P1: FYJ

Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

Sexual Abuse: A Journal of Research and Treatment, Vol. 14, No. 4, October 2002 (!
C 2002)

A DSM-IV Axis I Comorbidity Study

of Males (n = 120) With Paraphilias
and Paraphilia-Related Disorders
Martin P. Kafka1,3 and John Hennen2

One hundred and twenty consecutively evaluated outpatient males with paraphil-
ias (PAs; n = 88, including 60 sex offenders) and paraphilia-related disorders
(PRDs; n = 32) were systematically assessed for certain developmental vari-
ables and DSM-IV-defined Axis I comorbidity. In comparison with the PRDs,
the PA group was statistically significantly more likely to self-report a higher in-
cidence of physical (but not sexual) abuse, fewer years of completed education, a
higher prevalence of school-associated learning and behavioral problems, more
psychiatric/substance abuse hospitalizations, and increased employment-related
disability as well as more lifetime contact with the criminal justice system. In
both groups, the most prevalent Axis I disorders were mood disorders (71.6%),
especially early onset dysthymic disorder (55%) and major depression (39%).
Anxiety disorders (38.3%), especially social phobia (21.6%), and psychoactive
substance abuse (40.8%), especially alcohol abuse (30%), were reported as well.
Cocaine abuse was statistically significantly associated with PA males ( p = .03).
There was a statistically significant correlation between the lifetime prevalence of
Axis I nonsexual diagnoses and hypersexual diagnoses (PAs and PRDs). The preva-
lence of retrospectively diagnosed attention deficit hyperactivity disorder (ADHD)
was 35.8%, the third most prevalent Axis I disorder. ADHD ( p = .01), especially
ADHD-combined subtype ( p = .009), was statistically significantly associated
with PA status. ADHD was statistically significantly associated with conduct dis-
order, and both of these Axis I disorders were associated with the propensity for
multiple PAs and a higher likelihood of incarceration. When the diagnosis of
ADHD was controlled, the differences reported above between PAs and PRDs
either became statistically nonsignificant or remained as only statistical trends.

1 Department of Psychiatry, McLean Hospital, Belmont, Massachusetts.

2 Biostatistics Laboratory, McLean Hospital, Belmont, Massachusetts.
3 To whom correspondence should be addressed at Department of Psychiatry, McLean Hospital, 115
Mill Street, Belmont, Massachusetts 02478; e-mail: mpkafka@aol.com.

349

1079-0632/02/1000-0349/0 !
C 2002 Plenum Publishing Corporation
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

350 Kafka and Hennen

Thus, ADHD and its associated developmental sequellae and Axis I comorbidities
was the single most common nonsexual Axis I diagnosis that statistically sig-
nificantly distinguished males with socially deviant sexual arousal (PAs) from a
nonparaphilic hypersexual comparison group (PRDs). Sex offender paraphiliacs
were more likely to be diagnosed with conduct disorder, alcohol abuse, cocaine
abuse, and generalized anxiety disorder. The prevalence of any ADHD in the sex
offender paraphiliacs was 43.3%, and nearly 25% of offenders were diagnosed
with ADHD-combined subtype.
KEY WORDS: attention deficit hyperactivity disorder; depression; paraphilias; paraphilia-related
disorder; sex offender; sexual addiction.

INTRODUCTION

Paraphilias (PAs), as operationally defined in DSM-III (American Psychiatric

Association [APA], 1980), DSM-III-R (APA, 1987), and DSM-IV (APA, 1994), are
sexual disorders characterized by repetitive, socially deviant expressions of inten-
sified sexual arousal and associated behaviors. In DSM-IV, paraphilic behaviors
must persist at least 6 months and be accompanied by significant adverse personal
or social consequences to reach the threshold for a psychiatric diagnosis. The
most common PAs are exhibitionism, voyeurism, pedophilia, sexual masochism
and sadism, fetishism, transvestic fetishism, frotteurism, and telephone scatologia.
In contrast to the paraphilic disorders, a group of nonparaphilic hypersexuality
disorders has also been identified. These nonparaphilic conditions, not yet recog-
nized by a common diagnostic nomenclature, have been designated as nonpara-
philic sexual addictions (APA, 1987; Carnes, 1990), sexual compulsions (Coleman,
1992), or paraphilia-related disorders (PRDs; Kafka & Hennen, 1999). The major
diagnostic characteristic that distinguishes PAs from PRDs is that PAs are socially
deviant sexual behaviors whereas the latter conditions are persistent disinhibited
forms of socially sanctioned heterosexual and homosexual behavior. Like PAs,
PRDs must be associated with significant distress or impairment and persist for at
least 6 months to meet a diagnostic threshold. In males, the commonly identified
PRDs are compulsive masturbation, pornography dependence, protracted promis-
cuity, telephone sex dependence, and severe sexual desire incompatibility (Kafka
& Hennen, 1999). Recently, an additional PRD, cyber-sex dependence has also
been identified (Cooper, Scherer, Boies, & Gordon, 1999).
There are only a few studies directly comparing PA with PRD males in de-
velopmental difficulties, sexual behaviors, and Axis I comorbidity (Kafka, 1997;
Kafka & Prentky, 1992a, 1994, 1998). In these reports, both of these sexual disor-
der subtypes are characterized by increased and time-consuming sexual fantasies,
urges, and activities (i.e., clinical hypersexuality), similar Axis I comorbid condi-
tions and similar responses to pharmacological intervention (Kafka, 1994, 2000;
Kafka & Prentky, 1992b). In addition, PRDs are prevalent in the longitudinal
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 351

