Взаимосвязь между принудительным сексуальным поведением и педофилическим предпочтением в тюремной выборке лиц, осужденных за сексуальные преступления - последствия для лечения.

Introduction

The current paper seeks to explore the confluence of sexual compulsivity and paraphilic preference among convicted offenders who have sexually offended against children. Sexual compulsivity, or compulsive sexual behavior disorder, is a relatively new diagnosis that has been incorporated into the ICD-11 as an impulse control disorder, 1 though its forerunners have included terms such as sexual addiction, excessive sexual drive, and hypersexuality, 2 , 3 and Seto 4 notes that all of these terms overlap in describing chronic high sex drive that is associated with distress or impairment. Its presence, together with paraphilic disorder, has been found to be among the strongest factors in predicting recidivism among sex offenders, 5 and the confluence of the two often assumed among both researchers and practitioners. DSM-5’s diagnostic criteria for paraphilic disorder assumes the two to be interrelated, where a key criterion is “ recurrent and intense sexual arousal ” from the target of the paraphilia, and by virtue of the harmful consequences of actualizing such arousal outweighing its benefits, compulsivity may also be seen as a key component. In Seto’s 4 Motivation−Facilitation Model of Sexual Offending, the combination of paraphilia and sexual compulsivity is predicted to lead to the commission of sexual offenses, when facilitated by other trait (eg, personality) or state (eg, alcohol use) factors.

The assumed relationship between sexual compulsivity and sexual offending has also been adopted by many practitioners and governmental policies, 6–8 and while multi-dimensional models of sexual offending are often well validated and offer multiple treatment-relevant pathways, sexual compulsivity is often only one component of these models, such as the Motivation−Facilitation Model, 4 referenced obliquely in other models such as Ward & Beech’s 9 integrated theory of sexual offending, or more commonly not considered relevant at all, such as Ward et al.’s 10 highly influential self-regulation model, or Ward, Mann, & Gannon’s equally influential Good Lives Model of offender rehabilitation. Nonetheless, policymakers and professionals often justify treatments and legislation with reference to biological defects that are assumed to lead to heightened sex drive and sexual compulsivity, in particular, abnormal levels of testosterone. 11 As such, early interventions focused on reducing sexual drive, including surgical procedures, and later, hormonal treatments. For example, Denmark legislated medical interventions for sex offenders as early as 1929, 12 followed by other European countries. In the United States, 10 states have also introduced castration laws for high-risk sex offenders. 13 Chemical castration has been incentivized in countries like Poland, Korea, and the Republic of North Macedonia. 14 , 15 In the United Kingdom, prisoners convicted of sexual offenses specifically with evidence of sexual hyper-arousal (eg, frequent sexual rumination, sexual preoccupation, difficulties in controlling sexual arousal, high levels of sexual behavior) have been considered for pharmacological treatment to reduce sexual drive (Winder et al. 16 ). The focus on reducing sexual compulsivity as a strategy to lower recidivism has shaped guidelines for the pharmacological treatment of sex offenders. The World Federation of Societies of Biological Psychiatry recommends hormonal treatments for individuals at moderate or high risk of sexual violence. 17 A recent Cochrane review 18 showed that most pharmacological treatments reduce libido by suppressing testosterone, and the interrelationship between sexual compulsivity and sexual offending assumed to exist to the point that some have simply stated that the use of libido reducing medication is emerging as “ an effective treatment for severe paraphilia ” ( 19 p. 244).

However, the assumed link between heightened sexual compulsivity and offending behavior remains unclear, given both the changing conceptualization, criteria, and nosology of sexual compulsivity over the years and a paucity of empirical evidence regarding the confluence of sexual compulsivity and paraphilic preference among sex offenders. Thibaut et al. 17 note that most individuals convicted of sexual offenses do not suffer from paraphilic disorders, and many people with paraphilia do not offend, they report that there is some evidence for an association between what they term “ sexual hyperactivity ” and paraphilic disorder, although “ the relationships and boundaries between hypersexual behaviour and paraphilic disorders are poorly known ” (p. 8). The current study will briefly review the disparate findings regarding the relationship between sexual compulsivity and pedophilic preference among people who have sexually offended, before testing the hypothesis that compulsive sexual behavior, according to the most recent criteria, and pedophilic preference are related to each other among convicted sexual offenders against children.

