Multiple Paraphilic Diagnoses among Sex Offenders

Gene G. Abel, MD; Judith V. Becker, PhD; Jerry Cunningham-Rathner, BA; Mary Mittelman, DrPH; and Joanne-L. Rouleau, PhD

The psychiatric literature suggests that paraphiliacs can be expected to partic- ipate in only one type of deviant sexual behavior. Using self-reports gathered with assured confidentiality from 561 nonincarcerated paraphiliacs, we discovered that most paraphiliacs have had significant experience with as many as ten different types of deviant sexual behavior without regard, in many cases, to gender, age, and familial relationship of the victim. The relevance of these findings to our un- derstanding of paraphiliacs and their treatment is discussed.

Sex crimes are a major social problem. portant research. If the accelerating in- Despite changing cultural trends toward cidence of reported sex crimes and the open discussion of sexual behavior, in- attendant victimizations are to be re- formation about sex crimes and the in- duced, however, the psychopathology dividuals who commit them remains of the perpetrators must be examined. sketchy and inconclusive. Psychiatry, The best source of accurate infor- psychology, and sociology have tended mation about participation in deviant to avoid the study of sex offenders, per- sexual behaviors is the paraphiliacs haps because they are viewed with dis- themselves. However, a major factor dain by all levels of society. Instead, the inhibiting the collection of accurate in- focus has been on the victims of sex formation from paraphiliacs is the fear crimes-certainly worthwhile and im- of negative social and legal repercus- sions because of the lack of assured confidentiality. The paraphiliac be- Dr. Abel is Professor of Psychiatry and Dr. Rouleau lieves that valid reporting of his deviant is Assistant Professor of Psychology, Department of Psychiatry, Emory University, Atlanta, GA. Dr. behavior is likely to increase the prob- Becker is Associate Professor of Psychology, Ms. Cun- ningham-Rathner is Assistant Instructor of Psychol- ability of arrest for crimes unknown to ogy, and Dr. Mittelman is Associate Research Scien- others, to prolong his incarceration, or tist, Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NY. to jeopardize his probation status. Fur- Address reprint requests to Dr. Abel, Behavioral Med- thermore, most states have laws that re- icine Institute, 3193 Howell Mill Rd., Suite 202, At- lanta, GA 30327. quire the reporting of some sex crimes This research was supported by the National Institute (e.g., child molestation) to law enforce- of Mental Health, Grant ROlMH33678, "The Evalu- ation of Child Molesters," awarded to Dr. Gene G. ment authorities. Therefore, paraphi- Abel. The authors gratefully acknowledge the editorial assistance of Larry Brown in the preparation of this liacs are reluctant to discuss the true manuscript. scope of their deviant behavior with

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 153 Abel et a/.

others. The key issue, then, to obtain- tained from incarcerated paraphiliacs, ing valid and reliable information from these studies reported that each had paraphiliacs in order to facilitate as- participated, on the average, in fewer sessment and treatment is an assurance than two different types of paraphilia. of confidentiality. Information gathered under a Certifi- Information currently available con- cate of C~nfidentiality,',~.~however, re- cerning paraphiliacs has come generally vealed that paraphiliacs have often been from incarcerated offenders who are involved in many more paraphilias than also unlikely to report accurately their previously suspected. To gain a better various deviant sexual interests and be- understanding of the number of differ- haviors. It is well known among incar- ent paraphilias in which the paraphiliac cerated paraphiliacs that if the nature of may participate and how the various their crimes becomes known to fellow paraphilic diagnoses interrelate, the fol- inmates, the risk of violent reprisal will lowing study was undertaken. be substantially increased. Conse- Methods quently, what we know about sex of- fenders as a population has been lim- Subjects The study population in- ited. cluded 561 men seeking voluntary eval- Recent studies,'-3 relying on a Cer- uation andlor treatment for possible tificate of Confidentiality from the fed- paraphilia at the University of Tennes- eral government4 that protects the iden- see Center for the Health Sciences, tity and confidentiality of research Memphis, Tennessee, or at the New subjects (in this case, self-reports ob- York State Psychiatric Institute, New tained from paraphiliacs), have re- York City. At the former site, all cate- vealed a marked discrepancy between gories of paraphilia were evaluated; at information in the literature regarding the latter, subjects with a diagnosis of paraphiliacs and information gathered rape and/or child molestation, because from paraphiliacs with the above assur- of preselection, constituted the largest ance of confidentiality. The literature segment of the subject population. Ap- suggests that the average sex offender proximately one third of the subjects commits fewer than two crimes5-'; were referred through mental health more recent studies' indicate that a rap- routes, one third from legal or forensic ist may commit as many as seven rapes, sources, and one third from other that a pedophile molesting young boys sources.' may commit an average of 240 such Diagnostic Criteria Interviews were crimes, and that exhibitionists, frot- conducted over an eight-year period teurs, and voyeurs commit an average (1977 to 1985), during which time the of over 500 paraphilic acts each. Diagnostic and Statistical Manual of Previous studies5**have also under- Mental Disorders of the American Psy- estimated the number of different types chiatric Association was revised from of paraphilia in which sex offenders DSM-I1 to DSM-111. Both DSM-I1 and may participate. With information ob- DSM-I11 describe nearly all of the char-

