International Journal of Impotence Research (2000) 12, Suppl 1, S64±S69 ß 2000 Macmillan Publishers Ltd All rights reserved 0955-9930/00 $15.00 www.nature.com/ijir

CLINICAL RESEARCH Effects of alpha-2 blockade on sexual response: experimental studies with Delequamine (RS15385)

J Bancroft1

1The Kinsey Institute, Indiana University, USA

The role of alpha-2 receptors and alpha-2 antagonists in central and peripheral mechanisms of sexual response are discussed. It is concluded that the predominant role of the alpha-2 antagonist centrally is to increase arousal which in certain circumstances, is sexual. It is further concluded that the predominant role of the alpha-2 antagonist peripherally is to modulate (block) the (NE)-induced contractility in the of the penis. How the central arousal mechanisms are speci®cally linked to sexual response is not understood. Experimental studies with a selective alpha-2 antagonist, delequamine, are brie¯y reviewed and their complex results discussed. Evidence from sleep studies was consistent with delequamine having both central excitatory and inhibitory effects, dependent on dosage. The possibility that men with psychogenic might have increased central alpha-2 tone was considered. The apparent loss of responsiveness to the alpha-2 antagonist in older dysfunctional men was discussed. More questions are raised than answered; further research is needed in this area. International Journal of Impotence Research (2000) 12, Suppl 1, S64±S69