history of PA males (Black, Kehrberg, Flumerfelt, & Schlosser, 1997; Kafka &
Hennen, 1999; Langevin et al., 1985; Prentky et al., 1989).
In a series of studies examining Axis I comorbidity disorders in male paraphil-
iacs and males with PRDs, the published data clearly suggest that Axis I comorbid
diagnoses are common. In particular, mood disorders, including major depres-
sion, bipolar affective disorder, dysthymic disorder, and anxiety disorders, espe-
cially phobic disorders including social phobia, are frequently diagnosed (Black,
1998; Kafka & Prentky, 1994, 1998; McElroy et al., 1999; Raymond, Coleman,
Ohlerking, Christenson, & Miner, 1999). In addition, psychoactive substance abuse
disorders, especially alcohol and cocaine abuse, and impulse disorders not oth-
erwise specified, including the atypical impulse disorders compulsive shopping
(Black, 1998) and reckless driving (Kafka & Prentky, 1998) have been identified
(Black, 1998; Kafka & Prentky, 1994, 1998; McElroy et al., 1999; Raymond et al.,
1999). In some reports, attention deficit hyperactivity disorder (ADHD) or conduct
disorder has been identified, particularly in association with males with socially de-
viant sexual arousal (Galli et al., 1999; Kafka & Prentky, 1998; Kavoussi, Kaplan,
& Becker, 1988; Vaih-Koch & Bosinski, 1999). The identification of such disor-
ders is potentially important inasmuch as there are clinical data suggesting that
pharmacological treatment addressing Axis I comorbid conditions can ameliorate
PAs and PRDs (Greenberg & Bradford, 1997; Kafka, 2000; Kafka & Hennen,
2000).
Kafka and Prentky (1998) reported that DSM-III-R-defined ADHD (retro-
spectively assessed) and cocaine abuse were statistically significantly more preva-
lent in paraphilic males in comparison to PRDs. In addition, PA males were more
likely than PRD males to report a childhood history of physical/sexual abuse,
lower educational achievement, more school-related learning, and behavioral prob-
lems, more extensive involvement with the criminal justice system, a history of
psychiatric hospitalization, and lower current income. In that report (Kafka &
Prentky, 1998), all but one of these variables (the exception being physical/sexual
abuse) were statistically significantly associated with the retrospective diagnosis
of ADHD.
In this study, we extend the previous research by ascertaining the prevalence
of lifetime Axis I disorders in a different and larger sample of PA and PRD males
utilizing DSM-IV criteria. In addition to the evaluation of standard mood, anxiety,
psychoactive substance abuse, psychotic and impulse disorders, NOS disorders,
conduct disorder, ADHD, and its DSM-IV-derived subtypes, were retrospectively
assessed. In this manuscript, the term male hypersexuality disorders is a synony-
mous reference to PAs and PRDs (Kafka, 1997).
Two a priori hypotheses were tested. First, it was hypothesized that PA males
would differ significantly from the PRDs specifically in the same factors identified
in the previous study (see discussion above; Kafka & Prentky, 1998). Second, it
was hypothesized that conduct disorder would be statistically significantly more
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

352 Kafka and Hennen

frequently associated with PA status rather than with PRD status and specifically
be associated with sex offender PAs.

METHOD

Data were collected prospectively from 120 consecutively evaluated, volun-

tary outpatient males (age = 17–65) seeking treatment for principal diagnoses of
PAs or PRDs. The Axis I diagnostic assessments of males in this sample have
not been reported in any previous studies by this investigator (Kafka & Prentky,
1994, 1998). Informed consent was obtained from all study participants. Exclu-
sion criteria included the presence of a neurological condition or diagnosis, age
greater than 65 years, a history of significant head injury with sustained loss of
consciousness, current psychoactive substance abuse, or noncompliance with the
psychiatric evaluation. The 120 males who completed the diagnostic evaluation
were seeking treatment for either PAs (n = 88), including nonoffender paraphilic
(n = 28) and sex offender paraphilic (n = 60) subgroups, or PRDs (n = 32).
All participants completed a semistructured Intake Questionnaire that col-
lected demographic, medical, developmental, and psychiatric treatment data
(Kafka & Prentky, 1994, 1998). The Intake Questionnaire contained an exten-
sive checklist of symptoms designed to detect the lifetime prevalence of DSM-IV
Axis I mood, anxiety, psychoactive substance abuse, impulse NOS, and conduct
disorder (unpublished inventory, available on request).
The retrospective assessment of ADHD and its subtyping were ascertained
utilizing the ADHD Rating Scale (DuPaul, 1991) modified for DSM-IV (Findling,
Schwartz, Flannery, & Manas, 1996) and the Wender Utah Retrospective Scale
(WURS; Ward, Wender, & Reimherr, 1993). The modified ADHD Rating Scale
consists of the 18 diagnostic criteria identified as core symptoms of DSM-IV
ADHD; identifying persistent inattentiveness (six or more of nine-criterion items)
and/or persistent hyperactivity/impulsivity (six- or more of nine-criterion items).
To establish a threshold for the diagnosis of ADHD, participants retrospectively
self-rated each criterion on a 0–3 Severity scale. At least 12 items rated 2, 6 or more
from both the inattentiveness diagnostic cluster and the hyperactive/impulsive clus-
ter, were required for a retrospective diagnosis of ADHD-combined subtype (APA,
1994; Findling et al., 1996).
The 25-item WURS cumulative score was used to supplement the ADHD
Rating Scale, distinguishing participants with ADHD from controls. The WURS
has excellent concurrent validity. Ward et al. (1993) reported that a WURS score
of 46 or higher correctly identified 86% of the participants with attention deficit
disorder, 99% of the normal controls, and 81% of a comparison group of depressed
participants (Ward et al., 1993). In their group of males with DSM-III-R-defined
ADHD (equivalent to DSM-IV ADHD-combined subtype), the mean WURS score
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 353

was 60.3 ± 14.2, whereas the mean unipolar depression group score was 34.2 ±
18.0 and the mean normal group score was 17.9 ± 11.0. It is important to note that
Ward et al. stipulate that the WURS does not “diagnose” ADHD, rather, a score
of 46 is correlated with ADHD-combined subtype.
The lifetime prevalence of Axis I diagnoses were assigned on the basis of two
follow-up psychiatric interviews conducted by the first author, a Board certified
psychiatrist, incorporating data from the Intake Questionnaire.
Lifetime sexual diagnoses (PAs and PRDs) were assessed utilizing semistruc-
tured Sexual Inventories (unpublished inventories, available on request) and the
aforementioned follow-up interviews (Kafka, 1997; Kafka & Hennen, 1999; Kafka
& Prentky, 1994, 1998). For the determination of Axis I diagnoses, the diagnostic
hierarchy and exclusionary rules as described in DSM-IV were maintained with the
following exceptions: (a) dysthymia: primary versus secondary dysthymia were
not differentiated, (b) there was no distinction made between substance abuse
versus substance dependence.
Additional repetitive impulsive behaviors, including speeding/reckless driv-
ing and repetitive theft (not distinguished from kleptomania), were scored as
impulse disorders NOS. Problems associated with reckless driving have been
specifically reported in adults with ADHD (Barkley, Guevremont, Anastoploulos,
DuPaul, & Shelton, 1993; Murphy & Barkley, 1996). Participants diagnosed with
reckless driving had to meet the following criteria: self-assessment as having
driven recklessly repetitively, accompanied by at least two of the following—
multiple speeding tickets, multiple motor vehicle accidents, license suspension or
loss, repetitive “road rage,” or forced enrollment in driver’s safety classes. If these
behaviors were predominantly attributed to driving while under the influence of
psychoactive drugs, including alcohol, a reckless driving diagnosis was not scored.
All paraphilic diagnostic categories were assigned utilizing DSM-IV criteria.
PRDs were classified according to criteria previously published by the first author
(Kafka, 1997; Kafka & Hennen, 1999; Kafka & Prentky, 1994, 1998). Males in
the PRD group reported the presence of at least one PRD but no lifetime PAs.
Males in the PA group reported repetitive PA behavior but could also have current
or past PRDs. Sex offender paraphiliacs included all males who had repetitively
engaged in paraphilic behavior that included an unwilling or unsuspecting victim.
This could also include fetishism if paraphilic enactment included repetitive theft
to obtain fetish objects. In this study, sex offender–PA diagnoses included ex-
hibitionism, pedophilia, voyeurism, fetishism, frotteurism, telephone scatologia,
sexual sadism, and rape. Although rape is not currently considered a PA, the males
who committed rape (n = 3) all had comorbid paraphilias so their sexual coercion
was included in the paraphilic group. Nonoffender paraphilic diagnoses included
fetishism, transvestic-fetishism, sexual masochism, sexual sadism, and PAs NOS.
Lifetime prevalence rates of DSM-IV Axis I disorders are reported as percent-
ages. Continuous variables are summarized with means and associated standard
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

354 Kafka and Hennen

deviations or confidence intervals. The chi-square (χ 2 ) test with associated degrees

of freedom is used to compare prevalence rates among the categorical variables.
When necessitated by small cell frequencies, Fisher’s exact test is used instead
of chi-square. Contrasts of continuous variables are carried out using unpaired,
two-tailed Student t tests for multiple group comparisons. An alpha ( p) value
of ≤0.05 was considered as statistically significant.

