Content uploaded by Erick Janssen
Author content
All content in this area was uploaded by Erick Janssen on Oct 16, 2017
Content may be subject to copyright.
The dual control model of male sexual response: a theoretical approach to
centrally mediated erectile dysfunction
John Bancroft*, Erick Janssen
The Kinsey Institute for Research in Sex, Gender and Reproduction, Morrison Hall 313, Indiana University, 1165 East Third Street, Bloomington,
IN 47405-3700, USA
Abstract
A theoretical model of dual control of male sexual response is considered, based on the balancing of central excitation and inhibition, with
individuals varying in their propensity for both sexual excitation and inhibition of sexual response. A questionnaire method for measuring
propensities for sexual excitation and inhibition has been developed (SIS/SES questionnaire), resulting in one excitation factor (SES) and two
inhibition factors (SIS1 and SIS2). Evidence for the existence of both inhibitory and excitatory tone is discussed. The first inhibition factor
(SIS1) may be related to level of inhibitory tone and is associated with fear of performance failure. The second inhibition factor (SIS2) may
be related to external threats (e.g. from within the sexual relationship). The implications for the treatment of centrally mediated erectile
dysfunction are discussed, with predictions that high SIS2 individuals will respond to psychological treatment, whereas high SIS1 individuals
will respond better to pharmacological methods of treatment. q2000 Elsevier Science Ltd. All rights reserved.
Keywords: Sexual response; Erectile dysfunction; Inhibition; Theoretical model; Treatment implications; Prognosis
1. Introduction
This paper considers a new theoretical model of dual
control of male sexual response, involving both excitatory
and inhibitory mechanisms in the brain and explores how
the model might help to throw light on the etiology of
psychogenic erectile dysfunction. The basic theoretical
model, together with a detailed review of the evidence
pertaining to central inhibitory mechanisms relevant to
sexual response, has been published elsewhere [4]. This
model postulates that central inhibitory mechanisms are
adaptive and that individuals vary in their propensity for
inhibition of sexual response as well as propensity for
sexual excitation. For the majority, the presence of a fairly
typical level of inhibition proneness is adaptive, helping to
keep the individual out of trouble. Those whose propensity
for central inhibition of sexual response is too high have
increased vulnerability to sexual (e.g. erectile) dysfunction.
For those whose inhibitory propensity is too low, an
increased likelihood of engaging in high risk sexual
behavior may result.
It is also assumed that the adaptiveness of central
inhibition of sexual response would be of general signifi-
cance across species, and in the earlier paper consideration
was given to the various functions which inhibition of
sexual response might serve [4]. Four principal functions
were proposed:
1. Where a sexual situation is perceived as threatening and
inhibition of sexual response facilitates avoidance of that
threat;
2. Where the perceived threat is not sexual but requires an
avoidance response for which inhibition of other inap-
propriate distracting behavioral patterns such as eating or
sexual activity is a necessary part;
3. Where sexual response, (e.g., ejaculation in the male)
inhibits further sexual arousability to ensure that the
pursuit of sexual pleasure does not become excessive
(resulting in either reduced fertility or maladaptive
preoccupation with sexual rewards);
4. Where sexual and reproductive behavior is inhibited by
chronic stress (which, in many social species, has the
effect of reducing population overcrowding by selectively
inhibiting reproductive behavior in those animals which,
as a consequence of being low in the social hierarchy, are
more susceptible to chronic stress).
It was concluded that whereas a variety of specific neuro-
physiological mechanisms have been identified, it is difficult
Neuroscience and Biobehavioral Reviews 24 (2000) 571–579PERGAMON
NEUROSCIENCE AND
BIOBEHAVIORAL
REVIEWS
0149-7634/00/$ - see front matter q2000 Elsevier Science Ltd. All rights reserved.
PII: S0149-7634(00)00024-5
www.elsevier.com/locate/neubiorev
* Corresponding author. Tel.: 11-812-855-7686; fax: 11-812-855-
8277.
E-mail address: jbancrof@indiana.edu (J. Bancroft).
to link specific mechanisms to specific behavioral patterns
of inhibited sexual response/behavior. The evidence
suggests a more complex interactive system. However,
this complex system can be considered at two levels: (i)
the central arousal system and how it is recruited to produce
arousal of a specifically sexual kind (or inhibition of it), and
(ii) the inhibitory pathways from the brain stem (e.g. from
the nucleus paragigantocellularis and the locus coeruleus)
and peripheral genital manifestations of these ‘downstream’
inhibitory mechanisms.
Whereas the central arousal system is fundamentally
based on norepinephric mediation, its recruitment for sexual
purposes probably depends on disinhibition of dopamin-
ergic mechanisms and, in some as yet unidentified way,
on testosterone-dependent mechanisms. Inhibition of
central sexual arousal, on the other hand, probably involves
neuropeptidergic as well as serotonergic mechanisms. The
peripheral inhibitory pathways clearly involve serotonin and
norepinephrine, and quite possibly other mechanisms in
addition (see Ref. [4] for review).
When we consider the four postulated functions and
consider their relevance to humans, a distinction can be
made between the first two, both of which can be regarded
as ‘context’ or ‘stimulus’ specific, and the fourth which
would appear to involve a chronic inhibitory ‘state’ rela-
tively independent of specific context or stimulus at the
time. In this paper, we consider some of our early evidence
from measuring inhibition proneness in humans which is
consistent with this distinction. The third function, the
concept of post-ejaculatory refractoriness, is clearly rele-
vant to humans as well as other species, although as yet
we have very little evidence of the mechanisms underlying
this phenomenon in humans. But studies of ‘sexual exhaus-
tion’ in the rat, a state which appears to be an experimental
extension of the refractory period, suggest that a wide range
of inhibitory mechanisms, including inhibition of central
sexual arousal, are involved (see Ref. [4] for review).
