ABSTRACT Recent studies indicate that child sexual abuse is associated with multiple short-term and long-term psychological difficulties. Most of these studies have used correlational designs and retrospective reports of abuse, with the result that few studies have been theory-driven and are thus unable to inform clinical practice. This essay attempts to explore why the experience of child sexual abuse might lead to psychological difficulties by examining three models proposed in recent literature.
One of the most commonly used explanations for the development of sexual abuse sequelae is Post-traumatic Stress Disorder.Although this model adequately explains particular ‘anxiety effects’ associated with a history of child sexual abuse, it has difficulty in accommodating many of the symptoms observed and has thus more recently been replaced by the Traumagenic Dynamics model, Finkelhor & Browne, 1985. Four traumagenic dynamics; traumatic sexualization, betrayal, stigmatization and powerlessness, are identified as the core of the psychological injury inflicted by abuse, which provides a useful framework for understanding psychological and behavioural effects.The third model, proposed by Alexander 1992, uses attachment theory as a conceptual framework in which to understand the familial antecedents and consequences of child sexual abuse.
Whilst the Traumagenic Dynamics model posits experiential areas that can affect schema development, it does not address factors that mediate the development of each dynamic. It is argued here that combining attachment theory with the traumagenic dynamics model provides a better explanation of the consequences observed in the literature, as well as providing a conceptual framework with which to guide both further research and clinical practice.WHY MIGHT THE EXPERIENCE OF CHILD SEXUAL ABUSE LEAD TO MENTAL HEALTH PROBLEMS IN ADULTHOOD? Introduction The most widely used definition of sexual abuse is that of Schechter & Roberge 1976 which refers to the sexual exploitation of children as: “the involvement of dependent, developmentally immature children and adolescents in sexual activities that they do not fully comprehend, are unable to give informed consent to, and that violate the social taboos of family roles. A review by Peters 1988 found reported prevalence figures ranging from between 6 per cent and 62 per cent of female samples and between 3 per cent and 31 per cent of male samples.
There is, however, no universal definition of child sexual abuse, with the result that estimates of incidence or prevalence in any given sample can vary widely. Research is thus complicated by the range and breadth of definitions of child sexual abuse and the experiences that are said to constitute it.These can differ according to whether physical contact is involved, whether the perpetrator is an adult or another child, whether a single incident or series of incidents is involved, as well as according to the nature and degree of coercion and the nature of the relationship of the perpetrator to the child. The term ‘child abuse’ was first officially used in Britain in a 1980 government circular DHSS, 1980.
It was not until 1988 that sexual abuse was included as a separate category of child abuse. By 1991 the number of children on child protection registers in England and Wales sexual abuse category was 5,600 DOH, 1992.A MORI poll survey, on the other hand Baker and Duncan, 1985, found that the prevalence of intra-familial sexual abuse was 13 per thousand. This suggests that in the majority of cases child sexual abuse goes unchecked and, furthermore, that as clinicians, we will be faced with the associated mental health problems in generations to come.
It was not until the 1970’s and, it is argued, as a result of the feminist movement Herman, 1981, that childhood sexual experiences began to be seriously considered as a potential cause of serious psychological distress with harmful effects on both development and mental health throughout the life span.There is, however, no evidence of a single diagnosis or pattern of symptoms unique to a history of child sexual abuse. Among the problems and symptoms that have been associated repeatedly with a childhood sexual abuse history are symptoms of post traumatic stress, low self esteem, guilt, anxiety, depression, somatisation, dissociation, interpersonal dysfunction, eating disorders, sexual dysfunction, substance abuse and suicidality Briere, 1992; Browne ; Finkelhor, 1986.These findings indicate the powerful harmful effects associated with the experience of child sexual abuse but can tell us little about the pathways from the experience to specific mental health problems.
Two commonly used explanations for the development of sexual abuse sequelae will therefore be examined; Post Traumatic Stress Disorder after Lindberg ; Distad; 1985 and Traumagenic Dynamics after Finkelhor ; Browne;1985. Both of these models, however, have difficulty in explaining how a significant proportion of individuals who have experienced child sexual abuse do not go on to have mental health problems in adulthood.Browne and Finkelhor 1986 concluded that when studied as a whole, survivors of sexual abuse show impairment compared with their non-victimised counterparts, but under a fifth show serious psychopathology. A review by Finkelhor 1990 concluded that as many as one third of all children who are sexually abused remain symptom free.
