[Introduction]

Theories of the Effects of Child Sexual Abuse
Effects of Child Sexual Abuse

[Page 17]

CHILD SEXUAL ABUSE (CSA) has been reported to affect between 4% and 50% of children and adolescents with an average prevalence of approximately 15% to 20% (Violato & Genuis. 1993). This is an alarmingly high prevalence rate. The great variation in prevalence from study to study is due to a number of factors, including differences in definition, measurement, samples, and reporting methods. The inconsistencies in reports of both the short- and long-term sequelae

[Page 18]

of CSA have created confusion about the impact and severity of CSA on human development. Considerable research on the sequelae of CSA has now been conducted and published. The major purpose of the present study was to conduct a meta-analysis of this published research.

Some researchers have argued that the traumatic impact of CSA has been greatly overstated (Brongersma, 1984, 1991; Chiswick, 1983; Sandfort, Brongersma, & van Naerssen, 1990); others contend that victims of CSA experience numerous detrimental effects and disturbances (Conte & Shuennan, 1987). Many studies have been conducted, but the effects, sequelae, and prevalence of CSA still remain unclear (Violato & Genuis, 1993, 1994). Several of these studies are characterized by serious design and measurement problems, including variability in definitions of CSA, biased sampling, the use of varying and inadequate out- come measures, and the absence of appropriate control groups (Collings, 1995; Conte & Schuernan, 1987; Dubowitz, Black, Harrington, & Verschoore, 1993; Green, 1993; Violato & Genuis, 1993).

The impact on human development that may result from CSA experiences is of increasing concern to researchers, clinicians, and policymakers (Frazier & Cohen, 1992; Kolko & Moser, 1988; Tong, Oates, & McDowell, 1987). The present review and meta-analytic investigation focused on six specific effects of child sexual abuse on human development and behavior. Specifically, the out- comes studied include

	posttraumatic stress disorder (PTSD),
	depression, suicide,
	sexual promiscuity,
	victim-perpetrator cycle, and
	poor academic achievement.

Theories of the Effects of Child Sexual Abuse

There are several theoretical explanations of how CSA affects children's development. Green (1993) and Kendall-Tackett, Williams, and Finkelhor (1993) provided succinct reviews of two main theoretical conceptualizations of how CSA affects development.

The first, core-symptom theories, holds that victims of CSA display a conspicuous syndrome of symptomatology including sexualized behavior and PTSD. Although Kendall- Tackett et al. (1993) did not find support for this contention, proponents of this perspective have argued that damage to self-image and symptoms of sexualized behavior and PTSD are more common in sexually abused children than in other clinical groups.

Specifically, using structured interviews among 26 sexually abused children and 23 non-sexually abused children referred for psychiatric outpatient evaluation, McLeer, Callaghan, Henry, and Wallen (1994) found that although the groups did not differ significantly in the number of diagnoses, the prevalence of PTSD among sexually abused children was significantly greater. According to McLeer et al. (1994), the known stability of PTSD across time; its disruptive effect on behavior, affect, and social development; and its resistance to traditional treatments leave children who develop PTSD at considerable risk for prolonged dysfunction.

The second group of theories has been termed multifaceted models of traumatization (Kendall- Tackett et al., 1993). According to this view, the impact of sexual abuse is complicated because it produces multifaceted effects. As a result, some theorists have argued that distinct mechanisms and processes operate to account for the variety of outcomes (e.g., Finkelhor & Browne, 1985). Other theorists have posited that family dysfunction or a general maltreating environment is the root of CSA trauma (e.g., Conte & Schuerman, 1987). Still others have described sexual abuse as a generalized stressor after which the child victim is likely to develop problems in whatever area she or he may have had a prior vulnerability (Kendall- Tackett et al., 1993). Given its heterogeneous nature, Green (1993) recommended defining sexual abuse as an event rather than as a psychiatric syndrome or disorder. Moreover, Green urged that researchers and clinicians integrate a variety of data through careful assessments of both the victimized child and his or her parents.

Effects of Child Sexual Abuse

Despite a lack of theoretical and empirical support, proponents of child and adult sexual relationships have argued that sexual interests and behaviors of adults with children should be considered acceptable, normal, and healthy expressions that affirm a child's sense of personal worth, independence, and sexual power (Brongersma, 1984; Chiswick, 1983; Sandfort et al., 1990). These .'diverse" experiences are believed to enable children to construct their own sexuality from greater possibilities (Sandfort et al., 1990) and thereby foster healthy child development.

Conversely, numerous empirical studies investigating the developmental impact of sexual abuse on children and adolescents indicate that many symptoms of maladjustment and mal-adaptation are associated with the experience of sexual abuse. Most frequently reported are problems of depression, anxiety, and other internalizing disorders, as well as externalizing problems such as dissociation, conduct disorders, aggressiveness, and inappropriate or early sexual behavior and activity

(Bagley, 1991; Bagley & McDonald, 1984; Caffaro-Rouget, Lang, & van Santen, 1989; Dutton & Hart, 1992; Fromuth & Burkhart, 1989; Higgins & McCabe, 1994; Kolko & Moser, 1988; Mayall & Gold, 1995; Mullen, Martin, Anderson, Romans, & Herbison, 1994; Swanston, Tebbutt, O'Toole, & Oates, 1997; Trickett, McBride-Chang, & Putnam, 1994; Widom & Ames, 1994).

