Introduction

On July 31, 2008, I submitted a report on transvestism to the work group charged with revising the diagnoses concerning sexual and gender identity disorders for the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM). That report is reproduced in the remainder of this paper, beginning in the next section. I have made no changes to the original text, except for updating the references.

The original report included my proposal for a revised set of diagnostic criteria. In the year since I submitted my report, these diagnostic criteria have been further modified by input from the Paraphilias Subworkgroup of the Sexual and Gender Identity Disorders Work Group and from official Advisors to the Paraphilias Subworkgroup. Thus, the diagnostic criteria presented later in this paper are somewhat different from the diagnostic criteria currently being considered by the Paraphilias Subworkgroup, and they are likely different from the criteria that will eventually be approved by the DSM-V Task Force and the Board of Trustees of the American Psychiatric Association. I have included them because they were part of my original report, and because they help to document the evolution of the diagnostic criteria that will eventually form part of the DSM-V.

Report on Transvestic Fetishism

There are four key elements in the syndrome of Transvestism (later called Transvestic Fetishism) as described in the DSM. These four elements are: (1) cross-dressing (2) associated with sexual arousal (3) in a biological male (4) with a heterosexual orientation. There are, of course, cross-dressers who fall outside this definition: homosexual men who cross-dress without sexual arousal and perhaps rare women who cross-dress with sexual arousal. The existence of these other groups has no necessary bearing on whether the combination of male sex, heterosexual orientation, cross-dressing, and sexual excitement constitutes a distinct syndrome. The consensus of expert clinicians, for almost a century, has been that it does. This clinical consensus is supported by the available epidemiological data (Långström & Zucker, 2005).

In matters other than the key elements, the diagnostic criteria in successive versions of the DSM have varied. The diagnostic criteria for DSM-III (American Psychiatric Association, 1980), DSM-III-R (American Psychiatric Association, 1987), DSM-IV (American Psychiatric Association, 1994), and DSM-IV-TR (American Psychiatric Association, 2000) are given in the Appendix. In the first part of this report, I review differences among these DSM versions as a convenient way to illustrate problems with the nomenclature and uncertainties in the descriptive pathology of transvestism. I conclude this part by proposing a revised set of diagnostic criteria, including a new set of specifiers.Footnote 1 In the second part, I present a secondary analysis of a pre-existing dataset––an analysis intended to investigate the utility of the proposed specifiers.

Review

DSM-III

The diagnostic criteria in DSM-III state that cross-dressing is used “for the purpose of sexual excitement, at least initially in the course of the disorder” (emphasis added). The phrase was presumably added so that the diagnosis would capture the many patients who would have met all the diagnostic criteria for transvestism at one stage of their lives, but who state that cross-dressing now produces only feelings of comfort and relaxation, not sexual arousal (e.g., Benjamin, 1966; Buhrich & Beaumont, 1981; Buhrich & McConaghy, 1977a; Person & Ovesey, 1978; Wise & Meyer, 1980). An extensive discussion of self-reports of diminishing arousal and their possible meaning can be found in Blanchard, Racansky, and Steiner (1986).

Blanchard et al. (1986) examined this self-report phenomenon using phallometric testing, an objective technique for quantifying erotic interests in human males. In this psychophysiological procedure, the individual’s penile blood volume is monitored while he is presented with a standardized set of laboratory stimuli depicting potentially erotic objects or situations. Increases in the examinee’s penile blood volume (i.e., degrees of penile erection) are taken as an index of his relative responsiveness to different classes of stimuli. The abstract of that article states:

We examined whether an erotic response to cross-dressing fantasies could be detected in heterosexual male cross-dressers (HCDs) who verbally denied any erotic arousal in association with cross-dressing for at least the past year. Subjects were 37 HCD patients and 10 paid heterosexual controls. HCDs were divided into groups according to their response to a questionnaire item asking the proportion of occasions that cross-dressing was erotically arousing during the past year and offering response options from always to never. Penile blood volume was monitored while subjects listened to descriptions of cross-dressing and sexually neutral activities. All HCD groups responded significantly more to cross-dressing than to neutral narratives (p < .01); controls did not (p. 452).