Keywords: alpha-2 blockade; delequamine

Role of locus coeruleus (l.c.) and the lateral tegmental nucleus (l.t.n.).2 Each nucleus has with ascending and descending projections. Of the two, The pharmacological agent with the longest history the l.c. system is most likely to be relevant to sexual for enhancing sexual response is , origin- arousal and response; it is activated by novel 1 ally available as a plant (bark) extract. Yohimbine's sensory input to produce a state of central arousal principal pharmacological action is as an alpha-2 which may be associated with aversive or rewarding adrenoceptor antagonist. Activation of presynaptic situations. The ascending projections of the l.c. alpha-2 receptors increases re-uptake of norepi- terminate, quite widely, in the dorsal thalamus, nephrine (NE) at the , hence reducing NE hypothalamus, basal forebrain, including hippocam- transmission. Antagonism of presynaptic alpha-2 pus, and the frontal cortex. The descending projec- receptors will therefore have the principal effect of tions from the l.c. go to the spinal cord (mainly increasing NE transmission. Antagonism of post- ventral horn) and to sensory nuclei in the synaptic alpha-2 receptors will inhibit NE action at stem. The l.c. neurons are populated with alpha-2 the smooth muscle. Both central and peripheral receptors.3 effects of such a drug need to be considered. Davidson's group showed that yohimbine and other alpha-2 antagonists such as , could substantially restore sexual behavior in the castrated The role of the norepinephrine system male rat and that the sexually enhancing effect of yohimbine was independent of any direct effect on Central effects erection, leading them to conclude that the drug's principal effect was on central arousal.4 There are good reasons for expecting an alpha-2 antagonist to have central effects. The NE system in Peripheral effects the brain originates in two brain stem nuclei, the Some attention has been paid to the presence and Correspondence: Dr J Bancroft, The Kinsey Institute, possible function of alpha-2 receptors in the erectile 5 Morrison Hall 313, Bloomington, IN 46164. tissue. Saenz de Tejada et al concluded that Received 2 September 1999; accepted 15 November 1999 adrenergic activity is modulated, in part, by pre- Effects of alpha-2 blockade on sexual response J Bancroft S65 synaptic alpha-2 receptors in the corpus caverno- satiation or exhaustion in rats. This state, which sum; in those circumstances an alpha-2 antagonist results from repeated copulation, is characterized by would be expected to increase NE transmission and a prolonged period of sexual unresponsiveness, hence smooth muscle contractility. Andersson6 lasting for several days before recovery.13,14 This suggested that such alpha-2 receptor activity may unresponsive state can be partially corrected with a play an important role in the cavernous wall, variety of pharmacological manipulations including but not in the trabecular smooth muscle. However, the opioid antagonist, naloxone (indicating that Traish et al7 identi®ed post-synaptic alpha-2 recep- opioid inhibition is probably involved), 8-OH- tors on the trabecular smooth muscle, though the DPAT, a 5HT1a (indicating that serotonergic function of these receptors remains unclear.8 The inhibition is involved), and yohimbine. However, very limited evidence of the effects of injecting the effects of the ®rst two require an intact NE alpha-2 antagonists into the corpus cavernosum system, indicating complex interactions between indicates very little effect.9 If a drug produces a peptide and systems.15±17 predominantly presynaptic antagonistic effect on Thus, in those circumstances where central alpha-2 receptors in the erectile tissues, the con- arousal is elicited by an aversive situation, we sequence would be an increased release of NE and should not be surprised if this NE system produces a resulting increase in contractile tone. The central arousal of `anxiety' type and peripheral possibility of a post-synaptic effect of alpha-2 inhibition of sexual response at the same time. on inhibition of NE binding on Conversely, when central arousal is in response to the trabecular smooth muscle has to be considered, sexual stimuli we might expect a reduced pre- with two such contrasting effects possibly cancel- synaptic peripheral inhibition of genital response, ling each other out. although there does not appear to be any evidence directly relevant to that point. The integrating component of the system is presumably the l.c., The relationship between the central and with both its upwards and downwards projections. peripheral effects of the NE system It may be relevant that of the two sources of neuronal input to the l.c., one of them is from the nucleus paragigantocellularis (nPGi) which has We thus have a paradox; central effects of NE on direct, serotonergically mediated inhibitory in¯u- sexual response are excitatory and presynaptic ences on the erectile response.18 peripheral effects are probably inhibitory. The ®rst In conclusion, whereas it is clear that the central fundamental point in tackling this paradox is to NE system is of fundamental importance to sexual recognize that the excitatory central effects of the NE arousal and response, we remain largely ignorant of system are not con®ned to sexual arousal. The how it functions in that respect. upward projections of the l.c. are involved in general arousal, and in many circumstances it is entirely appropriate that elicitation of general central arousal should require localized peripheral inhibition of genital response. Thus if the NE system generates `arousal' in response to the threat of Pharmacological studies in humans punishment, an emotional state which in those circumstances Gray10 would call `anxiety', then inhibition of other response patterns, such as sexual A series of clinical studies of the effect of yohimbine response, which might interfere with an appropriate in men with erectile dysfunction have been carried avoidance response would be adaptive. However, out.19±22 Most of these studies have been affected by one of the principal mysteries of the central NE methodological problems of various kinds, but a system is how it is mobilized to serve sexual consistent though weak average bene®cial effect on function in some situations and avoidance behavior erectile function has been reported, presumably due in others.11 Testosterone may be relevant to this to central effects of the drug. In experimental selective process, because of the capacity for alpha-2 studies, yohimbine induces anxiety23 and combined antagonists to restore sexual behavior in castrated with naloxone, it produces both anxiety and penile male rats (see above), though the mechanism of this erection.24 Anxiety has not been a problem as a side hormonal effect is far from clear. Further complexity effect in the treatment studies cited above, and it is suggested in a study by Sala et al,12 showing a remains a question of some interest why anxiety has curvi-linear dose response relationship between an been evoked in such experimental studies, but not alpha-2 antagonist and central excitation, with in treatment studies. It may be a matter of acute middle range doses producing maximal excitatory dosage contrasted with chronic dosage; or it may be effects and high doses inhibitory effects. dependent on the context in which the drug is The relevance of the central NE system to sexual taken. Given its theoretical importance, further arousal is further indicated from studies of sexual research should be directed at this issue.

