Classif ying

Hypers exual Disorders :

Compulsive, I mpulsive,
a nd Addictive Models
Dan J. Stein, MD, PhDa,b,*
KEYWORDS
 Hypersexual disorder  Sexual addiction  Compulsive
 Impulsive

There is a range of variation in sexual activity. People with clinically excessive sexual
thoughts or behaviors have been categorized as suffering from a compulsive, impulsive, or addictive sexual disorder. Such terms reflect key theoretical assumptions about
the nature of the behavior. Thus, some have described compulsive sexual symptoms
and addressed their relationship to obsessive-compulsive disorder1. Others have emphasized the role of impulsivity, and the spectrum of impulse control disorders, in conceptualizing such symptoms.2 Finally, DSM-III-R used the term non-paraphilic sexual
addiction,3 an approach consistent with one that emphasizes the addictive nature of
these symptoms, but inconsistent with the decision not to include this term in DSM-IV.
Similar considerations apply to the symptoms of a range of conditions included in
the DSM category of impulse control disorders not otherwise classified. These disorders include intermittent explosive disorder, kleptomania, pathological gambling,
pyromania, and trichotillomania. Thus, for example, clinically excessive gambling
has been described as compulsive, impulsive, or addictive.4 Similarly, clinically excessive hair-pulling has been characterized as compulsive, impulsive, and addictive.5
Thus it seems that the phenomenology and psychobiology of these conditions remains incompletely understood, with a range of different nosological terminology
and conceptual models still under active consideration by the field.
In this article we consider the nosological implications of advances in understanding
the phenomenology and psychobiology of the non-paraphilic sexual addictions and
the impulse control disorders not otherwise classified. We argue that each of the compulsive, impulsive, and addictive approaches seems to offer only a partial view of
these conditions. We go on to review research suggesting that key components of
such conditions include affective dysregulation (A), behavioral addiction (B), and cognitive dyscontrol (C). We then use this research to argue for a revision of the
a

University of Cape Town, Private Bag X3, Rondebosch 7701, Cape Town, South Africa
Mt. Sinai School of Medicine, 5 East 98th Street, New York, NY 10029, USA
* University of Cape Town, Private Bag X3, Rondebosch 7701, Cape Town, South Africa.
E-mail address: dan.stein@uct.ac.za
b

Psychiatr Clin N Am 31 (2008) 587591

doi:10.1016/j.psc.2008.06.007
0193-953X/08/$ see front matter 2008 Elsevier Inc. All rights reserved.

psych.theclinics.com

588

Stein

nomenclature of these conditions, and to advocate for additional work on their classification and pathogenesis.
COMPULSIVE-IMPULSIVE-ADDICTIVE VIEWS OF SEXUAL SYMPTOMS

Obsessive-compulsive disorder (OCD) is perhaps the paradigmatic compulsive disorder. It is characterized by obsessions (intrusive thoughts, images, or impulses), and
compulsions (repetitive thoughts or actions that act to neutralize the obsessions). Typically obsessions increase anxiety, while compulsions decrease anxiety. Compulsive
sexual symptoms are also characterized by repetitiveness, with an increase in tension
before the behavior, and a sense of release at the time of their execution. Although OCD
behavior is typically characterized by dysphoria, and sexual behavior by gratification,
there are inter-individual differences (some OCD patients describe a just right feeling,
and many patients with compulsive sexual symptoms describe guilt and regret).6
Impulsive personality disorders are perhaps the paradigmatic impulsive condition.
They are characterized by various kinds of impulsivity, including hyperresponsiveness
to stress, inability to delay gratification, and impulsive decision making. Similarly, patients with clinically excessive sexual thoughts and behaviors may have an increase in
symptoms in response to stress, may demonstrate an inability to delay sexual gratification, and may have impulsive decision making in a range of situations including
those involving sex. As noted below, a range of other impulse control disorders may
also demonstrate these elements; the temper outbursts of intermittent explosive disorder, for example, may be exacerbated at times of stress, may be associated with
inability to delay gratification, and may reflect poor executive control.7
Substance dependence is the paradigmatic addictive disorder. Patients with substance dependence are preoccupied with obtaining more substances, attempt to
consume more and more of a particular substance to get the same high, and
demonstrate withdrawal symptoms after abrupt discontinuation of their symptoms.
Similarly, patients with clinically excessive sexual thoughts or behaviors are preoccupied with their concerns, may demonstrate an escalating pattern of symptoms, and
experience dysphoria when they attempt to discontinue their behaviors.8 It is not surprising, then, that a number of authors have conceptualized such symptoms as addictive in nature, and that 12-step programs based on models for addressing substance
use disorders have been proposed for their treatment.
There are important problems with each of these conceptual approaches to clinically excessive sexual thoughts or behaviors. First, there are subtle but important distinctions between these symptoms and those seen in the paradigmatic compulsive,
impulsive, and addictive disorders. The typical behaviors of OCD do not involve reward (unlike sexual behavior), the impulsive personality may be unable to plan carefully to achieve goals (such as sexual encounters), and substance use disorders
crucially involve a substance (rather than simply a rewarding behavior like sex). Second, there may be important distinctions in the psychobiology of these disorders; for
example, whereas OCD is mediated by cortico-striatal-thalamic circuitry, the role of
these pathways in excessive sexual symptoms is less clear. Third, the standard treatments for some of these conditions (eg, exposure and response prevention in OCD)
may differ from those thought to be effective in excessive sexual symptoms.
AN A-B-C MODEL OF IMPULSE CONTROL DISORDERS