In other words, Blanchard et al. (1986) concluded that transvestites (HCDs) who deny recent or past erotic arousal in association with cross-dressing or applying make-up still tend to respond with penile tumescence to fantasies of such activities. They discussed three possible explanations of the discrepancy between self-reported and objectively measured sexual response. The first explanation (which Blanchard et al. did not favor) is that patients intentionally attempt to mislead clinicians about the persistence of sexual arousal to cross-dressing. The second explanation is that some transvestites are actually unaware of mild and transient penile tumescence accompanying cross-dressing. The third explanation is more easily quoted verbatim than summarized:

A third possibility is that some HCDs’ erotic response to their usual cross-dressing activities has been extinguished through repeated exposure, so that they are strictly accurate when they report that putting on women’s attire or make-up produces no discernable penile erection. If this were the case, then the erotic response to cross-dressing fantasies in the present study could be partly a function of these fantasies’ novelty. The ability of the phallometric narratives to elicit penile tumescence, then, could be considered analogous to the animal investigators’ “Coolidge Effect” (Wilson, Kuehn, & Beach, 1963), a term used to denote the reactivation of copulatory preparedness in sexually exhausted males by a novel receptive female (p. 461).

In summary, there are many men who report that cross-dressing was once sexually arousing but that it has ceased to be so, and these reports may be accurate (especially if one holds a narrow view of erotically motivated behavior as behavior that is necessarily accompanied by penile erection). The DSM-III phrase at least initially in the course of the disorder avoided the absurd possibility that a man could outgrow the diagnosis of transvestism in later life simply by failing to experience (or attend to) sexual excitement when he cross-dresses. This was a more nuanced formulation than that used in later versions of the DSM.

The DSM-III also addressed the phenomenon of diminishing sexual response to cross-dressing in the text: “In some individuals sexual arousal by the clothing tends to disappear, although the cross-dressing continues as an antidote to anxiety” (p. 269). The notion that cross-dressing functions as an “antidote to anxiety” was repeated in the text of DSM-III-R (p. 288); in the texts of DSM-IV and DSM-IV-TR, this became “an antidote to anxiety or depression” (p. 531 and p. 574, respectively). The notion that cross-dressing has anxiolytic effects may have originated with, or have been transmitted through, the work of Ethel Person and Lionel Ovesey. Ovesey and Person (1976) make the statement, “there is a tendency in some transvestites for the sexuality to drop away, although cross-dressing continues as an antidote to anxiety” (p. 221). This statement is repeated almost word-for-word in Person and Ovesey (1978), including the phrase “antidote to anxiety” (p. 307). In any event, it is unclear whether the term anxiety, as repeatedly used in the DSM, is meant to denote a sense of fearful apprehension—as most people would consciously experience it—or some emotion specific to transvestites, for which anxiety is the best available description. In survey research on transvestites, the respondents do not usually report anxiety-reduction as a motivation for cross-dressing, although some indicate that they cross-dress to reduce tension or stress (e.g., Buhrich, 1978; Buhrich & McConaghy, 1977b; Croughan, Saghir, Cohen, & Robins, 1981; see also Docter & Prince, 1997). It is possible, of course, that some non-paraphilic heterosexual or homosexual men (or women) use sex in the same way.

The diagnostic criteria for transvestism in DSM-III included the separate criterion, “Intense frustration when the cross-dressing is interfered with.” This was probably intended as a core sign or symptom of transvestism rather than a distress or impairment criterion. It is not the same as impaired interpersonal or social functioning as a consequence of transvestism, and it is obviously quite different from remorse or discontent with being transvestic.