RESULTS

The developmental and demographic characteristics of the sample are re-

ported in Table I. The typical participant was a 37-year-old, white, married male
who had graduated from college (completed 16 years of education) and was earning
a middle-class income (mean = $58,200). Slightly more than half of the partic-
ipants (51.6%) reported some lifetime contact with the criminal justice system
(e.g., arrest, incarceration), although this contact was not necessarily exclusively
associated with sex offending behaviors. Nearly 41% (49/120) of the sample had

Table I. Developmental and Demographic Variables in a Sample of Males With Paraphilias and
Paraphilia-Related Disorders
Combined sample PA sample PRD sample
Variable n Mean ± SD % n Mean ± SD % n Mean ± SD %

Sample size 120 88 32

Age 37.1 ± 9.5 36.9 ± 10.2 37.5 ± 7.4
Ethnicity
Caucasian 117 97.5 86 97.7 31 96.8
Marital history
Ever married 65 54.1 50 56.8 15 46.8
Abuse history
Any abuse 32 26.6 26 29.5 6 18.7
Physical∗ 16 13.3 15 17.0 1 3.1
Sexual 21 17.5 16 18.1 5 15.6
Both 5 4.1 5 5.6 0 0
Educational history
Years completed∗∗∗ 15.7 ± 3.2 15.2 ± 3.3 17.1 ± 2.7
Years completed (median) 16 16 17
Any school problem∗∗∗ 39 32.5 36 40.9 3 9.3
Truancy∗ 17 14.1 16 18.1 1 3.1
Repeated grade 24 20.0 21 23.8 3 9.3
Suspended/expelled∗∗∗ 27 22.5 27 30.6 0 0.0
Employment history
Current income ($1000s) 58.2 ± 47.3 52.6 ± 48.2 71.6 ± 42.9
Currently employed 97 80.8 68 77.2 29 90.6
Unemployed/disabled∗ 16 13.3 15 18.0 1 3.3
Criminal justice history
Ever arrested∗∗∗ 60 50.0 54 61.3 6 18.7
Ever incarcerated∗∗ 15 12.5 15 17.0 0 0.0
Psychiatric hospitalization?∗∗ 23 19.1 22 25.0 1 3.1
∗p ≤ .05. ∗∗ p ≤ .01. ∗∗∗ p ≤ .005.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 355

been arrested for inappropriate sexual behavior. About one quarter (26.6%) of the
sample reported a history of physical or sexual abuse, and almost one third (32.5%)
reported significant school-associated learning (e.g., repeated grades) or behavior
problem (e.g., truancy, suspension/expulsion from primary or secondary school).
The PA group was statistically significantly different from the PRDs in the
incidence of physical abuse (χ 2 = 3.9, df = 1, p = .04), educational achieve-
ment (years of education completed; t = 2.9, df = 118, p = .004), and the pres-
ence of school-associated learning/behavioral problems, including truancy (χ 2 =
4.3, df = 1, p = .03), suspension/expulsion (χ 2 = 12.6, df = 1, p = .004) and a
trend (PA > PRD) was reported for repeated grades (χ 2 = 3.0, df = 1, p = .07).
The PA group also reported a statistically significantly higher incidence of
psychiatric/substance abuse hospitalization (χ 2 = 7.2, df = 1, p = .007) and an
increased prevalence of current unemployment or disability status (χ 2 = 3.9, df =
1, p = .04). There was a statistical trend indicating that the currently employed
PAs had a lower current income (t = 1.8, df = 94, p = .07) in comparison to the
employed PRDs. Last, the PA males self-reported statistically significantly more
lifetime contact with the criminal justice system, including both arrests (χ 2 =
17.0, df = 1, p = .0001) and incarceration (χ 2 = 6.2, df = 1, p = .01). The PA
and PRD groups did not differ significantly in the prevalence of sexual abuse
(18.1% vs. 15.6%).
The distribution of the sexual diagnoses in the combined sample is listed in
Table II. The mean number of lifetime hypersexual disorders (PAs + PRDs) in the
sample was 3.3 ± 1.7 (median = 3). In the PA group, 73.8% (65/88) males had
at least one PRD, most commonly compulsive masturbation. The PA group had
statistically significantly more lifetime hypersexual disorders than the PRDs (3.5 ±
1.8 vs. 2.7 ± 0.9; t = 2.3, df = 118, p = .02). On the other hand, the PRD group
reported statistically significantly more lifetime PRD diagnoses (2.7 ± 0.9 vs.
1.8 ± 1.2; t = 3.4, df = 118, p = .0008).
The Axis I lifetime diagnoses of the sample are listed in Table III, both for the
PA and PRD groups as well as the combined sample. Nine percent of the sample,
all from the PRD group, reported no lifetime nonsexual Axis I disorders, a statis-
tically significant difference (PA vs. PRD: χ 2 = 4.4, df = 1, p = .03). Although
the participants in the PA group were diagnosed with more lifetime nonsexual
Axis I disorders in comparison with the PRDs, the difference was nearly statisti-
cally significant with a p value at a trend ( p = .07). In the combined sample, there
was a robust statistical correlation between the lifetime total Axis I nonsexual and
Axis I sexual diagnoses (Spearman correlation coefficient Z = 3.9, p = .0001).
In both groups, the most prevalent Axis I disorders were mood disorders
(71.6%), especially dysthymic disorder (69.1%). Fifty-five percent of the combined
sample reported dysthymic disorder, early onset subtype; this was the single most
prevalent Axis I condition. Thirty-nine percent reported at least a single episode
of major depression, the second most common diagnosis. The PA and PRD groups
did not differ in the lifetime prevalence of any specific mood disorder diagnosis.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

