreover, serum f T levels were a more sensitive index than tT for identifying impotent men who can be successfully treated with androgens. KEY WORDS: testosterone; free testosterone; impotence; testosterone undecanoate; luteinizing hormone.
~Department of Endocrinology, University of Modena, Policlinico, Via del Pozzo, 71, 41100, Modena, Italy. 2Department of Metabolic Diseases, University of Modena. ~Department of Obstetrics, University of Siena, Italy. ~To whom correspondence should be addressed. 223 0004-0002/90/0600-0223506.00/0© 1990PlenumPublishingCorporation
224
Carani et aL
INTRODUCTION Various studies have suggested that hypogonadism in males leads to a reduction in sexual interest and to an increase of erectile difficulties. A marked reduction in frequency of sexual activity is a consequence of low androgen levels, with frequency improved by androgen replacement therapy (Schiavi and White, 1976; Franchimont et aL, 1978; Davidson et aL, 1979; Luisi and Franchi, 1980; Maisey et al., 1981; Skakkebaeck et al., 1981; Bancroft and Wu, 1983; Della Casa et al., 1983; Zini et al., 1985), Androgen replacement therapy represents an effective therapeutic approach in hypogonadal impotent men, and Davidson et al. (1979) and Skakkebaeck et al. (1981) demonstrated that exogenous androgen administration is superior to placebo. Two questions are outstanding regarding the relation between testosterone (T) and male sexual behavior: (i) Does a relation exist between circulating levels of T and sexual behavior? (ii) Is there a minimum T level that allows the expression of male sexuality? In that case, above this threshold T would exert its effects of sexuality fully, whereas subthreshold levels would cause a deficit in sexuality, which would be improved by androgen replacement therapy. In this respect, Damassa et al. (1977), in a study on male rats, found that serum T levels below 1.0 ng/ml compromised sexual behavior, whereas T levels in the normal reference range were not correlated with behavior. The dose-response curve between T levels and copulatory behavior seems markedly below the threshold value which, in turn, is below normal values and appears to shift with age (Gray et al., 1981). The relative T threshold for the different components of sexuality has not yet been established. In man, the dose-response curve between serum total testosterone (tT) levels and sexual behavior does not coincide with the normal range of values of tT (about 3-10 ng/ml, according to Pirke and Doerr, 1973) but seems to hover around the lower values of the normal range. In this regard, studies have shown that oral administration of testosterone undecanoate (TU) may stimulate sexual behavior in hypogonadal patients without elevating tT values to normal (Skakkebaeck et al., 1981); whereas Klinefelter patients, with androgen levels within the lower normal range, may show a certain degree of sexual fuction (Wu et al., 1982). On the other hand, it is difficutt to find significant correlations between androgens and sexual behavior in eugonadal men (Brown and Monti, 1978; Rubin et al., 1979; Lange et aL, 1980; Tsitouras et al., 1982; O'Carrol and Bancroft, 1984). According to Salminies et al. (1982), a threshold for tT, below which sexual function is impaired, also exists in man. On the other hand, there is wide variation between individuals with regard to this threshold, which varies from 2.0 to 405 ng/ml. Then, below 2.0 ng/ml, patients respond to testosterone replacement therapy.
Free Testosterone and Sexuality
225
Another question relating to the importance of serum T for male sexual behavior is the relation between serum free testosterone (fT) levels, the active fraction of circulation T, and sexual behavior. Davidson et al. (1983) found that male sexual behavior was more strongly correlated to serum fT levels and to the fT/luteinizing hormone (LH) ratio than to tT levels. This is not surprising, since it is well known that the biological effects of T are exerted via its free fraction, either directly or after aromatization to estradiol or 5-alpha reduction to dihydrotestosterone (Griffin and Wilson, 1985). Our preliminary data on the endocrine and behavioral effects of TU administration indicated that male sexual response seems to be influenced more by fT levels than by tT levels during androgen therapy (Zini et al., 1985; Carani et al., 1987). On the basis of this work, we performed a double-blind cross-over controlled study in important male patients to verify the sexual effects of administration of T in relation to basal serum fT levels.
SUBJECTS AND METHODS Subjects
The study was performed on a group of 14 mildly hypogonadal patients, 37 _ 6.5 years old (mean _ SD), who presented tT levels at the lower range of normal and a low t T / L H ratio, which is a more sensitive index of Leydig cell function than the absolute levels of either of these hormones (Della Casa et al., 1983). Of the 14 patients, a group of 6 presented serum fT levels below 6.6 ng/ml, which represents the lower limit of the range of fT, by our method. The second group had fT levels lying within the normal range (6.6-19.8 ng/ml). Table I shows the clinical data of the patients studied. The clinical characteristics of impotence in the two groups were homogeneous and suggested progressively deteriorating organic impotence, in agreement with Cooper (1972) (impotence generally persistent and progressively worsening, absence of, or reduced, morning and/or spontaneous erections and absence of or diminished erotic desire in sexual situations). The control group for hormone level data consisted of 57 male subjects, matched for age (mean +_ SD = 38 _+6.2 years). The sexual behavior profile was assessed in 18 of the control subjects (35 +_8.2 years). Thirteen of 18 controls had a regular sexual partner; of these 10 were married. None complained of sexual dysfunctions. No subject had endocrine, liver, or kidney disease, nor were they taking drugs that interfered with their endocrine profile.