154 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses acteristics of paraphilias appearing in paraphilia by relying upon the DSM-I1 this study, i.e., unusual or bizarre im- or DSM-I11 criteria, therefore, would agery or acts that tend to be insistently undermine any investigation into the and involuntarily repetitive, generally coexistence of multiple paraphilic di- involving the preference for the use of agnoses in a single subject. Regardless a nonhuman object for sexual arousal, of what proportion of sexual arousal re- repetitive activities with humans in- sulted from paraphilic interests or fan- volving real or simulated suffering or tasies, the victimization and the attend- humiliation, or repetitive sexual activi- ant consequences still occurred. This ties with nonconsenting partners. All study describes individuals who, at subjects reported recurrent, repetitive times, participated in nonparaphilic be- urges to carry out these deviant sexual haviors without relying on paraphilic behaviors; subjects were not included thoughts, and, at other times, used par- simply because they had committed the aphilic thoughts to develop fantasy, behavior. erection, and/or behaviors. Thus, our Diagnostic problems arose, however, diagnostic criteria in this one respect when the criteria for diagnosis using varied with the diagnostic criteria out- DSM-I1 and DSM-I11 indicated that the lined in DSM-I1 and DSM-111. subject's predominating sexual activity Based upon the subjects' reported life must involve paraphilic behavior history, we categorized each paraphilic (DSM-11), or that sexual excitement is diagnosis by gender and age of target possible only when the individual fan- (less than 14, 14 to 17, and more than tasizes or uses paraphilic images or be- 17 years of age) except in paraphilias haviors to become sexually excited where such divisions were irrelevant (DSM-111). We discovered that these (e.g., bestiality, coprophilia, some fe- latter criteria were inconsistent with our tishes). The majority of targets of the clinical experience with individuals who various paraphilic interests were fe- repetitively carry out paraphilic acts. male, except in cases of child moles- The majority of paraphiliacs in our sub- tation involving assaultive behavior. ject population could become involved Categories were subsequently col- with adult partners without relying lapsed across genders except in the cat- upon paraphilic fantasies and behav- egory of child molestation. The familial iors. Many indicated that they preferred relationship of the perpetrator to the paraphilic fantasies or behaviors to non- target was also initially categorized but paraphilic sexual behaviors, but both was subsequently ignored for all diag- paraphilic and nonparaphilic sexual be- noses except child molestation. The havior clearly coexisted in most of the number of pedophiles involved in in- subjects. cestuous activities was quite significant In most cases, the DSM-I1 or DSM- and appeared to warrant discrimination I11 definition of paraphilia excludes the from nonincestuous child molestation. possibility of multiple, concomitant The behaviors characterizing the 21 paraphilias. To determine a diagnosis of categories of paraphilias (see Table 2)

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, Abel et a/. are described in DSM-I1 and DSM-111. verse consequences. Subjects were The public masturbator differed from given the opportunity to discuss their the exhibitionist in that the former mas- concerns and to ask questions before turbated in a public setting but made no signing the consent form. attempt to expose his penis to his target. All subjects underwent a structured The exhibitionist, by contrast, became clinical interview'~1° focusing on spe- maximally aroused by exposing his cific demographic characteristics, num- penis to an unsuspecting target. Uro- bers and types of deviant acts, and num- lagnia involved sexual gratification as ber of victims. The subjects were again the perpetrator urinated on his victim reminded of the confidentiality of the or was urinated upon by others. Two data and the voluntary nature of their subjects were seen who were aroused participation, and told that they were by specific odors associated with men free to withdraw at any time for any rea- or women and appeared to be distinct son. If subjects were easily able to recall from coprophiliacs in that the odor to and describe the variety and complexity which they were particularly attracted of their paraphilic interests, the struc- did not involve feces. tured clinical interview lasted approxi- Three nonparaphilic categories of mately one hour. If subjects experi- sexual behavior (rape of adult women, enced difficulty, the interview lasted up transsexualism, and ego-dystonic ho- to five hours. mosexuality) were included in the sam- Numerous problems arose during the ple to investigate whether these cate- clinical interviews. First, it was some- gories of sexual behavior correlated times difficult for subjects to describe with any traditional paraphilic behav- the frequency of their involvement with ior. By including these three nonpara- paraphilias that had developed much philic categories, it was hoped that fur- earlier in their lives. To assist the sub- ther scientific data could be gathered to jects with the chronology of their de- substantiate or refute the inclusion of viant arousal and to improve the valid- these three categories of behavior in the ity of their reports, interviewers paraphilia category. attempted to associate important events Procedure In an effort to minimize in the subjects' lives with the onset and the subjects' attempts to conceal their frequency of their paraphilias. deviant behavior, each subject viewed Second, some subjects reported sev- a one-hour, videotaped presentation eral paraphilias occurring simultane- that explained the human investigation ously but at different frequencies. Such aspects of the study, the confidentiality reports tended to indicate that, as one of the data, and the protection afforded paraphilia rose to become the predom- to him under the Certificate of Confi- inating deviant behavior, others would dentiality. Subjects were informed that become less dominant. This wave effect their participation was strictly volun- continued as another paraphilia rose to tary and that they were free to withdraw dominance and other deviant interests from the study at any point without ad- diminished. Special care was taken