Measuring propensities for excitation and inhibition.In
order to explore our theoretical model, a questionnaire has
been developed which has good psychometric properties
[23]. An important feature of this questionnaire is that all
questions are aimed at establishing typical response
patterns, in terms of sexual arousal or genital response, to
two types of situation—one involving non-threatening
sexual situations and the other sexual situations which
involve some form of threat. This questionnaire was first
completed by 408 male students, together with a number
of other questionnaires used for establishing discriminant
and convergent validity. Factor analysis resulted in three
higher order factors, one related to excitation (SES), and
two to inhibition (SIS1 and 2). This factor structure was
further established in a confirmatory factor analysis of the
questionnaire responses from a second sample of 459 male
students. Based on the questions loading on the two inhibi-
tion factors, we called SIS1 “inhibition due to threat of
performance failure” (e.g. losing one’s arousal easily,
concerns about pleasing one’s partner and the effects of
external sources of distraction), and SIS2, “inhibition due
to threat of performance consequences” (e.g., threat of
unwanted pregnancy or sexually transmitted infection, risk
of ‘being caught’ or concerns about causing or experiencing
physical pain). As both of these first two samples involved
predominantly young men (mean ages 22.8 and 20.9 years,
respectively) a third sample of 313 men with a mean age of
46.2 years also completed the 45-item questionnaire.
The correlations among the three scales were similar in
each sample (see Table 1) with no correlation observed
between either inhibition scale and the excitation scale,
and a modest correlation between the two inhibition scales.
In the two younger samples there were no significant corre-
lations between scale scores and age. In the older sample
age did relate significantly to SES (r20.24) and SIS1
(r10.34) scores but not SIS2 (r10.13).
A psychophysiological experiment was subsequently
used to validate the scales [23]. Grouping subjects into
high or low SES (with the groups being similar on the two
SIS scales), high and low SIS1 and high and low SIS2,
genital and affective responsesto two types of erotic stimuli
(‘threatening’, i.e. coercive, and ‘non-threatening’, i.e.
consensual) were recorded. We found that the high SES
group showed generally higher response levels, both genital
and affective. The two SIS2 groups did not differ from each
other in response to non-threatening erotic stimuli, but the
high SIS2 group showed significantly less genital response
to the threatening erotic stimuli, although the two groups did
not differ in degree of negative affect associated with the
threatening stimuli (i.e. the stimuli were no less threatening
in the group which showed the greater genital response).
Using distraction and performance demand conditions, we
failed to discriminate between our high and low SIS1
groups. Thus we found convincing psychophysiological
validation of our SES and SIS2 scales, but we had not
found the right experimental manipulation to adequately
validate our SIS1 scale.
We can also look at our preliminary data relating SIS/SES
scores to frequency of sexual activity and also erectile
dysfunction [23]. In the first sample we found a weak nega-
tive relationship between SIS2 and frequency of sexual
intercourse (
b
2.18; p,0.002) and between SIS1 and
frequency of any form of sexual interaction with the partner
(
b
2.12; p,0.05) and a much more marked positive rela-
tionship between SES and frequency of masturbation
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579572
Table 1
Correlations between sexual excitation (SES) and sexual inhibition. SIS1
and 2 scales (Sample 1). (n408; mean age 22.8 years, range 20–
38 years)
SIS1 SIS2
SES 20.07 20.11
SIS1 10.26
(p,0.001)
(
b
1.43; p,0.0001). A similar pattern was found in the
older sample, and in this group there was also sufficient
reporting of erectile dysfunction to allow us to examine
this variable. Subjects were asked first whether they had
ever had difficulties in obtaining or keeping an erection in
sexual activities with their partner. Nearly half said they had
never ever had difficulty, with 43% saying ‘occasionally’,
4% ‘less than half the time’ and 4% ‘most of the time’. They
were also asked whether they had had such difficulties in the
past 3 months; 75% reported ‘never’, 18% ‘occasionally’
2% ‘less than half the time’ and 5% ‘most of the time’.
Thus, it was the difference between ‘never’ and ‘occasion-
ally’ which distinguished these two assessments most mark-
edly. Taking these frequency assessments as dependent
variables, multiple regression analysis was carried out
using our excitation score, two inhibition scores and age
as independent variables. In the analysis of the ‘ever had
difficulty’ variable, SIS1(
b
10.38; p,0.0001) SIS2
(
b
10.13; p,0.02) and age (
b
10.15; p,0.006)
were all significantly related. SES did not figure in the
equation. In the analysis of ‘past 3 months’, SIS1 again
featured strongly and positively, as did age (
b
10.36
and 10.25, respectively; p,0.0001). This time SES was
involved, though negatively and weakly (
b
20.12;
p,0.02), whereas SIS2 did not feature. Thus SIS1 was
clearly related to both dysfunction frequency measures;
this is consistent with SIS1 reflecting a trait inhibition
characteristic which, for each subject, is and has been for
some time, relevant to vulnerability to erectile failure. The
association with age is consistent with the well-established
fact that erectile function is less efficient as men get older.