~
~
g
~
0 0 0 0 0 0 0 0
Z Z Z Z Z Z Z Z o
0 0 0 0 0 0
0
0
0
0
0
0
0
0
Z Z Z Z Z Z Z Z
ZZZZZZ
0 0 0 0 0 0
0
0
0
0
0
0
0
Z Z Z Z Z Z Z Z
ZZZZZZ
0
~ ft., gen
r~
g
6
N~N.~.~
I
~
a~ a~ a~
:~ ~ ' ~ g o~ ~ ~ ~ .~ ~ o~ -~ .~ ~ ~o o~
I "Un
i:~.o
11 "~ 1t
v~
v~
Free Testosterone and Sexuality
227 Design
Subjects were seen as outpatients in our Department of Endocrinology. All gave informed consent before beginning the study. The same double-blind cross-over design was used with both impotent patients having low fT levels and with the group having normal fT levels. Three successive hormonal assays and behavioral assessments were performed: (i) at the initial phase of treatment, (ii) in two successive 6-week periods in which one-half of the two groups was orally administered first placebo and than TU (80 mg at 7-8 AM, 80 mg at 7-8 PM), and the other half of the two groups first TU and then placebo. Hormones were assayed at the end of each period. Behavioral assessment was carried out through daily diaries and weekly self-ratings recorded by the patients from the 15th day after the beginning of replacement therapy or placebo and for 4 consecutive weeks. This was necessary to avoid the possible influence of the previous treatment procedure. A sexological assessment was made by the same physician at the end of the third period.
Assessment
Behavioral
(a) Sexual activity (with partner or masturbation) was recorded daily on a diary form. A daily diary was kept by each subject at home over a 4-week period of study. A record was made of the occurrence of (i) walking erections, (ii) sexual acts (masturbation or sexual intercourse with partner), and (iii) whether ejaculation occurred. (b) Sexual interest was measured by weekly self-ratings of the following parameters: (i) frequency of sexual thoughts (0-4 scale; 0 = no sexual thoughts, 4 = sexual thoughts several times daily) and (ii) the extent to which those thoughts were associated with sexual excitement (0-2 scale), according to Skakkebaeck et al. (1981) and Bagcroft and Wu (1983). (c) Interview data: At the end of each study period (no treatment, treatment with TU and then placebo, or treatment with placebo and then TU), evaluations for sexual desire and erectile difficulty were made by the same investigator. Ratings were assessed as follows: (i) sexual desire (0-4 scale: 0 = never achieves sexual excitement, 4 = achieves sexual excitement in the majority of occasions), (ii) erectile function (0-4 scale: 0 = never ejaculates when awake, 4 = can ejaculate during vaginal intercourse on more that 50% of occasions).
228
Carani et aL
Hormonal Morning intravenous blood samples were collected at 7:45, 8:00, and 8:15 AM; each determined hormone level represents the mean of the three values. The determinations of serum levels of tT, fT, and LH were performed by previously described RIA methods (Zini et al., 1985). RIA of fT was performed after an ultrafiltration step, according to Vlahos et al. (1982). RIA of both fT and tT was preceded by ether extraction and celite column chromatography and used a charcoal-dextran bound/free separation. Serum LH concentrations were determined by a double-antibody RIA technique. The standard employed was a pituitary preparation calibrated by RIA against a second IRP HMG. Intra- and interassay coefficients of variation were below 10070 for all hormonal assays, and sensitivity for all hormones was acceptable.
Statistical Analysis Levels of tT, fT, and LH and the t T / L H ratio and f T / L H ratio were compared by means of analysis of variance. Behavioral data were evaluated with Mann Whitney's U test for nonparametric data.
RESULTS Table II illustrates the hormonal effects of TU and placebo administration in mildly hypogonadal males with different levels of fT. Of interest is that tT was not different in the two groups of patients with different levels of fT. There was a significant correlation of fT levels with tT levels in the group having iT levels within the normal range (r = 0.640, p < 0.01), whereas in the group with low fT levels there was no correlation (r = -0. 06, ns). Treatment with TU resulted in a marked increase of serum fT and tT levels and of the t T / L H and f T / L H ratios which reached upper normal limits in both study groups. A significant improvement in all the studied behavioral parameters of sexual function paralleled the improved hormone profiles in subjects with low fT (Figs. 1-3). On the other hand, treatment with TU in the subjects with normal fT levels led to a hormone profile comparable to that in the first study group, but it did not ameliorate sexual function. No relationship between fT levels within the normal range and sexological parameters was found.
Free Testosterone and Sexuality
229
0 r~
+1
+1
+i
+1+1+1+1 ~ O ~ O
0
0 0 0 0 4-1 +1 +1
+l+l+l+l
o
o
o
~
"~" ~
~: ~
r~ o
o
~
~t 0 ~ 1 0 ~
~1"~
~
~
o=
.!
0
~
~
+1
+1 +1
+1
o ~
+1
+1
+1
+,~
~ o o o ~
.~.~ ~ ~ .~0
~
v~
.~
~
v~
°
o
~VVVvV
230
Carani et aL
Dno
therapy
p