We have proposed that key components of impulse control disorders, such as trichotillomania, include affective dysregulation (A), behavioral addiction (B), and cognitive
dyscontrol (C).7 We briefly describe each of these components next.

Classifying Hypersexual Disorders

Affective dysregulation appears to be an important trigger of many of the symptoms

of the impulse control disorders not otherwise specified. Thus patients with trichotillomania note that hair-pulling is often triggered by negative affects. Similar associations
are described in other impulse control disorders (ICDs) such as pathological gambling.
There is growing interest in the psychobiology of stress-induced impulsivity, and it can
be hypothesized that amygdala activation may play a particularly important role. It is
possible that selective serotonin reuptake inhibitors are effective in some ICDs precisely because they act on such circuitry, and help to modulate affect dysregulation.
Many of the ICDs can be described as behavioral addictions insofar as patients
demonstrate a preoccupation with their symptoms, there may a gradual increase in
symptom severity consistent with tolerance, and there is considerable dysphoria
should symptoms be discontinued. It is notable that ventral striatal circuitry, and the
nucleus accumbens in particular, may play an important role in mediating not only
substance use disorders, but also a range of symptoms seen in impulse control disorders. It is possible that dopamine blockers are effective in some ICDs precisely because they act on such circuitry, and help to modulate behavioral addiction.
Many of the ICDs may be characterized by cognitive dyscontrol. At a clinical level,
patients seem unable to fully cognitively and affectively process the medium and longterm consequences of acting on their impulses. At a neuropsychological level, there is
often evidence of impaired executive control on systematic testing. At a biological
level, there may be evidence of decreased prefrontal activation in patients with such
cognitive dyscontrol, supporting the evidence that symptoms result from impaired executive control. It is possible that anticonvulsant (or neurostabilizer) medications are
effective in some ICDs precisely because they act on such circuitry, and help to reduce cognitive dyscontrol.
An A-B-C model of the ICDs suggests that instead of taking a single theoretical
approach to the classification and understanding of these conditions (whether
compulsive, impulsive, or addictive), it may instead be useful to explore the different
phenomenological and psychobiological components that underpin these conditions.
First, instead of reducing symptoms to a single kind of category, it would allow an
exploration of a range of phenomena seen in these disorders, as well as their comorbidity. Second, instead of explaining symptoms using only a particular set of neurocircuitry and related neurochemistry, it would allow various pathways to be explored.
Third, by outlining a range of different target phenomena, it would encourage the exploration of a number of different therapeutic interventions for these conditions.
AN A-B-C MODEL OF HYPERSEXUAL DISORDER

These considerations suggest that an A-B-C model of clinically excessive sexual

thoughts and behaviors may be useful and such symptoms should be described in
a theoretically neutral way. The term hypersexual disorder, for example, does not
rely on the compulsive, impulsive, or addictive models, and so allows each of these
approaches to be explored in relation to sexual symptoms. In the absence of clear
agreement on how best to conceptualize the phenomenology and psychobiology of
hypersexual disorder, a theoretically neutral term may be particularly useful in encouraging a range of approaches in both clinical settings and in the research laboratory.9
Furthermore, various data suggest that affective dysregulation may play an important role in hypersexual disorder. First, symptoms may be triggered or exacerbated
during times of increased stress and affective arousal. Second, many patients with hypersexual disorder have comorbid mood and anxiety disorders.1 Third, the selective
serotonin reuptake inhibitors (SSRIs) that are useful in modulating affect, and in

