An Ecological Analysis of Child Sexual Abuse Disclosure: Considerations for Child and Adolescent Mental Health

Abstract

Objectives:

Research continues to indicate a concerning number of children and youth, between 60–80%, withhold disclosure until adulthood suggesting that many children endure prolonged victimization or never receive necessary intervention. The study aim was to qualitatively identify factors that impede or promote child sexual abuse (CSA) disclosure.

Methods:

Using a phenomenological design, forty adult survivors of CSA were interviewed about their disclosure experiences to provide retrospective accounts of their childhood and adolescent abuse experiences, disclosure attempts, and meaning-making of these experiences.

Results:

Findings show that disclosure is multiply determined by a complex interplay of factors related to child characteristics, family environment, community influences, and cultural and societal attitudes. An ecological analysis is offered to understand these complexities. Unless barriers to disclosure are eradicated, negative effects of CSA can persist manifesting in serious mental health issues.

Conclusions:

Practitioners can expect to work with children, adolescents and adults who have withheld disclosure or attempted to tell over time having experienced a wide range of responses. Multi-level intervention is recommended at the individual, community and macro-levels. Future investigations should focus on how to identify and measure the impact of community and macro level factors on disclosure, aspects that have received much less attention.

While there have been recent significant advances in public awareness efforts along with prevention and intervention programs for child sexual abuse (CSA), disclosure of sexual victimization remains a difficult undertaking. By some estimates between 60–80% of CSA victims withhold disclosure suggesting that many children and adolescents endure prolonged victimization and do not receive necessary therapeutic intervention (Hébert, Tourigny, Cyr, et al., 2009; Jones, 2000; Paine & Hansen, 2002). Studies that examine latency to disclosure report a mean delay from 3–18 years (Hébert, Tourigny, Cyr, et al., 2009; Smith, Letourneau, Saunders, et al., 2000). In the most recent Canadian study on prevalence and disclosure of CSA (Hébert, Tourigny, Cyr, et al., 2009), researchers concluded that almost 58% of victims delayed disclosing for five years or more, and 20% of their sample never disclosed to anyone. Of primary concern are the immediate and long-term effects of CSA which have serious and persistent mental health implications if left untreated. Most common negative effects of sexual abuse in childhood manifest in emotional and behavioural problems, post-traumatic stress symptoms, depression, suicidal ideation and self-harm behaviours, anxiety, substance abuse, aggression, self-esteem issues, academic problems, and sexualized behaviors (Beitchman, Zucker, Hood, daCosta, & Akman, 1991; Kendall-Tackett, Williams, & Finkelhor, 1993; Martin, Bergen, Richardson, et al., 2004; Putnam 2003; Walrath, Ybarra, Sheenan, et al., 2006). Long-term effects indicate increased risk for individuals with a sexual abuse history (vs. non-abused individuals) for major depression, suicide, addictions, post-traumatic stress disorder, anxiety disorders, antisocial personality disorder, dissociation, and sexual dysfunction (Dubé, Anda, Whitfield, et al., 2005; Gold, Lucenko, Elhai, et al., 1999; Hébert, Tourigny, Cyr, et al., 2009; Putnam, 2003; Spataro, Mullen, Burgess, et al., 2004; Walrath, Ybarra, Sheenan, et al., 2006). Clearly, if victims are able to disclose when the sexual victimization is occurring, there is a greater likelihood that the abuse will be stopped and appropriate help will be received leading to better mental health outcomes.

However, barriers and facilitators of disclosure are still not that well understood. The aim of this study was to qualitatively identify dynamics that impede or promote disclosure by examining a range of factors including child and family features, environmental influences, cultural considerations, and societal attitudes. Identification and removal of barriers to disclosure means that victimization can be stopped and treatment can be offered to ameliorate short and long-term negative effects.

Method

Design

The research questions were formulated on a comprehensive review of the research literature as follows: 1) What individual and environmental influences impact disclosure of CSA and; 2) What contextual factors promote or impede disclosure in childhood? These questions suited an inductive, discovery-oriented inquiry using qualitative methodology. The study employed a phenomenological design -the Long-Interview method developed by McCracken (1988) compatible for collecting and analysing data involving process issues and for uncovering complex processes. This method lends voice to the ‘lived experience’ of CSA survivors and closely parallels the method developed by phenomenologist van Manen (1997). The study received approval from the investigator’s institutional ethics review board, Health Sciences I at the University of Toronto.

