ISSN 1862-2941

WelcomeOnline-Issues2-20191-20191/2-20182-20171-20172-20161-20162-20151-20152-20141-20142-20131-20132-20121-20122-20111-20112-20101-20102-20091-20092-20081-20082-20071-20073-20062-20061-2006Guidelines to authorsAbout usLegal NoticePrivacy



Online-Issues » 2-2006 » Lea H. Studer, A. Scott Aylwin

Sexual Offender Subtyping: The Incest Offender Question

Lea H. Studer, A. Scott Aylwin
Phoenix Program, Alberta Hospital Edmonton and Department of Psychiatry,
University of Alberta, Edmonton, Alberta, Canada

[Sexual Offender Treatment, Volume 1 (2006), Issue 2]


Aim/Background: Current wisdom in the risk assessment of sexual offenders is that incest offenders have a small chance of re-offending.  Previous research has shown that re-offense rates even for untreated offenders who choose victims from within the family, range from 4-10%.  Flowing from this belief are further assumptions that they are less dangerous and do not require intensive treatment, if they require any at all.
Material/Methods: The paper summarizes the findings of two prior reports which examined a sample of in-patient child molesters attending treatment at the Phoenix Program.
Results: Findings reveal a great deal of overlap between categories with incestuous offenders having numerous non-incestuous victims, and non-incestuous offenders also having
incestuous victims. There was also a great deal of overlap between these two groups in terms of their erotic preference testing responses. This was the finding even when biologically related fathers were examined separately, and when exclusively incestuous fathers were considered.
Conclusions: If there were a “pure” form of incest offender who poses little risk to other children, it is virtually impossible to discriminate who fits this category at the outset of treatment. Neither index offense nor erotic preference testing are very helpful in distinguishing this group from extrafamilial offenders. Beliefs about treatment requirements for incest offenders and their potential risk to victims
outside their own family need to be reconsidered.

Key words: Incest, pedophilia, child molester, intrafamilial, extrafamilial, risk assessment, treatment, psychotherapy.


