Unmastered Masturbation Conflict

Unmastered Masturbation Conflict
<< Young School-Age Sexually Aggressive Children >>
William N. Friedrich, M.D.¹, William J. Luecke, M.D.²

¹Mayo Clinic, Rochester, Minnesota
²Good Samaritan Mental Health Center, Puyallup, Washington

Professional Psychology: Research and Practice [Vol. 19 (2), pp. 155-164. April, 1988. Received Dec 22, 1986, revision received June 11, 1987]

As the full spectrum of sexual abuse sequelae is becoming known, a subset of child victims has been noted to exhibit sexually aggressive behavior. Their sexual behavior far exceeds the mutual exploratory behavior normally seen in young children and resembles more closely the behavior of older sex offenders. We present data from the psychological evaluation of twenty-two 4- to 11-year-old children referred for sexually aggressive behavior, including intellectual, behavioral, projective, and parent-child relational quality. Behavior problem data from these children are contrasted with data from twenty-two 5- to 13-year-old boys who completed a sexual abuse treatment program. Some differences appear related to the development of sexual aggression (e.g., aggression level, nature of the abuse, and family functioning) and are discussed.

INTRODUCTION

Individuals working with sexual offenders have suggested a link between childhood sexual abuse and later sexual aggression. In fact, clinicians have noted that 32% of incarcerated adult sex offenders report a childhood history of sexual victimization (Groth, 1979). This percentage is even higher with "fixated" (as opposed to "regressed") adult offenders. Recently, adolescent sex offenders have been studied, and the indications are that 18% have a history of sexual victimization (Fehrenbach, Smith, Monastersky, & Deisher, 1986). Apparently, sexual victimization in childhood is somehow related to subsequent sexual offense, at least for a subset of adults and adolescents.

At this time, the literature on sexually aggressive behavior before adolescence is extremely scant. However, in a recent study of adolescent offenders, Fehrenbach et al. (1986) identified 22% of their sample as ranging in age from 11 to 13. This finding certainly suggests that there are younger sexually aggressive children. We were able to identify four recent papers that focused exclusively on aggressive sexual behavior in the younger child. Sexually aggressive behavior was identified in a subset of 16 sexualized children (Pomeroy, Behar, & Stewart, 1981). This behavior included forcing younger children to undress and to kiss and sexual experimentation with peers and siblings. However, a direct linkage to sexual abuse was seen only rarely by the authors; the primary contributor was seen as aggressiveness. Yates (1982) described 3 child incest victims who actively behaved in a sexual manner toward other children, adolescents, and adults. The suggestion was that these children had been conditioned to behave in a sexual manner. The same formulation was seen with a case of sister-sister incest, in which the older sister, who had been previously sexually abused, used her younger sister "to get my sexual fix" (Fortenberry & Hill, 1986). Last, an 8-year-old boy was reported to have sexually assaulted his mother. This child had an earlier history of aggressive behavior with his younger brothers (Arroyo, Eth, & Pynoos, 1984).

Recent research on the behavioral problems of sexually abused young children has indicated that young sexually abused boys differ from conduct-disordered boys primarily on the variable of sexual behavior problems (Friedrich, Beilke, & Urquiza, 1988). The same is true for sexually abused children aged 3-12 when contrasted with children receiving outpatient psychiatric therapy or children who ostensibly are normal (Friedrich, Beilke, & Urquiza, 1987). Sexual behavior problems in sexually abused children appear to be related to the child's age, who the perpetrator was, the number of perpetrators, the frequency and duration of the abuse, and the length of time since the last abuse (Friedrich, Urquiza, & Beilke, 1986). More specifically, Friedrich et al. (1986) found that younger children more recently abused by a parent or by multiple perpetrators, and whose abuse was frequent and long term, were more likely to behave in a sexualized manner.

The sexual behavior problems that were assessed in these studies were excessive masturbation, sexual preoccupation, and gender disturbance. Sexually aggressive behavior was not assessed, and this has prompted the development of the Child Sexual Behavior Inventory (Purcell, Beilke, & Friedrich, 1986), a 42-item measure of a wide variety of sexual behavior problems in children, including sexually aggressive behaviors. In an effort to understand sexually aggressive behavior in children, we examined literature on adult pedophiles, fully realizing that a direct comparison would not be possible. Clinical experience with adult pedophilic sex offenders has implicated four identifying characteristics of this population (Lanning, 1986): a long-term pattern of behavior, children as preferred sexual objects, evidence for well-developed techniques in obtaining victims, and sexual fantasies that focus primarily on children. We certainly are not advocating the sole use of adult-derived criteria for understanding sexually aggressive children. The two groups are quite different developmentally. However, these characteristics quite accurately describe the first sexually aggressive school-age child treated by the first author. He was the 9-year-old son of a single-mother daycare provider and had acknowledged sexually molesting at least 4 children in his mother's care. These offenses spanned 3 years (33% of his life); he preferred very young children (less than 2 years old), used sweets and personal attention to obtain time with victims, and reported that his biggest decision each afternoon when returning home from fourth grade was whether to work his paper route or to molest a child. He had been sexually abused by his mother's boyfriend between the age of 4 and 7 years. This behavior was significantly different from that of other children we had seen, who behaved in sexual but not aggressive or coercive ways and usually with children very close in age. Although this latter group frequently had a sexual abuse history, their sexual behavior was not the dominant force in their life that it was with this young boy or with other children whom we treated.