In DSM-III, Transsexualism was an exclusionary criterion for Transvestism. This is consistent with a common taxonomic view at that time, namely, that transsexualism, transvestism, and homosexuality constitute three, completely different entities with mutually exclusive etiologies.

DSM-III-R

In this version of the DSM, the name of the diagnosis was changed from Transvestism to Transvestic Fetishism. This was probably an attempt to disambiguate the term transvestism, which was then, as now, sometimes used to denote cross-dressing homosexual men (“drag queens”), and which had historically also been used to denote transsexuals (e.g., Hamburger, Stürup, & Dahl-Iversen, 1953), before the term transsexual became standard for that group.

In my opinion, the name choice of Transvestic Fetishism was counter-heuristic. It stresses one frequent feature of transvestism (erotic interest in the material properties of women’s clothing) at the expense of another (erotic arousal at the thought or image of oneself as a woman).Footnote 2 There are certainly transvestites with strong interests in specific articles of feminine attire, and for whom the differential diagnosis of transvestism vs. fetishism may be difficult to make (see Freund, Seto, & Kuban, 1996). There are also, however, transvestites for whom the physical properties of women’s attire appear secondary, and for whom the most important objective involves presenting themselves as women. I will return to this point later in this report.

A minimum duration of six months was added to the diagnostic criteria for all the paraphilias, including Transvestic Fetishism. It is unclear if this (obviously arbitrary) criterion was added simply to reduce the probability of false positive diagnoses by requiring more clinical evidence, or whether it reflects some notion that deviant sexual interests can occur as acute phenomena.

The phrase, “at least initially in the course of the disorder,” was dropped from the diagnostic criteria, perhaps to avoid the seeming paradox of a paraphilia without sexual arousal. The implication that a transvestite’s diagnosis should change if his sexual response to cross-dressing wanes beyond the point of subjective awareness is confirmed by this statement in the text:

In some people sexual arousal by clothing tends to disappear, although the cross-dressing continues as an antidote to anxiety. In such cases the diagnosis should be changed to Gender Identity Disorder of Adolescence or Adulthood, Nontranssexual Type [GIDAANT]. A small number of people with Transvestic Fetishism, as the years pass, want to dress and live permanently as women, and desire surgical or hormonal sex reassignment. In such cases the diagnosis should be changed to Transsexualism. (DSM-III-R, pp. 288–289)

The DSM-III-R was the only version of the DSM to offer the diagnosis of GIDAANT. It was therefore the only version to suggest this alternative diagnosis for patients whose sexual response to cross-dressing has disappeared while the cross-dressing itself continued at the same or even higher rate.

In DSM-III-R, the criterion, “Intense frustration when the cross-dressing is interfered with,” was dropped. The reasons for this are not apparent.

The writers of DSM-III-R added the beginnings of a separate distress/impairment criterion (Criterion B) to all the paraphilias; “The person has acted on these urges, or is markedly distressed by them.” In the present context, this meant that sexual urges to cross-dress and sexual fantasies of cross-dressing were not sufficient for a diagnosis of Transvestic Fetishism; the patient must either have tried on women’s apparel or else be distressed by his desires to do so. Thus, an adolescent male, who had unconflicted sexual desires to cross-dress but who lived at close quarters with his family and had literally no opportunity to satisfy these desires, could not be diagnosed with Transvestic Fetishism.

DSM-IV

In this version of the DSM, cross-dressing behavior was moved from Criterion B to Criterion A, where it became simply another sign or symptom or transvestism. Criterion B was now purely a distress/impairment criterion, written in precisely the same language for all the paraphilias: “The fantasies, sexual urges, or behaviors cause clinically significant distress or impairment in social, occupational, or other important areas of functioning.”