156 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses under these circumstances to separate Results the occurrences and frequencies of the The age range of the study population various paraphilias. of 561 males was 13 to 76 years, with a Third, the validity of the frequency mean age of 31.5. Of these, 67 percent of past deviant behaviors was a major fell into the age range of 20 to 39 years concern. Verification of the subjects' of age. Approximately half of the study reported frequency of paraphilic acts population were single and the remain- with arrest records proved to be inef- ing half either were married, had been fective because the subjects' reported married, or had formed a significant, frequencies of deviant behavior were "living with" relationship with an adult substantially higher than the number of partner. The ethnic distribution was actual arrests. For example, the ratio of 62.1 percent white, 23.8 percent black, arrest to reported commission of the vi- and 11.2 percent Hispanic. The major- olent crimes of rape and child moles- ity of the participants had completed tation was approximately 1 :30 and the high school and 40 percent had com- ratio of arrest to reported commission pleted at least one year of college. of the less violent crimes of exhibition- Nearly two thirds (65%) of the subjects ism and voyeurism was approximately were fully employed and earned annual 1: 150. When the subject reported a incomes in the range of $7,500 to range in the frequency of his deviant be- $25,000. Approximately 30 percent havior (e.g., three to five times a were referred for evaluation through month), the lowest value was selected legal or judicial sources, and 30 percent so that the data would reflect minimal were referred by mental health profes- frequencies of deviant behavior. Only sionals. The remainder were referred by incidents of deviant behavior as re- other sources, including self-referral. ported by the subject himself were in- To determine a paraphilic diagnosis, cluded. For example, if a family mem- the data were analyzed according to the following four subcategories: (1) gender ber reported five incidents of deviant of target, (2) age of target, (3) inces- behavior but the subject reported only tuous versus nonincestuous behavior, two, then the interviewer recorded two and (4) assaultive versus nonassaultive acts. Conversely, if the subject reported behavior. All paraphilic diagnoses were ten acts of deviant behavior but his ar- classified by each of these subcatego- rest record reflected only three, then ries and the frequency of occurrence of the interviewer recorded ten acts. each subcategory was determined. It Fourth, the possibility of overreport- should be noted that, in order to be in- ing was addressed. In the few instances cluded in a diagnostic category, a sub- of suspected overreporting, the inter- ject must have reported an overt act in view was repeated until data were con- that category; deviant arousal alone sistent. If repeated interviews failed to was not sufficient. yield consistency, the subject's data Gender of Target In our total popu- were not included in the study. lation of 561 subjects, we observed that

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 157 Abel et a/.

Table 1 (8.4%) participated in deviant behavior Target Age that was not classifiable according to No. of % of age of target, e.g., fetishism. Analysis Subjects Subjects of these data reveals that 239 subjects Child, adolescent, (42.3%) targeted victims in at least two adult Chiid and age groups. adolescent Incestuous versus Nonincestuous Be- Child and adult havior It is commonly thought that Adolescent and adult paraphiliacs offend either against their Child only family members or against nonfamily Adolescent only Adult only members, but rarely against both. In Not applicable our total population of 561 subjects, we Total observed that 3 15 (56.1%) participated in nonincestuous deviant behavior only and 68 (12%) participated in incestuous 377 (67.2%) targeted only females and deviant behavior only. Again, 47 sub- 67 (11.9%) targeted only males. Five jects participated in deviant behavior subjects participated in deviant behav- that was not classifiable in this subcat- ior that was not classifiable in this sub- egory, e.g., fetishism. A total of 131 category, e.g., fetishism. A total of 112 (23.3%) subjects offended against both (20%) subjects offended against both family and nonfamily targets, indicating male and female targets, indicating that that this subgroup of the study popu- this subgroup of the study population lation participated in deviant behavior participated in deviant behavior irres- irrespective of familial relationship. pective of gender of target. Assaultive versus Nonassaultive Behav- Age of Target Another important ior Subjects were next categorized by area of concern was age of the victims, assaultive or nonassaultive paraphilic i.e., whether a paraphiliac who offends behavior to determine whether para- against a young child would also com- philiacs who committed assaultive acts mit acts against an adolescent and/or an (child molestation, rape, frottage) also adult, and vice versa. Subjects were di- carried out nonassaultive paraphilic be- vided into three categories of age of tar- haviors (public masturbation, voyeur- get: children less than 14 years of age, ism, exhibitionism) and vice versa. In adolescents 14 to 17 years of age, or our total population of 561 subjects, we adults (more than 17 years of age) observed that 331 (59%) participated in (Table 1). In our total population of 561 assaultive deviant behavior only and 84 subjects, we observed that 275 subjects (14.9%) participated in nonassaultive (49%) targeted victims in only one age deviant behavior only. However, 146 group, 176 subjects (3 1.3%) targeted (26%) subjects participated in both as- victims in two age groups, and 63 sub- saultive and nonassaultive deviant be- jects (11.2%) targeted victims in all havior indicating that this subgroup of three age groups. Forty-seven subjects the study population offended against