Our understanding of why some sexually abused children behave in a sexually aggressive manner that is over and above either the mutually exploratory behavior seen in young children who have not been sexually abused or the acute hyper-sexualization of the recently sexually abused young child is greatly aided by Finkelhor and Browne's (1985) conceptualization regarding the sources of trauma inherent in sexual abuse. They identified four sources of trauma: betrayal, powerlessness, stigmatization, and traumatic sexualization. Each of these sources of trauma can vary from one child to the next, depending on who the perpetrator is, the type of sexual abuse experienced, and whether force was used. A child experiences significant powerlessness when force is used with the abuse, and profound betrayal is presumed to be operative when the perpetrator is emotionally important to the child. A sense of stigmatization comes from the child's feeling "damaged" or "broken" and when the child is not supported emotionally by his or her family. Traumatic sexualization is presumed to occur simply because the child has been prematurely introduced to sexual behavior and because the duration of the sexual abuse and the use of force combine to make it even more traumatic. Our clinical impression is that children who behave in a sexually aggressive manner are likely to have been affected very deeply across each of these four sources of trauma. The manner in which a child copes with the sexual abuse and adapts in a nonaggressive and nonsexual manner to this trauma is a function of the integrity of the family and the consistent support that can be provided to the child. This support begins with the presence of a nonblaming and emotionally supportive significant caregiver, usually the mother. Her caretaking abilities cannot be compromised by her own psychopathology or chemical dependency. In addition, it is essential that these children have a history of secure attachment with their parents so that the child can experience optimal parental support.

For a specific child, characteristics that would mitigate against sexually aggressive behavior would be good social skills, the ability to inhibit and modulate behavior, and the cognitive capabilities to allow school to be a positive source of emotional support and an area in which the child can experience a much-needed sense of mastery over his or her life. However, these initial formulations about the development and maintenance of sexually aggressive behavior in young children are quite tentative, and to this date there exists minimal clinical literature that could help us to understand this phenomenon.

A further aspect that must be considered in the assessment and treatment planning with these children is developmental level. This must be considered both in terms of the age of onset of victimization, the duration of victimization, the age at which the child's own acting out began, and the age at which the child presents to the clinician. For example, a very young child who is recently sexually victimized and acts out aggressively almost immediately afterwards appears, in a punitive manner, to identify with the aggressor in an effort to undo the victimization or return to a state of personal control. This is quite different than an older child who has a history of school failure, behavior problems, and ongoing sexualized but nonaggressive behavior that results in impaired self-efficacy and identification as a "bad child," and who then begins to behave in a sexual and aggressive manner. The significant social-cognitive developmental changes that transpire between the age of 3 to preadolescence are significant and very likely interact in a variety of different ways.

Over the past 3 years, we have been referred an additional 22 young children for a psychological consultation that was specific with regard to possible sexually aggressive behavior. However, not all of these children had a documented history of sexual abuse. Several of these children were subsequently involved in therapy with one or the other of us. The assessment and treatment of these children has provided a unique opportunity to begin an understanding of the contributors to sexual aggression in young children, to derive several conceptual formulations as to its etiology, and to report on its treatment. In this article we focus solely on the assessment of these children and contrast them with a group of young boys referred for psychotherapy after sexual abuse. This latter group was not considered sexually aggressive. Given the increase in reported sexual abuse in the past decade, an increase in our understanding of sexual aggression in young children, and its relationship to previous victimization, is essential.

METHOD

Between July 1983 and August 1986, we received 22 referrals of children suspected of sexually aggressive behavior. These children ranged in age from 4 to 11 (M = 7.3 years), and 18 were male. Their sexual behavior was the primary reason for referral in all but 1 case. This latter child was referred for psychotherapy after discovery of her having been sexually abused, and shortly after treatment began, she began to act out sexually.

Criteria used to determine whether the behavior was truly sexually aggressive, and hence could be labeled as coercive sexual behavior, was evidence (from the victim, the caregiver, and/or the offending child) that the behavior was persistent (had occurred two or more times) and involved coercion and genital contact and that the referred child was at least 2 years older than the victim(s). The sexual behavior had to be significantly more aggressive than simply the mutually exploratory behavior that is developmentally quite common in children in this age range. According to the assessment, 6 of the 22 children did not meet these criteria, but they are described in this study because of the contrast that they provide with the remaining 16 children judged to be sexually aggressive.

The usual assessment procedure was to interview the child and the parent(s), to assess the child's intelligence, and to obtain a behavior checklist, projective drawings, and other projective measures (Rorschach, Thematic Apperception Test [TAT], Roberts Apperception Test for Children). Sometimes, objective personality assessment of the parent was also obtained (through the Minnesota Multiphasic Personality Inventory, or MMPI). Given the variability of each referral and the clinical settings, we did not adhere to a standardized assessment format. Each child was assessed in regard to the needs of the referral source (e.g., social services, another therapist, public health nurses) or to provide pretreatment data for one of us.

The group of 22 children were from primarily lower- to lower-middle class families, and the mode for maternal education was 12 years. All but 1 of the sexually aggressive children were from single-parent families, and 2 were not living with their natural mother at the time of referral. Of the 6 children not sexually aggressive, 4 were from single-parent families. In Tables 1, 2, and 3 we present descriptive data on the 22 children referred for a consultation.