The foregoing change produced the absurd situation that a man could not be diagnosed with Transvestic Fetishism unless he was distressed or impaired by being a transvestite (Zucker & Blanchard, 1997). As I will show later, this problem could be solved without re-writing the diagnostic criteria at all, simply by re-naming the syndrome Transvestic Disorder—getting rid of an inapt name in the process—and by making a distinction between ascertaining transvestism (Criterion A) and diagnosing a Transvestic Disorder (Criterion B).

In DSM-IV, gender dysphoria/transsexualism is no longer an exclusionary criterion for the diagnosis. The diagnostic criteria now offer the specifier With Gender Dysphoria, which should be used “if the person has persistent discomfort with gender role or identity.” This amendment to the diagnostic criteria was consistent with data showing that gender dysphoria and erotic arousal to cross-dressing are not mutually exclusive. This had been demonstrated in a study by Blanchard and Clemmensen (1988). The abstract of their article states:

This study sought to determine the proportion of adult, male, heterosexual cross-dressers who acknowledge both gender dysphoria and at least occasional fetishistic response to cross-dressing. Subjects were 193 outpatients of the gender identity clinic or behavioral sexology department of a psychiatric teaching hospital. Questionnaire items were used to assess subjects’ current level of gender dysphoria and their recent history of sexual response to cross-dressing. Subjects who reported higher levels of gender dysphoria tended to report lower frequencies of sexual arousal with cross-dressing (r = −.56, p < .0001) and lower frequencies of masturbation with cross-dressing (r = −.62, p < .0001). About half of even the most strongly gender dysphoric subjects, however, acknowledged that they still become sexually aroused or masturbate at least occasionally when cross-dressing. These findings indicate a need for revision in the DSM-III-R’s diagnostic criteria for transvestism and gender identity disorders, which presuppose that gender dysphoria and fetishistic reactions are mutually exclusive (p. 426).

In the discussion section of their article, Blanchard and Clemmensen (1988) argued that the overlap of heterosexual gender dysphoria and fetishistic cross-dressing was more likely to have been underestimated from their data than overestimated.

The recognition that gender dysphoria and erotic arousal to cross-dressing are not mutually exclusive was an important step forward for DSM-IV. The implementation of the changes to the diagnostic criteria and to the text was awkward and confusing, however. The text for Transvestic Fetishism states:

Transvestic Fetishism is not diagnosed when cross-dressing occurs exclusively during the course of Gender Identity Disorder. (p. 531)

The text for Gender Identity Disorder states:

Males with a presentation that meets full criteria for Gender Identity Disorder as well as Transvestic Fetishism should be given both diagnoses. (p. 536)

Suppose that an adolescent male meets full criteria for Gender Identity Disorder before he begins to fantasize about cross-dressing and before his first episode of fetishistic cross-dressing. His cross-dressing, once initiated, is highly erotic and often culminates in masturbation to orgasm; however, its temporal course lies entirely within that of the Gender Identity Disorder. Should he be diagnosed with Transvestic Fetishism or not? The text for Gender Identity Disorder suggests yes; the text for Transvestic Fetishism suggests no.

The text for Gender Identity Disorder also contains the following statement:

If gender dysphoria is present in an individual with Transvestic Fetishism but full criteria for Gender Identity Disorder are not met, the specifier With Gender Dysphoria can be used (pp. 536–537).

There are two questionable points about this recommendation. First, it is not clear how someone could exhibit gender dysphoria, defined in the Transvestic Fetishism diagnostic criteria as “persistent discomfort with gender role or identity,” and not meet full diagnostic criteria for Gender Identity Disorder. Second, even if the patient somehow failed to meet full diagnostic criteria for Gender Identity Disorder, why not simply use the available diagnosis of Gender Identity Disorder Not Otherwise Specified?

In summary, the use of the specifier With Gender Dysphoria seems to have been an unnecessarily cumbersome and redundant way of doing something that could have been accomplished more simply with the addition of a second diagnosis—Gender Identity Disorder or Gender Identity Disorder Not Otherwise Specified. As a practical matter, the clinical consequences of gender dysphoria are at least as great as those of Transvestic Fetishism, and therefore the presence of gender dysphoria would better have been highlighted by a separate diagnosis rather than by noting its presence with a specifier.