International Journal of Impotence Research Effects of alpha-2 blockade on sexual response J Bancroft S66 Experimental human studies with a speci®c alpha-2 Two types of experimental procedures were antagonist, Delequamine therefore used, both involving intravenous infusion of the drug in two contrasting dosages (`high' and `low' dose) compared with a third placebo condi- About 10 years ago a new alpha-2 antagonist, tion. The ®rst procedure assessed effects of the drug RS15385 (delequamine) was developed by Syntex during sleep on nocturnal penile tumescence;32 the Pharmaceuticals as a potential prosexual drug second on subjective, erectile and cardiovascular (Figure 1). This compound has more speci®c and responses to visual erotic stimuli and erotic fantasy selective alpha-2 antagonist action than yohim- in the waking state.33,34 In each study, 12 `normal bine.25 This provided us with the opportunity to controls' and 24 men with psychogenic erectile carry out some experimental studies of the effects of dysfunction were involved (there was a small over- a selective alpha-2 antagonist on human male sexual lap of subjects in the two studies). response. We based these studies on a theoretical model derived from the effects of testosterone on Effects on nocturnal penile tumescence human male sexuality, and the capacity for alpha-2 antagonists to restore sexual response in testoster- one de®cient animals (see above). We worked on the Our hypotheses were largely con®rmed.32 Erections assumption that the alpha-2 antagonist would have during sleep were increased by the drug in both the similar effects in eugonadal men as testosterone controls and the dysfunctional men, and in the replacement would have in hypogonadal men. controls, spontaneous erections occurred with the Previous research had shown that erections occur- high dose of the drug just before sleep onset. ring during sleep (nocturnal penile tumescence or However, the effects during sleep were complex. In NPT) were very sensitive to testosterone and were the normal controls, there was a curvilinear dose± markedly diminished in hypogonadal men and response relationship; the low dose of the drug restored with testosterone replacement.26±29 By increased erections but mainly during non-rapid eye contrast, erections in response to visual erotic movement (non-REM) sleep. The high dose, apart stimuli (VES) were relatively independent of testos- from its positive effects on spontaneous erections terone. In a series of studies, maximum erectile before sleep onset, reduced erections during REM response to VES was not signi®cantly different in sleep (an effect which was characterized by a hypogonadal and eugonadal men.27,29,30 A more fragmentation of the erectile response). recent study showed that duration and to some In analyzing the results for the dysfunctional extent rigidity of erectile response, variables which men, we divided them into two age groups around had not been measured in the earlier studies, were the median split of 47 y. Although men in the older sensitive to testosterone withdrawal and replace- group were, as expected, more sensitive to the sleep ment.31 This indicated some degree of testosterone disrupting effect of the drug (diminution of total dependence, but much less clear cut than in the case sleep time), they showed no effect of the drug on of NPT. erections at either dose level. The younger dysfunc- The hypotheses that we set out to test were as tional men, on the other hand, showed a linear dose- follows; that delequamine would: (1) Enhance response relationship; the low dose showed no spontaneous erections and spontaneous feelings of signi®cant effect; the high dose of the drug increased sexual arousal (i.e. NPT or spontaneous erections erections during non-REM sleep, mostly between when awake, not in response to an external sexual stage 2 and the ®rst episode of REM. stimulus); or (2) have no effect on maximum erectile There was therefore an interesting contrast between response to visual erotic stimuli, but would prolong the controls and the younger dysfunctional men. In erectile responses to such stimuli. the controls, the high dose of the drug only stimulated spontaneous erections during the waking state at the time when sleep onset was delayed. In the younger dysfunctional men, the positive effects of the high dose was observed shortly after sleep onset, as if in their case inhibitory tone which was high in the waking state was reduced by the onset of sleep.

Effects in the waking state

Once again our hypotheses were broadly con- ®rmed.33; 34 In this study the age groups of the dysfunctional men were formed around the median split of 45 y. In the controls, there were signi®cantly

Figure 1 Delaquamine: Alpha2 selective antagonist. higher subjective ratings of sexual arousal and an