356 Kafka and Hennen

Table II. Sexual Diagnoses in 120 Outpatient Males With Paraphilias

and Paraphilia-Related Disorders
Sexual diagnoses n %

Sample size 120

Paraphilic disorders
Exhibitionism 32 26.6
Voyeurism 19 15.8
Pedophilia, any 20 16.6
Opposite sex 9 7.5
Homosexual 7 5.8
Bisexual 1 0.8
Incestuous 3 2.5
Telephone scatologia 15 12.5
Transvestistic fetishism 13 10.8
Masochism 13 10.8
Frotteurism 11 9.1
Fetishism 10 8.3
PA NOS 6 5.0
Sadism 5 4.1
Rape 3 2.5
Paraphilia-related disorders
Compulsive masturbation 87 72.5
Pornography dependence 57 47.5
Protracted promiscuity 53 44.1
Hetero- 25 20.8
Homo- 18 15.0
Bisexual 10 8.3
Telephone sex 30 25.0
Sex desire incompatibility 16 13.3
PRD NOS 9 7.5

Participants in the combined sample diagnosed with lifetime major depression or

dysthymic disorder were statistically significantly more likely to self-report more
lifetime hypersexual diagnoses compared to those without these mood disorder
diagnoses (major depression: t = 2.3, df = 118, p = .02; dysthymic disorder: t =
2.7, df = 118, p = .006).
Anxiety disorders were reported by 38.3% of the combined sample. The most
prevalent lifetime anxiety disorder was social phobia (21.6%), followed by gener-
alized anxiety disorder (9.1%). The PA and PRD groups were not statistically sig-
nificantly different in the incidence of any specific anxiety disorder, although there
was a statistical trend ( p = .07) for the PA group to report obsessive–compulsive
disorder (9% vs. 0.0%) more frequently. Posttraumatic stress disorder, diagnosed
in only 5.8% of the combined sample, was significantly associated with a de-
velopmental history of physical or sexual abuse (χ 2 = 13.2, df = 1, p = .0003).
No specific anxiety disorder diagnosis was associated with multiple hypersexual
disorders.
A lifetime psychoactive substance abuse diagnosis was reported by 40.8% of
the combined sample. Alcohol abuse, the most prevalent substance abuse diagnosis,
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 357

Table III. Lifetime Axis I Diagnoses in a Sample of Males With Paraphilias and Paraphilia-Related
Disorders
Combined sample PA sample PRD sample
Variable n Mean ± SD % n Mean ± SD % n Mean ± SD %

Sample size 120 88 32

Mean of diagnosesa 3.2 ± 2.3 3.4 ± 2.5 2.5 ± 1.0
No Axis I disorder∗ 11 9.1 11 12.5 0 0.0
Any mood disorder 86 71.6 63 71.5 23 71.8
Dysthymic disorder 83 69.1 61 69.3 22 68.7
Major depression 47 39.1 34 38.6 13 40.6
Bipolar 6 5.0 6 6.8 0 0.0
Any anxiety disorder 46 38.3 34 38.6 12 37.5
Social phobia 26 21.6 18 20.4 8 25.0
GAD 11 9.1 6 6.8 5 15.6
Panic 9 7.5 6 6.8 3 9.3
PTSD 7 5.8 5 5.6 2 6.2
OCD 8 6.6 8 9.0 0 0.0
Any psychoactive 49 40.8 37 42.0 12 37.5
substance abuse
Alcohol 36 30.0 28 31.8 8 25.0
Cocaine∗ 17 14.1 16 18.1 1 3.1
Marijuana 22 18.3 16 18.1 6 18.7
Polydrug abuse 23 19.1 18 20.4 5 15.6
Any impulsivity NOS 32 26.6 27 30.6 5 15.6
Reckless driving 25 20.8 21 23.8 4 12.5
Any ADHD∗∗ 43 35.8 37 42.0 6 18.7
Combined subtypea∗∗∗ 22 18.3 21 23.8 1 3.1
Inattentive subtype 21 17.5 16 18.1 5 15.6
Any conduct disorder∗∗∗ 20 16.6 20 22.7 0 0.0
Psychosis 5 4.1 5 5.6 0 0.0
a Two participants with ADHD hyperactive/impulsive subtype are coded as combined subtype in this
table.
∗ p ≤ .05. ∗∗ p ≤ .01. ∗∗∗ p ≤ .005.

was reported by 30% of the combined sample, followed by marijuana abuse

(18.3%). Of the males with any psychoactive substance abuse diagnosis, nearly one
fifth (19.1%) reported polysubstance abuse. The PA group (18.1%) was statisti-
cally significantly more likely than the PRD group (3.1%) to report cocaine abuse
(χ 2 = 4.3, df = 1, p = .03). There were no other statistically significant differ-
ences between PAs and PRDs in regards to psychoactive substance abuse diag-
noses. In the combined group, participants with alcohol abuse had statistically sig-
nificantly more lifetime hypersexual disorders (t = 1.9, df = 118, p = .05) than
did those without alcohol abuse.
Psychotic disorders (including mania with delusions) were reported by only
a small number of males in this sample (4.1%), all within the PA group.
Impulse disorders NOS were reported by more than one quarter of the sam-
ple (26.6%). The most common impulse disorder was reckless driving, reported
by 20.8% (25/120) of the sample. Other impulse disorder NOS conditions were
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

358 Kafka and Hennen

relatively uncommon. The PA group did not differ from the PRD group in the
prevalence of any specific impulse NOS disorder. Reckless driving was not as-
sociated with a statistically significant increase in the total number of lifetime
hypersexual disorders diagnosed.
The prevalence of ADHD in the combined sample was 35.8% (43/120), mak-
ing it the third most common Axis I disorder (behind dysthymic disorder and
major depression). The combined subtype (ADHD-C; 18.3% sample prevalence)
and the inattentive subtype (ADHD-I; 17.5% sample prevalence) were nearly
equally represented in the combined sample. ADHD was statistically significantly
more prevalent amongst PA males (42% prevalence) in comparison with PRDs
(18.7%; χ 2 = 5.5, df = 1, p = .01). This difference was even more pronounced
in comparing the prevalence of ADHD-C between the groups (23.8% vs. 3.1%;
χ 2 = 6.7, df = 1, p = .009). As expected, the mean WURS score of the ADHD
participants was statistically significantly higher in comparison to those partici-
pants without ADHD (mean 59.2 ± 13.8 vs. 26.5 ± 15.2; t = 11.5, df = 117, p =
.0001). The mean WURS score for the ADHD-I subgroup, however, was not
statistically significantly different from that of the ADHD-C subgroup.
In addition to ADHD, conduct disorder (combined sample prevalence rate of
16.6%) was statistically significantly more prevalent amongst PA males (20/88;
22.7%) in comparison to PRDs (0/32, 0.0%; χ 2 = 8.7, df = 1, p = .003).
Inasmuch as ADHD was both a prevalent diagnosis and a diagnosis that
distinguished PA from PRD males, additional statistical testing on demographic,
developmental, Axis I and sexual disorders was performed comparing males with
ADHD (n = 43) to those with no-childhood history of that diagnosis (n = 77).
Compared to participants without ADHD, males with/without a retrospective di-
agnosis of ADHD did not differ in age or marital status. In comparison to the non-
ADHD group, the ADHD group was statistically significantly more likely to report
any physical or sexual abuse (χ 2 = 3.8, df = 1, p = .05), although this difference
was not specific for either physical or sexual abuse alone. Males with ADHD com-
pleted less education (14.1 ± 2.8 vs. 16.6 ± 3.1 years; t = 4.2, df = 118, p =
.0001), reported more school-related behavioral problems such as truancy (χ 2 =
18.6, df = 1, p = .0001), suspension/expulsion (χ 2 = 26.6, df = 1, p = .0001),
and repeated grades (χ 2 = 6.6, df = 1, p = .01). Also, in comparison to non-
ADHD males, ADHD males reported a lower current income (39.0 ± 23.3 vs.
66.1 ± 52.3 thousand dollars/year; t = 2.6, df = 94, p = .01) were less likely to
be currently employed (χ 2 = 7.7, df = 1, p = .005) and more likely to be cur-
rently unemployed or disabled (χ 2 = 6.4, df = 1, p = .01). ADHD males were
statistically significantly more likely to report encounters with the criminal jus-
tice system, including being arrested (χ 2 = 8.1, df = 1, p = .004) and incarcer-
ated (χ 2 = 10.4, df = 1, p = .001). Last, ADHD males were statistically signifi-
cantly more likely to report a history of psychiatric/substance abuse hospitalization
(χ 2 = 5.2, df = 1, p = .02) than were the non-ADHD participants.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 359