158 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses targets involving both touching and in which our subjects could have been nontouching behaviors. involved. (For this and subsequent The foregoing analyses indicate that analyses, pedophiles targeting children a significant percentage of paraphiliacs less than 14 years of age and those tar- cross gender, age, familial, and assaul- geting adolescents aged 14-17 were tive/nonassaultive behavior boundaries combined into a single category.) during the commission of paraphilic The percentage of subjects in each of acts. the 21 diagnostic categories who had We next examined the interrelation- one or more paraphilic diagnoses ap- ship between paraphilic diagnoses. The pears in Table 2. Excluding infrequently full impact of the results of this study seen categories having less than 12 sub- emerges when the crossings of deviant jects per category (such as obscene behaviors are combined in order to de- mail, urolagnia, coprophilia, and attrac- termine multiple paraphilic diagnoses. tion to specific odors), subjects in all Number of Paraphilic Diagnoses by Di- other diagnostic categories had histo- agnostic Category Our clinical experi- ries of numerous other separate para- ence with this population of 561 sub- philic diagnoses. There were at least a jects indicates that, when multiple few subjects in each category who had paraphilias exist in the same subject, as many as 10 different paraphilic di- one paraphilia initially takes domi- agnoses. nance. A second paraphilia develops Because of preselection, it was as- and overtakes the first in dominance, sumed that the relative occurrence of and then continues for a number of paraphiliacs in the sample was unrepre- months or years, while the first contin- sentative of the general population, ex- ues at a greatly reduced intensity. On cept for the relative occurrence of the rare occasions, the initial paraphilia will subcategories of child molesters. To appear to lose its arousal properties en- avoid skewing of the data by overre- tirely for the patient and essentially dis- presentation of some groups (e.g., child appear. Because our role was to deter- molesters) and underrepresentation by mine the existence of more than one others (e.g., voyeurs) and to obtain a paraphilia in the same subject, we less biased appraisal of the number of elected to count all paraphilias that had different paraphilias in the "average occurred during his lifetime, even paraphiliac," the percentage of each of though some were no longer actively the 21 paraphilic classifications with arousing or erection producing. The one or more paraphilias was calculated temporal relationship among the var- and then averaged across all 21 para- ious paraphilic diagnoses awaits further philic categories. This averaged inci- analysis. dence of concomitant or nonconcomi- Clinical interviews revealed that a tant paraphilic diagnoses is reflected in number of subjects were involved in dif- Figure 1. Only 10.4 percent of these ferent paraphilias during their lives. "average paraphiliacs" had one diag- There were 21 categories of paraphilia nosis, 19.9 percent had two diagnoses,

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 159 Abel et a/.

Table 2 Percentage of Paraphiliacs with Multiple Paraphilias

% of Paraphiliacs Diagnosis I* 2* 3* 4' 5' 6* 7* 8' 9' 10'