Sexual compulsivity among offenders: definitions and findings

Early research on sexual compulsivity among offenders produced mixed results. Blanchard 20 and Carnes 21 reported “hypersexuality” rates of around 50% in their samples, but they did not provide a clear operational definition of hypersexuality. More robust studies, such as Kingston and Bradford, 22 operationalized hypersexuality as 7 or more orgasms per week (total sexual outlet [TSO] ≥7). They found that hypersexuality predicted recidivism among men convicted of sexual offenses against children. However, failure to control for age, a key factor in both sex drive and recidivism, 23 limits these findings.

Marshall et al. 24 raised methodological concerns about studying hypersexuality and sexual offending. They explored sexual addiction in 114 prisoners convicted of sexual offenses and found that 43.9% exhibited signs of sexual addiction. However, the relationship between sexual addiction and recidivism was not examined. A large online study 25 of German males (n = 8718) used a simple measure of sexual desire (TSO, consisting of number of orgasms per week), and found no relationship with sexual offending. More recent research has incorporated the ICD-11 diagnosis of compulsive sexual behavior disorder 26 to define excessive sexual drive and the distress and dysfunction that derive from it. Castro-Calvo et al. 27 and Castellini et al. 28 found little or no relationship between compulsive sexual behavior and pedophilic interest among normative samples, the former using multi-dimensional measures of compulsive sexual behavior, the latter assessed using PPG and self-reported interest. Hertz el al. 29 studied hypersexuality in offenders using proposed DSM-5 criteria, finding that 6.6% met the criteria and that hypersexuality made a small but significant contribution to recidivism risk. These mixed findings highlight the need for more refined research to clarify the relationship between sexual compulsivity and offending.

Why is this relationship important?

The rationale for reducing sexual drive (and thus the need to compulsively satisfy it) is based on testosterone's role in sexuality, as it affects arousal, motivation, and aggression. 30 , 31 However, meta-analyses, like that by Wong and Gravel, 11 have produced mixed results regarding testosterone levels in offenders. For example, some studies even suggest that child molesters may have lower testosterone levels than controls. 32

The Cochrane review by Khan et al. 18 failed to identify robust randomized controlled trials on the efficacy of hormonal treatment in reducing recidivism. Trials tend to be small, with short follow-ups and insufficient risk stratification. The focus on reducing recidivism may also be misleading, given that offenders may be offered hormonal treatment some years after they have last offended (either due to a delay between the offense and the conviction, or due to them being offered treatment after serving a long prison sentence), ignoring the finding that re-offending rates drop sharply after years of natural desistance. 33 Similarly, Lewis et al. 34 reviewed the use of gonadotrophin-releasing hormone (GnRH) agonists and found no solid evidence for their effectiveness in preventing recidivism. These treatments are among the most invasive interventions for people who have committed sexual offenses against children, with significant side effects including osteoporosis, mood disturbance, and weight gain, which have been documented. 35 Ethical issues also arise, particularly around coercion and the blurring of treatment and punishment. 14

While the ethical issues involved in the provision of such treatment to offenders are clear, it may be noted that there is preliminary evidence to suggest that some groups of sex offenders may take anti-libidinal medication without any legal or social coercion, and a small group of self-identifying pedophiles who were treated in the Dunkenfeld Project in Germany, where patients may self-refer for treatment without professionals being legally bound to report them, are described by. 36 This group was characterized by significantly higher levels of sexual compulsivity, as compared to other patients at the Project, together with significantly higher levels of self-awareness. However, it is important to note that this was an extremely small group of patients, with over half of them not being available for post-treatment assessment, and though promising, the limited results do not appear to counteract the significant empirical limitations or ethical issues that have been discussed.