It would seem, therefore, that there are two questions here; Question1: Why might the experience of child sexual abuse per se lead to mental health problems in adulthood?And Question 2: How might other factors in the child’s experience mediate between child sexual abuse and mental health problems in adulthood? Examination of the literature would suggest that protective factors are not properly accounted for in the models used to explain the pathways from child sexual abuse to mental health problems. One exception is Alexander’s 1992 application of attachment theory as a conceptual framework in which to understand the familial antecedents and consequences of child sexual abuse.It is argued here that combining Browne and Finkelhor’s 1985 model with Alexander’s 1992 model provides a useful framework with which to both understand the development of sexual abuse sequelae and guide further research. Question 1: Why might the experience of child sexual abuse per se lead to long term mental health problems in adulthood? Post traumatic stress disorder PTSD is a descriptive model frequently proposed to characterise the impact of child sexual abuse.
It has been argued that both the symptoms and symptom onset closely fit the diagnostic criteria for a severe stress reaction Lindberg ; Distad, 1985, as defined by the 1980 Diagnostic and Statistical Manual of Mental Disorders DSM-III; American Psychiatric Association, 1980. These symptoms, well documented in cases of war, terrorism and rape, include; anxiety, recurring nightmares or intrusive daytime imagery, insomnia, depression, anger, guilt and mistrust. Horowitz 1986 presents a compelling formulation of response to traumatic stress.He suggests that we have internal representations schemas of the world and the self that include beliefs, knowledge, images and expectations, operating at both a conscious and unconscious level.
According to Horowitz 1986, trauma presents the individual with information and experience that is discrepant with these schemas. The more acute the discrepancy the greater the affect it arouses. If the experience is sufficiently stressful so as to preclude integration, it is stored in ‘active memory’, leading to the symptoms of post-traumatic stress.One of my reservations about this model as an explanation for the long term effects of child sexual abuse is that it would seem to follow that if an individual’s schemas are compatible with the abusive event, it would be perceived as less stressful and thus be less traumatic than if it were more discrepant; so a person who was repeatedly abused from an early age would be less traumatised by a subsequent abusive event and have fewer mental health problems later on than someone who was not abused from an early age.
Whilst there is evidence of a greater effect of post-pubertal abuse than pre-pubertal abuse, age of onset is associated with several other abuse-specific variables, confounding the results. For example, older children and adolescents are more likely than younger children to be subjected to invasive abuse Gomes-Schwartz, Horowitz & Sauzier, 1985. On the other hand age of onset is likely to be confounded with duration of the abuse and younger children are more likely to be abused by a father or stepfather, which is reported to result in greater trauma than abuse by other perpetrators Finkelhor, 1979]. The PTSD explanation for the symptoms of child sexual abuse cannot accommodate other associated adult symptomatology such as sexual disturbance, revictimization and personality disorders see Polusny & Follette, 1995, for a review.
Even within the PTSD research there are some difficulties. Lindberg, Lois and Distad, 1985 specifically selected their sample using the DSM-III criteria for PTSD, so that an association between the symptoms of PTSD and sexual abuse is inevitable the incidence of child sexual abuse in the psychiatric population as a whole is estimated at between 30 and 47 per cent; Beitchman, Zucker, Hood, Da Costa, Arkman & Cassavia, 1992.A review by Polusny and Follette, 1995, found that the prevalence of PTSD in clinical populations reporting a history of child sexual abuse is between 33per cent and 86per cent. It is quite clear, therefore, that not all victims of child sexual abuse experience the symptoms of post traumatic stress.
Furthermore, the criteria set by Polusny and Follette 1995 was past or present experience of the symptoms of PTSD, so that the short and long term effects might well be confounded.Finally, Rodriguez, Ryan and Foy, 1992, found that a history of concurrent physical abuse was significantly related to whether subjects received a diagnosis of PTSD, suggesting that sexual abuse per se might not be sufficient to produce the symptoms of post traumatic stress. Whilst all of the symptoms indicated in the DSM-III criteria for PTSD have been reported singly and in combination by victims of child sexual abuse, it is apparent that PTSD is too limited a model to account for all the long term effects of child sexual abuse.Some of the symptoms reported by survivors of child sexual abuse such as; re-experiencing the trauma through recurrent intrusive recollections; avoidance of activities or intensification of symptoms in situations associated with the traumatic event; anxiety; as well as problems with memory or concentration; can be accounted for in the PTSD model but there is a whole host of problems associated with child sexual abuse that the PTSD model cannot explain:Difficulties in interpersonal relationships, sexuality and social functioning have all been associated with a history of child sexual abuse, Mullen, Martin, Anderson, Romans & Herbison, 1994, as have self-injurious behaviour, substance abuse, eating disorders, dissociation, somatisation and personality disorders, Polusny & Follette, 1995; Beitchman, Zucker, Hood, Da Costa, Arkman & Cassavia, 1992, which cannot be accounted for by the PTSD model.