Sampling

Purposive sampling, a method of qualitative sampling that targets participants on specific characteristics and experiences, was employed for this study (Creswell, 2007). The intent was to capture the experiences of both women and men, as well as those who disclosed during the abuse and those who withheld. Recruitment occurred through networking with social service agencies and advertising on a university campus. Challenges with sampling were encountered when trying to recruit male survivors. Since male participants were more difficult to locate snowball sampling was also employed, e.g. word of mouth where male participants were asked to refer other men they knew into the study. Over one-third (36%) of the final sample were men which is consistent with rates cited in incidence and prevalence studies (Cawson, Wattam, Brooker, & Kelly, 2000; Fergusson, Lynsky & Horwood, 1996; Putnam, 2003). Exclusion criteria included people under the age of 18 and anyone experiencing major mental health issues (e.g., actively psychotic or suicidal). Study participants reflected a mixed clinical and community sample although the majority had received treatment for their victimization at some point in their lives. Sampling was stopped at the point of saturation on the research questions posed. This is consistent with Creswell’s (2007) assertion that categorical saturation can be achieved on a sample of 20–30 cases.

Data Collection Instruments

An in-depth interview guide was developed by the author in collaboration with a research advisory committee comprised of child and family practitioners, and mental health specialists. The interview guide probed: individual and family history; family background and dynamics; sexual abuse onset, duration, type; strategies used by the perpetrator in maintaining secrecy; disclosure attempts and outcomes; (mis)perceptions and beliefs developed by the victim; nature and quality of primary/significant relationships in childhood and adolescence; and environmental contributors, and cultural influences. Questions were open-ended and allowed room for participants to add comments that may not have been asked of them. The majority of the interviews were conducted by the author.

Data Analysis

Interviews were transcribed and then imported into N*Vivo, computer software used for the analysis of qualitative data. Phenomenological reduction was the primary method of data analysis in which transcendental (descriptive) and hermeneutical (interpretive) analyses were conducted (Ray, 1994; van Manen, 1997). Initially the interviews were read for descriptive data and thematic development. Line-by-line review of the transcripts was conducted to inform a first level coding framework that was later used to assist in the interpretive analysis. In the second stage, the interviews were imported into N*Vivo, coded, and higher-level codes were also determined. The use of N*Vivo software aided in managing these data, grouping of categories, creating memos for theme evolution, and analysis within and across interviews.

Trustworthiness

Reliability issues were addressed through credibility, dependability, and transferability typically used to establish trustworthiness in qualitative investigation (Drisko, 1997; Lincoln & Guba, 1985; Padgett, 1998). In qualitative research transferability takes precedence over generalizability. Uncovering processes and understanding context are the main concerns in examining the phenomenon. Trustworthiness of the data was ensured through prolonged engagement with the topic (the author’s 20 years of experience in CSA counseling and program development), persistent observation through intensive interviewing and follow-up with individual participants, and peer debriefing of emerging themes with experts in the field. Dependability of the data occurred through verbatim transcription of the interviews to prevent secondary revision or distortion of the narratives. An independent coder analysed a sub-sample of transcripts to notice if similar or different themes were emerging and if all themes were exhausted. This process was undertaken until consensus was achieved. In addition, theoretical triangulation was conducted in interpreting the findings, achieved in this study through using an ecological theoretical framework for deeper analyses of these data.

Results

Sample Characteristics

Data were generated through in-depth interviews with fourteen male (14) and twenty six (26) female survivors of CSA for a total of 40 interviews. The participants ranged in age between 18 and 65 with a mean age of 40.1 years. About two-thirds of participants were female and a little over one third were male. The average age of onset of abuse was 5.3 years old and the vast majority of the perpetrators were biological fathers, step-fathers, mother’s partners and grandfathers. Participants were of diverse socio-economic background – middle-class, working poor, and homeless and 36% of the sample were non-white. Forty percent (42%) of the participants had disclosed the abuse during childhood; 32% had attempted some form of disclosure in indirect ways during childhood and; 26% had not disclosed because they had repressed the memory, or the abuse had occurred in pre-school years and they had difficulty with recall. Purposeful disclosure is slightly higher than reported in other studies because of the sampling attempts to purposefully locate disclosers.