Rationale for the current study

From the literature reviewed to date, it would appear that the assumption of confluence between sexual compulsivity and pedophilic preference among sexual offenders against children has not been sufficiently established. Nonetheless, highly invasive interventions, with a limited evidence base for their efficacy, potentially iatrogenic effects, and unsolved ethical dilemmas, are increasingly used and rest upon this assumption being an established one. The current study aims to examine whether sexual compulsivity and pedophilic preference are related in a sample of convicted sex offenders against children.

Method

Participants

Participants were recruited from 3 Israeli prison treatment units, with 97% agreeing to participate. The study included 95 male prisoners convicted of sexual offenses against minors. The average age was 43.57 years (SD = 16.59), and 30.5% were married. The mean education level was 11.7 years, and 14.7% had a prior prison sentence. The treatment units provide a community-based treatment program lasting approximately 18 months, based on relapse prevention and cognitive-behavioral skill approaches, and acceptance to the units is typically offered to prisoners during the later stages of their sentence. All participants were at various stages of their treatment program.

Procedure

Questionnaires were printed on hard copies. The questionnaires were authorized by the institutional ethics committees (academic and Israel Prison Service research committees). The researchers presented the rationale and the procedure for the research at the end of the weekly treatment unit community meeting, answering questions that participants may have, stressing anonymity, stating that non-participation would not be noted by prison staff or have any effect on their progress in prison and that they had the right to withdraw at any stage. Those who wished to participate were then provided with informed consent forms to sign.

Measures

The Individual-Based Compulsive Sexual Behavior Scale (I-CSB 37 ) was used to assess compulsive sexual behavior across 4 domains: unwanted consequences (eg, “I feel that my sexual fantasies hurt those around me”), lack of control (eg, “I waste lots of time with my sexual fantasies”), negative affect (“I feel bad when I don’t control my sexual urges”), and using sexual behavior for emotional regulation (eg, “I feel that my sexual fantasies provide a way to deal with my painful feelings”). The scale consists of 24 items, rated on a 7-point Likert scale. It has a strong factor structure, and significant correlations have also been obtained with numerous behavioral measures of sexual excitation, arousability, and hypersexual disorder. 37 The questionnaire has been used successfully in previous research differentiating clinical populations of Sexaholics Anonymous participants from normal (Efrati & Gola), 38 and from people convicted of sexual offenses. 39 In the current study, alpha coefficients were high (0.84-0.86 for the 4 scales, 0.94 for the total score). In their responses to this questionnaire, participants were asked to think about themselves throughout their lives, and not just at the present time in prison, where sexual fantasy and sexual behavior would likely be limited compared to their past lives.

The Screening Scale for Paedophilic Interests (SSPI 40 ) is a brief measure of sexual attraction to children based on 4 items taken from file information, though in the case of the current research, participants provided the information themselves due to reasons of anonymity, as specified by the Israel Prison Service ethics committee. Sexual attraction to children is measured on a scale between 0 and 5 based on the presence or absence of 4 behavioral items, with a score of 0 or 1 indicating the absence or presence of the item (or a score of 0 or 2 for the first item): male child victims, more than 1 child victim, prepubescent victims, and extra-familial victims. SSPI scores ranged from 0 to 5, with higher scores indicating greater pedophilic preference. In Seto et al.’s original research, 40 significant correlations were found between total scores and phallometric data, making the SSPI a proxy measure for pedophilic sexual arousal, and a very recent study involving the phallometric scores of 1953 offenders against children revealed consistently positive correlations between SSPI-2 scores and phallometric measures of pedophilia and pedohebephilia 41 In a review of the SSPI by Helmus et al., 42 they concluded that the “SSPI is useful as a brief measure of sexual interest in children. It performs equally as well as indicators of deviant sexual interests from other risk scales.” (p. 42). Similarly, in a review of methods of measuring pedophilic sexual interest, Carvalho et al. 43 referred to the SSPI as being “possibly the most recognized (behavioural) measure” of pedophilic sexual interest (p. 387). While the SSPI is still in widespread use in the research literature, a more recent version, the SSPI-2, which adds a single item regarding use of child pornography, has been found to have significant predictive validity in terms of sexual recidivism. 44 Distribution of SSPI scores in the current sample is displayed in Table 1 .