Given that age is accounted for in the equation, the associa-
tion between SES and the latest time period score is presum-
ably not simply an age-related loss of sexual arousability
and requires further explanation. The relevance of SIS2 to
the first variable but not the second is interesting. This is
consistent with erectile failure, on occasions (hence, contri-
buting mainly to the ‘occasional’ category) being a response
to a threatening situation which may have occurred from
time to time through the individuals life, but not on a regular
basis and not necessarily within the last 3 months.
Thus we have established a valid measure of excitation/
inhibition to explore our theoretical model with some inter-
esting preliminary results.
2. Are there two types of inhibition?
The main unexpected finding so far is the identification of
two inhibition scales rather than one. This had not been
anticipated in our original theoretical model, and requires
explanation. The principal purpose of this paper is to extend
the theoretical model to take this finding into account, and to
specifically consider the relevance of the three scales to
erectile dysfunction.
We can compare and contrast the threats typically
associated with ‘fear of performance failure’ (SIS1) and
‘fear of performance consequences’ (SIS2). The second
appears related to the perception of an external threat in
the specific situation, with variations in the SIS2 score
reflecting varying degrees of inhibitory responsiveness to
such threats. Clearly, such threats (e.g. fear of hostile
response from the partner) are relevant in a fair proportion
of cases of psychogenic ED, and may end up as important
foci for psychological interventions. The first scale (SIS1)
presents a more complex challenge. Here the threat is more
intrinsic, based on the learnt awareness that failure of
response, for that individual rather than that situation, is to
be expected. This can therefore tentatively be conceptua-
lized as the cognitive structuring of a trait reflecting ‘high
propensity for erectile failure’. In other words, the high SIS1
individual has learnt that he responds unreliably in sexual
situations and the anticipation of further failure augments
that trait characteristic. It is noteworthy how little attention
has been paid to ‘performance anxiety’, the hallmark of any
putative ‘fear of performance failure’. Clinical experience
suggests that often such individuals are aware of an unplea-
sant emotional state; they may not spontaneously describe it
as ‘fear of failure’ though they may readily agree to that
description if it is suggested to them.
We should remain open to the possibility that these two
scales are measuring two different types of inhibitory
mechanism. We would also like to propose an alternative
explanation; that whereas the SIS2 scale measures inhibi-
tory responsiveness to external threats, SIS1 is reflecting the
level of ‘inhibitory tone’ and that high SIS1 individuals
have high ‘inhibitory tone’. ‘Tone’ in this sense refers to
the level of inhibition that the system is set at when not
actively responding to a sexual stimulus or an external
sexual threat.
The heuristic value of this ‘inhibitory tone’ concept needs
to be considered carefully. There are various reasons for
believing that inhibition of sexual response is a “tonic”
state which requires to be reduced, or at least exceeded by
excitation, to allow sexual response to occur. This is clearly
evident in the tonic contraction of the smooth muscle of the
flaccid penis; the dependence of this state on ‘tonic’ inhibi-
tory signals from the brain is apparent from the effect of
spinal transection which, in both animals [19,29] and man
(see Ref. [21] for review) results in a lowering of the thresh-
old for ‘reflexive’ erections. More centrally, the sexual acti-
vation role of the medial preoptic area of the hypothalamus
may well depend on reduction of inhibitory tone [22]. For
any given level of baseline inhibitory ‘tone’ we would
expect that level to be increased in the presence of a
perceived external threat. Whether we would expect this
baseline level of inhibition to be lowered, in circumstances
where active appraisal of the situation concludes that there
is no threat, is less certain. It is possible that in such circum-
stances the inhibitory ‘tone’ remains unchanged and the
occurrence of a sexual response will depend on the activat-
ing effects of a sexual stimulus, with the propensity for
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579 573
sexual excitation being sufficient to overcome the inhibitory
‘tone’. But it is also possible that, typically, the level of
inhibitory tone is set to make sexual response unlikely
until active appraisal of low threat has been carried out,
when the lowering of the inhibitory tone may contribute
to the sexual response. Thus, in the typical case (i.e. the
individual with a typical level of inhibitory tone) appraisal
of the situation as non-threatening, with consequent reduc-
tion of inhibitory tone, at least to the level necessary for
sexual interaction, would be a prerequisite for sexual
response to occur. In the individual with low inhibitory
tone, there would be proportionally less need for reduction
of inhibition, and consequently a lower threshold for sexual
response. If such an individual’s propensity for sexual
excitation was also high, then we might expect occurrence
of a sexual response even in the presence of a threat. In the
converse situation, with the individual with a relatively high
inhibitory tone, particularly if his propensity for sexual
excitation is relatively low, we can predict that a proportion-
ally stronger sexual stimulus, and a lesser amount of threat
will be required before a sexual response will occur. We are
therefore conceptualizing inhibition in two ways, the level
of basal inhibitory tone characteristic of an individual, and
the degree of responsiveness to the perception of sexually
relevant threat (or its absence) with an increase (or decrease)
in the level of inhibition. With our two scales SIS1 and 2, we
are also conceptualizing two types of threat, one external
(SIS2), the other more intrinsic (i.e. the threat of perfor-
mance failure related to the tendency to fail due to the
high inhibitory tone).
Does ‘excitatory tone’ have similar heuristic value, or is
excitation ‘stimulus bound’, dependent on the existence of a
sexual stimulus and the prevailing potential for an excita-
tory response? The capacity for sexual excitation, or arou-
sability, clearly has intrinsic determinants. The best
illustration of this in the human context is the role of testos-
terone. In hypogonadal men, arousability is impaired.
