To protect patient anonymity, the case presented here includes features from two separate patients, and additional changes to details have been made to disguise identity.

This history immediately raises several different issues. From a phenomenological point of view, "problematic use" of the Internet has recently been described in the psychiatric literature (1, 2). Although this is a new category of psychopathology, pathologic use of pornographic materials as well as excessive masturbation have long been described (3, 4). The patient’s history immediately raises questions about the relationship of his excessive use of the Internet for viewing pornography and the return of a depressed mood. Similarly, there is the question of how best to diagnose the patient’s problematic sexual behavior.

From a pharmacological point of view, there is a small but clinically important literature on the return of depressive symptoms in patients who have successfully responded to an antidepressant and who have continued to be compliant with maintenance therapy (5). The reasons for this phenomenon are not well understood, but the possibility that an increase in stressors plays a role has obvious face validity. The optimal management of such patients has also not been well studied, although an increase in medication dose has some empirical support (5).

Although the optimal diagnosis and management of this patient may not have been immediately clear, there did seem to be an obvious need for intervention. Excessive use of the Internet at work for non-work-related reasons was, unsurprisingly, associated with decreased productivity. The patient was presumably at risk for facing legal action by his employer should his actions have come to light. The distress he experienced was in some ways fortunate, as it appeared to have contributed to his decision to seek treatment.

The patient’s lack of hypomania and mania is important, given that hypersexuality may be a symptom of these conditions. The apparent increase in hypersexual behaviors during periods of depressed mood is interesting in terms of previous suggestions that such behaviors may in fact be symptoms of depression and may respond to antidepressant medications (6). Ruling out substance abuse is also important, particularly given that cocaine use may result in hypersexual symptoms (7). Finally, patients with hypersexual symptoms may have a range of comorbid conditions, including obsessive-compulsive disorder (OCD) and Tourette’s disorder (8), so it is appropriate to rule these out.

In terms of pharmacotherapeutic intervention, the presence of atypical depressive symptoms has important implications. There is strong evidence that irreversible monoamine oxidase inhibitors (MAOIs) are more effective than tricyclic antidepressants in the treatment of such symptoms (9). Given the inconvenience of MAOI dietary precautions, selective serotonin reuptake inhibitors (SSRIs) are useful first-line medications. Certainly, their apparent efficacy in the treatment of this patient’s major depression is consistent with a presumed role of serotonin in hypersomnia and hyperphagia and with the findings of some previous reports that SSRIs are effective in treating atypical depression (10).

Sexual fantasy, along with dreams, has of course long been conceived of as one of the important roads to understanding the unconscious. A clinician would want to understand why domination played an important role in this patient’s psychic life. Although aggressive urges are perhaps universal, understanding this patient’s unique life history and consequent unconscious conflicts may have been useful in developing a treatment plan. It would have been pertinent to inquire about early sexual experiences as well as about childhood sexual abuse, which may be associated with later excessive sexual behavior (2).

It is interesting to note that cultural factors—development of the Internet—seem to have markedly contributed to the pathogenesis of this patient’s symptoms. Although the Internet may offer clinicians and their patients valuable opportunities for psychoeducation and support (11), it may also provide an opportunity for pathological gambling and other kinds of dysfunctional behavior (1, 2).

The phrase "just right," which the patient used to describe his search for arousing pornographic images, is reminiscent of a symptom of OCD. However, as noted earlier, this patient apparently demonstrated no evidence of suffering from any of the anxiety disorders. The lack of an association of sexual arousal with sadistic material rules out the paraphilia of sexual sadism. This point is important to emphasize, given that there is a high comorbidity between paraphilias and so-called paraphilia-related disorders (13).

Young (12) suggested that the subjugation schema may develop when childhood expression of anger is discouraged, and that adults with this schema are able to express this emotion only indirectly. Assertiveness training may be an initial intervention in order to help patients begin to overcome the subjugation schema. Referral for cognitive therapy to help change underlying early maladaptive schemas may also be considered. The relationship between schemas, stressors, symptoms, and mood does not simply involve one-directional causality but, rather, is likely to be complex.