In the initial analysis a thematic analysis was undertaken. Themes that emerged as having an impact on childhood disclosure included: individual and developmental factors; family dynamics; community context and societal attitudes. These themes closely paralleled the interview guide and supporting quotes for confirmability are presented below. Although these are presented as discrete thematic areas, they overlap in significant ways.

Individual Characteristics

Study participants were asked to describe themselves throughout their childhood and adolescence and in relation to disclosing the sexual abuse. Some quotes suggest that developmental factors entered into whether they comprehended what was happening, while temperament or personality traits also had some bearing on their ability to tell.

But I never had that thought as a child to tell. I just (pause) I don’t know how or why the fear was there but I, I didn’t know I could tell because we weren’t allowed to talk about anything.

(P9: f: 41).1

I am still a very introverted person. I always will be. I am able to still be very withdrawn. I wouldn’t have told you anything if I hadn’t trusted you.

(P7: f: 54).

In other cases, descriptions indicated that the words just were not there for them. Another participant stated ‘we tried to tell my mother’ yet concludes his statement by acknowledging that ‘I couldn’t come out and say it.’ (P2: m: 36).

Family Dynamics

Family themes that emerged indicated that CSA disclosure can be negatively affected by certain characteristics such as: rigidly fixed gender roles with dominating fathers; chaos and aggression; the presence of other forms of child abuse; domestic violence; dysfunctional communication and; social isolation. Notice that in this first quote, although the participant is talking about the past, the words come out as if the danger still exists in the present, even though his disclosure was highly publicized as part of uncovering a pedophile ring.

I do have the reality that I have death threats from my family for talking about the abuse, and I do have the reality that I could be killed.

(P18: m: 43).

Other study participants recounted closed communication patterns in their families:

But there was no communication, there was no talking, there was none and then my mother left us.

(P4: f: 61).

Many study participants also described high levels of aggression in the family unit which shut down possibilities of disclosure:

If he did some of the things he said he was going to do then he would have killed us.

(P14: f: 29).

He was abusive with me mostly, but with my mother yeah. Looking back now yeah he was very abusive with her.

(P15: f: 32).

Neighbourhood and Community

It was not surprising to find that participants spoke of the larger environmental impact they experienced as child victims, however the voluminous nature of detailed description was of particular interest. The vast majority of study participants noted how distressing it felt to experience a lack of empathy from neighbours. For example, one participant recounts how she was ostracized from neighbours and friends after the police came to their home responding to a domestic dispute call involving her father’s assault on her mother. Following this incident her best friend pulled away from her:

And she said, ‘I’m not allowed to talk to you any more. My mother said your family’s not nice.’ So I remember walking home and then after that I didn’t make any attempts to make friends.

(P4: f: 61)

She further explained a ‘snowball effect’ taking place whereby she already felt isolated because of this incident and then the thought of telling anyone about her father sexually abusing her was far too overwhelming.

Study participants also noted a marked lack of involvement from school personnel when they had the expectation that someone should take notice and reach out:

I used to hate recess because I had no friends and I used to walk back and forth along the fence all by myself, hands in my pockets, head down, and I never spoke to anybody. I was always, always by myself. If one teacher would have asked me is there something wrong, I mean it would have helped.

(P15: f: 32).

I know, in high school the teachers just used to think of me as a loner and quiet and shy and just leave me alone.

(P32: f: 45).

When teachers did notice they did not seem to know how to pursue troubling behaviour:

I remember grade ten through twelve I would be terrified to go home. I would stay until the doors closed at school and the teachers always wondered why are you staying here so late, why don’t you go home? I don’t want to go home.

(P9: f: 41).

Finally, there were many statements around simply anticipating not being believed:

I just don’t trust people or trust that people will respect it. They look down upon me or think there is something wrong or they won’t believe it. There are all kinds of reasons. People just don’t want to hear about it.