The terms “fixated” and “regressed” are rooted in classical psychodynamic theory. These terms were first applied to sexual offenders by Groth and Birnbaum in 1978, and have been used for several decades as dichotomous labels applied to pedophiles.  According to their theory, fixated sexual offenders are a group whose primary sexual attraction is toward children. Regressed sexual offenders are considered to be a distinct group whose primary sexual attraction is not toward children. Instead, they possess a primary sexual attraction toward adults, but act out against children sexually at times of stress. Characteristics about these two groups have been investigated, unfortunately, this may not have brought the field any closer to a valid or reliable typology.
Closely related to the fixated versus regressed dichotomy, coming from clinicians, is the intrafamilial versus extrafamilial distinction which is frequently used by the legal system. Intrafamilial molesters (i.e., the so-called incest offenders who limit their sexually abusive behavior to family members) are seen as being most representative of the regressed pedophile. Most of these men have partners to whom they presumably have some sexual access (and success), but in response to life stress, marital discord, or other similar condition, opt to act out sexually toward their own children. Alternatively, the extrafamilial molesters are most representative of fixated offenders, men who are likely to invest a significant amount of energy facilitating sexual activity with children, and have little sexual interest in adults.
Studies have examined differences between these various sexual offender subtypes, especially their risk to recidivate. It is now widely held that extrafamilial offenders are more likely to re-offend, are more dangerous to the public, show deviant sexual arousal on erotic preference testing, and are in need of more intensive treatment than incest offenders. It is also believed that once identified and convicted, incest offenders have a relatively small likelihood of re-offending (Marshall & Anderson, 1996; McGrath, 1991; Quinsey 1986).  An example of how firmly entrenched this has become, can be found in official position statements of the Association for the Treatment of Sexual Abusers (ATSA). This influential body clearly differentiates between pedophiles and child molesters along these very lines in their public policy statements (i.e., Pedophiles and child molesters: the differences; ATSA, 2006). Also, the following was formally adopted by the ATSA Executive Board of Directors in 1996: “…at present, the research literature indicates that re-offense rates for untreated sex offenders, who choose victims from within the family unit, range from four to ten percent.” (p. 2) The distinction between these two offender subtypes, however, has come under scrutiny (Abel, Becker, Cunningham-Rathner, Mittelman, & Rouleau, 1988; Barsetti, Earls, Lalumiere, & Belanger, 1998; Conte 1991; Seto, Lalumiere, & Kaban, 1999; Studer, Aylwin, Clelland, Reddon, & Frenzel, 2002; Studer, Clelland, Aylwin, Reddon, & Monro, 2000).
Conte (1991) reported that there is little rationale for interpreting the sexual behavior of intra and extrafamilial child molesters as distinctly different, and further suggested that the fixated--regressed dichotomy has lost its clinical value. Upon a review of relevant literature, Wilson (2004) argued that “…an inference of risk is merited for the majority of children in incestuous families.” (p. 152) Phelan (1986) reported that among 102 incestuous families referred to an out-patient treatment clinic, sexual abuse of more than one child was not uncommon. Approximately one third of the incest fathers had multiple victims within the same family. Among the biological fathers specifically, this occurred in 50% of 46 cases. In an examination of 40 female victims of incest, 28% knew of siblings that were also being molested, and another 25% had unconfirmed suspicions that this had occurred (Herman & Hirschman, 1981).
Barsetti et al. (1998) have reported rates of deviant sexual arousal among incest offenders that were higher than would be expected if these individuals were actually representative of the regressed offender subtype. Using phallometry (the procedure typically used in erotic preference testing for males) to compare sexual arousal responses among n = 19 intrafamilial molesters and n = 20 extrafamilial molesters, Barsetti et al. found that both groups responded differently (more deviant) than a group of normal controls. However, the groups were virtually indistinguishable from each other in terms of arousal response to pedophilic stimuli. That is, the intra and extrafamilial offenders responded with a virtually identical degree of sexual arousal to audio taped scenarios of child sexual abuse. Although being aware of the limitations of erotic preference testing, as identified by Marshall (2006; Marshall & Fernandez, 2000) there is some reliance on this procedure in the field, and therefore this data is included here. 
Two of our own previous studies have looked at this issue from different vantage points. Studer et al. (2000) examined the degree to which incestuous and nonincestuous offenders had victims outside these categories and considered the frequency with which these categories actually tended to overlap. Studer et al. (2002) examined differences in erotic preference testing results among incestuous and nonincestuous molesters.
Both studies involved samples derived from the cohort of voluntary treatment participants in the Phoenix Program at Alberta Hospital Edmonton. This treatment is provided in a secure psychiatric hospital and has been described elsewhere (Studer, Reddon, Roper, & Estrada, 1996). All patients were sexual offenders who transferred from correctional facilities to hospital for treatment.  All of the men in both studies committed offenses against children. These offenses may have been their index (or referring) crime, or they may have been crimes self-reported during treatment. There were 328 men available at the time of Studer et al. (2000) for inclusion in the studies. Of this group, erotic preference testing data was available for 217 patients and these results were reported in Studer et al. (2002).


In both studies we maintained identical parameters for defining incestuous and nonincestuous groups. The definition of incestuous encompassed all offenses against children over whom the offender had guardianship and were separated by a single generation (i.e., natural fathers, step-fathers, adoptive fathers, and foster fathers). Being aware of the complexities of the definition of incest, an arbitrary decision was made to include only the single generation caregiver, due to the dynamics of that relationship. From a clinical perspective, the dynamics in a direct parental relationship are, and should be, quite different from those between other family members (e.g., brother and sister). For this reason, all other victim-offender relationships (e.g., no relationship, nonrelatives, brothers, and grandfathers) were considered “nonincestuous”.
A father biologically related to his own children best represents general perceptions about what constitutes incest. So, for the sake of completeness and rigor, specific analyses of fathers biologically related to their children were also undertaken regarding both histories of other offenses and primary erotic preference.  It is possible that offenders fitting the very narrowest definition of incest might show differences due to the societal “taboo” of sexual activity with one’s own natural children. Interestingly, Seto et al. (1999) have even worked from the premise that overcoming the incest taboo should logically require a greater attraction toward children than that of extrafamilial molesters. A subset of offenders in our database who had only incestuous (but not necessarily biologically related) victims was also identified and examined as to their primary erotic preferences.
In terms of historical data, both prior offenses and self report of offenses were used to determine the number of additional victims acknowledged by any particular patient Studer et al., 2000).  The self report and legal convictions referred to victims under 14 years of age. Again, we wished to utilize the most restrictive definition of offenders against children, so as not to bias our results. People were included in the sample, regardless of length of time in the program (which may be correlated with willingness to self report additional victims).  That is, our data could still represent an under reporting of additional victims as some members of the sample were only in treatment a short while and may not yet have come to trust the group process or treatment staff sufficiently to reveal all their additional victims. 
Regarding the erotic preference testing results, the primary response was used, rather than secondary responses. It is quite common to have a range of stimuli which may cause some degree of arousal. The one age category which elicited a notably more elevated response than the others is referred to as the primary response. If the individual responded equally to all age categories of stimuli, this is referred to as pangynephilia. Four categories of primary response were delineated: pre-pubescent, pubescent/hebephilic, adult, and pangynephilic.  No one responded equally to all ages of males and therefore the term panandrophilic is not used (Studer et al., 2002). The type of testing changed during the study period with the majority having volumetric measurement and the remaining having the strain gauge technique.