Splitting therapy between a psychiatrist and a psychologist entails a number of potential problems; certainly in the case of symptoms that the patient finds embarrassing, the thought of having to disclose these to a new person may exacerbate matters. The response of depressed symptoms to an increased dose of fluoxetine is consistent with evidence from a previous report (5). Although SSRIs have been reported to be useful in decreasing excessive masturbation and similar symptoms, their effects are not always robust (6, 8, 14). Furthermore, in a controlled trial of clomipramine versus desipramine for such symptoms, efficacy was not found (15). Whether SSRIs may diminish the distress of loneliness in the absence of a threshold mood disorder is an interesting theoretical question, about which there are few data.

Psychotherapy has been reported a useful treatment for excessive masturbation and similar symptoms by many authors (3), and although there are a lack of controlled studies in this specific area, psychotherapy is certainly thought effective for commonly occurring comorbid axis I disorders (such as depression), as well as for certain axis II problems (such as difficulties with assertiveness). A couples intervention might also have been a consideration had there been evidence of marital dysfunction. It is also theoretically possible that pharmacotherapy and psychotherapy enhanced one another. Despite the generally positive outcome for this patient, it is notable that symptoms of excessive sexual behavior may often have a chronic course (2).

The patient here is redolent of Krafft-Ebbing’s description of "pathological sexuality" 100 years ago (16):

Of course, modern communications media provide a range of alternative modes for expression of psychopathology. The Internet, in particular, is likely to become an important location for expressing different symptoms, including "pathological sexuality."

Fairly recent studies have suggested that "pathological sexuality" is far from uncommon and can be associated with considerable morbidity (3, 17). The disorder appears more common in men, and patients may be seen with a range of different behaviors, including compulsive masturbation, excessive use of printed or telephonic pornography, and pathological use of the services of sex workers. As with impulse-control disorders, although the symptoms are gratifying, there is also typically an element of ego dystonicity. Comorbid diagnoses include mood disorders, anxiety disorders, and substance use disorders. Symptoms may severely affect family, social, and occupational function, and negative consequences include those of sexually transmitted disease. There is clearly a need for appropriate diagnosis and treatment of such patients.

Over the years, a range of different terms have been used to refer to such patients, including "Don Juanism" and "nymphomania" (18, DSM-III). Although the DSM-III-R section on sexual disorders not otherwise specified includes the term "non-paraphilic sexual addictions," this term was dropped from DSM-IV. The concept of "sexual compulsivity" (19, 20) is based on the idea that there is a phenomenological and psychobiological overlap between this entity and OCD. In contrast, others have used the term "sexual impulsivity" and emphasized the overlap with disorders of impulse control (21, 22). The notion of sexual addiction has also been proposed, again based on putative similarities with addictive disorders (3, 23). "Paraphilia-related disorder" has been suggested in view of the high comorbidity with, and phenomenological similarity to, paraphilias (13).

The lack of an agreed-on term has arguably contributed to the relative paucity of research in this area. Each of the different terms arguably has both advantages and disadvantages. Certainly, they suggest a range of different theoretical approaches to future research in this area. However, whatever the strengths and limitations of these approaches, we emphasize that there is a limited empirical literature in this area, making it difficult to endorse any single theoretical model (17, 24). In keeping with DSM’s emphasis on descriptive phenomenology rather than unsupported theory, the term "hypersexual disorder" is perhaps most appropriate.

"Hypersexual disorder" perhaps receives support from evidence that total sexual outlet, defined as number of sexual behaviors in a week that culminate in orgasm, is relatively high in this group of patients (13), although the degree to which symptoms involve physical orgasm (rather than, e.g., sexual fantasies and urges) varies from patient to patient. Crucially, however, the term focuses on observable phenomena and moves away from any possibly inadequate theoretical framework. The older alternative of "pathological hypersexuality" arguably sounds pejorative to the modern ear.

Is it possible to formulate diagnostic criteria that distinguish hypersexual disorder from behavior that is merely symptomatic of another disorder (such as depression), as well as from normal sexual behavior? It needs to be established, for example, that there is excessive preoccupation with nonparaphilic sexually arousing fantasies, urges, or excessive sexual behaviors over a notable length of time (e.g., 6 months). In addition, it needs to be determined that symptoms are not better accounted for by another axis I disorder (e.g., a manic episode or delusional disorder, erotomanic subtype) and the symptoms are not due to the direct physiological effects of a substance (e.g., a drug of abuse or a medication) or a general medical condition. Finally, the judgment that the sexual fantasies, urges, or behaviors are excessive (i.e., represent psychopathology) must take into account normal variation as a function of age (e.g., in teenagers, high levels of preoccupation with sexual fantasy may be normative) and subcultural values (e.g., in patients who value celibacy, the presence of some sexual urges and associated distress may be normative), as well as the degree to which the symptoms are the source of distress or interfere with important areas of functioning.