Additional statistically significant differences were noted in comparing the

ADHD group to non-ADHD group on variables related to nonsexual Axis I comor-
bidity. Males with ADHD self-reported more lifetime Axis I disorders (t = 5.0,
df = 118, p = .0001) and were more likely to meet diagnostic criteria for the
following specific Axis I diagnoses: dysthymic disorder (χ 2 = 14.5, df = 1, p =
.0001), reckless driving (χ 2 = 10.8, df = 1, p = .001), conduct disorder (χ 2 =
25.2, df = 1, p = .0001), polydrug abuse (χ 2 = 7.7, df = 1, p = .005), cocaine
abuse (χ 2 = 10.4, df = 1, p = .001), marijuana abuse (χ 2 = 4.1, df = 1, p = .04)
but not alcohol abuse.
The ADHD group of males was also statistically significantly more likely to
report more lifetime hypersexual disorders (4.1 ± 1.8 vs. 2.8 ± 1.4; t = 4.3, df =
118, p = .0001), especially multiple PAs (1.7 ± 1.4 vs. 0.9 ± 0.8; t = 4.2, df =
118, p = .0001). ADHD-positive males also reported more lifetime PRDs, but the
difference was not statistically significant (2.3 ± 1.1 vs. 1.9 ± 1.2; t = 1.8, df =
118, p = .07). The following hypersexual disorders were specifically associated
with a diagnosis of ADHD: frotteurism (χ 2 = 11.1, df = 1, p = .0008), fetishism
(χ 2 = 5.5, df = 1, p = .01), paraphilia NOS (χ 2 = 6.1, df = 1, p = .01), com-
pulsive masturbation (χ 2 = 6.1, df = 1, p = .01), and pornography dependence
(χ 2 = 4.5, df = 1, p = .03).
Although conduct disorder was not as prevalent (20/120; 16.6%) as ADHD
(35.8%) in this sample, the statistical association between conduct disorder and
paraphilic status (PA vs. PRD; χ 2 = 8.7, df = 1, p = .003) was more robust than
the association between ADHD ( p = .01) or ADHD-C ( p = .009) and paraphilic
status. Males with conduct disorder, in fact, were statistically significantly more
likely to self-report childhood ADHD, 17/20 males (85%); χ 2 = 25.2, df = 1,
p = .0001, especially combined subtype, 11/20 males (55%); χ 2 = 21.5, df =
1, p = .0001, in comparison with inattentive subtype (6/30 males; 30%; χ 2 =
2.5, df = 1, p = .10). As might be expected from the aforementioned comorbid
association between conduct disorder and ADHD, conduct disorder was not only
associated with multiple hypersexual disorders (t = 3.2, df = 118, p = .001) but
more specifically with multiple PAs (t = 4.9, df = 118, p = .0001).
When the groups were divided as sex offender paraphiliacs (n = 66) ver-
sus nonoffender males (n = 22 PAs and 32 PRDs), the sex offenders did not
differ in age, marital status, current employment/disability status, physical or
sexual abuse history, or the total number of lifetime hypersexual or nonsexual
Axis I diagnoses from nonoffender PAs and PRDs. The sex offenders, however,
were statistically different from the nonoffenders in variables related to educa-
tional history, legal history, and incidence of psychiatric hospitalization (F =
6.4, df = 1, 118, p = .01). Sex offenders achieved fewer years of completed edu-
cation (14.6 ± 3.2 vs. 17.0 ± 2.7; F = 19.0, df = 1, 118, p = .0001), were more
likely to repeat a grade (F = 4.9, df = 1, 118, p = .02), or be suspended/expelled
from school (F = 14.1, df = 1, 118, p = .0003). As would be expected, they
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

360 Kafka and Hennen

were significantly more likely to be arrested ( p = .0001) or incarcerated ( p =

.0001) for a sexual offense. The Axis I disorders that were statistically signifi-
cantly more prevalent amongst sex offender paraphiliacs included conduct dis-
order (χ 2 = 11.8, df = 1, p = .0006), generalized anxiety disorder (χ 2 = 3.7,
df = 1, p = .05), alcohol abuse (χ 2 = 4.3, df = 1, p = .03), and cocaine abuse
(χ 2 = 3.6, df = 1, p = .05). There was only a statistical trend ( p = .09) associ-
ation between ADHD and sex offending paraphilias, and there was no specific
association between sex offending diagnoses and ADHD subtypes. This lack of a
specific association between ADHD and nonoffender status, however, is mediated
by the nearly equal prevalence of ADHD in nonoffender paraphiliacs (39.2%) in
comparison to offender paraphiliacs (43.3%). Nearly one quarter of sex offenders
(15/66; 22.7%) were diagnosed with ADHD-C. As was the case for PA males
(which includes sex offender PAs), there was a statistical trend for sex offend-
ers to report physical abuse (χ 2 = 2.9, df = 1, p = .08), but not sexual abuse in
comparison to nonoffenders.
Inasmuch as many of the aforementioned statistically significant differences
distinguishing PAs from PRDs are also characteristic of the ADHD subgroup, a
recomparison of the statistically significant differences between PAs and PRDs
controlling for the diagnosis of ADHD (PA, n = 51; PRD, n = 26) was also car-
ried out. The purpose of this reexamination was to determine to what degree the
identified differences between PAs and PRDs could be mediated by the presence
of childhood ADHD.
In this comparison, there were no statistical differences in age, ethnicity,
marital status, abuse history, or current income between PAs and PRDs after con-
trolling for the presence/absence of childhood ADHD. There were only statistical
trends suggesting that PA males were more likely to be disabled/unemployed
(χ 2 = 3.3, df = 1, p = .06) and to have been psychiatrically hospitalized (χ 2 =
2.9, df = 1, p = .08). Education-related differences between PAs and PRDs were
reduced to trends suggesting that the PA males might be disadvantaged by less ed-
ucation completed (t = 1.8, df = 75, p = .07) and the presence of school-related
behavioral problems such as suspension/expulsion (χ 2 = 3.3, df = 1, p = .06).
Problems involving the criminal justice system, such as having been arrested for
any offense, remained statistically significant (χ 2 = 7.2, df = 1, p = .007). It is
of interest, however, that the prevalence of incarceration in PA males was no longer
statistically significant ( p = .14).
Controlling for the diagnosis of childhood ADHD had similar effects in re-
ducing the statistically significant differences between PA and PRD groups on vari-
ables associated with Axis I comorbidity. After controlling for childhood ADHD,
the PA group no longer reported statistically significantly more lifetime Axis I
diagnoses, conduct disorder or cocaine abuse (χ 2 = 2.7, df = 1, p = .09). On
sexuality-related variables, the PA group was not statistically significantly dif-
ferent from the PRDs in the total number of lifetime hypersexual disorders, but
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 361