The sample was first examined with regard to actual offending behavior (Studer et al., 2000). The sample of 328 child molesters was first divided into men whose index offense was incestuous (n = 150), and men whose index offense was nonincestuous (n = 178). The frequency with which each group reported having other victims was then determined from extensive file review. The frequency with which these groups committed other incestuous and other nonincestuous offenses is reported in Table 1. Among both groups of offenders it was unlikely that men would report having no other victims.


index offence (n=150)

index offence (n=178)

Other incest victims

22.0% (n=33)

12.9% (n=23)

Other nonincest victims

58.7% (n=88)

78.7% (n=140)

No other victims

33.3% (n=50)

18.5% (n=33)

Note. Columns do not sum to 100% due to overlap of sample members in rows 1 and 2. Adapted from “Re-thinking Risk Assessment for Incest Offenders” by L.H. Studer, S.R. Clelland, A.S. Aylwin, J.R. Reddon and A. Monro, 2000, International Journal of Law and Psychiatry, 23, p. 19. Adapted with permission of the authors.

Table 1: Prevalence of Incest and Nonincest Victims Among N=328 Child Molesters

The group of 150 men whose index offense was incestuous was further subdivided into those who were biologically related to their victims versus those who were not. As reported in Table 2 the majority of men in both these incestuous subgroups reported having nonincestuous victims in addition to their index offense(s).  The groups did not differ significantly, Chi2 (1, N = 150) = 1.173, p > .05, two-tailed.


victims (n=60) 

victims (n=90)

Other nonincest victims

53.3% (n=32)

62.2% (n=56)

Note. Percentages were compared using a point-biserial correlation coefficient (identical to chi-square); Chi2 (1, N = 150) = 1.173, p > .05, two-tailed. Adapted from “Re-thinking Risk Assessment for Incest Offenders” by L.H. Studer, S.R. Clelland, A.S. Aylwin, J.R. Reddon and A. Monro, 2000, International Journal of Law and Psychiatry, 23, p. 19. Adapted with permission of the authors.

Table 2: Prevalence of Other Nonincest Victims Among N=150 Incestuous Fathers

The primary sexual preferences of a subset of 217 child molesters is reported in Table 3 (Studer et al., 2002). The sample was first divided into incestuous and nonincestuous offenders. Those with an incestuous index offense were less likely to have a primary erotic preference to prepubescent children (p<.01), and more likely to have a primary erotic preference to adults (p<.01).  The two groups did not differ in terms of theprevalence of pangynephilia or pubescent/hebephilia as a primary preference (p>.05). If the incestuous offenders truly were a group of regressed pedophiles who were primarily attracted to adults, then one would expect the prevalence of primary attraction to any category other than adults to be negligible. Clearly this was not the case as only 36.9% of the incestuous offenders demonstrated a primary erotic preference to adults. 


Index offense

Primary preference





29.8% (n=34)

12.6% (n=13)



43.0% (n=49)

40.7% (n=42)



19.3% (n=22)

36.9% (n=38)



7.9% (n=9)

9.7% (n=10)


Note. Percentages were compared using Pearson’s chi-square. Overall Chi2 = 13.72, P = .003. * Denotes p < .01, two-tailed. Adapted from “Primary Erotic Preference in a Group of Child Molesters,” by L.H. Studer, A.S. Aylwin, S.R. Clelland, J.R. Reddon and R.R. Frenzel, 2002, International Journal of Law and Psychiatry, 25, p. 176.  Adapted with permission of the authors.

Table 3: Primary Erotic Attraction of 217 Child Molesters According to Index Offense

The group of incestuous fathers was further divided into a group of men who were biologically related to their victim(s), and a group who were not (i.e., step-fathers, foster fathers). The results of this subdivision on primary erotic preference are reported in Table 4. There were no statistically significant differences in primary sexual preference between incestuous fathers who were biologically related to their victim, and those who were not. 