The text of DSM-IV specifically mentions the existence of transvestites who cease to experience sexual arousal in association with cross-dressing; however, it no longer says anything about whether their diagnosis should be altered or to what it should be altered:

In some individuals, the motivation for cross-dressing may change over time, temporarily or permanently, with sexual arousal in response to the cross-dressing diminishing or disappearing. In such instances, the cross-dressing becomes an antidote to anxiety or depression or contributes to a sense of peace and calm. (p. 531)

In the foregoing passage, the DSM-IV introduced the notion (which was not explicitly stated in earlier versions) that the motivation for cross-dressing may change over time. This seemingly simple descriptive statement actually contains a lot of inference. How does one know that the fundamental motivation for cross-dressing has changed in these cases? It is, after all, the same people, doing the same thing. There is no obvious and objective change in incentives, as in the case of an amateur athlete becoming a paid professional. It is possible that the diminution in “sexual arousal” (which, in context, probably means a decrease in spontaneous penile tumescence or in the patient’s likelihood of masturbating while cross-dressed) simply reflects a developmental difference in the manifestation of transvestism.

DSM-IV-TR

The DSM-IV-TR diagnostic criteria for Transvestic Fetishism are identical to the DSM-IV criteria. Therefore the following problems remain:

  1. 1.

    A man cannot be identified as a transvestite—however much he cross-dresses and however sexually exciting that is to him—unless he is unhappy about this activity or impaired by it.

  2. 2.

    There is no clear diagnostic guidance regarding patients whose sexual response to cross-dressing diminishes or disappears while their frequency of cross-dressing remains the same or grows even higher.

  3. 3.

    The one available specifier (With Gender Dysphoria) is unnecessary and confusing, whereas other specifiers of potential usefulness are lacking.

The remainder of this report concerns my proposed solutions to these problems. Table 1 presents my proposed revision of the diagnostic criteria, and the following pages explain and justify the altered features.

Table 1 Proposed diagnostic criteria for Transvestic Disorder
Full size table

The most obvious difference between the proposed criteria and those used in DSM-IV and DSM-IV-TR is the replacement of the specifier “With Gender Dysphoria” by the specifiers “With Fetishism” and “With Autogynephilia.” The term autogynephilia (Blanchard, 1989a) denotes a male’s propensity to be erotically aroused by the thought or image of himself as a female. The frequent co-occurrence of fetishism and transvestism was reported by Wilson and Gosselin (1980). The frequent co-occurrence of autogynephilia and transvestism was reported by Blanchard (1991).

Commentary on the Proposed Criteria

Renaming the Diagnosis

The first problem with the DSM-IV-TR diagnostic criteria—erotic cross-dressing can be labeled transvestism only if the practitioner is distressed or impaired by it—can be solved simply by changing the name of the diagnosis from Transvestic Fetishism to Transvestic Disorder. Such a name change is consistent with my general proposal to distinguish between paraphilias and paraphilic disorders. On this view, a paraphilia is any powerful and persistent sexual interest other than sexual interest in copulatory or precopulatory behavior with phenotypically normal, consenting adult human partners (Cantor, Blanchard, & Barbaree, 2009). A paraphilic disorder is a paraphilia that causes impairment or distress. One would ascertain a paraphilia (determine whether it is present or absent according to common signs or symptoms) but diagnose a paraphilic disorder (determine whether the paraphilia is distressing the patient or impairing his psychosocial functioning). In my proposal for Transvestic Disorder, Criterion A is an ascertainment criterion that identifies the patient as transvestic according to traditional indicators of transvestism, and Criterion B is a diagnostic criterion that classifies the patient’s condition as a psychiatric disorder on the grounds of distress or impairment.