(P14: f: 29).

Cultural and Societal Attitudes

An enduring theme that surfaced through probing of larger structural and cultural influences was the richness of examples participants provided of messages they were exposed to through the media and in social circles. One woman whose abuse started at the age of 11 remembered commercials she viewed with young girls acting in seductive ways:

I don’t know if I was maybe 11 or but um, seeing Brooke Shields in a jeans commercial on TV and her lying down and the jeans being unzipped and finding out how old she was ...Calvin Klein using 12-year-old girls.

(P24: f: 31).

This was confusing for her as she was being abused by a teacher and wondered if this was how girls her age were expected to behave. She was also conflicted because she enjoyed the attention of an adult as she was living in an emotionally deprived family environment.

Male survivors of CSA described a host of conflicts about disclosing on two levels –men shouldn’t be victims and because they were abused by men this called into question their sexual orientation.

So in a way by saying to somebody you were sexually abused, is the fear of just having that ‘feminine’ thing attached to you.

(P31: m: 34).

I think it’s because growing up, guys are always bombarded with the whole jock mentality, jock image that this is how a man is.

(P26: m: 29).

For some men this spun into hyper-masculinity and highly sexualized behaviour.

I didn’t understand lovey-dovey feelings. I was strictly out for what I could get. Fulfill my own needs. I guess I was so sexualized that would make me feel more like a man and less confusion about homosexuality.

(P1: m: 58).

A pervasive theme was one of how they felt unheard as children. Participants at the age of 23 or 64 made references to cultural attitudes of children as being without voice:

However in sort of daily life there was a sort of um, a how do you call it? Sort of like, I think it was an element of sort of children being seen and not heard and not as full human beings.

(P8: m: 64).

This type of message (of not being heard) lead some participants to assume that if they told they would be dismissed:

Now when I get older it is the fear of telling because I am being blamed, well you should have done it then, maybe you don’t remember, how can you remember that many years ago?

(P14: f: 29).

Higher level analysis revealed important sub-themes that became apparent as these categories were collapsed and refined, and relationships among categories were explored. This stage of the analytic procedure identified important sub-themes related to the interactive dynamics of disclosure indicating how both victims and recipients of disclosure are impacted by greater structural and cultural forces. Just as the individual child or youth is affected by environmental, cultural and structural influences, significant people in their lives may also be impacted by these same factors in their response to disclosure. For those individuals who attempted disclosure at the time of abuse or shortly after onset (40%), responses were consistently insensitive. One participant explained:

I told my mom when I was in grade one about our neighbor’s daughters, that they were doing things with each other and to me, but it caused a whole lot of ruckus.

(P8: m: 52).

An unpleasant situation unfolded for him when his mother did take action and spoke to the parents of the girls, who immediately denied the events, with the parents accusing the boy of fabricating the story. His account was made more difficult to endorse because the girls corroborated each other’s stories. He further explained that he was ostracized and isolated at school and in his community after that.

The following participant was entering adolescence when an older cousin started to sexually molest him. Upon telling his mother she believed that her son should put a stop to it, emphasizing that he would if he really didn’t like it, as if he had enough personal agency to stop it.

I just described to her what he did to me one day (long pause) and she just told me to stay away from him. She said that if I really didn’t like it that I should stop him.

(P10: m: 28).

This participant believed that his mother did not want to disrupt and upset the family by confrontations around sexual abuse. He described her attempts to reduce his contact with his cousin but that he ultimately did not feel fully supported by her. He further explained that she was a single mother and did not want to alienate her family for the support they offered her.

In the following situation a young girl disclosed to her aunt after she had initially told her mother, who did not take steps to protect her from her step-father’s sexual abuse. However, because the family was in a socially prominent position, and her step-father held a high rank in the military, she believed she was silenced to preserve the family’s elevated status and his job.

So I finally decided to tell my aunt one day what was going on and she acted just like my mother and started to get abusive too.

(P7: f: 54).