Relationship to Incestuous Victim

Primary preference

Biological (N=44)

Nonbiological (N=59)



15.9% (n=7)

10.2% (n=6)



31.8% (n=14)

47.5% (n=28)



40.9% (n=18)

33.9% (n=20)



11.4% (n=5)

8.5% (n=5)


Note. Percentages were compared using Pearson’s chi-square. Overall Chi2 = 2.72, P = .436. Adapted from “Primary Erotic Preference in a Group of Child Molesters,” by L.H. Studer, A.S. Aylwin, S.R. Clelland, J.R. Reddon and R.R. Frenzel, 2002, International Journal of Law and Psychiatry, 25, p. 177. Adapted with permission of the authors.

Table 4: Primary Erotic Attraction of 103 Incestuous Fathers According to Biological Relationship With Victim

A further analysis was completed with men whose index, prior, or self-reported offenses included only incestuous victims. This subgroup showed a similar profile on erotic preference testing and the results are reported in Table 5.

Primary preference


12.5% (n=6)


39.6% (n=19)


43.8% (n=21)


4.2% (n=2)

Note. This group is wholly subsumed by the larger sample from which all other erotic preference test data was derived.

Table 5: Primary Erotic Attraction of 48 Male Sex Offenders With Conviction (Index or Prior) or Self-report of Incestuous Offenses Only


As first commented on by Abel and his colleagues in 1988, the paraphilias are far from exclusive entities. This also holds true within the paraphilia of pedophilia regarding the categories of incestuous and nonincestuous acts against children. Individuals who commit these types of offenses, may indeed be more prone to committing offenses of a similar type, but very often (most often in our studies) their incestuous or nonincestuous acts are simply a part of a larger overarching repertoire of deviant sexual behavior.
It is very apparent from this and other studies that index offense and prior sexual offenses are very poor indicators of the diversity and extent of actual deviant sexual behavior (Weinrott & Saylor, 1991). This is an understood truism for the category of extrafamilial child molester, as it is almost assumed that these offenders have additional victims. However, for incestuous offenders some people in corrections/offender management find this same result surprising or unlikely. Indeed, more than half of our sample with an incestuous index offense admitted to also having non-incestuous victims.
There was a significant finding in Studer et al. (2002) of incestuous offenders being more attracted to adults and nonincestuous offenders to pre-pubescent children, as current wisdom would predict. However, the overlap in sexual response between categories was substantial and therefore one cannot rely on erotic preference testing to differentiate these two groups.
Unexpectedly, the majority of both incestuous and nonincestuous offenders showed their primary preference to the adolescent hebephilic category. This may be expected for the incestuous group, but was more surprising for the nonincestuous group. The additional analysis of men who are only known to have offended incestuously (i.e., no nonincestuous victims reported), again showed similar erotic preference trends with slightly more than half the group showing some type of “deviant” primary attraction. 
Perhaps most surprising was that separating biological fathers by index offense as one group, yielded a similar result. These men still admitted to nonincestuous victims (or had a conviction for a nonincestuous victim) at a rate of >50%.  Further, in this group of men who theoretically best represented the regressed non-pedophilic molester 59.1% showed some type of “deviant” arousal pattern on erotic preference testing with either pre-pubescent, pubescent (hebephilic) or pangynephilic attractions.
In summary, there are significant problems with the distinction between incestuous and nonincestuous offenders against children. Perhaps a continuum of behavior is a more accurate postulation, with a multitude of factors contributing to its expression either within or outside the family unit. More important to us, given our position as treatment providers, are the presumptions and assertions that continue to be attributed to those labeled as incest offenders. Many court and case management decisions are influenced by this label, and these may not be appropriate in all cases. Being an incest offender makes one much more likely to be assessed as low-risk to recidivate based on the existing actuarial risk assessment instruments. In some regions (e.g., Canada), the case management decisions following from this are based on the premise that risk and treatment needs are directly related. That is, that an incest offender who is deemed “low risk” is also deemed “low needs” in terms of treatment. Unfortunately, an offender’s need or suitability for treatment is hardly related to their status as an intra or extrafamilial offender. Indeed, it could be argued that an incest offender who maintains a relationship with his family would be in greater need for comprehensive treatment so as to perpetuate the least amount of future harm to an already flawed family system. The limitations of the studies summarized here revolve mainly around how generalizable these results are to other groups. Our sample is somewhat self-selected as they are voluntary. They were serving prisoners, prior to their admission to our program. This might imply that their offenses were more serious, invasive, of longer duration, or more numerous than other samples. This caveat limits how applicable our findings could be to other samples. We also had to make an arbitrary distinction regarding definitions of incestuous behavior, and decided to make this on clinically relevant variables rather than legal distinctions.
The findings reviewed here should remind clinicians and researchers that we must remain open to the possibility that some of our current assumptions regarding risk may be faulty and require ongoing refinement.