The PTSD model has more recently been replaced by Finkelhor and Browne 1985 who have conceptualised the traumatic effects of sexual abuse in terms of traumagenic dynamics. These affect a child’s cognitive and emotional orientation to the world, creating trauma by distorting his/her self concept, world view and affective capacities.This model proposes four trauma causing factors; traumatic sexualization, betrayal, powerlessness and stigmatization, which have greater scope in accounting for the effects of child sexual abuse mentioned above: Traumatic sexualization refers to a process in which the child’s sexuality is shaped in a developmentally inappropriate and interpersonally dysfunctional way as a result of sexual abuse. Betrayal refers to the dynamic by which children discover that someone on whom they were vitally dependent has caused them harm.
Powerlessness refers to the process in which a child’s will, desires and sense of efficacy are continually contravened. Stigmatization refers to the negative connotations that are communicated to the child around the experiences and which become incorporated into the child’s self image.One of the advantages of this model is that each category might contain opposite but related sequelae: The dynamic of Traumatic sexualization occurs when a child is repeatedly rewarded by the offender for sexual behaviour inappropriate to his or her level of development, through the exchange of affection, attention, privileges and gifts. It occurs when certain parts of a child’s anatomy are fetishized and given distorted importance and meaning, by transmitting misconceptions and confusions about sexual behaviour and sexual morality and through the association of sexual activity with negative affect such as fear.
According to this model, sexual abuse experiences can vary dramatically in terms of the amount and type of traumatic sexualization provoked, leading to a variety of psychological and behavioural effects.The psychological impact might therefore include an aversion to sex or intimacy, an association of negative affect with arousal, the confusion of sex with love, the increased salience of sex, as well as confusion about sexual identity and sexual norms. The behavioural impact of traumatic sexualization might include a preoccupation with sex, compulsive and /or aggressive sexual behaviour, sexual dysfunction, prostitution, avoidance of sexual activity and sexualization of parenting. Betrayal might occur through a child later realising that a trusted individual has manipulated their trust and vulnerability.
Children can experience betrayal not only at the hands of the abuser but on the part of the wider family who did not protect them or who responded to the disclosure by disbelieving, blaming or ostracising them.Psychologically, the effects of betrayal might include anger, grief and depression, mistrust and impaired judgement about the trustworthiness of others. Behaviourally, the effects of betrayal might manifest themselves in isolation through an aversion to intimate relationships, clinginess or a desperate search for a redeeming relationship, as well as antisocial behaviour and aggression. Powerlessness is the dynamic involved in a child’s body being invaded and the abuser’s use of coercion, manipulation or force.
This might then be reinforced by failed attempts to halt the abuse, by being dependent on the abuser and by the realisation of the consequences of disclosure.Psychologically, the effects of this dynamic might manifest themselves in fear and anxiety, despair and depression, dissociation, somatic complaints, an impaired sense of self-efficacy and ability to cope as well as a need to be in control. Behaviours associated with this dynamic might include nightmares, phobias, clinginess, suicidal behaviour, helplessness or a need to control or dominate. Stigmatization is the dynamic involved when negative connotations such as badness, shame and guilt are communicated to the child, either directly from the abuser, both overtly and covertly, and indirectly through attitudes that the victim infers from society.
This might then be reinforced by the person’s experience of disclosure.Psychologically, the effects of this dynamic might manifest themselves in guilt, shame, low self-esteem and a sense of being different, whilst behaviourally the effects might be to gravitate to other stigmatised groups in society, leading to drug and alcohol abuse, criminal activity and prostitution, as well as self-injurious behaviour Finkelhor and Browne, 1985. The arguments proposed by the traumagenic dynamics model are not entirely convincing with regard to self-injurious behaviour, dissociation and substance abuse. Shapiro 1987 has hypothesised that self mutilation may function to terminate dissociative states that were previously used by the survivor to cope with the overwhelming stress associated with the experience of victimisation.
Van der Kolk, Perry and Herman 1991 have noted that; “many patients report feeling numb and ‘dead’ prior to harming themselves. They often claim not to experience pain during self injury and report a feeling of relief afterward. ” p. 1665.