International Journal of Impotence Research Effects of alpha-2 blockade on sexual response J Bancroft S67 increased likelihood of spontaneous erections occur- the low dose of the drug there was an increase in ring with the high dose of the drug before the erectile response, noticeable during both REM and presentation of visual erotic stimuli or use of erotic non-REM sleep but only signi®cant during non- fantasy. Response to VES, on the other hand, REM. With the high dose of the drug, we found a sig- showed no effect of the drug on penile circumfer- ni®cant reduction in erectile response, compared to ence, but in the controls and the younger dysfunc- the low dose during REM-sleep. In some way our tional men there was a signi®cant prolongation of lower dose was activating the NE excitatory system, erectile response to VES with the high dose of the producing effects which were most noticeable drug. during non-REM sleep, presumably because during There were further interesting ®ndings in relation REM the drug, particularly at the higher dose level, to the cardiovascular system. There was a blunting was also activating the inhibitory system, `turning of blood pressure (BP) and (HR) response on' the `switched off' l.c. neurons. to the erotic stimuli in the younger dysfunctional An interesting comparison is to be made with the men during the placebo condition, which was effects of trazadone on NPT reported by Saenz de partially normalized by the high dose of the drug. Tejada et al.37 Trazadone increased NPT in a dose The cardiovascular responses of the older dysfunc- related fashion, but this effect was largely con®ned tional subjects were not different from those of the to the period following REM related erections. In controls, whereas their erectile responses were other words, when the centrally derived peripheral impaired, suggesting some form of dissociation inhibition was `switched on' again following a REM between cardiovascular and erectile responses in related erection, the drug blocked the resulting older dysfunctional men. detumescence, prolonging the erection. This was not the pattern with delequamine which increased erection during non-REM sleep predominantly in Interpretations and questions raised by these the period between sleep onset and ®rst REM. Nor results was there any evidence of trazadone disrupting NPT during REM sleep, though there might conceivably The sleep study have been with a higher dose. It is also relevant that The evidence from the sleep study is consistent with delequamine had a marked effect in delaying sleep there being a double role for NE in the central onset which trazadone did not. These contrasting nervous system, both excitatory and inhibitory, effects are consistent with having a each of which could be elicited by this alpha-2 peripheral alpha-1 antagonist effect and delequa- antagonist, depending on the dose. One possible mine a central alpha-2 antagonist effect. Trazadone interpretation of the curvilinear dose±response has a complex pharmacological action which also relationship in the controls is the consequence of a includes serotonergic effects, and it remains a dose related shift from predominantly excitatory to possibility that these effects may have been respon- predominantly inhibitory action. However, the sible or contributed to the pattern observed. linear dose response relationship in the younger At this stage we can only go so far in explaining dysfunctional men was consistent with an increased these complex effects of delequamine during sleep. alpha-2 inhibitory tone in those cases. According to Are there l.c. cells which are responsible for the that interpretation, higher doses of the drug than downstream peripheral inhibition and others res- those used in this study would produce a curvi- ponsible for the upstream arousal effects? Is there a linear picture in the dysfunctional men also.35 If change of balance of control of these systems during there is any validity in such an interpretation it REM sleep which makes response to pharmacologi- raises the possibility of central `increased alpha-2 cal agents unpredictable at that time or less relevant tone' as being a causal factor in certain cases of to their effects at other times? psychogenic erectile dysfunction. But how does this shift from `excitation' to `inhibition' work? One widely accepted explanation During the waking state for this type of curvi-linear dose±response curve is Given that our controls and younger dysfunctional a shift from predominantly pre-synaptic to post- men showed an increase in the duration of erections synaptic effect. But it is dif®cult to see how that in response to VES with delequamine, it is reason- would explain these ®ndings. An important factor to able to conclude that, while both central and take into account is the physiological signi®cance of peripheral effects may have been involved, the REM sleep. This has been described as a state balance was in favor of the central alpha-2 `excita- involving virtual cessation of NE activity tory' effects, and, at least with the doses used, with within the l.c.36 This offers a partial explanation for no suggestion of a curvi-linear dose-response rela- why erections occur during REM sleep; i.e., there is tionship. Of particular interest are the blunted a `switching off' or at least reduction of the centrally cardiovascular responses in the younger dysfunc- derived peripheral inhibitory tone in the erectile tional men. We should consider the possibility that tissues. What we found in our controls was that with such blunting may be part of a more generalised