the PRD group was still more likely to report more PRDs in comparison to PAs
(t = 3.6, df = 75, p = .0005).

DISCUSSION

In this sample of 120 consecutively evaluated males with PAs (n = 88) and
PRDs (n = 32), the most common hypersexual disorders were the PRDs compul-
sive masturbation (72.5% sample prevalence), pornography dependence (47.5%),
and protracted promiscuity (44.1%) and the PAs, especially exhibitionism (26.6%),
pedophilia (16.6%), and voyeurism (15.8%). The frequency distribution of these
sexual diagnoses is consistent with previous data from Axis I comorbidity studies
by this investigator utilizing other samples (Kafka & Prentky, 1994, 1998). The PA
group self-reported more lifetime hypersexual disorders but the PRDs self-reported
more lifetime PRDs.
The PA group differed significantly from the PRD group in demographic and
developmental variables including the incidence of physical abuse ( p = .04) but
not sexual abuse, lower educational achievement ( p = .004), and the presence
of school-associated behavior problems such as truancy ( p = .004) and suspen-
sion/expulsion ( p = .004). The PA group was significantly more likely to report
a history of psychiatric/substance abuse hospitalization ( p = .007) and contact
with the criminal justice system, including arrest ( p = .0001) and incarceration
( p = .01). Last, PA males had current unemployment/disability status ( p = .04)
and a trend toward lower current earnings ( p = .07). These statistical differences
between PA and PRD males are the same as those reported in a previous compar-
ative study that included the retrospective diagnosis of DSM-III-R-defined ADHD
(Kafka & Prentky, 1998).
The most commonly diagnosed comorbid DSM-IV Axis disorders were mood
disorders (71.6% lifetime prevalence) especially dysthymic disorder early onset
subtype (55%) and major depression (39.1%). Psychoactive substance abuse was
diagnosed in 40.8% of the participants, and alcohol abuse (30%) was the most
prevalent substance abuse disorder in this sample. Any anxiety disorder was diag-
nosed in 39.1% of the sample, and social phobia (21.6%) was the most common
anxiety disorder diagnosis. Impulsivity NOS disorders were diagnosed in 26.6%
of the sample, but this prevalence was predominantly accounted for by the inclu-
sion of an atypical impulsivity disorder, reckless driving (20.8%). In all of the
aforementioned categories, the only diagnosis that was statistically significantly
more prevalent amongst the PA group in comparison to PRDs was cocaine abuse
( p = .03). The frequency distribution of the major aforementioned Axis I disorders
is very similar to the frequency distribution of Axis I disorders reported utilizing
DSM-III-R criteria in comparing PA to PRD males in some prior reports (Kafka
& Prentky, 1994, 1998). In addition, there was a robust significant correlation
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

362 Kafka and Hennen

between lifetime number of nonsexual and sexual Axis I disorders. That is, males
with the fewest Axis I nonsexual disorders were more likely to also be diagnosed
with fewer hypersexual disorders.
By extending the boundaries of impulsivity disorders to include ADHD (sam-
ple prevalence 35.8%) and conduct disorder (16.6%) as well, these latter diagnoses
were statistically significantly more prevalent ( p = .01; p = .003 respectively)
in the PA group than in the PRD group. When ADHD was subdivided into its
two major DSM-IV-derived subtypes, inattentive and combined, it was the com-
bined subtype but not the inattentive subtype that was most commonly associated
with both persistent socially deviant sexual arousal (PA status; p = .009), con-
duct disorder ( p = .0001), and multiple hypersexual disorders ( p = .0004), in
particular, multiple paraphilias ( p = .0001). The comorbid association between
ADHD and conduct disorder has been reported by several investigators, and this
comorbid association also predicts more severe comorbid psychopathologies in-
cluding antisocial behavior, conduct disorder, and polysubstance abuse (Abikoff
& Klein, 1992; Lynam, 1996; Schubiner et al., 2000b). In this study, frotteurism
( p = .0008), fetishism ( p = .01), paraphilia NOS ( p = .01), and the PRDs com-
pulsive masturbation ( p = .01) and pornography dependence ( p = .03) were sta-
tistically significantly associated with a retrospective diagnosis of ADHD.
Sex offender paraphiliacs differed from nonoffender PA and PRD males in de-
mographic/developmental variables, such as lower educational achievement, more
school-associated behavioral problems, an increased incidence of psychiatric hos-
pitalization and, as expected, more arrests and incarcerations associated with sexual
behavior. Sex offender paraphiliacs were similar to nonoffender PAs/PRDs on all
other demographic/developmental variables included in this study. Sex offender
PAs were specifically more likely to report conduct disorder ( p = .0006), gen-
eralized anxiety disorder ( p = .05), cocaine abuse ( p = .05), and alcohol abuse
( p = .03) but not ADHD, including ADHD-combined subtype. It is noteworthy,
however, that 43.3% of sex offender paraphiliacs were retrospectively diagnosed
with any ADHD and that nearly 25% were diagnosed with ADHD-combined
subtype.
Inasmuch as ADHD was the third most prevalent lifetime Axis I diagnosis
and had been previously reported specifically associated with PA status, a stratified
reanalysis of the sample excluding those with ADHD was undertaken (n = 77).
With this stratification, almost all of the reported statistically significant differences
between PAs and PRDs no longer attained statistical significance criteria (although
several remained as statistical trends). The only statistically significant differences
distinguishing PAs from PRDs that remain after such an reanalysis were that PAs
were still more likely to have been arrested and have fewer lifetime PRDs. Thus,
in this sample of 120 hypersexual males, the presence of DSM-IV ADHD was the
most frequently diagnosed Axis I disorder that distinguished socially deviant (PA)
from nonparaphilic sexual arousal (PRD) and accounted for the major demographic
and developmental variables that distinguished the PA from the PRD group.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 363