Although we should not overlook the possibility that testos-
terone has a role in maintaining genital response mechan-
isms at the spinal level, it is clear that testosterone has a
crucial role centrally in relation to sexual arousal. The most
robust, predictable testosterone-related impairment of arou-
sability is manifested in the reduced spontaneous erections
during REM sleep (nocturnal penile tumescence or NPT). In
hypogonadal men, NPT occurs with much the same
frequency as in eugonadal men, but with a markedly
reduced degree and duration of response [12,13,15,25,33].
In comparison, erectile response to visual erotic stimuli
(VES) discriminates much less between hypogonadal and
eugonadal states [8,12,26]. At one stage, we had interpreted
this finding as indicating that there were at least two
response systems in the CNS, one of them testosterone
dependent, the other testosterone independent. But in view
of the most recent evidence [13] indicating that whereas
erectile responses to VES do not differ between hypogona-
dal and eugonadal states in terms of maximum erectile
response, they do differ in terms of duration (and probably
rigidity), we should consider other explanations for this
discrepancy (e.g. that the capacity for a sexual response to
continue even beyond the duration of the erotic stimulus
may depend on the testosterone dependent level of arousa-
bility, whereas the response to the initial maximum impact
of the stimulus may not).
The hypogonadal impairment of NPT is unlikely to be
caused by increased inhibitory tone alone since during REM
sleep, the NE neurons in the locus coeruleus, the down-
stream projection of which are responsible, at least in part,
for the peripheral inhibitory tone in the erectile tissues, are
effectively switched off [34]; hence the assumption that
erections occur during REM sleep because of the reduction
of inhibitory tone. But clearly the occurrence of only partial
erections during REM in the hypogonadal state indicates
that normal NPT cannot be explained solely in terms of
REM related disinhibition. On the other hand, there is no
clear reason to assume that relevant sexual stimulation is
involved in NPT. This leaves us with the possibility that
some central state of ‘excitatory tone’, dependent on testos-
terone, does exist, and when peripheral inhibition is
sufficiently reduced, as during REM, then spontaneous
erections will occur. The lesser discrimination between
hypogonadal and eugonadal states in response to VES
may be caused by the fact that VES is a relatively powerful
external stimulus. The comparison of hypogonadal and
eugonadal states in terms of erectile response to internal
erotic imagery (i.e. fantasy) is less consistent than is the
case with VES [9,13], suggesting a more variable depen-
dence on androgen (or a more variable potency of such
imagery as sexual stimuli). Auditory erotic stimuli, or
non-moving erotic visual stimuli (i.e. relatively weaker
sexual stimuli) have not been tested in this paradigm. This
nevertheless reminds us of the fundamental role of informa-
tion processing in the excitatory component—the process
by which sexual meanings are associated with external
stimuli.
Individual variability in propensities for excitation and
inhibition. Given that the capacity for sexual excitation
varies with testosterone level, at least below a certain
threshold level of the hormone, it is not unreasonable to
assume that there will be other sources of individual varia-
bility in sexual excitation or arousability. Similarly, we can
postulate that the propensity for inhibition as well as the
levels of inhibitory tone show individual variability.
Whether in either case such variability derives from earlier
learning, genetic determinants or a combination of the two
must remain unanswered for the time being. However, from
a clinical perspective we should consider what pathological
or otherwise atypical conditions might alter these states.
Using testosterone deficiency as a paradigm, we can
consider pharmacological effects (e.g. dopamine antago-
nists, such as butyrophenones, that impair ‘excitatory’
dopaminergic activity; e.g. Ref. [38], and certain types
of brain lesion as having the potential for lowering the
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579574
propensity for excitation [27]. Such states are likely to be
experienced principally as a loss of sexual interest (or
arousability) rather than as a propensity for erectile dysfunc-
tion. It is more difficult to find convincing examples of
pathologically induced increase in central inhibition, and
the equivalent pharmacological effects of centrally acting
drugs (e.g. in increasing inhibitory serotonergic activity)
are more clearly manifested as an inhibition of orgasm
or ejaculation than as an inhibition of sexual interest or
arousability.
Depression, however, requires special consideration. The
most usual, but by no means invariable, sexual pattern
associated with depression is loss of sexual interest and/or
erectile dysfunction [1,9]. NPT is also commonly impaired
in depressive illness, suggesting a biochemical impairment
of ‘excitatory tone’ [35,39]. On the other hand, depression
is commonly associated with overactivity of the
hypothalamo–pituitary–adrenal axis [16], a state sharing
some features in common with the ‘general adaptation
syndrome’ response to chronic stress. As mentioned earlier,
chronic stress is commonly associated with suppression of
reproduction and sexual behavior in non-human species, an
effect which may involve alteration of serotonergic activity
(e.g. Refs. [10,28]), and/or an increase in neuropeptidergic
activity, especially with
b
-endorphin [20]. The increase in
corticotrophin releasing factor (CRF) which accompanies
many depressive states, probably accounts for increased
levels of
b
-endorphin, which have been shown to be raised
in depressive illness [17]. The
b
-endorphin could be con-
tributing to an increase in inhibitory tone. On the other hand,
loss of sexual interest or responsiveness is not an invariable
feature of depressed states and some individuals even
experience an increase in sexual interest when they are in
a negative mood state. This may be more likely when the
negative mood is associated with increased arousal, as with
anxiety (e.g. Ref. [32]). In such cases a low propensity for
inhibition of sexual response may in some way allow
‘excitation transfer’ of the arousal because of the negative
mood state to augment arousal responses to sexual stimuli.