International Journal of Impotence Research Effects of alpha-2 blockade on sexual response J Bancroft S68 pattern consistent with general anxiety. Recent centrally, which would reduce or impair the normal evidence has shown that reduced responsiveness is alpha-2 modulation of NE release, resulting in a characteristic of general anxiety disorder, and is at increased NE tone in the smooth muscles? least in part due to loss of vagal control.38 However, the partial normalization of the cardiovascular responsiveness by the alpha-2 antagonist in our Conclusions study suggests that in these men with erectile dysfunction it was not solely a matter of vagal control. This does raise some interesting research Given the complex ®ndings with delequamine in possibilities, however. Younger men with psycho- these acute dosage studies, this paper has raised genic dysfunction should be compared with men more questions than it has provided answers. with general anxiety disorder in their patterns of However, most of the questions raised are research- cardiovascular responsiveness to a variety of stimuli, able, and are of potential relevance to de®ning the including sexual. At the same time, the distinction characteristics of the man with erectile dysfunction between `increased inhibitory tone', whether alpha-2 who is likely to bene®t from this type of drug. The or otherwise, and an acute inhibitory response to tendency in the pharmaceutical industry when some threatening aspect of the sexual situation is of evaluating drugs of this kind is to look for a broad fundamental theoretical importance. Are we dealing, therapeutic impact in large scale phase 3 studies. at least in some cases, with a high basal inhibitory More consideration should be given to formulating tone, which is part of a more general personality theoretical models which may enable us to predict trait, which makes the individual vulnerable to likely responders, who may be relatively speci®c erectile dysfunction, or is it a matter of the degree sub-groups, and to structure treatment outcome of inhibitory response to some perceived external studies accordingly. sexual threat that makes him vulnerable? In the former case pharmacological treatment may hold particular appeal. In the latter the analysis of sexual References threats may lend itself to exploitation by psycholo- gical methods of treatment. The use of sleep studies, 1 Crenshaw TL, Goldberg JP. Sexual pharmacology: Drugs that when it is assumed that the individual is not react- affect sexual function. New York: W.W.Norton, 1996. ing to speci®c threats in his environment, provides 2 Role LW, Kelly JP. The brain stem: Cranial nerve nuclei and one model for studying `inhibitory tone'. But the the systems. In: Kandel ER, Schwartz JH, Jessell TM, eds, Principles of Neural Science, 3rd ed. Appleton investigation of cardiovascular (and electrodermal) & Lange, Norwalk. 1991, pp. 683±699. responsiveness to a variety of external stimuli, 3 Stahl SM. Essential Psychopharmacology. Boston: Cambridge including aversive, sexual and non-sexual positive, University Press, 1998, p159. in younger men with psychogenic erectile dysfunc- 4 Clark JT, Smith ER, Davidson JM. Testosterone is not required for the enhancement of sexual motivation by yohimbine. tion, may throw additional light on this issue. Physiol Behav 1985; 33: 517±521. 5 Saenz de Tejada IS, et al. Regulation of adrenergic activity in penile corpus cavernosum. J Urol 1989; 142: 1117±1121. The age effect 6 Anderson K-E. Pharmacology of lower urinary tract smooth muscles and penile erectile tissues. Pharmacol Rev 1993; The impact of age on the effects of delequamine, in 45(3): 253±308. both the sleep and waking studies, is also of 7 Traish AM, Moreland RB, Huang Y-H, Goldstein I. Expression theoretical interest. It was not expected and as a of functional alpha 2- subtypes in human consequence we did not have aged matched controls corpus cavernosum and in cultured trabecular smooth muscle for our dysfunctional subjects. The controls were cells. Recept Sign Transduct 1997; 7: 55±67. 8 Traish AM, Kim N, Moreland R, Goldstein I. Role of alpha slightly younger than the younger dysfunctional adrenergic receptor blockade in erectile dysfunction. Int J Imp men. It would be interesting to assess the effects of Res 1999 (in press). this drug in a group of older functional men. The 9 Brindley GS. Pilot experiments on the actions of drugs results, nevertheless, suggest that there is a decline injected into the human corpus cavernosum penis. Br J in responsiveness to central NE stimulation, at least Pharmacol 1986; 87: 495. 10 Gray JA. The psychology of and stress. Cambridge: that mediated by alpha-2 receptors, with increasing Cambridge University Press, 1987. age (although no lack of interference with sleep in 11 Bancroft J. Central inhibition of sexual response in the male: the older men). Lerner et al39, on the basis of in vitro A theoretical perspective. Neurosci Biobehav Rev 1999; 23: studies, concluded that there is an increase in 763±784. 12 Sala M, et al. Central effect of yohimbine on sexual behavior peripheral norepinephrine inhibitory tone in the in the rat. Physiol Behav 1990; 47: 165±173. erectile tissues in older men, contributing to the age 13 Beach FA, Jordan L. Sexual exhaustion and recovery in the related increase in vulnerability to erectile dysfunc- male rat. Q J Exp Psychol 1956; 8: 121±133. tion. Can their ®ndings be reconciled with our age- 14 Larsson K. Conditioning and sexual behaviour in the male related loss of response to alpha-2 antagonists? Is it albino rat. Stockholm: Almquist & Wiksell, 1956. 15 Rodriguez-Manzo G, Fernandez-Guasti A. Reversal of sexual possible, for example, that there is a loss of alpha-2 exhaustion by serotonergic and noradrenergic agents. Behav receptor responsiveness in the periphery as well as Brain Res 1994; 62: 127±134.