Thus dissociation may arise through the dynamic of powerlessness, as Finkelhor and Browne 1985 suggest but Shapiro 1987 provides a more convincing argument for the role of behavioural strategies i. e. , substance abuse and self-injurious behaviour in alleviating negative internal abuse-related experiences, rather than being a direct result of the dynamic of stigmatization. Overall, the traumagenic dynamics model allows for flexibility in the combination and degree to which dynamics are provoked, but it does have difficulty in accommodating factors that might mediate in the development of each dynamic.
Question 2: How might other factors in the child’s experience mediate between child sexual abuse and mental health problems in adulthood?Recent investigations have called attention to the fact that sexual abuse tends to occur in the context of other family stressors Nash, Husley, Sexton, Harralson ; Lambert; 1993. Spaccarelli 1994 argues that the experience of child sexual abuse predicts poor mental health outcome because of a combination of stressful events; abuse events, abuse-related events and disclosure-related events. In a sense, the traumagenic dynamics model can adequately account for the abuse events and disclosure-related events, but it fails to address the abuse-related events i. e.
, circumstances surrounding the abuse specified by Spaccarelli as family dysfunction, marital separation and loss of social contacts.Bryer, Nelson, Miller ; Krol 1987 found that individuals with a history of child sexual abuse were more likely than nonabused controls to come from single-parent families or families with a high level of marital conflict. Similarly, Peters 1988 found that the quality of maternal warmth was the strongest predictor of psychological difficulty in adulthood, accounting for 25 per cent of the variance, [although sexual abuse accounted for additional variance]. Mullen, Martin, Anderson, Romans and Herbison 1994 found that women who grew up in a non nuclear family with one or both parents absent were twice as likely to have been abused and three times as likely to report abuse involving intercourse, compared to women from an intact nuclear family.
Parental separation and frequent moves during childhood were also associated with increased risk of abuse, as were having a mother who was perceived as providing low levels of care and concern but at the same time being intrusive or punitive. This last finding must be interpreted with caution since Mullen et al. , 1994 used a correlational design, with the women providing retrospective accounts of both their abuse experiences and family functioning. We know that women who have been sexually victimised as children often report feelings of hostility towards their parents, Herman 1981, so that it would be imprudent to infer a causal link between retrospective accounts of family functioning and reported abuse experiences.
In Finkelhor’s 1984 Child abuse Risk Factor Checklist, the eight strongest independent predictors are : having a stepfather, having lived without the mother, not being close to the mother, mother never having finished secondary school, having a sex-punitive mother, no physical affection from the father, low family income and having two friends or less in childhood. Family context may therefore influence the risk of child sexual abuse and / or play a role in determining the nature of any associated long term difficulties, but with an absence of theories to predict the pathways from family dysfunction to sexual abuse, our understanding of the relationship between these factors remains ambiguous and open to misinterpretation.Formuth 1986 found that the difference between abused and nonabused college women became non-significant once parental supportiveness was controlled for. She concluded that the Parental Support Scale was a better predictor of later psychological adjustment than sexual abuse.
Interestingly, replication of this study in a sample of college men, Formuth & Burkhart; 1989, failed to find any correlation between either parental support or parental attachment. There are, however, methodological issues to be considered when interpreting these findings, quite apart from its retrospective design, which may also play a part in the discrepancy between studies. Both samples were unrepresentative of the population as a whole; university students with an average age of twenty.This selection might thus be an unusually highly functioning and ‘healthy’ sample, and as Formuth and Bukhart 1989 themselves point out, the age of the samples might mask difficulties which may emerge later in life.
In addition, broad band standardised measures of adjustment were used, which, not being theoretically-driven, might not be sensitive to the specific effects of child sexual abuse. Lastly, the manner in which the questionnaires were administered a large group in one room, might have affected the individuals’ responses, since they may have been inhibited by the prospect of having to then leave the room together and, presumably, discuss the contents of the questionnaires.In spite of the problems with interpreting the current research findings, there is considerable evidence for the role of family variables in mediating between the experience of child sexual abuse and later mental health problems; the effects of sexual abuse are difficult to differentiate from those of emotional and physical abuse as well as from other family variables and there is no evidence of a specific diagnosis or pattern of symptoms unique to a history of child sexual abuse. Alexander 1992 uses attachment theory as a conceptual framework in which to understand the antecedents and consequences of child sexual abuse.
She uses Bowlby’s 1988 concept of an internal working model of relationships as a framework for understanding the onset and outcome of child sexual abuse.Briefly, the essence of attachment theory is that on the basis of early experience with the attachment figure, the infant develops expectations about; i his or her own role in relationships worthy and capable of getting others’ attention versus unworthy and incapable of getting needed attention, and ii others’ roles in relationships, trustworthy, accessible, caring and responsive versus untrustworthy, inaccessible, uncaring and unresponsive. These expectations are internalised and govern how incoming interpersonal information is attended to and perceived, as well as the affect experienced, the memories evoked and mediating behaviour with others in important relationships. The internal working model is therefore affected by and comes to affect the types of interpersonal experiences that are encoded into the concept of the self.