This pattern, which we have recently started to study, is of
considerable potential interest in explaining patterns of
compulsive sexual behavior and deserves closer attention
in a separate paper.
3. Other clinical evidence relevant to the concept of
‘inhibitory tone’
Three experimental paradigms are considered. The first
exploits the smooth muscle relaxant effects of intracaverno-
sal injections (ICI) of compounds such as prostaglandin E1
or papaverine. Whereas response to such injections was
initially regarded as a peripheral target organ effect inde-
pendent of higher influences, it soon became apparent that a
substantial proportion of men with psychogenic erectile
dysfunction, whose peripheral erectile mechanisms should
be relatively intact, responded poorly to such injections
[6,11], suggesting that there was some central inhibitory
signal counteracting the effect of the injection. Kim and
Oh [24] found that in men with psychogenic ED who
showed that type of poor response to ICI, there was an
increased level of NE in the penile blood, consistent with
a high inhibitory NE tone in the erectile tissues. Granata et
al. [18] showed that NE levels in the general circulation
were significantly lower in such ‘high inhibition’ men;
they also showed higher trait and state anxiety scores than
the ‘low inhibition’ group, although state anxiety was not
increased by the ICI. Combining the results from these two
studies, we have evidence that in men with psychogenic ED
who respond poorly to intracavernosal smooth muscle relax-
ants, there is increased NE tone in the erectile tissues and
low circulating NE combined with high state anxiety scores
consistent with central inhibition of sexual arousal and
peripheral inhibition of erection.
In the second experimental paradigm, the effects of acute
dosage with delequamine, a selective alpha-2 adrenoceptor
antagonist, were evaluated [30,31]. An alpha-2 antagonist
principally acts on presynaptic norepinephric (NE) autore-
ceptors, blocking their re-uptake function. Hence this type
of drug can be seen to increase available NE at the synapse
in the central nervous system, and possibly in the periphery
also. Centrally, this NE effect is probably associated with
arousal in general; the locus coeruleus, which is involved in
central arousal states, has many alpha-2 receptors. Periph-
erally, the effects of alpha-2 antagonists are not yet clear
(see Ref. [5] for fuller discussion), but may involve both
pre-synaptic inhibitory effects and post-synaptic excitatory
effects which cancel each other out. The central effects may
be linked to sexual arousal by a testosterone-dependent
mechanism as yet not understood; alpha-2 antagonists
partially restore the loss of sexual arousability that follows
castration in male animals [14]. It should be emphasised that
there is much about the central NE mechanisms and their
association with arousal that we do not understand. But it
does appear that, with delequamine, we have a drug where
the central effects could contribute to sexual arousal.
The effects of the delequamine were evaluated in compar-
ison with a placebo in men with psychogenic erectile
dysfunction. Younger dysfunctional men in this study
showed both impairment of erectile response and blunting
of cardiovascular response to visual erotic stimuli during the
placebo condition; differences from the functional controls
which were substantially reduced by the drug. Thus, by
antagonizing the central alpha-2 activity and increasing
available NE centrally, delequamine increased peripheral
non-genital (i.e. cardiovascular) manifestations of sexual
arousal as well as increasing erectile response. These results
are to some extent consistent with those from the ICI study
described above.
The third experimental paradigm involved acute dosage
with delequamine during sleep [8]. Here we find some
interesting effects which are clearly suggestive of increased
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579 575
inhibitory tone based on alpha-2 receptor activity. In the
normal controls, the high dose of the drug only increased
erections in the waking state, in fact just before the onset of
sleep. In the younger psychogenic dysfunctional cases, the
high dose of the drug increased erections after the onset of
sleep, in fact mainly between onset of stage 2 and the first
REM episode. One interpretation was that with the onset of
sleep there was a reduction of inhibition in the dysfunctional
men that allowed the high dose of the drug to have an effect.
But even during sleep, the dysfunctional men showed a
different dose–response relationship to the normal controls
consistent with their having higher inhibitory levels of
alpha-2 activity [3,5].
To what extent do the results of these three experimental
paradigms support the concept of ‘increased inhibitory tone’
rather than an inhibitory response to an external-threatening
stimulus? In the first case, the ICI could be regarded as an
external threat, and the impairment of the response to ICI as
an inhibitory response to that external threat. However,
psychological as well as neuroendocrine indicators of a
‘response to threat’ were not clearly present and the
evidence was more consistent with a ‘high inhibitory
tone’ model. In the second paradigm, external sexual stimuli
were clearly involved, and these may have been associated
with some threatening meaning, most probably related to
‘fear of failure to respond’, although once again the pattern
of psychophysiological responses was also consistent with a
‘high inhibitory tone’ model. In the third paradigm, invol-
ving genital responses during sleep, the relevance of an
external inhibition-provoking threat can be ruled out, and
the results are most convincingly supportive of the ‘high
inhibitory tone’ model.
The age factor. A striking finding in the two delequamine
studies [7,30,31] was a virtual absence of effects of the drug
in the older dysfunctional men, both in the waking study and
the sleep study. Their erections, which were impaired in
response to the visual erotic stimuli, were unaffected by
the drug. There appeared to be a loss of responsiveness
to the alpha-2 blockade. Interestingly, cardiovascular
responses to the erotic stimuli, which were blunted in the
younger dysfunctional men, were no different in the older
dysfunctional men to those in the young functional controls;
there was some form of dissociation between cardiovascular
and erectile response to erotic stimuli in this older dysfunc-
tional group. How can we account for this age effect?