In addition to the comparison group composed of the 6 children who did not meet the criteria for sexually aggressive behavior, a second comparison group was used. It consisted of 22 boys between the ages of 5 and 13 (M = 8.9 years), who were all the boys who had completed an agency-run treatment program for behavior problems related to their own history of sexual abuse. Both of us provided clinical and research consultation to this treatment program, and Child Behavior Checklist, family functioning, school performance, and sexual abuse history data for each of these 22 boys are available. These boys were almost exclusively from lower- to lower-middle-class families, and the majority were from single-parent families. As far as we are aware, none of these boys had exhibited behavior that we would classify as sexually aggressive. They were an appropriate comparison group in that they were also referred because their behavior was of significant magnitude to warrant treatment in the same way that the behavior of the 22 boys referred for sexually aggressive behavior was of sufficient magnitude to warrant an initial consultation. The presence of this additional comparison group does allow for some additional contrasts that can help us to understand further the nature of sexually aggressive behavior.

RESULTS

The results in Tables 1, 2, and 3 are presented so that Patients 1-16 are those who we decided had been sexually aggressive (SA), whereas patients 17-22 were not (NSA). This allows for some direct comparisons regarding the characteristics of the children in these two groups.

Tables 1, 2, 3
Assessment Data; Abuse and Victimization Experiences; Parent-Child Relations

Patient

IQ
(a)

Sex

Race

Age
(years-
months)

Primary diagnosis
(DSM-III)

Social
skills
(b)

Sexual
preoccu-
pation
(c)

School
problems
(d)

Previous
sexual
abuse

Perpetrator

Most severe abuse

Victims

Type of Offense
(most severe
documented)

History of
maternal
absence/neglect

Projective
identification

Emotional support

Maternal
chemical
dependency

Patient

Whether
given

Overall
quality

Sexually aggressive

106

4-4

Oppositional

Yes

—

Yes

Uncle, father

Oral

Yes

Fair

111

4-6

Conduct

Yes

—

Yes

Father

Oral, anal

Yes

Fair

Yes

4-11

Adjustment

Yes

—

Uncle

Anal

Fondle

Yes

—

Absent

—

139

5-1

Adjustment

Yes

—

Yes

Uncle

Oral

Yes

Good

5-2

Oppositional

Yes

—

Yes

Babysitter,
uncle

Oral

Yes

Poor

Yes

114

6-11

Oppositional

Yes

Father

Oral

Yes

Fair

Yes

101

7-1

Conduct

Yes

—

Fondle, undress

Yes

Poor

Yes

7-11

Conduct

Yes

Stepfather

Oral, anal

Yes

Poor

Yes

116

8-2

Conduct

Yes

Grandfather

Oral, vaginal

Vaginal

Yes

Fair

Yes

108

8-3

Schizophrenia

Yes

Father

Oral, anal

Oral

Yes

Poor

—

9-0

Conduct

Yes

Uncle

Anal

Fondle/
simulate intercourse

Yes

Poor

—

9-3

Conduct

Yes

—

Fondle, undress

Yes

Fair

101

9-4

Oppositional

Yes

Stepfather

Oral

Yes

Fair

10-2

Dysthymia

Yes

Neighbor

Oral, vaginal

Oral

Yes

Poor

10-6

Conduct

Yes

2 adolescent
neighbors

Oral, anal

Oral

Yes

Poor

Yes

10-7

Conduct

Yes

Father

Oral, anal

Yes

Poor

Yes

Not sexually aggressive

105

4-9

Adjustment

Yes

—

Yes

Father

Fondling

—

Yes

Good

100

5-0

—

Yes

—

Yes

Good

102

5-11

Adjustment

Yes

—

Yes

Good

109

6-1

Adjustment

Yes

Father

Fondling

—

Yes

Good

108

6-11

—

Yes

—

Yes

Good

121

11-0

Oppositional

Yes

—

Yes

Fair

Note. M = male, F = female, Bl = Black, Wt = White. For DSM-III, see American Psychiatric Association (1980).
       ^(a)Obtained with the Wechsler Preschool and Primary Scale of Intelligence (WPPSI), the Wechsler Intelligence Scale for Children-Revised (WISC-R), the Stanford-Binet Intelligence Scale, and the Peabody Picture Vocabulary Test-Revised (PPVT-R).
       ^(b) Yes = child has age-appropriate social skills.
       ^(c) Sexual preoccupation indicated on drawings or other testing material.
       ^(d) Behavior and/or academic problems reported by parent.

With regard to descriptive statistics, 12 of the sixteen SA children were male; 3 were Black (Nos. 4, 10, and 11), and the remaining were White. Their mean age was 7.6, which is somewhat older than the NSA group’s mean age of 6.6. The SA group had a mean IQ of 97.8, whereas the NSA group’s mean IQ was 107.5. No intelligence test data was available on 1 SA child.* The results suggest that the two groups differ somewhat in regard to cognitive abilities, although both groups fell within the average range. Almost half of the SA children were in the low-average range or borderline range, which is over twice what would be expected in a normally distributed sample.
      * No formal data were available, but it was estimated that he was in the low-average/average range cognitively. He was only in the second grade and was 9 years old.

Previous Sexual Abuse History

NSA group. Patients 17-22 were not included in the final sample because they did not meet the criteria of sexual aggression. For example, Patient 17 was accused by a mildly retarded 7-year-old boy of having had anal sex with him. This was discounted because the alleged victim, who had been abused 1 year earlier by his brother, admitted to lying about the abuse, the boys were supervised together, and the 11-year-old did not indicate deception on a polygraph that was done at his parents’ instigation. In addition, the alleged victim had, in the previous year, made identical unsubstantiated accusations of 2 other neighbor boys. Patient 18 was discovered by his stepmother-to-be in a closet with her 4-year-old daughter. Both had their underwear off and admitted to “touching” each other, and both denied additional contact. The stepmother requested an evaluation to address her own concerns about her stepson-to-be. Patient 19 was the son of a daycare operator; 18 months earlier, he had been observed pulling on the penis of a 6-year-old boy and, in the previous month, had been observed throwing leaves on a 2-year-old girl’s crotch while they were playing in the back yard. His mother’s anxiety about losing her license led to her requesting an evaluation. Patient 20 was overheard asking a 2½-year-old girl to “lick my peeper,” and when his mother came into the bathroom, both were urinating on the floor. His mother mentioned this to her caseworker who was in charge of licensing her daycare, and as a precaution, this boy was referred for an evaluation. Patients 21 and 22 were brothers whose father had molested them; their mother was concerned that they still would grab each other’s penises and buttocks, although close to 1 year had elapsed since their father’s visitation rights were terminated. No other victims could be identified, and their sexualized play appeared to be confined to themselves.