The immediate consequence of re-naming the diagnosis is that the patient does not have to be subjectively distressed or objectively impaired by his transvestism to be identified as a transvestite; he simply has to manifest the syndrome of repeated dressing in women’s apparel with sexual excitement (Criterion A). An ego-syntonic, well-adjusted transvestite could be classified as a transvestite for research or descriptive purposes without being diagnosed with a disorder. This eliminates the paradox implied by a literal reading of the DSM-IV-TR criteria for Transvestic Fetishism, according to which a man cannot be a transvestite unless he is upset or handicapped by being a transvestite.

Duration of Signs and Symptoms

I have not suggested any alteration of the qualifying phrase, “over a period of at least 6 months,” but I will note that it might be better applied to Criterion B than to Criterion A. Penile erection and masturbation in connection with the act or fantasy of changing into women’s apparel rarely or never occur as transient phenomena in adult men. There does not, therefore, seem to be any particular need to stress the duration of signs and symptoms in Criterion A. Some duration condition might actually make more sense in Criterion B, because the distress occasioned by transvestism could fluctuate according to circumstances (whether the patient is married or single, for example), and according to levels of self-acceptance that could change as the patient ages.

“Post-Erotic” Transvestites

My proposed approach to the second problem with DSM-IV-TR—the diagnosis of patients who report that sexual responding to cross-dressing has disappeared—is to deal with this in the text rather than in the diagnostic criteria. I suggest that patients who have been clearly ascertained as transvestic retain that label whether or not they report that cross-dressing continues to be accompanied by penile erection or subjective feelings identifiable as sexual excitement.

Specifiers

As I have previously indicated, I do not see any need for the DSM-IV-TR specifier “With Gender Dysphoria.” If the patient has gender dysphoria in addition to Transvestic Disorder, he can simply receive the additional diagnosis of Gender Identity Disorder or Gender Identity Disorder Not Otherwise Specified (or their equivalents in DSM-V). The clinical consequences of gender dysphoria are at least as great as those of transvestic disorder, and therefore the presence of gender dysphoria would better be highlighted by a separate diagnosis.

There is, on the other hand, a need to distinguish different types of transvestism according to the foci of the patient’s erotic interest. Transvestites vary greatly in their overt behavior and in their mental content during sessions of cross-dressing. Some seem quite similar to simple fetishists in their preference for very specific garments (e.g., white panties) and report no conscious thoughts of themselves as female even while dressed in multiple pieces of female underwear (panties, brassieres, and stockings). Some men of this type have particular rituals (e.g., tearing their women’s undergarments to shreds at some point during their masturbatory routine) that make them seem much more similar to fetishists than to other transvestites. Other transvestites, whom I have called autogynephiles, are most aroused by the thought or image of themselves as women. For these men, the material properties of women’s garments may seem secondary. The most exciting act is appearing to oneself and to others as a woman. There are many variations on this theme. A very common one is forced feminization, a fantasy scenario in which a man is coerced into wearing feminine attire by a dominant woman or group of women (e.g., an improbably motivated college sorority). This scenario is very common in fiction written by transvestites for transvestites (Beigel & Feldman, 1963; Buhrich & McConaghy, 1976; see also Veale, Clarke, & Lomax, 2008).Footnote 3 In this fiction, the (male) protagonist inevitably turns out to make an astonishingly beautiful and convincing woman as soon as his wig, make-up, and so on are in place, and he may live happily thereafter in a romantic relationship with the woman who precipitated this discovery.

As a practical matter, the autogynephilic type seems to have a higher risk of developing gender dysphoria. There is, at present, no way besides specifiers to capture the distinction between notably fetishistic and notably autogynephilic transvestites. This cannot be done simply by assigning patients a second paraphilic diagnosis, because there is no specific DSM diagnosis of autogynephilia. Therefore my solution to the third problem with the DSM-IV-TR diagnosis of transvestism—the lack of meaningful specifiers—is to add the specifiers “With Fetishism” and “With Autogynephilia.” The usefulness of these specifiers is evaluated in the next section of this report.