Pedophilia 15.2 23.7 19.2 14.7 9.4 4.5 6.7 3.1 1.3 2.2 (nonincestuous), female target Pedophilia 19.0 26.8 19.6 12.4 4.6 3.9 6.5 3.9 .7 2.6 (nonincestuous), male target Pedophilia 28.3 25.8 17.0 5.7 8.2 3.8 5.0 1.9 .6 3.8 (incestuous), female target Pedophilia 4.5 15.9 20.5 18.2 13.6 6.8 9.1 2.3 .O 9.1 (incestuous), male target Rape 27.0 17.5 19.0 12.7 7.1 3.2 7.9 1.6 1.6 2.4 Exhibitionism 7.0 20.4 22.5 15.5 7.0 7.0 9.2 4.9 2.8 3.5 Voyeurism 1.6 9.7 27.4 14.5 12.9 8.1 11.3 8.1 3.2 3.2 Frottage 21.0 16.1 12.9 16.1 11.3 3.2 12.9 3.2 .O 3.2 Obscene mail .O 33.3 66.7 .O .O .O .O .O .O .O Transsexualism 51.7 31.0 13.8 3.4 .O .O .O .O .O .O Transvestitism 6.5 29.0 29.0 9.7 .O 6.5 12.9 .O 6.5 .O Fetishism .O 15.8 21.1 15.8 26.3 5.3 10.5 .O 5.3 .O Sadism .O 17.9 28.6 14.3 14.3 3.6 3.6 3.6 7.1 7.1 Masochism .O 41.2 11.8 5.9 11.8 5.9 5.9 5.9 5.9 5.9 Homosexuality 25.0 41.7 25.0 4.2 .O .O .O 4.2 .O .O Obscenephonecalling 5.3 5.3 21.1 21.1 5.3 10.5 15.8 5.3 5.3 5.3 Public masturbation 5.9 17.6 .O 17.6 17.6 17.6 5.9 5.9 5.9 5.9 Bestiality .O 28.6 7.1 14.3 14.3 7.1 14.3 .O 14.3 .O Urolagnia .O .O .O 25.0 .O 25.0 .O .O 25.0 25.0 Coprophilia .O .O .O .O .O .O .O 50.0 25.0 25.0 Arousal to odors .O .O 50.0 .O .O .O .O 50.0 .O .O

Refers to number of paraphilias.

20.6 percent had three diagnoses, and highest percentage (52%) of individuals 11.5 percent had four diagnoses. The re- with only one paraphilia were those in- maining 37.6 percent were concomi- volved with transsexualism. In each of tantly or nonconcomitantly involved in the other 17 categories of paraphilias five to ten different paraphilic behav- presented, less than 30% of the subjects iors. confined their deviant behavior to only The percentage of cases in each di- one paraphilia. In 10 categories of par- agnostic category with only one para- aphilia, less than 10 percent of subjects philia is indicated in Table 3. Categories had participated in only one type of par- containing fewer than 12 subjects were aphilic behavior. It was especially im- excluded because the number of sub- pressive that no cases of fetishism, sad- jects was considered too small to rep- ism, masochism, or bestiality were seen resent such paraphilias reliably. The in which an individual had only one par-

160 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses

havior, the likelihood of the possible presence of other paraphilias relative to that indicated in column 1 is indicated in Table 4, a cross-diagnosis table. (Note: Table 4 should be read from left to right and not from top to bottom.) To ensure that the number of paraphiliacs who had multiple paraphilic diagnoses indeed reflected more than a simple de- viant sexual arousal, data reflect the number of subjects who had reported actual commission of different para- philic acts. Table 4, Column 1, includes the possible paraphilic diagnoses. Col- umns 2 through 22 represent additional O1llllaatatl 012345678910 concomitant or nonconcomitant para- Number of Paraphilias philic behaviors that subjects had com- Figure 1. Average number of paraphilias. mitted. By reading across row 1, for ex- ample, one sees in Column 2 that 100 percent (224 subjects) represents the aphilic diagnosis. These results clearly total subsample of men involved with indicate that paraphiliacs with only one girls outside the home. Reading further paraphilia are rare. to the right, 35 percent of these indi- The average number of different par- viduals were (or had been) also involved aphilias found in the histories of the 561 in male nonincestuous pedophilia, 35 subjects in our study population is percent in female incestuous pedophi- shown in Table 3. Except for those with lia, 12 percent in male incestuous pe- a diagnosis of transsexualism or ego- dophilia, 25 percent in rape, 29 percent dystonic homosexuality, the average in exhibitionism, 14 percent in voyeur- number of paraphilias by diagnosis was ism, 11 percent in frottage, and so forth. in the range of three to five paraphilias Therefore, from the clinical standpoint, per diagnostic category. Especially re- 35 percent of paraphiliacs involved in vealing was that multiple paraphilias nonincestuous deviant behavior with were very common in individuals tra- female targets have a high probability ditionally considered to have only one of having been involved with noninces- paraphilia, i.e., female-targeted incest tuous deviant behavior with male tar- pedophilia and male-targeted incest pe- gets. By examining each row, one can dophilia. see the frequent history of cross-diag- To understand which paraphiliacs nostic behavior that exists in paraphi- have histories of having committed liacs. For those diagnoses with fewer other specific types of paraphilic be- than 12 subjects per category, i.e., ob-

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 161 Abel et a/.