We judged none of these 6 boys to be sexually aggressive, as defined by the aforementioned criteria. They did serve as a useful comparison group. Neither group was randomly selected, and both are consecutive referral convenience samples, with all the limitations inherent in this type of study.

SA group. Of the 16 sexually aggressive children, 13 had very definite histories of sexual abuse. These histories were usually validated by at least two of the following: the child’s report, admission of the offender, medical/physical evidence, and an eyewitness report. Two of the remaining children did not appear to have a history of sexual abuse per se, although No. 9’s mother had a sexualized relationship with him. No. 5 had a definite history of physical abuse and social services involvement. In addition, his younger sister had been molested by an adolescent neighbor approximately 1 year earlier, and she was one of his two victims. Of the last of the 3 cases without a definite history of sexual abuse, No. 11 is the one about which we were most unclear. He did report witnessing sexual aggression against his mother by one of her boyfriends, but he denied experiencing sexual abuse.

For each of the 13 with very definite histories of sexual abuse, this abuse appeared to be severe and of fairly long duration. Each of these children had experienced oral, anal, and/or vaginal intercourse. Sometimes it was accompanied by violence. For purposes of illustration, we describe in more detail the abuse experienced by the first 3 of these 13 children.

The first child was forced to commit fellatio with his maternal uncle at least 12 times, according to his uncle’s report, and his father was convicted of molesting him on at least four separate occasions. This latter abuse also involved oral-genital contact. No. 2 was a boy who was sodomized and forced to commit fellatio with two older adolescent neighbors of his grandmother. This apparently occurred whenever he visited his grandmother. A conservative estimate of the number of sexual contacts between this boy and these two adolescents during the I½-year period is in excess of 50. No. 3 observed his father having oral-genital contact with his younger sister, and he in turn had oral-genital contact with his father or at least two separate occasions. Less clear were allegations that the father had intercourse with his male lover within the boy’s view.

In summary, the sexual abuse these 13 had experienced typically involved genital contact and intercourse, usually oral or anal. This frequency of severe sex abuse is greater than what would be expected in a random sample of sexually abused children. It appears that severe sexual abuse contributes to a predilection for sexually aggressive behavior. Table 2 also indicates that the type of sexual aggression that the sexually abused children committed parallels their own victimization. For example, No. 4, who committed fellatio and was sodomized by his stepfather, forced both of his victims (one male and one female) to have oral-genital contact with him and inserted the end of a broom handle into the rectum of his half sister, one of his two victims.

Treatment group. The sexual abuse experienced by the 22 boys in the treatment group (TG) differed from that of the 13 SA children with a history of sexual abuse. The treatment group boys had experienced as their most severe abuse the following: genital fondling (22.7%), forced vaginal intercourse (9.1 %), and oral/anal intercourse (68.2%). In addition, 36.3% of the perpetrators were between the ages of 12 and 16. This is in contrast to the SA children: Of those with a history of sexual abuse, 100% had experienced oral/anal vaginal intercourse, and 100% of the primary perpetrators were 17 or older.

Characteristics of Sexually Aggressive Children

In this section we focus on those characteristics of the sexually aggressive children that we considered to be most important in regard to understanding their behavior. These include individual and parental variables.

Concurrent psychiatric diagnoses. Table 1 indicates that all 16 SA children satisfied criteria for DSM-III diagnosis (American Psychiatric Association, 1980). The diagnosis was usually characterized by aggressiveness; there were 8 diagnoses of conduct disorder and 4 of oppositional disorder. Two of the children were exhibiting adjustment disorders, 1 child was dysthmic, and 1 was schizophrenic. This last child, No. 8, had either a grandparent or an aunt on both his mother and father’s sides of the family with a positive history for schizophrenia, and thus he clearly had an underlying tendency for a severe psychiatric disorder. He exhibited auditory hallucinations, he was very self-destructive; he had had bicycle-car accidents, had jumped out of an upstairs window, and had set the kitchen on fire. Four of the 6 NSA children also had a DSM-III diagnosis, but none of these were conduct disordered.

It seems very possible that the sexual aggressiveness seen in many of the SA group was closely related to an overall aggressive style. Available reports were that physical force was frequently a component of the sexual behavior. No. 6’s sexual aggressiveness came to light when he severely bit his stepbrother’s penis, and No. 2 would choke his victims, who were usually 2-3½ years old at most.

Table 4
Child Behavior Checklist Data: Mean T Scores

Behavior	SA boys (n = 9)	SA girls (n = 3)	NSA boys (n=5)	Treatment group boys (n=22)
Anxious/obsessive	69.9	70.3	59.3	67.8
Depression	69.6	78.3	64.0	69.2
Somatic complaints	67.6	65.6	61.4	62.7
Hyperactive	73.4	68.6	61.3	67.0
Aggressive	75.8	71.3	63.2	69.4
Delinquent	75.3	66.6	63.0	68.0
Sex problems*	5.9	6.8	1.1	1.3

Note. SA = sexually aggressive; NSA = not sexually aggressive.
* Total raw score for Items 5, 59, 60, 73, 96, and 110 on the Achenbach Child Behavior Checklist. Range = 0-12.