Empirical Study of the Proposed Specifiers: Frequency of Use and Relation to Gender Dysphoria

I have, in the past, conducted several studies on the relations among transvestism, fetishism, autogynephilia, and gender dysphoria (e.g., Blanchard, 1991, 1993). Relevant research has also been conducted by other investigators (e.g., Wilson & Gosselin, 1980). None of the prior studies, however, bears directly on the potential frequency of the use of both proposed specifiers (fetishism and autogynephilia) in an ascertained sample of transvestites, and none of them has examined the relations of fetishism and autogynephilia to gender dysphoria within an ascertained sample of transvestites. I therefore carried out a secondary analysis of a pre-existing dataset to investigate the utility of the proposed diagnostic specifiers for Transvestic Disorder.

Method

Subjects

The pool of potential subjects consisted of 427 adult male outpatients who reported histories of dressing in women’s garments, of feeling like women, or both. This group of patients was originally studied by Blanchard (1992). The raw data from Blanchard (1992) had been archived in a separate computer file, and they were therefore a convenient dataset for the present purpose.

These patients had presented from 1980 to 1990 at one of two departments of the Clarke Institute of Psychiatry (now the Centre for Addiction and Mental Health, Toronto, Ontario, Canada): the Research Section of Behavioural Sexology (now the Kurt Freund Laboratory) or the Gender Identity Clinic for adults. Because the data analyzed by Blanchard (1992) and reanalyzed here are from questionnaires, all these patients were necessarily literate in English.

In 392 cases (92% of the sample), the patient’s presenting complaint was gender dysphoria or transvestism, and the questionnaire materials were administered in the course of assessing these conditions. For most of the remaining 35 cases, the recorded presenting complaint (usually the most serious condition present) was masochism, sadism, fetishism, or the courtship disorder cluster (voyeurism, exhibitionism, toucheurism–frotteurism, and preferential rape).

Materials

The questionnaire measures in this reanalysis assessed sexual orientation, transvestism, autogynephilia, fetishism, the desire for sex reassignment surgery, and feminine gender identity. All of these measures, including those that are multi-item scales, were dichotomized for this study. As a strategy to avoid “over-fitting” the data, I used the same cutting score for the variables, sexual orientation, transvestism, autogynephilia, and fetishism, as I had used in an earlier analysis (Blanchard, 1991). For the same reason, I tried only one way of dichotomizing the items for surgery and female identity, rather than making any attempt to adjust such recoding to obtain the strongest or the cleanest results.

Sexual Orientation

This was assessed with the Modified Androphilia-Gynephilia Index (Blanchard, 1985). Sample items from this scale are: “About how old were you when you first felt sexually attracted to males?” “In your sexual fantasies, are females age 17–40 always, or almost always, involved?” Patients who obtained scores less than 10 were classified as heterosexual and those with scores greater than or equal to 10 were classified as homosexual.

Transvestism

This was assessed with the Cross-Gender Fetishism Scale (Blanchard, 1985). Its items include the following: “Has there ever been a period in your life of one year (or longer) during which you always or usually masturbated if you put on female underwear or clothing?” “Have you ever felt sexually aroused when putting on women’s perfume or make-up, or when shaving your legs?” A subject was classified as transvestic if he endorsed any item on this scale.

Autogynephilia

This trait was measured with the Core Autogynephilia Scale (1989b). The scale’s items include these: “Have you ever become sexually aroused while picturing yourself having a nude female body or with certain features of the nude female form?” “Have you ever been sexually aroused by the thought of being a woman?” A patient was classified as autogynephilic if he obtained a score of three or higher on this instrument.