There is wide support for this theory as well as evidence to suggest that a child’s attachment to a particular caregiver is relatively stable over time Bretherton; 1985, and that there is continuity of attachment from childhood into adulthood Main, Kaplan ; Cassidy; 1985. Alexander argues that sexual abuse is frequently associated with the intergenerational transmission of insecure attachment styles, citing evidence Haft ; Slade; 1989 that mothers’ attachment styles correlate highly with maternal attunement to the infant’s affect and that the relative comfort or discomfort of the mother with certain kinds of emotional states can influence the infant’s subsequent access to those same emotions at a very early age.She goes on to postulate how specific organising themes rejection, avoidance, role reversal/parentification and fear/unresolved trauma, associated with the three insecure attachment categories, might both precede the onset of abuse and interfere with its termination see Alexander; 1992, pp. 188-189.
Finally, Alexander 1992 argues that the application of attachment theory can account for many of the factors found in the literature which point to a good prognosis following the experience of child sexual abuse; the fact that those who remain symptom-free are more likely to have been abused for a shorter period of time without force or penetration, by someone who is not a father figure, and more likely to have received acknowledgement and support from the family following disclosure after Finkelhor; 1990.Alexander’s approach therefore leads to interesting predictions and testable hypotheses regarding the prognosis for individuals with a history of child sexual abuse. Furthermore, the application of attachment theory has implications for therapeutic interventions, see p. 13 below.
Discussion Combining Alexander’s model with the traumagenic dynamics model would seem to be more fruitful than using either model in isolation. Doing so might help to understand the formation of early maladaptive schemas Young; 1990. For example, the dynamic of sexualization might lead to the formation of a schema such as people will hurt me, the dynamic of betrayal to no one can be trusted, the dynamic of powerlessness to I am helpless and the dynamic of stigmatization to I am worthless and bad.The concept of the self would be mediated by the child’s relationship with a significant other, so that rejection by that person would exacerbate the effects of the dynamics above by contributing a schema such as I am unlovable, and role reversal might add other peoples’ needs come first and so on.
In contrast, the experience of having one’s needs met would contribute more adaptive schema such as other people are trustworthy, caring and responsive and I am loveable, to a person’s sense of self. According to such a model, the maladaptive schemas might be limited to the abuse experience itself, if the attachment figure or other significant relationships provided conflicting information.One would predict, therefore, that a ‘positive’ working model of most relationships would significantly reduce the interpersonal difficulties associated with the abuse experience, and would affect a person’s vulnerability to mental health problems. Not only does such a model provide testable hypotheses with which to guide further research into the long term effects of child sexual abuse, it also provides a useful framework for therapeutic interventions.
Clinical Implications Firstly, given that sexual abuse is frequently associated with the intergenerational transmission of insecure attachment styles Haft ; Slade; 1989and that mothers of insecurely attached infants have been shown to have insecure attachment histories themselves Main et al.1985, therapeutic interventions should specifically focus on attachment issues, using the principles set out by Bowlby 1989; i to provide a secure base that permits the exploration of painful experiences which are otherwise too difficult to explore, ii to encourage consideration of current relationships, iii to encourage consideration of the relationship to the therapist, iv to explore how current perceptions, feelings and actions relate to the client’s feelings towards his/her parents v to enable the client to recognise that current models of self and others may or may not be appropriate, and to help the patient to imagine alternatives.Secondly, therapeutic interventions should focus on changing early maladaptive schemas using standard cognitive techniques such as identifying the schemas and the triggers for them, educating the client about schemas and their role in moderating affect and behaviour, re-interpreting old schemas, activating pre-existing conflicting schemas if available and constructing new schemas, Young; 1990.Thirdly, therapeutic interventions need to take into consideration the fact that many of the long term effects of child sexual abuse involve difficulties in interpersonal relationships, which the client will naturally bring to the therapeutic relationship.
Conclusion Examination of the literature indicates that child sexual abuse is associated with multiple short-term and long-term psychological difficulties. Most of these studies have used correlational designs and retrospective reports of abuse, with the result that few studies have been theory-driven and are thus unable to inform clinical practice. Combining attachment theory with the traumagenic dynamics model provides a better explanation of the consequences observed in the literature than either model in isolation, as well as providing a conceptual framework with which to guide both further research and clinical practice.