Lerner et al. [26] have reported evidence of increased
inhibitory tone in penile smooth muscle in older men. If,
as seems likely, peripheral alpha-2 receptors modulate NE
inhibitory tone in the penile smooth muscle, then one expla-
nation for this increase of muscle tone could be an age-
related loss of alpha-2 receptor responsiveness (i.e. loss of
response to both agonist and antagonist), which in the
periphery could result in increased inhibition of erectile
response.
Obviously, a variety of mechanisms could be involved
in the general age-related decline in sexual interest and
arousability (see Ref. [36] for review). However, the
decreases in both free testosterone and NPT (frequency,
degree and duration) that typically accompany aging in
men [37], are of potential relevance to our model, consistent
with the idea of an age-related decline in the central exci-
tatory mechanisms. The evidence already discussed
suggests that there is an age-related loss of alpha-2 receptor
responsiveness which is general, i.e. affecting both central
and peripheral alpha-2 receptors. The loss of responsiveness
in the central alpha-2 receptors may conceivably be asso-
ciated with an age-related loss of central arousability.
On the basis of the alpha-2 receptor system we can
therefore postulate three age-related mechanisms (or ‘trait’
characteristics). In younger men, increased alpha-2 tone
centrally would result in inhibition of central arousal
mechanisms. Whether this active mechanism would also
apply in the periphery in those circumstances is not clear,
but because it is an active mechanism one can postulate that
it is selectively confined to central processes. In older men,
central loss of alpha-2 responsiveness could be associated
with loss of central arousability or excitation, whereas
peripheral loss of alpha-2 responsiveness could result in
increased inhibitory tone in the penile smooth muscles,
i.e. inhibited erection. Both of these effects in older men
could help to explain the age-related decline in NPT.
4. The relationship between SIS1 and SIS2
We have argued the case for the usefulness of the ‘inhi-
bitory tone’ concept. We can now reconsider the relation-
ship between our two inhibition scales, keeping in mind the
fact that there is a modest correlation between these two
scales, and hence presumably some overlap in what is
being measured. We are postulating that the essential
feature of SIS1 is the level of inhibitory tone, and of SIS2
the degree of responsiveness of the inhibitory system to
external threats. Thus, a high SIS1 individual will have a
high level of inhibitory tone and a high SIS2 individual will
have a propensity to respond to external threats with marked
increases in inhibitory level. The common ground between
the two scales we can see as the particular threat incurred by
recognition of the consequences of ‘high inhibitory tone; i.e.
the individual with high tone will learn that he has a
tendency to respond poorly in sexual situations. The antici-
pation of failure thus becomes a threat, his response to
which will increase further his already high inhibitory
tone. This illustrates how our theoretical model involves
both neurophysiological mechanisms and the role of
learning.
5. Implications of the model for treatment of erectile
dysfunction
The theoretical interpretation that we have presented sees
‘threat of performance failure’, the hallmark of SIS1, as
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579576
more intrinsic, and less context specific than the ‘threat of
performance consequences’. This could have considerable
implications for treatment. If, in a particular case, erectile
failure is largely dependent on the SIS2 mechanism (i.e.
inhibitory response to the perception of a threat in the sexual
situation), then it is conceivable that psychological treat-
ment might alter or reduce that threat. If, on the other
hand, erectile failure is a manifestation of a chronic ‘high
inhibitory tone’ state, whatever the cognitive expression of
that state may be, changing the inhibitory tone by
psychological treatment may be much more difficult, or
at least require a different approach. In those circum-
stances, a pharmacological agent that counteracts the
high inhibitory tone may be more appropriate than
psychological treatment.
6. Putting the model to the test
So far we have speculated vigorously on the basis of very
limited evidence. The model, however, does lend itself to
further testing, particularly now that we have apparently
valid measures of individual variability.
6.1. Potential experimental studies
1. Explore the relationship between SIS1 scores and
evidence of ‘inhibitory tone’ (e.g. by replication of the
sleep study with an alpha-2 antagonist given to men with
high and low SIS1 scores; by relating SIS1 scores to
circumference of the flaccid penis, and controlling for
size of erect penis—the higher the inhibitory tone the
smaller the flaccid penis).
2. Explore different types of threatening sexual stimuli that
discriminate between high and low SIS2 individuals,
looking in particular for threatening sexual stimuli that
may be of relevance to psychogenic ED (e.g. certain
types of partner response patterns).
3. Look for tests which would enable us to discriminate
between our putative age related patterns, with high
central inhibition associated with younger age and high
peripheral inhibition with older age, using parameters
other than age.
4. Explore the extent to which our putative ‘high inhibitory
tone’ is specific for sexual response or more generalized.
Investigate the emotional state as well as psychophysio-
logical characteristics of younger individuals with high
SIS1 scores. Compare them with subjects with other
inhibited response patterns such as Generalized Anxiety
Disorder (GAD), while responding to both sexual and
non-sexual stimuli. In this way we plan to establish
whether the lack of responsiveness in the sexual situa-
tion occurs in other non-sexual contexts as well (i.e. to
what extent is this type of erectile dysfunction a form
of GAD).