Fetishism

As in Blanchard (1991), a patient was classified as fetishistic if he responded positively to the single questionnaire item, “Do you think that certain inanimate objects (velvet, silk, leather, rubber, shoes, female underwear, etc.) have a stronger sexual attraction for you than for most other people?”

Desire for Sex Reassignment Surgery

The patient was classified on this variable according to the single questionnaire item, “Have you ever wanted to have an operation to change you physically into a woman?” Patients who endorsed the response-option “Unsure” were grouped with those who responded “No.”

Female Gender Identity

This attribute was assessed with the single questionnaire item, “Have you ever felt like a woman?” Patients were scored positively if they endorsed the response-option, “At all times and for at least one year.” Patients were scored negatively if they indicated that they had never experienced such feelings or that they experienced them intermittently.

Results

Nine of the 427 patients were missing data on sexual orientation. Of the remainder, 292 were classified as heterosexual, and 126 as homosexual. A history of transvestism was admitted by 247 (85%) of the heterosexual patients and denied by the other 45 (15%). There were no obvious demographic differences between the heterosexual patients who admitted and those who denied transvestism. The mean age of the admitters was 33.38 years (SD = 9.53), and that of the deniers was 33.22 years (SD = 9.10), t(290) = 0.11, ns. The mean education of the admitters was 5.19 (SD = 1.23), where “5” equaled “at least 12 grades completed but no university,” and that of the deniers was 4.93 (SD = 1.12), where “4” equaled “more than 8 grades completed but less than 12,” t(290) = 1.29, ns.

The numbers of admittedly transvestic patients who acknowledged histories of fetishism, autogynephilia, both, or neither are shown in Table 2. There is nothing in these results to suggest that either of the proposed diagnostic specifiers should be eliminated. Only 7.7% of cases denied both fetishism and autogynephilia. Almost half of this sample (49%) acknowledged histories of both fetishism and autogynephilia.

Table 2 Numbers and percentages of transvestites who acknowledged fetishism or autogynephilia
Full size table

Two binary logistic regression analyses were used to investigate the potential clinical significance of these specifiers. These were carried out on 244 patients, because 3 patients were missing the questionnaire section pertaining to gender dysphoria.

The criterion variable in the first analysis was whether the patient had ever wanted sex reassignment surgery. The two predictors were the patient’s self-reported history of autogynephilia and his self-reported history of fetishism. Both predictors were entered directly into the equation. The results are shown in Table 3.

Table 3 Prediction of the desire for sex reassignment
Full size table

These results show that transvestic patients who acknowledged autogynephilia had almost five times higher odds of reporting past or current desires for sex reassignment than transvestic patients who denied autogynephilia. The opposite result was found for fetishism, that is, transvestites who reported fetishism were less likely to report a desire for sex reassignment. It is noteworthy that these predictors were independent to a large extent. Both were highly significant when the other was controlled for.

The criterion variable in the second analysis was whether the patient had an unwavering female identity. The predictor variables were the same as in the first analysis, and the equation was built in the same way. The results are shown in Table 4.

Table 4 Prediction of constant female identity
Full size table

The results were similar to those from the first analysis. Transvestic patients who acknowledged autogynephilia had over four times higher odds of reporting unwavering female identities than transvestic patients who denied autogynephilia. Transvestites who reported fetishism were less likely to report constant female identities. Fetishism was a stronger predictor than autogynephilia even though it was assessed with a single questionnaire item.

Discussion

There is no doubt that questionnaire data canvassing gender identity, desires for sex reassignment, fetishism, and so on are strongly influenced by many extraneous factors, for example, patients’ desires to present themselves in a favorable light for obtaining sex hormones or reassignment surgery, embarrassment regarding unusual or bizarre sexual practices, and misunderstanding or incomprehension of questionnaire items. On the other hand, the results obtained with these crude and obvious measures were quite strong, and they did clearly suggest that the addition of the proposed specifiers to the diagnosis of transvestic disorder could provide clinically meaningful information as well as data useful for research.