589

590

Stein

treating a range of mood and anxiety disorders, appear to also be useful in decreasing
symptoms of hypersexual disordereven in patients with comorbidity of these
disorders.10,11
A number of arguments can also be put forward to support the hypothesis that behavioral addiction is relevant to hypersexual disorder. In particular, as described earlier, patients with hypersexual disorder may be preoccupied with their sexual desires,
may demonstrate an escalating pattern of acting on such desires, and may exhibit significant dysphoria when they attempt to cut back on their behaviors. Although the relative paucity of psychobiological research on hypersexual disorder makes it difficult to
conclude that neurocircuitry relevant to behavioral addiction plays a role in mediating
this condition, it is notable that pro-dopaminergic drugs may increase sexual behavior.
Finally, there is at least some evidence that cognitive dyscontrol may play a role in
hypersexual disorder. Although there is again a paucity of systematic psychobiological research on this condition, there is evidence that executive functions are impaired
in patients with paraphilias. Certainly, there is anecdotal evidence that patients with
hypersexual disorder are unable to process optimally the consequences of their actions. Although additional research is clearly needed, such considerations raise the
question of whether anticonvulsant or neuromodulator agents that act on glutamatergic and GABAergic circuitry, and may be useful in a range of impulse control disorders,
may also be useful in hypersexual disorder.
SUMMARY

In closing, we argue for two conclusions. First, there are advantages to using theoretically neutral terms (such as hypersexual disorder) that go beyond the compulsiveimpulsive-addictive distinctions.9 Although the notion of theory-neutral observation
cannot be defended, it is important not to rely on any particular theoretical framework
before all the evidence is in. Our current nosology employs a range of contradictory
terms and frameworks (eg, impulse control disorder, compulsive gambling and buying, trichotillomania, and kleptomania). In keeping with the approach taken in other
DSM categories, it may be useful to find a more theory-neutral term that can cut
across these conditions.
Second, any conclusions drawn here about the nosology of hypersexual disorder
must be tempered by the relative lack of rigorous psychobiological and systematic
treatment data. A better understanding of the psychobiology of hypersexual disorder
might provide greater confidence in one or the other theoretical model. The A-B-C
model proposed here is tentative at best, given the relative absence of supporting
data. Further, a richer assessment and treatment literature would allow clearer conclusions about the clinical utility of different nosological approaches. We emphasize the
need for much additional work to characterize the phenomenology and psychobiology
of hypersexual disorder and other conditions characterized by affective dysregulation,
behavioral addiction, and cognitive dyscontrol, in the hope that such research would
ultimately lead to improved assessment and management.
ACKNOWLEDGMENTS

Dr. Stein is supported by the Medical Research Council of South Africa.

REFERENCES

1. Black DW, Kehrberg LL, Flumerfelt DL, et al. Characteristics of 36 subjects

reporting compulsive sexual behavior. Am J Psychiatry 1997;154:2439.

Classifying Hypersexual Disorders

2. McElroy SL, Phillips KA, Keck PE Jr. Obsessive compulsive spectrum disorder.
J Clin Psychiatry 1994;55(Suppl):3351.
3. American Psychiatric Association. Diagnostic and statistical manual of mental
disorders. 3rd (revised) edition. Washington, DC: American Psychiatric Association; 1987.
4. Grant JE, Potenza MN. Pathological gambling: a clinical guide to treatment.
Washington, DC: APPI; 2004.
5. Stein DJ, Mullen L, Islam MN, et al. Compulsive and impulsive symptomatology in
trichotillomania. Psychopathology 1995;28:20813.
6. Lochner C, Stein DJ. Does work on obsessive-compulsive spectrum disorders
contribute to understanding the heterogeneity of obsessive-compulsive disorder?
Prog Neuropsychopharmacol Biol Psychiatry 2006;30:35361.
7. Stein DJ, Chamberlain SR, Fineberg N. An A-B-C model of habit disorders: hairpulling, skin-picking, and other stereotypic conditions. CNS Spectr 2006;11:
8247.
8. Goodman A. Sexual addiction: An integrated approach. Madison, CT: International Universities Press; 1998.
9. Stein DJ, Black DW, Shapira NA, et al. Hypersexual disorder and preoccupation
with Internet pornography. Am J Psychiatry 2001;158:15904.
10. Stein DJ, Hollander E, Anthony D, et al. Serotonergic medications for sexual obsessions, sexual addictions, and paraphilias. J Clin Psychiatry 1992;53:26771.
11. Kafka M: Psychopharmacological treatments for nonparaphilic compulsive sexual behaviors. CNS Spectr 2000;5:4959.

591