Mean scores on the six Child Behavior Checklist factors common to this age range (i.e. anxiety, depression, somatic complaints, hyperactivity, aggressiveness, and delinquency) were obtained for the boys in the SA and NSA groups. T scores greater than or equal to 70 are considered clinically significant. Although a complete data set was not available, Table 4 does reveal some differences among the groups. The data were obtained through parent report and have all the problems inherent in that method. Yet the summary scores can be illuminating. For example, the elevations on both depression and aggression in the 3 girls are poignantly portrayed in a statement by No. 13. She was being interviewed with regard her feelings at the time of her sexual offenses and stated, “I felt like killing myself that day, but I just went and played with [the victim].”

In regard to DSM-III diagnoses of the 22 boys in the treatment group, the majority would be described as adjustment disorders with mixed features. However, we did not have the capacity to establish, with certainty, a diagnosis for each boy and can conclude only that it is our impression that the TG boys were not as severely impaired.

Socialization. In reference to social skills, we evaluated each child with regard to involvement with same-age playmates (usually from the social competence factor of the Achenbach Child Behavior Checklist) and the quality of their interaction with the examiner during the interview(s). We judged 8 of the 16 SA children and all of the NSA children to have age-appropriate social skills. At least 3 of the 22 TG boys exhibited clear problems with social skills. The mean Social Competence-Social score for the TG boys on the Achenbach was just within normal limits (M = 35.1, SD = 13.1).

In regard to the social difficulties of the SA children, several examples are appropriate. No. 2, for example, reported as his only friends the 2 adolescents who abused him, and a 2-year-old boy whom he had molested. During the interview, he had difficulty maintaining eye contact, wandered about the room, picked in his nose frequently, interrupted the examiner, passed gas, belched, and was routinely unable to answer simple questions. No. 4 attempted to steal several toys and actually succeeded in deliberately smashing the examiner’s hand in a desk drawer. His response to the examiner’s pain was a gleeful laugh. No. 16 could not be left unsupervised for long with classmates in his behaviorally disordered class because of his frequent aggressive outbursts. It seemed that the socialization problems noted in these children also reflected a defect in empathy; their dehumanizing experiences interfered with their capacity for appropriate object relations.

Sexual preoccupation. Sexual preoccupation was present in most of the children. Twelve of the 16 exhibited clear signs of sexual preoccupation, as measured according to spontaneous drawings (Draw-a-Person or Kinetic Family Drawings Test) of genitalia or sexual acts, sexual content on the Rorschach, the TAT, or the Roberts test, and/or an elevated sexual problems score on the Child Behavior Checklist. For example, when asked to take the Kinetic Family Drawings Test, No. 1 drew a picture of an elongated object, which he labeled “pee-pee,” and then drew heavy lines through it in an effort to cover it up. Case No. 2 gave sexual content in 12 of his 15 Rorschach responses (e.g., “butthole,” “titties”). To Card 4 on the Thematic Apperception Test, his entire response was, “They are going to kiss and then go to bed and do sex.” This card depicts a man and woman standing together, with the man looking away as if to leave. No. 6 would masturbate while having bowel movements and drew genitalia on his first human figure drawings. Case Nos. 7, 12, and 13, all female, were reported to be chronic masturbators, even in public. Case No. 11 spontaneously remarked, “Look, a titty” when he was making circles on paper. On several occasions, there was explicit sexual content in his play that was clearly inappropriate (e.g., puppets engaged in oral sex). These behaviors were observed in the first two evaluation sessions. Case No. 13 produced a relatively appropriate series of responses to the previous Roberts cards until Card 15. This card depicts the side view of a nude woman in a bathtub, and show a young boy looking through a partly open door at the woman, who is unaware of his presence. The child’s complete response was, “Uh, oh, she’s naked in the tub and that man is looking at her. Is that a man or a boy? It’s a boy. She’s going to get out of the tub and abuse him. He’ll feel very sad. She won’t feel anything.” When she was questioned, she indicated the “abuse” meant sexual abuse, but she was not clear what the woman would do. Case No. 13 was molested by her stepfather over a 6-month period that began when her mother was pregnant with her younger sister. He masturbated her and forced her to commit fellatio at least four times. He also had oral-genital contact with her. Once the abuse became known, her mother divorced her husband and successfully pushed criminal charges. Yet, despite this show of maternal support, this child was caught twice and admitted to several other instances of committing fellatio with a 4-year-old neighbor boy.

The relative level of sexual behavior problems in the SA boys, SA girls, NSA boys, and TG boys are indicated in Table 4 with the variable “sex problems.” This variable has been used in research with other samples of sexually abused children (Friedrich et al., 1988; Friedrich et al., 1986) and is the sum of raw scores of six items from the Achenbach Child Behavior Checklist (5, 59, 60, 73, 95, and 110). Children in both SA groups were clearly far more sexualized than children in either the NSA or treatment groups.

School problems. Related to the socialization difficulties is the presence of school problems, either behavioral or academic. All 11 of the school-aged SA children had coexisting school problems. Eight were in learning-disabled or other special education programs and frequently were also reported as having behavior problems. Yet neither of the 3 school-age NSA children had school problems, but 42-50% of the TG boys were receiving below-average or failing grades in at least one of the following subjects: reading, writing, spelling, and math. Consequently, despite the fact that as a group they had average intelligence, school difficulties in the SA group were routine and reflect the degree to which these children’s lives are disrupted.