These considerations and the wording used here are consistent with proposals in the literature (17, 24). Thus, establishing that the symptoms are sexual fantasies, urges, and behaviors that are nonparaphilic follows from the DSM-IV definition of paraphilias; these are recurrent, intense sexually arousing fantasies, sexual urges, or behaviors generally involving nonhuman objects, the suffering or humiliation of oneself or one’s partner, or children or other nonconsenting persons. In effect, the logic here is that in hypersexual disorder, the symptoms are those seen in normative arousal patterns.

Similarly, it is clearly important to determine when hypersexual symptoms are better explained by other psychiatric or general medical conditions than by a specific diagnosis of hypersexual disorder. As noted earlier, for example, patients with mania or cocaine use may display hypersexual behavior. Furthermore, hypersexual behavior may be seen in a number of different neurological conditions (7). In the case presented here, there was no evidence that the symptoms could be solely accounted for by a mood or other disorder, although mood (and possibly lack of assertiveness) may have exacerbated the sexual symptoms and in turn been exacerbated by them.

Finally, there is the conceptually difficult task of delineating normal variation from psychopathology (25). The wording used above emphasizes that clinical judgments about psychopathology should take into account both normal variation and the harm caused by symptoms. Thus, for example, intense sexual fantasies in teenagers or distress caused by sexual urges in individuals striving to be celibate are typically not psychopathological.

There is, of course, a rich philosophical literature that attempts to define medical and psychiatric disorders and their borders with normality more precisely (26–28); the problem of delineating normal variation from psychopathology is particularly difficult when, as in the case of hypersexual disorder, the form of the phenomenology is (by definition) normative. The wording used here is consistent with the views of many authors who assert that clinical diagnosis involves evaluative judgments regarding cultural norms (27, 28).

Although it would be theoretically possible to include "hypersexual disorder" in the DSM section on impulse control disorders, it seems most to belong in the section on sexual disorders. This is consistent with the classification of analogous entities such as bulimia (which has impulsive characteristics but is categorized as an eating disorder).

The recent emergence of a range of different behaviors under the rubric "problematic Internet usage" raises the question whether this, too, should be a psychiatric diagnosis (29, 30). Two studies (1, 2) have indicated that the consequences of such use can indeed be far reaching, with many subjects going without sleep, being late for work, ignoring family obligations, and suffering financial and legal consequences. The typical subject in these studies was in his or her low- to mid-30s, had at least some college education, spent about 30 hours per week on "nonessential" Internet use, and had a mood, anxiety, substance use, or personality disorder. Given that the Internet allows rapid access to sexual material and even sexual partners (31), sexual behavior in this context is particularly pertinent (32). It seems reasonable to suggest that a history of Internet behavior be included as part of the standard psychiatric interview. Nevertheless, given that such symptoms may often be understood in terms of existing diagnoses (including hypersexual disorder), there is reason to be cautious of simply making a diagnosis of problematic Internet usage. Consensus on a diagnostic term and criteria for hypersexual behavior would encourage further research that would help us better understand these patients and, it is hoped, provide better care. Although a range of hypotheses have been put forward about the etiology of hypersexual disorder (3, 17), there are relatively few empirical data to support any particular theory. A number of medications have been suggested useful, with much of the focus on SSRIs in particular, but there is a dearth of controlled trials. Similarly, psychotherapy is routinely advocated despite limited research support. Nevertheless, clinicians who specialize in working with hypersexual disorder are optimistic that many patients can be helped with appropriate clinical care (33).

Received July 24, 2000; revisions received Jan. 19, April 13, and May 22, 2001; accepted May 23, 2001. From the Department of Psychiatry, University of Stellenbosch; the Department of Psychiatry, University of Iowa, Iowa City; the Department of Psychiatry, University of Florida, Gainesville; and the New York State Psychiatric Institute, Department of Psychiatry, Columbia University, New York. Address reprint requests to Dr. Stein, Unit on Anxiety Disorders, Medical Research Council, Department of Psychiatry, University of Stellenbosch, P.O. Box 19063, Tygerberg 7505, Cape Town, South Africa; djs2@gerga.sun.ac.za (e-mail).Dr. Stein is supported by the Medical Research Council of South Africa.