In another case when child welfare authorities found out about the abuse and removed the teen from her family, the participant’s mother later asked why she hadn’t told her after which her mother blatantly admitted she wouldn’t have believed her disclosure anyway. At the time the sexual abuse had been discovered, the participant was a rebellious teenager who was drinking heavily and engaging in other substance abuse. She had also started running away from home so that the mother-daughter relationship was very strained. Any of these behaviours could have been attributed to the abuse which had not yet been disclosed.

At that point I had disclosed after to my mother and she said, ‘why didn’t you tell me before’ and I said ‘because I didn’t think you would have believed me’ and she said ‘you know you are right. I may not have at the time’. I mean I love her for her honesty (laughs) but you know?

(P22: f: 38)

Professionals were also noted to have responded ineffectively by involving the family and increasing the risk to the abused child who in the following case had disclosed to a teacher at the age of seven:

Then the teacher found out and I was told to go to the principal’s office. At the principal’s office, we didn’t have a telephone, my mom and dad and I ended up with my grandmother finding out and she locked me up for two weeks, saying I was lying and I was terrible and I was telling all of these lies.

(P36: m: 47).

Others responded in biased and discriminatory ways, in this case racist, when a young adolescent disclosed to her physician:

He actually, he said to me – and he’s still a doctor there- um, he said to me ‘well that’s not sexual assault, you were old enough [13 or 14]. You probably enjoyed it, you know. You girls in those countries enjoy those things anyway, that’s not sexual assault’.

(P24: f: 31).

Discussion

Interview data from this study show that disclosure of child sexual abuse is multiply determined by factors related to child characteristics and history, family dynamics, community context, and larger cultural and societal attitudes. It is a complex process, not easily defined and is shaped by what each individual brings to it and their unique set of circumstances. As well, these data confirm that disclosure is an on-going process not a single event (Bradley & Wood, 1996; Sorensen & Snow, 1991; Summit, 1983). While the study objectives set out to identify both barriers and facilitators of disclosure, unfortunately barriers far outweighed facilitators.

When trying to make meaning of these data an ecological theoretical framework was used for organizing and analyzing the vast amount of information brought forward through the participants’ narratives. In-depth analyses of the study data demonstrate a close fit with ecological theory. Drawing from Bronfenbrenner (1979), and Belsky’s (1980) ecological theoretical framework initially used for understanding the interplay of factors contributing to child abuse, it became clear that this framework can also be extended to disclosure of CSA for a fuller understanding of child victims. The ecological framework rests on the premise that individual behavior can only be understood by taking into account factors at each of: ontogenic, micro-system, exo-system and macro-system levels (Bronfenbrenner, 1979). Ontogenic development deals with the personal history and characteristics of the individual; micro-systems involve the immediate context of the individual (e.g., family and other intimate relationships); exo-systems refer to formal and informal structures such as neighbourhood, community, social networks and their availability and; macro-systems include cultural values and belief systems that impact on the individual.

Ontogenic Level Factors – Individual Characteristics: The age of onset of abuse had a significant impact on CSA disclosure in this sample. The younger the child was when the abuse started, the more difficult the disclosure. Obvious developmental constraints related to cognitive and verbal functions, and recall ability, hindered intentional disclosure. Behavioural indicators were most often cited as ‘signals’ of victimization. Study participants described behaviors ranging from clinging and temper tantrums as younger children, to angry outbursts as adolescents, serious substance abuse, withdrawal, eating problems, avoiding being at home or running away to alert adults to problems (see Alaggia, 2004 for expanded typologies of disclosure).

An intervening variable was the temperament or personality of the child. Adults in this sample retrospectively recounted personality characteristics and described themselves as either withdrawn or introverted children or as acting out and disconnected from feelings. It is difficult to tell whether these were internalizing or externalizing effects as a result of CSA, manifesting after the onset of abuse, or whether these were traits of the child before the onset of abuse that both affected their ability to disclose and were vulnerabilities for being abused.