  1. Abel, G.G., Becker, J.V., Cunningham-Rathner, J., Mittelman, M., & Rouleau, J.L. (1988). Multiple paraphilic diagnosis among sex offenders. Bulletin of the American Academy of Psychiatry and Law, 16, 153-168.
  2. Association for Treatment of Sexual Abusers. (1996). Reducing sexual abuse through  treatment and intervention with abusers. Beaverton, Oregon: Author.
  3. Association for Treatment of Sexual Abusers. (2006). Pedophiles and child molesters: The differences. Retrieved April 3, 2006,
  4. Barsetti, I., Earls, C.M., Lalumière, M.L., & Belanger, N. (1998). The differentiation of  intrafamilial and extrafamilial heterosexual child molesters. Journal of Interpersonal Violence, 13, 275-286.
  5. Conte, J.R. (1991). The nature of sexual offenses against children. In C.R. Hollin, & K. Howells (Eds.), Clinical approaches to sex offenders and their victims (pp.11-34). Toronto: Wiley.
  6. Groth, A.N., & Birnbaum, H.J. (1978). Adult sexual orientation and attraction to underage persons. Archives of Sexual Behavior, 7, 175-181.
  7. Herman, J., & Hirschman, L. (1981). Families at risk for father-daughter incest. American Journal of Psychiatry, 138, 967-970.
  8. Marshall, W.L. (2006). Clinical and research limitations in the use of phallometric testing  with sexual offenders. Sexual Offender Treatment, 1, 1-18.
  9. Marshall, W.L., & Anderson, D. (1996). An evaluation of the benefits of relapse prevention programs with sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 8, 209-221.
  10. Marshall, W.L., & Fernandez, Y.M. (2000). Phallometric testing with sexual offenders: Limits to its value. Clinical Psychology Review, 20, 807-822.
  11. McGrath, R.J. (1991). Sex offender risk assessment and disposition planning: A review of  empirical and clinical findings. International Journal of Offender Therapy and  Comparative Criminology, 35, 328-350.
  12. Phelan, P. (1986). The process of incest: Biologic father and stepfather families. Child  Abuse and Neglect, 10, 531-539.
  13. Quinsey, V.L. (1986). Men who have sex with children. In D.N. Weisstub (Ed.). Law and  Mental Health: International perspectives (Vol. 2, pp. 140-172). New York:  Pergamon Press.
  14. Seto, M.C., Lalumière, M.L., & Kaban, M. (1999). The sexual preferences of incest offenders. Journal of Abnormal Psychology, 108, 267-272.
  15. Studer, L.H., & Aylwin, A.S. (in press). Pedophilia: The Problem with Diagnosis and Limitations of CBT in Treatment. Medical Hypotheses.
  16. Studer, L.H., Aylwin, A.S., Clelland, S.R., Reddon, J.R., & Frenzel, R.R. (2002). Primary erotic preference in a group of child molesters. International Journal of Law and Psychiatry, 25, 173-180.
  17. Studer, L.H., Clelland, S.R., Aylwin, A.S., Reddon, J.R., & Monro, A. (2000). Rethinking risk assessment for incest offenders. International Journal of Law and Psychiatry,  23, 15-22.
  18. Studer, L.H., Reddon, J.R., Roper, V., & Estrada, L. (1996). Phoenix: An in-hospital treatment program for sexual offenders. Journal of Offender Rehabilitation, 23, 91-97.
  19. Weinrott, M.R., & Saylor, M. (1991). Self-report of crimes committed by sex offenders. Journal of Interpersonal Violence, 6, 286-300.
  20. Wilson, R.F. (2004). Recognizing the threat posed by an incestuous parent to the victim’s siblings: Part I: Appraising the risk. Journal of Child and Family Studies, 13, 143-162.

Author address

Lea H. Studer, M.D.
Phoenix Program, Unit 3-3
Alberta Hospital Edmonton
17480 - Fort Road
Box 307, Edmonton, Alberta
Canada, T5J 2J7
Phone: (780) 472-5251
Fax: (780) 472-5351