Perhaps because of a paucity of contemporary empirical data describing Axis I

comorbidity in male sex offenders, there has been no integration of these diag-
nostic data into actuarial models of risk assessment. From the available studies
however, (combined n300 male adolescent and adult sex offenders (Galli et al.,
1999; Kafka & Prentky, 1994, 1998; Kavoussi et al., 1988; McElroy et al., 1999),
Raymond et al., 1999; Vaih-Koch & Bosinski, 1999), mood disorders, and impul-
sivity disorders (ADHD and conduct disorder) have been repetitively identified as
prevalent in male sex offenders. These diagnoses could be significant contributing
variables associated with both static and dynamic risk factors associated with sex
offender recidivism (e.g., antisocial impulsivity, sexual deviance, sexual preoc-
cupation, lower educational achievement, and multiple PAs; Hanson & Bussiere,
1998; Hanson & Harris, 2000).
Dysthymic disorder and ADHD can be easily overlooked during the psychi-
atric assessment of adults, yet both are treatment-responsive conditions, in partic-
ular, amenable to pharmacotherapy. Last, it is possible that the underrecognition
and lack of treatment of concurrent Axis I comorbidity in sex offenders could be
associated with treatment disengagement, a dynamic risk factor associated with
sex offender recidivism (Hanson & Harris, 2000). Based on these data, systematic
and thorough psychiatric evaluation of all sex offenders appears warranted and the
identification and treatment of comorbid Axis I disorders, in particular, appears
justified.
There are several methodological limitations associated with the conclusions
reported here. First, the diagnosis of ADHD in adults was retrospective and did not
include informants other than the study participants. Some authors have noted that
males with ADHD can have poor self-observational skills (Mannuzza & Gittelman,
1985), whereas others have reported that adults can retrospectively self-rate child-
hood ADHD accurately (Murphy & Schachar, 2000). No corroborative current
neuropsychological testing was performed to assess residual inattention or im-
pulsivity in these participants and study participants, at a mean age of 37 years,
were expected to reliably recall childhood domestic and school-associated behav-
iors. Despite these limitations, the diagnosis of ADHD in this sample was derived
by utilizing validated diagnostic and correlative rating scales and the retrospec-
tive diagnosis of ADHD was associated with many of the vicissitudes that have
been reported as sequellae to ADHD, with these sequellae including learning and
school-associated problems, polydrug abuse, mood disorders, conduct disorder,
and reckless driving (Barkley et al., 1993; Biederman et al., 1993, 1995; Eyestone
& Howell, 1994; Mannuza, Klein, Bessler, Malloy, & LaPadula, 1993; Mannuza
et al., 1991; Schubiner et al., 2000a).
Second, this study did not utilize validated structured diagnostic instruments
to assess Axis I comorbidity and instead relied on self-report and an Intake Ques-
tionnaire assessing symptoms and syndromes as defined by DSM-IV. Ultimately,
the primary investigator (the first author) in collaboration with the participant
and the Inventory data ascertained diagnostic categories and thresholds. The
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

364 Kafka and Hennen

veridicality of Axis I diagnoses derived in this manner could thus be compro-

mised by the possible bias of the primary investigator.
Third, there are no rating instruments with documented validity and reliability
available to diagnose both PAs and all PRDs. Although this investigator has utilized
these Sexual Inventories in previous studies, ascertainment bias regarding sexual
diagnoses reported must be considered as a possible limitation of the conclusions
reported.
Despite these limitations, it remains certainly plausible that socially deviant
sexual arousal (i.e., PA status) could be correlated with other Axis I comorbid
disorders that have been associated with socially deviant behavior such as mood
disorders, substance abuse, ADHD, and conduct disorder. Inasmuch as ADHD and
early onset dysthymic disorder are significant psychopathologies that can indepen-
dently impact childhood and adolescent development, it has been speculated that
this comorbid combination could have a particularly pernicious effect on develop-
ing male sexuality (Kafka & Prentky, 1998). Such a concurrence could conduce to
increased sexual appetitive behavior, thrill seeking, social deviancy, and impaired
impulse control. Last, the successful pharmacological treatment of such concurrent
conditions, including ADHD and mood disorders in hypersexual males, has been
reported to significantly ameliorate PAs as well as PRDs (Greenberg & Bradford,
1997; Kafka, 2000; Kafka & Hennen, 2000).

REFERENCES

Abikoff, H., & Klein, R. G. (1992). Attention-deficit hyperactivity and conduct disorder: Comorbidity
and implications for treatment. Journal of Consulting and Clinical Psychology, 60, 881–892.
American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders
(3rd ed.). Washington, DC: Author.
American Psychiatric Association. (1987). Diagnostic and statistical manual (3rd ed., revised).
Washington, DC: Author.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th
ed.). Washington, DC: Author.
Barkley, R. A., Guevremont, D. C., Anastoploulos, A. D., DuPaul, G. J., & Shelton, T. L. (1993).
Driving-related risks and outcomes of attention deficit hyperactivity disorder in adolescents and
young adults. Pediatrics, 92, 212–218.
Biederman, J., Faraone, S. V., Spenser, T., Wilens, T., Norman, N., Lapey, K. A., et al. (1993). Patterns of
psychiatric comorbidity, cognition, and psychosocial functioning in adults with attention deficit-
hyperactivity disorder. American Journal of Psychiatry, 150, 1792–1798.
Biederman, J., Wilens, T., Mick, E., Milberger, S., Spenser, T. J., & Faraone, S. V. (1995). Psychoactive
substance use disorders in adults with attention deficit hyperactivity disorder (ADHD): Effects of
ADHD and psychiatric comorbidity. American Journal of Psychiatry, 152, 1652–1658.
Black, D. W. (1998). Compulsive sexual behavior: A review. Journal of Psychiatry and Behavior, 4,
217–229.
Black, D. W., Kehrberg, L. L. D., Flumerfelt, D. L., & Schlosser, S. S. (1997). Characteristics of 36
subjects reporting compulsive sexual behavior. American Journal of Psychiatry, 154, 243–249.
Carnes, P. (1990). Sexual addiction. In A. Horton, B. L. Johnson, & L. M. Roundy (Eds.), The incest
perpetrator: A Family member no one wants to treat (pp. 126–143). Newbury Park, CA: Sage.
Coleman, E. (1992). Is your patient suffering from compulsive sexual behavior? Psychiatric Annals,
22, 320–325.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