6.2. Potential treatment studies
We can at this stage make crude but testable treatment-
based predictions:
1. In treatment studies using psychological methods of
treatment, high SIS2 individuals are more likely to
respond than high SIS1 individuals.
2. Younger age, high SIS1 patients will respond favorably
to inhibition reducing drugs. Because of the uncertain
effects of simple alpha-2 blockade on peripheral mechan-
isms, a combined alpha-1 and alpha-2 antagonist, such as
phentolamine, might do better than a pure alpha-2
antagonist.
3. In older, high SIS1 patients, treatment may be based on
the ratio of SES and SIS1 scores. Relatively high SES
and high SIS1 individuals might respond better with a
disinhibiting drug such as phentolamine, or a more speci-
fic alpha-1 antagonist. The lower the SES score in rela-
tion to the SIS1 score, the greater the likelihood of an
excitation facilitator such as sildenafil or apomorphine
being effective. Where there is a combination of low
SES and high SIS1, then a combination of sildenafil (or
apomorphine) and phentolamine might have the best
chance of success.
7. Conclusions
The decade between the mid-1970s and mid-1980s saw a
number of controlled treatment outcome studies aimed at
evaluating the effectiveness of different types of psycholo-
gical treatment for sexual dysfunctions (see Ref. [2], pp.
498–502). A recurring theme was the failure to demonstrate
superiority of one psychological approach over another. In
an earlier paper [40], it was suggested that the principal
explanation for this apparent failure was the considerable
degree of prognostic variability, sufficient to overwhelm the
variance attributable to the treatment method. The answer, it
was proposed, was to establish relevant prognostic indica-
tors and control for them appropriately in future treatment
outcome studies. In this paper, we have presented some
relatively novel theoretical ideas about the central control
mechanisms of erectile response and the interface between
such mechanisms and cognitive mechanisms, as well as the
relationship between central and peripheral mechanisms.
We see this as highly relevant to ‘prognostic variability’,
and with the questionnaire measure now available [23],
believe that we have the ability to control for an important
component of prognostic variability in future treatment
outcome studies. At the same time, we believe that this
theoretical approach offers the possibility of a new research
agenda for understanding psychogenic erectile dysfunction.
It has the added virtue of generating predictions relevant to
both psychological and pharmacological methods of treat-
ment. Its relevance to more organic forms of erectile
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579 577
dysfunction should also not be underestimated. Men with
greater degrees of peripheral organic impairment of erectile
function are still likely to vary in the relevance of these
central mechanisms to their erectile dysfunction, and such
factors may well account for a considerable amount of
the variance in treatment outcome. We look forward to
reporting more evidence based on this model in the near
future.
References
[1] Araujo AB, Durante R, Feldman HA, Goldstein I, McKinlay JB. The
relationship between depressive symptoms and male erectile dysfunc-
tion: cross-sectional results from Massachusetts male aging study.
Psychosom Med 1998;60:458–65.
[2] Bancroft J. Human sexuality and its problems. 2nd edn. Edinburgh:
Churchill Livingstone, 1989.
[3] Bancroft J. Are the effects of androgens on male sexuality noradre-
nergically mediated? Some consideration of the human. Neurosci
Biobehav Rev 1995;19(2):325–30.
[4] Bancroft J. Central inhibition of sexual response in the male: a
theoretical perspective. Neurosci Biobehav Rev 1999;23:763–84.
[5] Bancroft J. Effects of alpha-2 blockade on sexual response: experi-
mental studies with Delequamine (RS15385). Int J Impotence Res
1999 (in press).
[6] Bancroft J, Malone N. The clinical assessment of erectile dysfunction:
a comparison of nocturnal penile tumescence monitoring and intra-
cavernosal injections. Int J Impotence Res 1995;7:123–30.
[7] Bancroft J, Munoz M, Beard M, Shapiro C. The effects of a new
alpha-2 adrenoceptor antagonist on sleep and nocturnal penile tumes-
cence in normal male volunteers and men with erectile dysfunction.
Psychosom Med 1995;57:345–56.
[8] Bancroft J, Wu FCW. Changes in erectile responsiveness during
androgen therapy. Arch Sex Behav 1983;12:59–66.
[9] Beck AT. Depression. Clinical, experimental and theoretical aspects.
London: Staples Press, 1967.
[10] Brotto LA, Gorzalka BB, Hanson LA. Effects of housing conditions
and 5-HT 21 activation on male rat sexual behavior. Physiol Behav
1998;63(4):475–9.
[11] Buvat J, Buvat-Herbaut M, Lemaire A, Marcolin G, Quittelier E.
Recent developments in the clinical assessment and diagnosis of
erectile dysfunction. Ann Rev Sex Res 1990;1:265–308.
[12] Carani C, Bancroft J, Granata A, Del Rio G, Marrama P. Testosterone
and erectile function, nocturnal penile tumescence and rigidity,
and erectile response to visual erotic stimuli in hypogonadal and
eugonadal men. Psychoneuroendocrinology 1992;17(6):647–54.
[13] Carani C, Granata ARM, Bancroft J, Marrama P. The effects of
testosterone replacement on nocturnal penile tumescence and rigidity
and erectile response to visual erotic stimuli in hypogonadal men.
Psychoneuroendocrinology 1995;20(7):743–53.
[14] Clark JT, Smith ER, Davidson JM. Testosterone is not required for the
enhancement of sexual motivation by yohimbine. Physiol Behav
1985;33:517–21.