Specifically, sexual maladjustment, interpersonal problems, educational difficulties, acute anxiety neuroses, self-destructive acts, somatic symptoms, loss of self-esteem, prostitution and delinquent criminal behavior, depression, and actual or attempted suicide are some of the difficulties reported by CSA victims (Browne & Finkelhor, 1986; Conte & Schuerman, 1987; Tong et al., 1987). Although the list of psychological sequelae that researchers have found to be related to CSA is long, PTSD symptoms and sexualized behaviors are among the

[Page 20]

effects reported most frequently (Browne & Finkelhor, 1986; Higgins & McCabe, 1994; Kendall-Tackett et al., 1993). Still, no specific sexual abuse syndrome has been confirmed as a sequela to CSA. Numerous reviews, primarily narrative in nature, have been conducted in the area of child sexual abuse. For instance, Browne and Finkelhor (1986) reviewed studies that attempted to confirm empirically the effects of CSA within the clinical literature. They found that sexual abuse perpetrated by father figures and involving genital contact and force was the most damaging CSA experience. .Browne and Finkelhor's intentional decision to limit their review only to female victims of CSA does not assist us in understanding how CSA affects males. Furthermore, this narrative review did not employ the rigors of selection, data coding, and analysis that are possible in meta-analysis.

Another well-known review that includes both a narrative and an empirical synthesis was reported by Kendall- Tackett et al. ( 1993). They reviewed 45 quantitative studies of sexually abused children (i.e., 18 years and under) that had been published between 1988 and 1992. The majority of these studies had sample sizes between 25 and 50 that were drawn primarily from sexual abuse evaluation or treatment programs.

Kendall- Tackett et al. found that sexually abused children had more symptoms (e.g., fears, PTSD, behavior problems, sexualized behaviors, and poor self-esteem) than non-abused children, with abuse accounting for 15% to 45% of the variance in symptomatology differences. Although analyses revealed symptoms specific to certain ages, no one symptom appeared to characterize a majority of sexually abused children. Kendall- Tackett et al. (1993) argued that there is not any specific syndrome in children who have been sexually abused and no single traumatizing process. The Kendall-Tackett et al. (1993) review made an important contribution, but it is limited in several ways.

	First, only studies from a very limited time frame (1988 to 1992) were included, which may reflect a research inclusion bias.
	Second, studies with very small sample sizes (e.g., as few as 8 participants) were included, thereby reducing statistical power.
	Third, the effects of sexual abuse only on persons aged 18 or older were studied. Accordingly, we cannot determine whether specific symptoms do become particular to CSA victims later in life. A discussion of the long-term effects and trends of CSA is limited within the Kendall- Tackett et al. (1993) review, given the restricted age range.
	Fourth, although many important intervening variables, such as age at the time of assessment, identity of the perpetrator, and time elapsed between abusive incident and assessment were discussed, inferential statistical analyses were not conducted on these variables. Instead, Kendall- Tackett et al. ( 1993) simply recorded, when possible, the number of studies that reported a significant difference in impact and the direction of the findings.

To date, there is only one published study that has attempted to empirically summarize the existing state of research regarding the consequences of CSA. Jumper's (1995) meta-analysis of the effects of CSA on adult adjustment was based on the specific outcomes of psychological symptomatology, depression,

[Page 21]

and self-esteem. Jumper's meta-analysis included 26 published studies. Effect size estimates were computed for each of the dependent variables.

The results indicated statistically significant relationships for the experience of CSA and subsequent difficulties in psychological adjustment as measured by psychological symptomatology (r_u = .27), depression (r_u = .22). and self-esteem (r_u = .17). Effect size estimates in all three analyses suggested a significant increase in pathologies among those experiencing CSA across a variety of sampling methods, definitions of sexual abuse, date of publication, and gender of participants. Jumper's (1995) meta-analysis, however, is limited in several ways.

	First. it was focused on rather narrow outcome variables.
	Second, the number of studies included was quite limited.
	Third, it was focused specifically on adult outcomes. thereby neglecting short- or medium-term outcomes.

Thus. there is still considerable disarray in the study of the effects of CSA. Nonetheless, to come to a better understanding of the chaotic and inconsistent research, greater efforts such as those of Jumper and Kendall- Tackett et al. ( 1993) are required. There is a need to transform the results of all the relevant CSA studies to a common statistical metric using meta-analysis to provide an accurate estimate of the strength of the effects of CSA on victims.

In the present meta-analysis, we attempted to overcome the limitations of previous work. The main aim of the present study, therefore. was to conduct an empirical integration of the published research and to clarify some of the conflicting information about six specific outcomes of CSA (PTSD, depression, suicide, sexual promiscuity, victim-perpetrator cycle, and poor academic performance). The possible impact of a number of mediating variables on each of the outcomes was also investigated.