Parent-child relations. The quality of parent-child relations were also assessed. This was done through interview in the SA and NSA groups and through therapist report and family assessment questionnaire data in the TG group. Although one child (No. 11) was in foster placement, we usually had a conjoint interview with the parent and child, and we interviewed each, individually, about their relationship. Eight of the SA relationships were rated as poor. To arrive at this rating, we had to see (a) lack of support of the child, which included blaming, (b) a history of scapegoating and projection, and (c) a history of neglect, even periods of abandonment. In some cases, a fourth feature, that of an eroticized mother-son relationship, was also noted. No. 9 typifies this relationship. He had lived away from his mother, an alcoholic, for 2 years. She alternatively ridiculed her son, describing him as “probably possessed,” or praised him lavishly. He slept with her nightly, and he routinely saw her when she was nude. His sexual aggression was directed at both children and adult women.

The parent-child relationship for No. 11 was also described as poor. His mother began prostitution at age 12, and his father was her pimp. Although she had custody of her 2 younger children, her oldest son, No. 11, lived with his maternal grandmother. It was at his grandmother’s house that he experienced brutal sexual abuse, including anal penetration, by his mentally retarded uncle. No. 8’s mother reported seeing her husband force her son to have oral-genital contact with him when the child was 4 months old. She denied ever seeing other abuse until he was 6½ years old. However, in treatment, her husband admitted to frequent, ongoing abuse of their son during those first 6 years, and she was ambivalent about divorcing him. She involved her son in therapy for a 1-year period but was emotionally neglectful of him, missing parent-teacher conferences at school and leaving him alone for long periods of the day. No. 15 was severely abused over a 4-year period by a neighbor, and her mother “knew something funny was going on,” but never intervened. When she was hospitalized, her mother rarely visited and refused to accept her home because her boyfriend would not allow it. Not one of 12 Roberts Apperception Test responses indicated even one instance of maternal support or nurturing.

The relationships described as fair also were problematic. No. 5 had been physically abused by his father, with a subsequent social services investigation and confirmation. Their relationship was still negatively charged, but his mother, though behaving inconsistently, was supportive and pursued therapy for at least 5 sessions. Case No. 6 was the only child born to this very young mother. She was 19 at the time of the evaluation. Although described as having moderately severe alcohol problems and leaving her son for overly long periods with her mother, she stuck with a very lengthy treatment process (over 1 year) when the abuse was discovered and began to manage her drinking. Her behavior was also more consistent with her son, but she would still leave him for long periods with her mother. These periods were invariably followed by an increase in aggressiveness by her son.

Only 1 of 16 SA cases could be described as having a good parent-child relationship, whereas 5 of 6 NSA cases could be thus described. Case No. 12’s mother pursued a complete course of therapy for her daughter, of whom she was very supportive. Initially guilt ridden about her daughter’s abuse, she reacted inconsistently to her daughter’s tantrums and general aggressiveness until encouraged to be more consistent. Their relationship was primarily positive and characterized by mutual affection.

Marked pathology in the parent is also frequently evident. MMPIs were obtained from the mothers of 7 of the SA children (Nos. 1, 7, 8, 9, 13, 15, and 16). Five were significantly elevated (T ≥ 70); a 4-8 code type was represented in 3 of those 5 patients, and 4-3, 4-6, and 6-8 code types were present in 3 other parents. Anger is a predominant feature in each of these code types, as is family discord, impulsivity, impaired interpersonal relationships, and alienation.

In regard to the quality of parent-child relationships and parental pathology in the TG group, we do not have the richness of detail for these families that we do for the SA group. Of the 22 families, 8 (36.3%) fell into an extremely pathological range on the FACES II, a measure of family adaptability and cohesion (Olson et al., 1982). Three additional children (13.6%) were a frequent focus in supervision with regard to their extremely pathological family relationships. Thus one half of the 22 TG families gave us evidence, of one type or another, that parent-child relations were very problematic. The only measure of maternal psychological distress in the TG group that we used was the Beck Depression Inventory (Beck, 1967), and at intake the mean depression score for all of the mothers was in the mildly depressed range, although an examination of individual scores indicated that approximately two thirds of the mothers did not report depression. Last, family quality, particularly the two variables of “unconditional support of the child” and “overall conflict in the family,” were rated by therapists on 5-point scales on which 1 was most positive; the mean initial ratings were 2.1 (SD = 1.3) and 2.4 (SD = 1.0), respectively. This suggested generally positive family qualities in the majority of the TG families.

In summary, a history of long-standing parent-child problems, existing before the abuse, was more often the case with these sexually aggressive children than it appeared to be with a treatment sample that served as a comparison group. They clearly lacked the parental nurturance essential to emotional health (Winnicott, 1975). Chemical dependency in the parents was common, as was chronic depression. Only 1 of the 13 SA children (No. 5) was in a home situation with both biological parents, but that relationship too was unstable. Projective identification (Skynner, 1976), in which the child was viewed as deviant or as a “bad seed,” like the offending parent, was very common. This tendency for mothers of abused children to perceive a greater number of personality problems in their children has been empirically validated (Oates, Forrest, & Peacock, 1985). Even with No. 12, who had the one good parent-child relationship, the mother maintained a belief that some of her daughter’s emotional difficulties were attributable to the fact that schizophrenia was present in her husband’s immediate family. The result is that these children were infrequently emotionally supported.