Table 1 Distribution of SSPI scores.
Score

Percent (N)
1 6.3 (6) 2 9.5 (9) 3 29.5 (28) 4 27.4 (26) 5 27.4 (26)

Distribution of SSPI scores.

Results

SSPI and on the I-CSB scores were initially analyzed in order to investigate whether there was a correlation between pedophilic preference and compulsive sexual behavior (general CSB), in line with the main hypothesis of the current research. Given the finding that the relationship was significant, further analysis of its relationship with the different components of compulsive sexual behavior was conducted in order to better understand whether the relationship is moderated by 1 or more of its component factors. Finally, the relationship between the I-CSB scales and the component items of the SSPI were explored, as some studies have found that only the relationship between the SSPI and other variables is differentially moderated by its component items depending on the subject under study, such as recidivism in Helmus et al.’s 42 research, or denial of offending in a study by Stephens et al. 45

SSPI scores (ie, extent of pedophilic preference) and their association with compulsive sexual behavior (sexual-related unwanted consequences, negative affect, lack of control, affect dysregulation, and general CSB) are presented in Table 2 . The analysis indicated that the SSPI scores of the participants in the current sample were significantly and negatively associated with compulsive sexual behavior. However, the results indicate that the association was primarily due to a lack of negative affect among those with higher SSPI scores. Further analysis, as displayed in Table 3 , indicated a further and more specific association between lack of negative affect and specific SSPI components, namely, multiple victims and victims below the age of 11 years. It is also noted that 2 other components (unwanted consequences and affect dysregulation) were not significantly related to SSPI scores, but a non-significant negative trend did emerge.

Table 2 Intercorrelations between SSPI and compulsive sexual behavior.   SSPI (n = 95) Unwanted consequences −0.19~ Negative affect −0.23 a Lack of control −0.07 Affect dysregulation −0.19~ CSB −0.22 a a
P  < .05; ~ P  = .06.

Intercorrelations between SSPI and compulsive sexual behavior.

P  < .05; ~ P  = .06.

Table 3 Links between SSPI clusters and compulsive sexual behavior.   SSPI male victim SSPI 2 + victims   Male (n = 81) Female (n = 14) – 2+ (n = 73) 1 (n = 22) –  
M

SD

M

SD

t(93)

d

M

SD

M

SD

t(93)

d
UC 2.23 1.23 2.34 1.16 0.32 0.09 2.14 1.19 2.61 1.25 1.62 0.39 NA 3.11 1.96 2.98 2.22 -0.22 -0.06 2.83 1.87 3.95 2.16 2.37 a 0.57 LC 2.10 0.99 2.29 1.11 0.62 0.19 2.06 0.94 2.36 1.19 1.25 0.30 AD 1.99 1.46 2.54 1.39 1.59 0.46 1.98 1.18 2.36 1.22 1.38 0.32 CSB 2.28 1.03 2.49 1.15 0.69 0.19 2.19 0.99 2.71 1.13 2.10 a 0.51
SSPI unrelated victim

SSPI victim < age 11

Unrelated (n = 48)

Related (n = 47)

<Age 11 (n = 59)

>Age 11 (n = 36)


M



SD



M



SD



t(93)



d



M



SD



M



SD



t(93)



d

UC 2.14 1.22 2.36 1.23 0.85 0.17 2.08 1.23 2.53 1.16 1.73 0.37 NA 2.81 1.84 3.37 2.11 1.39 0.29 2.74 1.90 3.65 2.03 2.21 a 0.47 LC 2.17 0.99 2.08 1.03 -0.43 -0.08 2.12 0.98 2.13 1.05 0.03 0.01 AD 2.04 1.07 2.11 1.31 0.27 0.05 1.98 1.14 2.21 1.28 0.87 0.18 CSB 2.23 1.01 2.39 1.07 0.94 0.16 2.18 1.02 2.52 1.05 1.59 0.38 Abbreviations: d , Cohen’s d ; UC, unwanted consequences; NA, negative affect; LC, lack of control; AD, affect dysregulation; CSB, CSB total score. a
P  < .05.