[15] Cunningham GR, Karacan I, Ware JC, Lantz CD, Thornley II. The
relationship between serum testosterone and prolactin levels on
nocturnal penile tumescence in impotent men. J Androl
1982;3:241–7.
[16] Goodwin G. Mood disorder. In: Johnstone EC, Freeman CPL, Zealley
AK, editors. Companion to psychiatric studies, Vol. 6. Edinburgh:
Churchill Livingstone, 1998. p. 399–430.
[17] Goodwin GM, Austin MP, Curran SM. The elevation of plasma
beta-endorphin levels in major depression. J Affect Disord
1993;29:281–9.
[18] Granata A, Bancroft J, Del Rio G, Carani C. Stress and the erectile
response to intracavernosal prostaglandin E1 in men with erectile
dysfunction. Psychosom Med 1995;57:336–44.
[19] Hart BL. Sexual reflexes and mating behavior in the male rat. J
Comparative Physiol Psychol 1968;65:453–60.
[20] Herbert J. Sexuality, stress, and the chemical architecture of the brain.
Ann Rev Sex Res 1996;7:1–43.
[21] Higgins GE. Sexual response in spinal cord injured adults: a review.
Arch Sex Behav 1979;8:173–96.
[22] Hull EM, Lorrain DS, Du J, Matuszewich L, Bitran D, Nishita JK,
Scaletta LL. Organizational and activational effects of dopamine on
male sexual behavior. In: Ellis L, Eberty L, editors. Male/female
differences in behavior: toward biological understanding, Westport,
Con: Greenwood Press, 1998.
[23] Janssen E, Vorst H, Finn P, Bancroft J. The sexual inhibition (SIS)
and sexual excitation (SES) scales: measuring individual differences
in the propensity for sexual inhibition and excitation in men.
Submitted for publication.
[24] Kim SC, Oh MM. Norepinephrine involvement in response to intra-
corporeal injection of papaverine in psychogenic impotence. J Urol
1992;147:1530–2.
[25] Kwan M, Greenleaf WJ, Mann J, Crapo L, Davidson JM. The nature
of androgen action on male sexuality; a combined laboratory and self-
report study in hypogonadal men. J Clin Endocrinol Metab
1983;57:557–62.
[26] Lerner SE, Melman A, Christ GJ. A review of erectile dysfunction:
new insights and more questions. J Urol 1993;149:1246–55.
[27] Lundberg PO. Sexual dysfunction in patients with neurological
disorders. Ann Rev Sex Res 1992;3:121–50.
[28] McKittrick CR, Blanchard DC, Blanchard RJ, McEwen BS, Sakai
RR. Serotonin receptor binding in a colony model of chronic social
stress. Biol Psychiatry 1995;37:383–93.
[29] Meisel RL, Sachs BD. Spinal transection accelerates the developmen-
tal expression of penile reflexes in male rats. Physiol Behav
1980;24:289–92.
[30] Munoz M, Bancroft J, Beard M. Evaluating the effects of an alpha-2
adrenoceptor antagonist on erectile function in the human male. II.
The erectile response to erotic stimuli in men with erectile dysfunc-
tion, in relation to age and in comparison with normal volunteers.
Psychopharmacology 1994;115:471–7.
[31] Munoz M, Bancroft J, Turner M. Evaluating the effects of an alpha-2
adrenoceptor antagonist on erectile function in the human male. I. The
erectile response to erotic stimuli in volunteers. Psychopharmacology
1994;115:463–70.
[32] Nofzinger EA, Thase ME, Reynolds III CF, Frank E, Jennings R,
Garamoni GL, Fasiczka AL, Kupfer DJ. Sexual function in depressed
men: assessment by self-report, behavioral and nocturnal penile
tumescence measures before and after treatment with cognitive beha-
vioral treatment. Arch Gen Psych 1993;50:24–30.
[33] O’Carroll RE, Shapiro C, Bancroft J. Androgens, behaviour and
nocturnal erections in hypogonadal men: the effect of varying the
replacement dose. Clin Endocrinol 1985;23:527–38.
[34] Parmeggiana PL, Morrison AR. Alterations in autonomic functions
during sleep. In: Loewy AD, Spyer KM, editors. Central regulation of
autonomic functions, New York: Oxford University Press, 1990. pp.
367–86.
[35] Roose SP, Glassman AH, Walsh BT, Cullen K. Reversible loss of
nocturnal penile tumescence during depression: A preliminary report.
Neuropsychobiology 1982;8:284–8.
[36] Schiavi RC. Aging and male sexuality. Cambridge: Cambridge
University Press, 1999.
[37] Schiavi RC, White D, Mandeli J, Schreiner-Engel P. Hormones and
nocturnal penile tumescence in healthy aging men. Arch Sex Behav
1993;22:207–15.
[38] Tennent G, Bancroft J, Cass J. The control of deviant sexual behavior
by drugs: a double-blind controlled study of benperidol, chlorproma-
zine, and placebo. Arch Sex Behav 1974;3:261–70.
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579578
[39] Thase ME, Reynolds CF, Glanz LM, Jennings JR, Sewitch DE, Kupta
DJ, Frank E. Nocturnal penile tumescence in depressed men. Am J
Psychiatry 1987;144:89–92.
[40] Warner P, Bancroft J. Sex therapy outcome research: A reappraisal of
methodology. II. Methodological considerations. Psychol Med
1986;16:855–63.
J. Bancroft, E. Janssen / Neuroscience and Biobehavioral Reviews 24 (2000) 571–579 579