DISCUSSION

The sexually aggressive children in this small sample differ on a number of important dimensions from two small comparison samples. Their relative level of personal and parental pathology was particularly striking, as was the severity of the sexual abuse that they experienced. Without exception, these children had been exposed to physical violence, sexualized adult behavior, and/or sexual abuse that involved intercourse. This high percentage of intercourse far exceeds what is usually reported in large samples of sexually abused children (Finkelhor, 1984; Friedrich et al., 1986; Gomes-Schwartz, Horowitz, & Sauzier, 1985; Russell, 1985). Physical violence or threats of violence commonly accompanied the abuse, making more likely the link between aggression and sexual behavior. The extreme deviance in our sample of SA children suggests that childhood sexual aggressiveness is usually related to significant pathology. Researchers in larger studies may find more variability in these children than we did.

The traumatic nature of the abuse is increased when bodily penetration and aggressiveness are combined, a common occurrence in our sample. This increases the likelihood that the experience is “split off’ (denied) by the child because of its “not-me” nature (Guntrip, 1969; Sullivan, 1953). When this splitting-off process occurs, the child is far less able to integrate the experience and make sense of it, and the events are less amenable to therapy. Yet the child’s behavior is dictated at least in part by the event, in a way similar to the reliving of the event that is seen in individuals experiencing a posttraumatic stress reaction. Furman (1986) wrote that after a catastrophic event, the child is faced with the need to bind or master this excess of stimuli. This binding process takes the form of repetition compulsion. In the case of the SA children in this study, their sexual behavior, in which they repeated the acts perpetrated on them, was a type of repetition compulsion. However, the repetition of a traumatic event enables many children to vary it so as to allow for its being assimilated or altered in an ego-enhancing way. This was not the case with the SA children. Also of relevance to our focus is Furman’s (1986) statement that the traumatic event became sexualized in the children she was discussing, each of whom had lost a parent by death. The presence of sexualization in a sexually abused group is even more likely.

In addition, the nature of the sexual offenses that these children committed is more extreme than that seen in Fehrenbach et al.’s (1986) large sample of adolescent sexual offenders. For example, 23% of their sample was involved in rape, whereas the majority of our sample’s offenses could be classified as rape. Using the term rape to describe their behavior may seem inappropriate to some readers, but a review of their aggressiveness, described earlier, makes the point quite clear. This difference is attributable in part to the fact that it took more deviant behavior for these children to come to our attention, since children’s aberrant behavior is more likely to be dismissed as transient. Another contributor is that these children came from families for which teachers and social workers did the majority of the referring, and more serious behavior is more likely to warrant a referral.

It is clear that the traumatic components of sexual abuse are related to the behavioral symptoms noted (Finkelhor & Browne, 1985). These children routinely felt powerless and betrayed and were traumatically sexualized. However, the relationship between the traumatic elements and the outcome is not linear, and the impact of the trauma is modified by a variety of buffering variables. In fact, an adaptational perspective is the most appropriate model to use in understanding the relationship between the stressor of sexual abuse and the subsequent behavioral response (Friedrich, 1988). Examination of the presence of moderator variables in these children’s lives could illuminate the relationship between severe sexual abuse and sexual aggressiveness. Moderator variables that have been identified in the literature on child psychopathology and coping include both individual variables, such as cognitive abilities and absence of significant psychopathology, and interpersonal variables, such as peer relations, school adaptation, and parent-child relations (Rutter, 1983).

As a group, these children were in the average range cognitively, but a larger percentage than expected were in the low-average and borderline range. The relationship between intellectual abilities and antisocial behavior has been explored frequently, and in a recent study of adolescents, lower IQ was associated with a greater likelihood of delinquent behavior (Moffitt, Gabrielli, Mednick, & Schulsinger, 1981). Reduced cognitive abilities make school less rewarding, which thus cuts these children off from ego-enhancing experiences and contributes to a reduced sense of competency. Language abilities, in particular, are related to the child’s developing ability to inhibit their behavior, and inhibition seems to be in short supply in many of these children. Treatment can also be more difficult with a child of lower cognitive abilities and impaired inhibition. Play and verbal behavior may be more immature and may not lend itself to expressing and understanding one’s behavior and reducing internal conflict. Thus a certain subset of SA children do not have available to them the moderating influences that accompany adequate cognitive abilities.

Significant psychopathology is certainly evident in the majority of the SA children, particularly that in which a primary feature is aggressive and undersocialized components. It is clear to us that with few exceptions, premorbid family features, including inconsistent and overly primitive parenting, lack of parental empathy, and covert reinforcement of aggressive behavior, made conduct or oppositional disorders an expected outcome. The sexually abusive experience served primarily to add a sexualized channel to the aggressiveness that was already emerging. The adjustment disorders noted were also related to conduct disturbances. In an interesting turn, several of the conduct-disordered children seemed more amenable to behavior change than No. 15, diagnosed with dysthymia, who was extremely passive, who had few ego resources, and whose lack of aggressiveness suggested a defeated stance.

However, we noted additional and profound deficits in moderator variables related to interpersonal functioning. Research with adult sex offenders documents their impaired social cognition (e.g., misreading interpersonal cues), and the same appears to be true with many of these children (Baxter, Barbaree, & Marshall, 1986). Peer relations are markedly impaired, and parents routinely report significant sibling conflict. These children are frequently rejected by their peers, seem to have great difficulty with interpersonal boundaries, and tend to interact with their peers in aggressive and sexualized ways. Of central importance in understanding the genesis of these interpersonal difficulties is the existence of long-standing parent-child problems. From the very beginning, in many cases, these children were neglected and/or abused by their parents. Parental absence was commonplace, and the long duration of abuse in many of these children seemed to be related to a breakdown in parental availability or protection. It also appears that after the abuse, appropriate parenting was the exception, rather than the rule, and this further exacerbated the trauma of sexual abuse, making it more likely for sexually aggressive behavior to occur.