Table 3 Percentage of Subjects with Only One Diagnosis and Average Number of Paraphilias per Subject

No. of % with only Average No. Total Diagnosis Subjects' 1 Diagnosis of Paraphilias Paraphilias

Pedophilia (nonincestuous), 806 female target Pedophilia (nonincestuous), 520 male target Pedophilia (incestuous), 493 female target Pedophilia (incestuous), 198 male target Rape 41 6 Exhibitionism 596 Voyeurism 298 Frottage 236 Transsexualism 49 Transvestitism 118 Fetishism 84 Sadism 129 Masochism 75 Homosexuality 55 Obscene phone calling 97 Public masturbation 87 Bestiality 67 Total 4,324

* A subject is included in each diagnostic category in which he reported a completed act of paraphilic behavior. Therefore, overlapping of subjects across categories occurs.

scene mail, urolagnia, coprophilia, and jects who reported involvement with fe- arousal to specific odors, these data male incestuous pedophilia, 49 percent should be interpreted cautiously. With had histories of also having been in- the exception of these four categories volved in female nonincestuous pedo- and transsexualism, all categories of philia, 12 percent in male nonincestuous paraphilia had large percentages of sub- pedophilia, and 12 percent in male in- jects who had also participated at one cestuous pedophilia. Of the 44 subjects time or another in other types of para- who reported involvement with male in- philic behavior. cestuous pedophilia, 61 percent had his- Especially impressive were results in tories of also having been involved with the categories of child molestation. Of female nonincestuous pedophilia, 68 the 153 subjects involved with boys out- percent with male nonincestuous pe- side the home, 51 percent had histories dophilia, and 43 percent with female in- of also having been involved with girls cestuous pedophilia. From these data, outside the home, 12 percent with girls it becomes apparent that child molest- within the home, and 20 percent with ers have a very high incidence of de- boys within the home. Of the 159 sub- viant behavior with both family and

162 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses nonfamily targets. Furthermore, these other deviant behaviors in this subsam- data suggest that, contrary to traditional ple of paraphiliacs. belief, incestuous child molesters are or Voyeurs, frotteurs, and fetishists all have been involved very frequently revealed histories of paraphilic interests with children outside the home. in other major categories of paraphilia. Rapists also demonstrated a high in- Sadists and masochists also appear to cidence of concomitant or nonconcom- have or have had experiences with itant cross diagnosis. Of the 126 sub- other types of paraphilia. Of greatest jects who had raped an adult woman, concern is that 46 percent of sadists re- 44 percent had also been involved in fe- ported involvement with rape behavior, male nonincestuous pedophilia, 14 per- the highest percentage of cross-diag- cent in male nonincestuous pedophilia, nosis into the rape category of any other and 24 percent in female incestuous pe- category of paraphilia. dophilia; 28 percent had histories of ex- Ego-dystonic homosexuals, obscene hibitionism, 18 percent of voyeurism, phone callers, and public masturbators and the remainder had been involved, all had significant histories of involve- to a lesser degree, in other types of par- ment with other paraphilias. It is diffi- aphilia. cult to draw conclusions from infre- Exhibitionists had a high degree of quently seen paraphiliacs (e.g., senders other concomitant or nonconcomitant of obscene mail) because these cate- paraphilic behaviors in addition to ex- gories of paraphilia occurred at a low hibitionism. Forty-six percent had been frequency in the study population and, involved in female nonincestuous pe- therefore, these findings need corro- dophilia, 22 percent in male noninces- boration from studies of larger subject tuous pedophilia, 22 percent in female populations. incestuous pedophilia, 25 percent in In summary, examination of Table 4 rape, 28 percent in voyeurism, and 16 suggests that our traditional view of percent in frottage. Smaller proportions paraphiliacs has been somewhat naive. of exhibitionists had been involved in With the exception of transsexuals, the other categories of paraphilia. An there is a significant incidence of cross- impressive aspect of these findings was ing of deviant sexual behaviors. that, contrary to some traditional texts, Conclusions the exhibitionists evaluated frequently The principal conclusion to which were involved in a large variety of other this study points is that paraphiliacs fre- paraphilic behaviors at one time or an- quently participate in a variety of dif- other, some overtly aggressive. Once ferent paraphilias and that the paraphi- again, these data do not indicate that all liac with a history of only one paraphilia exhibitionists have also been involved is rare. These assertions are at variance in other paraphilias, but it certainly sug- with the traditional view of the para- gests that there is a much higher like- philiac, i.e., as one who becomes fix- lihood of a history of involvement in ated on one type of paraphilia to the ex-

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 163 Table 4 Percentage of Cross-Diagnosis by Paraphilia