Links between SSPI clusters and compulsive sexual behavior.

Abbreviations: d , Cohen’s d ; UC, unwanted consequences; NA, negative affect; LC, lack of control; AD, affect dysregulation; CSB, CSB total score.

P  < .05.

Discussion

This study is 1 of the first to examine the relationship between pedophilic preference and compulsive sexual behavior in incarcerated offenders. The findings suggest that offenders with high levels of pedophilic interest do not experience the key symptoms of CSB, and there is some evidence to suggest that they are in control of their sexual behavior, are less likely to feel distress around, it, less likely to engage in it as a way of coping and are aware of its consequences (and by extension may make a rational choice to engage in it). They would appear to be qualitatively different to offenders against children who have lower levels of pedophilic interest, where the motivations for the latter may be rooted in personal and interpersonal crises, emotional identification with children, and will tend to be accompanied by high levels of distress (especially, as the current study found, when they offend against older minors, a victim group associated with offenders with lower levels of pedophilic interest). As such, treatment needs may be different according to the level of pedophilic interest, but the targeting of compulsive sexual behavior per se would not appear to be warranted for either group.

In practice, it may be the case that the reduction of sexual drive as a target for treatment may only be appropriate for those who genuinely suffer from compulsive sexual behavior, regardless of their level of specific paraphilic interest. Lievesley et al. 46 studied 10 adult males serving a prison sentence for sexual offenses who were receiving anti-androgen medication on a voluntary basis, many of whom were sentenced for their first time, had not been convicted of contact sex offenses, or had not offended against children, but had been found to demonstrate “behavioural facets of Problematic Sexual Arousal, e.g. compulsive or hypersexual behaviour.” Their participants reported complex, multi-faceted and ambivalent factors underlying their consent, but strikingly these included many intrinsic motivations and positive experiences of treatment, where for many of the participants their high level of sexual interest and drive was experienced as being ego-dystonic and risky, and their reduction a welcome relief. The results would therefore imply that specifically targeting offenders with a high level of pedophilic interest in the absence of evidence of compulsive sexual behavior would be counterproductive, especially if there is an expectation that pharmacological treatment would be sustained over the long term in order for it to be effective in reducing recidivism.

The findings of the current study, in particular with regard to the lack of negative affect or unwanted consequences among the participants with higher levels of pedophilic interest, would support the overwhelming evidence for deficits in social cognition among sex offenders 47 as their lack of emotional response may in part derive from the beliefs such as children being sexual beings who are not harmed by their behavior. 48 As the current study found, negative affect was particularly lower for items of the SSPI, which would appear to be most criminogenic (more than 2 victims and prepubescent victims), and this may explain why interventions focusing on increasing consequential thinking and victim empathy have been found to be effective for high-risk offenders in particular. 49 Finally, the non-significant trend toward a lack of emotional dysregulation among those with higher levels of pedophilic interest would suggest that their urge to engage in sexual activity does not represent a way of escaping from difficult emotions or solving problems. This would appear to largely obviate the importance of emotional regulation deficits as an established risk factor for sex offenders 50 for those with high levels of pedophilic preference, meaning that treatment goals based on this factor may be more appropriate for those whose offending does not derive primarily from pedophilic interest.