Significant parental pathology, including characterological problems, chemical dependency, and a history of sexual abuse, made the parent-child relations even more difficult to change than that of a more typical conduct-disordered child. Parents could not be expected to empathize with their child, provide consistency, or put their child’s needs first long enough for the relationship to change in a more positive direction. With many of these children, we had to face the fact that we were dealing with a child who, more probably than not, would continue for many years to have significant problems in the area of sexual aggressiveness towards children.

That was not uniformly the case, and there exists enough heterogeneity within sexually aggressive children for some to have more favorable long-term prognoses. The developmental considerations mentioned earlier (e.g., age of victimization; the interval between victimization and acting out/aggression) appear to contribute to the heterogeneity in ways that are not clearly worked out at this point. Some treatment outcomes are positive. But, in closing, it does seem that sexually aggressive children present to the clinician a very difficult treatment challenge, compounded by the severity of the abuse and the lack of positive individual, parental, and relational variables that could contribute to a more optimistic outcome. They do appear to represent a group of children whose treatment needs, and the concomitant limitations of their family members, exceed that of the majority of sexually abused children.

REFERENCES

   American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders. Washington, DC: Author.
   Arroyo, W., Eth, S., & Pynoos, R. (1984). Sexual assault of a mother by her preadolescent son. American Journal of Psychiatry, 141, 1107-1108.
   Baxter, D. J., Barbaree, H. E., & Marshall, W. L. (1986). Sexual responses to consenting and forced sex in a large sample of rapists and nonrapists. Behavior Research and Therapy, 24, 513-520.
   Beck, A. T. (1967). Depression: Causes and treatment. Philadelphia: University of Pennsylvania Press.
   Fehrenbach, P. A., Smith, W., Monastersky, C., & Deisher, R. W. (1986). Adolescent sexual offenders. American Journal of Orthopsychiatry, 56, 225-233.
   Finkelhor, D. (1984). Child sexual abuse. New York: Free Press.
   Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal ofOrthopsychiatry, 55, 530-541.
   Fortenbeny, J. D., & Hill, R. F. (1986). Sister-sister incest as a manifestation of multigenerational sexual abuse. Journal of Adolescent Health Care, 7, 202-204.
   Friedrich, W. N. (1988). Behavior problems in sexually abused children: An adaptational perspective. In G. Wyatt & G. Powell (Eds.), The lasting effects of child sexual abuse. Beverly Hills, CA: Sage.
   Friedrich, W. N., Beilke, R. L., & Urquiza, A. J. (1987). Children from sexually abusive families: A behavioral comparison. Journal of Interpersonal Violence, 2, 391-402.
   Friedrich, W. N., Beilke, R. L., & Urquiza, A. J. (1988). Behavior problems in young sexually abused boys. Journal of Interpersonal Violence, 3, 1-12.
   Friedrich, W. N., Urquiza, A. J. & Beilke, R. J. (1986). Behavior problems in sexually abused young children. Journal of Pediatric Psychology, 11, 47-57.
   Furman, A. (1986). On trauma. Psychoanalytic Study of the Child, 41, 191-208.
   Gomes-Schwartz, B., Horowitz, J. M., & Sauzier, M. (1985). Severity of emotional distress among sexually abused preschool, school-age, and adolescent children. Hospital and Community Psychiatry, 36, 503-508.
   Groth, N. A. (1979). Sexual trauma in the life histories of rapists and child molesters. Victimology, 4, 10-16. Guntrip, H. (1969). Schizoid phenomena, object relations and the self. New York: International Universities Press.
   Lanning, K. V. (1986). Child molesters: A behavioral analysis. Washington, DC: National Center for Missing and Exploited Children.
   Moffitt, T. E., Gabrielli, W. F., Mednick, S. A., & Schulsinger, F. (1981). Socioeconomic status, IQ, and delinquency. Journal of Abnormal Psychology, 90, 152-156.
   Oates, R. K., Forrest, D., & Peacock, A. (1985). Mother of abused children: A comparison study. Clinical Pediatrics, 24, 9-13.
   Olson, D. H., McCubbin, H. I., Barnes, H., Larsen, A., Muxen, M., & Wilson, M. (1982). Family inventories. St. Paul, MN: Family Social Science, University of Minnesota.
   Pomeroy, J. C, Behar, D., & Stewart, M. A. (1981). Abnormal sexual behavior in pre-pubescent children. British Journal of Psychiatry, 138, 111-125.
   Purcell, J., Beilke, R., & Friedrich, W. N. (1986, August). Assessing sexual behavior in children: The development of the Child Sexual Behavior Inventory. Paper presented at the annual meeting of the American Psychological Association, Washington, DC. Russell, D. (1985). Sexual exploitation. Beverly Hills: Sage.
   Rutter, M. (1983). Stress, coping, and development. In N. Garmezy & M. Rutter (Eds.), Stress, coping, and development in children (pp. 1-14). New York: McGraw-Hill.
   Skynner, A. (1976). Systems of family and marital psychotherapy. New York: Brunner/Mazel.
   Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York: Norton.
   Winnicott, D. (1975). Through pediatrics to psychoanalysis. New York: Basic Books. Yates, A. (1982). Children eroticized by incest. American Journal of Psychiatry. 139, 482-485.

>>