DSM-IV Axis I Comorbidity 365

Cooper, A., Scherer, C. R., Boies, S. C., & Gordon, B. L. (1999). Sexuality on the internet: From sexual
exploration to the pathological expression. Professional Psychology, 30, 154–164.
DuPaul, G. J. (1991). Parent and teacher ratings of ADHD symptoms: Psychometric properties in a
community sample. Journal of Clinical Child Psychology, 20, 245–253.
Eyestone, L. L., & Howell, R. J. (1994). An epidemiological study of attention-deficit hyperactivity
disorder and major depression in a male prison population. Journal of the American Academy of
Psychiatry and the Law, 22, 181–193.
Findling, R. L., Schwartz, M. A., Flannery, D. J., & Manas, M. J. (1996). Venlafaxine in adults with
attention deficit hyperactivity disorder: An open clinical trial. Journal of Clinical Psychiatry, 57,
184–189.
Galli, V., McElroy, S. L., Soutello, C. A., Kizer, D., Raute, N., Keck, P. E., et al. (1999). The psychiatric
diagnoses of twenty two adolescents who have sexually molested other children. Comparative
Psychiatry, 40, 85–87.
Greenberg, D. M., & Bradford, J. M. W. (1997). Treatment of the paraphilic disorders: A review of
the role of the selective serotonin reuptake inhibitors. Sexual Abuse: Journal of Treatment and
Research, 9, 349–360.
Hanson, R. K., & Bussiere, M. T. (1998). Predicting relapse: A meta-analysis of sexual offender
recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348–362.
Hanson, R. K., & Harris, A. J. R. (2000). Where should we intervene? Dynamic predictors of sexual
offense recidivism. Criminal Justice and Behavior, 27, 6–35.
Kafka, M. P. (1994). Sertraline pharmacotherapy for males with paraphilias and paraphilia-related
disorders: An open trial. Annals of Clinical Psychiatry, 6, 189–195.
Kafka, M. P. (1997). Hypersexual desire in males: An operational definition and clinical implications for
men with paraphilias and paraphilia-related disorders. Archives of Sexual Behavior, 26, 505–526.
Kafka, M. P. (2000). Psychopharmacological treatment for nonparaphilic compulsive sexual behaviors.
CNS Spectrums, 5, 49–59.
Kafka, M. P., & Hennen, J. (1999). The paraphilia-related disorders: An empirical investigation of non-
paraphilic hypersexuality disorders in 206 outpatient males. Journal of Sex and Marital Therapy,
25, 305–319.
Kafka, M. P., & Hennen, J. (2000). Psychostimulant augmentation during treatment with selective
serotonin reuptake inhibitors in males with paraphilias and paraphilia-related disorders: A case
series. Journal of Clinical Psychiatry, 61, 664–670.
Kafka, M. P., & Prentky, R. (1992a). A comparative study of nonparaphilic sexual addictions and
paraphilias in men. Journal of Clinical Psychiatry, 53, 345–350.
Kafka, M. P., & Prenky, R. A. (1992b). Fluoxetine treatment of nonparaphilic sexual addictions and
paraphilias in men. Journal of Clinical Psychiatry, 52, 351–358.
Kafka, M. P., & Prentky, R. A. (1994). Preliminary observations of DSM-III-R Axis I comorbidity in
men with paraphilias and paraphilia-related disorders. Journal of Clinical Psychiatry, 55, 481–
487.
Kafka, M. P., & Prentky, R. A. (1998). Attention deficit hyperactivity disorder in males with paraphilias
and paraphilia-related disorders: A comorbidity study. Journal of Clinical Psychiatry, 59, 388–
396.
Kavoussi, R. J., Kaplan, M., & Becker, J. V. (1988). Psychiatric diagnoses in adolescent sex offenders.
Journal of the American Academy Child and Adolescent Psychiatry, 27, 241–243.
Langevin, R., Bain, J., Ben-Aron, M. H., Coulthard, R., Day, D., Handy, L., et al. (1985). Sexual
aggression: Constructing a predictive equation: A controlled pilot study. In R. Langevin (Ed.),
Erotic preference, gender identity, and aggression in men (pp. 39–72). Hillsdale, NJ: Erlbaum.
Lynam, D. (1996). Early identification of chronic offenders: Who is the fledgling psychopath? Psycho-
logical Bulletin, 120, 209–234.
Mannuzza, S., & Gittelman, R. (1985). Informant variance in the diagnostic assessment of hyperactive
children as young adults. In J. Barrett & R. Rose (Eds.), Mental disorders in the community:
Progress and challenge: Proceedings of the American Psychopathological Association (pp. 243–
254). New York: Guilford.
Mannuza, S., Klein, R. G., Bessler, A., Malloy, P., & LaPadula, M. (1993). Adult outcome of hyperactive
boys: Educational achievement, occupational rank, and psychiatric status. Archives of General
Psychiatry, 50, 565–576.
P1: FYJ
Sexual Abuse: A Journal of Research and Treatment [saj] pp526-sebu-375650 August 23, 2002 10:20 Style file version June 4th, 2002

366 Kafka and Hennen

Mannuza, S., Klein, R. G., Bonagura, N., Malloy, P., Giampino, T. L., & Addali, K. A. (1991). Hyper-
active boys almost grown up. V: Replication of psychiatric status. Archives of General Psychiatry,
48, 77–83.
McElroy, S. L., Soutello, C. A., Taylor, P., Nelson, E. B., Beckman, D. A., Brusman, L. A., et al. (1999).
Psychiatric features of 36 convicted sexual offenses. Journal of Clinical Psychiatry, 60, 414–420.
Murphy, K., & Barkley, R. A. (1996). Attention deficit hyperactivity disorder in adults: Comorbidities
and adaptive impairments. Comprehensive Psychiatry, 37, 393–401.
Murphy, P., & Schachar, R. (2000). Use of self-ratings in the assessment of symptoms of attention
deficit hyperactivity disorder in adults. American Journal of Psychiatry, 157, 1156–1159.
Prentky, R. A., Burgess, A. W., Rokous, F., Lee, A., Hartman, C., Ressler, R., Douglas, J. (1989).
The presumptive role of fantasy in serial sexual homicide. American Journal of Psychiatry, 146,
887–891.
Raymond, N. C., Coleman, E., Ohlerking, F., Christenson, G. A., & Miner, M. (1999). Psychiatric
comorbidity in pedophilic sex offenders. American Journal of Psychiatry, 156, 786–788.
Schubiner, H., Tzelepis, A., Milberger, S., Lockhart, N., Kruger, M., Kelley, B., et al. (2000a). Preva-
lence of attention deficit/hyperactivity disorder and conduct disorder among substance abusers.
Journal of Clinical Psychiatry, 61, 244–251.
Schubiner, H., Tzelepis, A., Milberger, S., Lockhart, N., Kruger, M., Kelley, B. J., et al. (2000b). Preva-
lence of attention deficit/hyperactivity disorder and conduct disorder among substance abusers.
Journal of Clinical Psychiatry, 61, 244–251.
Vaih-Koch, S. R., & Bosinski, H. A. G. (1999, June 23–26). Childhood attention deficit hyperactivity
disorder and conduct disorder in 121 sex offenders. Paper presented at The Twenty Fifth Annual
Meeting of the International Academy of Sex Research. Stony Brook, New York.
Ward, M. F., Wender, P. H., & Reimherr, F. W. (1993). The Wender Utah Rating Scale: An aid in the
retrospective diagnosis of childhood attention deficit hyperactivity disorder. American Journal of
Psychiatry, 150, 885–889.