-I2 0 (I) 3 3 (I) -03 p2om a, f? Kgz$ .- .- c 0 .-KO$3 .-". -c 0 C u .-5 E,, 5 (I]=la r .so mo g*:~ .-E.z - 5.5 c .- (I] % .E = -O .- Ez a,.i 2.: .n.~~~g~r g o.? aza--a .-+;psm>-- (I)r!+J OK=.-0"O: kZz2 (I) 0 a,o 0 a,o a, ?2 L :: g % " " (I)5 a xoPo FFa, 0 425 a, L 0 2 22CT w > u o I- u V) I O~LEm 2 o a Female nonincestuous pedophilia Male nonincestuous pedophilia Female incestuous pedophilia Male incestuous pedophilia Rape

Exhibitionism

Voyeurism Frottage N 241210 414231462 0 0 137 0 0 4 2 100 1 ZE % 391916 7233723100 0 0 2 511 0 0 7 3 2 0 0 2 5 Obscene mail N 101011003000001000000 -o_ %3303303333001000000033000000 " Transsexualism N 0010010002990007000000 % 0040040001003100024000000 $ Transvestitism N 74726113109312346314210 % 23 13 23 7 19 36 10 3 0 29 100 7 10 13 19 10 3 13 7 3 0 -.- Fetishism N10653544300219122010100 % 533226 162621 21 16 0 0 11100 5 11 11 0 5 0 5 0 0 &In Sadism N117911367700312850422220 5 % 392532 446212525 0 011 410018 014 7 7 7 7 0 0 Masochism N 6651361000425172001220 g % 35 35 29 6 18 35 6 0 0 0 24 12 29100 12 0 0 6 12 12 0 Homosexuality N 1312030017620224001100 % 413 4 8 013 0 0 42925 8 0 8100 0 0 4 4 0 0 Obscene phone N 83507129400304001922010 calling % 42 16 26 0 37 63 47 21 0 0 16 0 21 0 0100 11 11 0 5 0 Publicmasturbation N 5 2 4 1412 5 2 0 0 1 1 2 0 0 217 1 1 10 % 291224 624712912 0 0 6 612 0 012100 6 6 6 0 Bestiality N10163356100402112114110 % 71 743 21 21 3643 7 0 0 29 0 14 7 7 14 7100 7 7 0 Urolagnia N 212202100021221011420 % 50 25 50 50 0 50 25 0 0 0 50 25 50 50 25 0 25 25 100 50 0 Coprophilia N 423103100010220111240 % 100 50 75 25 0 75 25 0 0 0 25 0 50 50 0 25 25 25 50 100 0 Arousal to odors N 000011210000000000002 % 0 0 0 0505010050 0 0 0 0 0 0 0 0 0 0 0 0100 Abel et a/.

clusion of other kinds of deviant sexual the scope of the various deviant sexual behavior. activities of each subject. Because these results challenge our A third and extremely important fac- traditional view of the paraphiliac, one tor is the issue of confidentiality. Be- might question the validity of these cause the majority of prior research on findings. Why have traditional inter- paraphilias has been conducted within views failed to reveal a variety of par- the prison setting, where confidentiality aphilias? One reason might be the is limited, it is not surprising that the amount of time required to obtain a sub- subjects reported a low incidence of ject's full history. When histories are multiple paraphilias. Using the various taken in the rush of forensic evaluation, means of maintaining confidentiality the interviewer is likely to focus upon described earlier, we found that sub- behaviors that have come to his or her jects would more honestly report their attention. The interviews used in this various paraphiliac behaviors if they study were time consuming (from one felt assured that negative repercussions to five hours in duration) but provided as a result of such revelations would not the subject with the opportunity to es- be forthcoming. Where confidentiality tablish rapport with the examiner and is minimal or nonexistent, concealment to acquire some degree of comfort with is common, and therefore histories will the interview. be less valid. A second factor might be the lack of The finding that histories of multiple specificity in the questions that inter- paraphilias are common and cross-di- viewers pose. Our early experience in- agnoses are a frequent occurrence should be of importance to clinical judg- dicated that suspected paraphiliacs did ments regarding paraphilias. For ex- not volunteer information relative to all ample, it is customarily believed that of their deviant sexual activities. Ini- cases of incest simply result from dis- tially, interviews focused on the behav- torted family dynamics and that indi- ior that had come to our attention. As viduals involved in incest are unlikely we gained greater experience and to be involved in other types of para- greater knowledge about the frequent philias. This conclusion, however, is in- occurrence of multiple deviant behav- consistent with the findings described iors, the interviewers were more aware above. These results and prior psycho- of the need to ask specific questions physiological investigation of inces- about other types of paraphilias. As in- tuous and nonincestuous pedophiles2." terviews became more structured, in- suggest that incest cases frequently in- terviewers worked from a listing of all volve individuals who participate in possible paraphilias and asked each nonincestuous activities in addition to subject about his participation in each. incestuous behaviors. Similarly, indi- Although the latter technique was viduals involved with children outside rather routine and nonspontaneous, it the home frequently have concurrent was critical for an accurate recording of involvement with children within the