International Journal of Impotence Research Effects of alpha-2 blockade on sexual response J Bancroft S69 16 Rodriguez-Manzo G, Fernandez-Guasti A. Participation of the 30 Bancroft J, Wu FCW. Changes in erectile responsiveness central noradrenergic system in the reestablishment of during therapy. Arch Sex Behav 1983; 12: 59±66. copulatory behavior of sexually exhausted rats by yohimbine, 31 Carani C, Granata ARM, Bancroft J, Marrama, P. The effects of naloxone, and 8-OH-DPAT. Brain Res Bull 1995; 38: 399±404. testosterone replacement on nocturnal penile tumescence and 17 Rodriguez-Manzo G, Fernandez-Guasti A. Opioid antagonists rigidity and erectile response to visual erotic stimuli in and the sexual satiation phenomenon. Psychopharmacology hypogonadal men. Psychoneuroendocrinology 1995; 20: 743± 1995; 122: 131±136. 753. 18 McKenna KE. Central control of penile erection. Int J Imp Res 32 Bancroft J, Munoz M, Beard M, Shapiro C. The effects of a new 1998; 10(Suppl 1): S25±S34. alpha-2 adrenoceptor antagonist on sleep and nocturnal penile 19 Morales A, et al. Is yohimbine effective in the treatment of tumescence in normal male volunteers and men with erectile organic impotence? Results of a controlled trial. J Urol 1987; dysfunction. Psychosom Med 1995; 57: 345±356. 137: 1168±1172. 33 Munoz M, Bancroft J, Beard M. Evaluating the effects of an 20 Reid K, et al. Double-blind trial of yohimbine in treatment of alpha-2 adrenoceptor antagonist on erectile function in the psychogenic impotence. Lancet 1987; ii: 421±423. human male. II. The erectile response to erotic stimuli in men 21 Riley AJ, Goodman RE, Kellet JM, Orr R. Double blind trial of with erectile dysfunction, in relation to age and in comparison yohimbine hydrochloride in the treatment of erection inade- with normal volunteers. Psychopharmacology 1994; 115: 471± quacy. Sex Marit Ther 1989; 4: 17±26. 477. 22 Riley AJ. Yohimbine in the treatment of erectile disorder. Br J 34 Munoz M, Bancroft J, Turner M. Evaluating the effects Clin Pract 1994; 48: 133±136. of an alpha-2 adrenoceptor antagonist on erectile function 23 Charney DS, Heninger GR, Redmond DE. Yohimbine induces in the human male. I. The erectile response to erotic anxiety and increases noradrenergic function in humans: stimuli in volunteers. Psychopharmacology 1994; 115: 463± Effect of diazepam and . Life Sci. 1983; 33: 19±29. 470. 24 Charney DS, Heninger GR. Alpha2-adrenergic and opiate 35 Bancroft J. Are the effects of on male sexuality receptor blockade. Synergistic effects on anxiety in healthy noradrenergically mediated? Some consideration of the hu- subjects. Arch Gen Psychiat 1986; 43: 1037±1041. man. Neurosci Biobehav Rev 1995; 19: 325±330. 25 Brown CM, et al. The pharmacology of RS-15385-197, a potent 36 Parmeggiana PL, Morrison AR. Alterations in autonomic and selective alpha 2 adrenoceptor antagonist. Br J Pharmacol functions during sleep. In: Loewy AD, Spyer KM, eds, Central 1993; 108: 516±525. regulation of autonomic functions. New York: Oxford Uni- 26 Cunningham GR, et al. The relationship between serum versity Press, 1990, 367±386. testosterone and prolactin levels on nocturnal penile tumes- 37 Saenz de Tejada I, et al. Pathophysiology of prolonged penile cence in impotent men. J Androl 1982; 3: 241±247. erection associated with trazodone use. J Urol 1991; 145: 60± 27 Kwan M, et al. The nature of androgen action on male 64. sexuality; a combined laboratory and self-report study in 38 Thayer JF, Friedman BH, Borkovec TD. Autonomic character- hypogonadal men. J Clin Endocrinol Metab 1983; 57: 557±562. istics of generalized anxiety disorder and worry. Biol Psychiat 28 O'Carroll RE, Shapiro C, Bancroft J. Androgens, behaviour and 1996; 39: 255±266. nocturnal erections in hypogonadal men: the effect of varying 39 Lerner SE, Melman A, Christ GJ. A review of erectile the replacement dose. Clin Endocrinol 1985; 23: 527±538. dysfunction: New insights and more questions. J Urol 1993; 29 Carani C, et al. Testosterone and erectile function, nocturnal 149: 1246±1255. penile tumescence and rigidity, and erectile response to visual erotic stimuli in hypogonadal and eugonadal men. Psycho- neuroendocrinology 1992; 17: 647±654.

International Journal of Impotence Research