There is increasing evidence for other foci for treatment that do not involve direct interventions for sexual compulsivity. Early childhood abuse or predatory tendencies have been found to underlie pedophilic preference, 51 and maladaptive early childhood schemas that have been found to be related to compulsive sexual behavior among sex offenders 39 are also likely to be underlying the personality disorders that have been found to be of high prevalence among sex offenders (Eher et al.). 52 Attachment difficulties have also been found to be present among many sex offenders, where the need for intimacy and difficulties in obtaining it in a pro-social manner are considered to be criminogenic needs, 53 do not directly involve sexual functioning or drive, and are of particularly high prevalence among female sex offenders. 54 Emerging models like the Good Lives Model and Circles of Support and Accountability require further evaluation but would appear to be effective in increasing motivation for high-risk offenders to desist from offending. 55 , 56 Generalized impulsivity, in particular, ADHD, has been found to be of high prevalence among men with paraphilic disorders (Soldati et al.), 57 and this may also suggest a promising new area in the search for less invasive pharmaceutical intervention. Additionally, psychopathy, a strong predictor of recidivism, 5 is not necessarily considered to be related to pedophilic interest, though the finding in the current study that negative affect is negatively correlated with pedophilic interest would suggest that psychopathy-targeted interventions may be beneficial for this group.

Limitations

While the absence of major mental illness was a criterion for the subjects’ admission to the treatment units from where they were drawn, psychiatric comorbidity was not assessed for the purpose of the current research, thus undetected anxiety and mood disorders may have had some influence on the self-report of subjects (in particular with reference to the affective components of the I-CSB, although the prevalence in the literature of such disorders among incarcerated sex offenders, while higher than the normal population, is not overwhelming [7.7% for anxiety disorders and 7.6% for mood disorders in Eher et al.’s 52 sample]. The reliance on self-report data may have introduced bias, although it may be noted that the wide distribution of SSPI scores, the majority of which were at the higher end of the scale, would appear to reflect a willingness to openly admit to severe offending characteristics, and would likely signify that self-report was not significantly compromised by socially desirable responding, possibly due to the anonymity of participants and their involvement in treatment. While this is the case, the study sample consisted of offenders engaged in treatment and who admit their offenses, which limits the generalizability to non–treatment-seeking populations. Finally, the current research treated pedophilic interest as dimensional, and indeed, the obtained scores on the SSPI were evenly distributed among the participants. While some researchers have supported the notion of sexual preference for children being a dimensional construct (eg, Mackaronis et al.), 58 recent research has indicated that pedophilic preference is likely to be a taxon, as opposed to a dimensional construct. 59 This was not operationalized in the current research, due to difficulty in obtaining a cut-off score for the SSPI measure that was used, Brankley et al. 59 used the SSPI-2, which includes a further item of child pornography that was not relevant for the subjects of the current study, but further research utilizing taxonomic criteria for pedophilic interest may reach different results that would have implications for better identifying offenders in need of treatment. However, while Brankley et al. 59 found that 48% of their sample met their criteria for the pedophilia taxon, a recent paper by Bergner-Koether et al. 60 found rates of exclusive pedophilia ranging from 17% to 36%. Such disparities would therefore indicate that analysis of data using a taxonomic rather than a dimensional approach would be premature at this stage.

Conclusion

This study disentangles the relationship between compulsive sexual behavior and pedophilic preference, showing that they are not inherently linked in incarcerated offenders. The findings do not counter-indicate the use of libido suppressants for individuals convicted of sexual offenses but do question the current trend of targeting sexual compulsivity through hormonal treatments among high-risk offenders and may suggest that more nuanced, individualized treatment strategies are needed. This would also support a recent initiative that disentangles risk from prescribing, ensuring that libido suppressants are given to those who would most benefit (such as those with very high level of sexual activity or those who find it difficult to control their actions), with provision given for informed consent and ensuring that medication is always prescribed in conjunction with psychological therapies. 61 The current study has also sought to highlight the multiple psychological mechanisms underlying high risk sexual offending (whether due to heightened pedophilic interest or a compulsive need to act on it), which can be the focus of therapeutic intervention, and have been found to significantly lower recidivism in their own right, or in conjunction with testosterone-lowering medication, as compared to the use of the latter alone. 62

Supplementary Material