166 Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Multiple Paraphilic Diagnoses home. Treatment plans and assessment a result of the first deviant act may rein- strategies must therefore investigate force the rationalization or acceptability both of these possibilities. The presence of the act for the perpetrator. Thus, he of more serious paraphilias in individ- may feel less inhibited about acting uals with "benign" paraphilias should upon other paraphilic fantasies. also be of concern. Some professionals Regardless of how one might inter- have considered exhibitionists, voy- pret the theoretical implications of the eurs, and fetishists as being rather be- results of this study, a number of factors nign, nuisance paraphiliacs. These re- suggest that these findings are repre- sults, however, suggest that some (but sentative of paraphiliacs seeking psy- not all) of the benign paraphilias may chiatric or psychological evaluation and actually lead to very aggressive behav- treatment. The large subject population iors and should not be viewed as reli- was collected at two different sites in ably benign. Further clinical relevance the United States, both offering assess- of these results is the indication that ment and treatment services for indi- treatment must be modified to incor- viduals seeking help in controlling their porate all of the various paraphilias that sexual behavior. Both sites emphasized an individual might have. If an individ- treatment for assaultive sex offenders ual involved in incestuous pedophilia is (child molesters and rapists), which is also involved with children outside of the major treatment emphasis for most the home, then treatment must focus on sex offender treatment programs the full scope of the individual's pedo- throughout the country. Furthermore, philic interests including the possibility the considerable consistency with of other, unsuspected paraphilias. which paraphiliacs reported involve- Without a thorough investigation of an ment in multiple paraphilias (regardless individual's deviant interests and the of an initial category of paraphilia) ap- formulation of a comprehensive treat- pears to support the validity of these ment program, recidivism is much more new findings. likely. Prevention of sexual violence and de- Finally, the theoretical relevance of viant sexual acts necessitates a better these findings must be considered. Why appreciation of the perpetrator-who do paraphiliacs have histories of so he is and what his deviant interests many different paraphiliac interests and are-so that appropriate services may behaviors? If a specific conflict ac- be provided to control sexually deviant counted for one discreet type of para- behavior and to prevent further victim- philia, how could it explain the exis- ization. tence of multiple paraphilias in the same individual? Because paraphilic behav- References ior is, by and large, a secretive event, Abel GG, Becker JV, Mittelman M, et al: and because most paraphilic acts re- Self-reported sex crimes of non-incarcer- ated paraphiliacs. Journal of Interpersonal main unreported, it may be that failure Violence 2:3-25, 1987 to experience aversive consequences as Abel GG, Mittleman MS, Becker JV: Sexual

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988 Abel et a/.

offenders: results of assessment and rec- Washington, DC: U.S. Government Printing ommendations for treatment, in Clinical Cri- Office. 1977- 1980 minology: Current Concepts. Edited by 8. Frisbie LV, Vanasek FJ, Dingman HF: The Ben-Aron MH, Hucker SJ. Webster CD. self and the ideal self: methodological study Toronto, M & M Graphics, 1985 of pedophiles. Psycho1 Rep 20:699-706. 3. Abel GG, Rouleau JL: Sexual disorders, in 1967 Medical Basis of Psychiatry. Edited by Wi- 9. Abel GG, Rouleau J, Cunningham-Rathner nokur G, Clayton P. Philadelphia: Saunders J: Sexually aggressive behavior. in Modem (in press) Legal Psychiatry and Psychology. Edited by 4. Protection of Identity-Research Subjects. Curran W, McGany AL. Shah SA. Phila- 40 Fed Reg 234 (December 4, 1975) delphia, Davis, 1985 5. Gebhard PH;Gagnon JH. Pomeroy WB, et 10. Abel GG: A clinical evaluation of possible al: Sex Offenders. New York. Harper & sex offenders, in The Incest Offender, The Row, 1965 Victim, The Family: New Treatment Ap- 6. Flanagan TJ, van Alstyne DJ, Gottfredson proaches. White Plains, NY, The Mental MR (eds): Sourcebook of Criminal Justice Health Association of Westchester County, Statistics-1981. U.S. Law Enforcement 1985 Assistance Administration, National Crim- 11. Abel GG, Becker JV, Murphy WD, et al: inal Justice Information and Statistics Ser- Identifying dangerous child molesters, in Vi- vice. Washington, DC, U.S. Government olent Behavior: Social Learning Ap- Printing Office, 1982 proaches to Prediction, Management, and 7. U.S. Department of Justice: Uniform Crime Treatment. Edited by Stuart R. New York, Reports (Federal Bureau of Investigation). BrunnerIMazel, 1981

Bull Am Acad Psychiatry Law, Vol. 16, No. 2, 1988