Micro-System Level Factors – Family Environment: Family characteristics were probed to understand the impact of family environment in the context of disclosure but not for causal CSA factors. However, the families of survivors were largely described as chaotic, ruled by aggression, plagued by substance abuse, closed systems that were uncommunicative and socially isolated –all factors that silenced the child victims. The predominance of these characteristics in this study sample can be explained by turning to the work of Alexander and Schaeffer (1994) who studied typologies of incestuous families. They found “Cluster 3 families” to be the most pathological and dysfunctional of these families marked by high levels of conflict and control, violence in the parental unit, less emotional bonding, and the presence of other forms of child abuse such as physical abuse. In the current study two thirds of the sample described family environments very similar to “cluster 3” descriptions. The families in the study sample were characterized by high levels of violence, such as threats of murder, and poor parental response to disclosure indicating less emotional bonding. This may account for the preponderance of non-disclosures and indirect disclosures (see Alaggia & Kirshenbaum, 2005 for detailed description).

Finally, another contributor to non-disclosure was the presence of intimate partner violence in this sample with the adult victim most often being the mother. The co-occurrence of woman abuse and child sexual abuse has been documented elsewhere and is remarkably high. This family dynamic also has serious implications for both disclosure and response (Alaggia & Turton, 2005; Kellogg & Menard, 2003). Clearly, disclosure became more complicated for child victims in the face of their mothers’ victimization. When both the child and a parent are being victimized by one perpetrator the prospect of telling becomes more frightening.

Exo-System Level Factors – Neigbourhood and Community Influences: Contextual aspects have been much more difficult to identify and measure in terms of their impact on disclosure of CSA. A qualitative approach is a starting point for unearthing these factors and the themes brought forward through this study give guide-posts for areas of further investigation. Overall, participants described their communities as uncaring or unknowledgeable in how to handle their behaviours or their disclosures. It should be taken into account that the sample mean age reflected a generally older cohort which indicates that they may not have been exposed to newer sexual abuse awareness and prevention programs. Future research could utilize stratified sampling to include adults who have been recipients of in-school prevention programs and more recent child welfare legislation.

Macro-System Level Factors – Cultural and Societal Attitudes: As with exo-system factors, little is known about the impact of cultural and societal attitudes. One area that becomes clearer through these data is how socialized gender roles and subsequent expectations pervaded through cultural messaging. Men reported being profoundly affected by prevailing attitudes about masculinity and images of men as supposed to be immune to victimization or as being perceived as inadequate when victimized. They feared being identified as homosexual and being the target of unfair treatment if they disclosed abuse by a male perpetrator. A few described adopting hyper-masculine stances to counter these fears. Male victimization is beginning to be more closely examined but needs to be further explored in terms of disclosure and on larger samples (Alaggia & Millington, 2008; Gartner, 1999). Women were equally as affected by a culture of sexist and patriarchal attitudes inhibiting their ability to disclose. They were more likely to fear being blamed and described feeling more responsibility for their victimization and for repercussions of telling. They anticipated being disbelieved, and blamed for the abuse (or for not stopping the abuse). As well, both women and men described being raised with attitudes that children do not count. The phrase “children should be seen but not heard” was frequently recited by participants as commonly heard throughout their childhoods.

Study Limitations

Using a retrospective sample is one constraint of the study that should be noted. Recounting events that occurred in childhood is subject to memory failure, distortion and revision of events. Although the range of age of participants was broad the sample mean age was 40.1 and therefore the experiences of a younger cohort are not as well represented. The sample is also skewed to those who did not disclose, or attempted to disclose on childhood/adolescence but were met with disbelief or minimization. Therefore they recounted more negative factors in their narratives. Future studies should focus on securing a sample of disclosers who experienced positive responses in childhood to better identify and understand the facilitators of disclosure.

Implications for Practitioners

Although CSA disclosure is hoped for at the onset of victimization, our best research to date is still showing a consistent trend of disclosures being commonly delayed or withheld. The present study has documented that individual level factors partly account for non-disclosure, but that a host of environmental factors are also implicated in withholding disclosure. Based on the narratives of the study participants, practitioners should be prepared for disclosures of historic sexual abuse given that so many victims only disclose childhood victimization in adulthood. Patients will often present in therapy with other complaints or problems such as depression, addictions, eating disorders, relationship issues, sexual dysfunction, aggression and so forth, before revealing sexual trauma in childhood.

Practitioners can expect to work with children, adolescents and adults who have attempted to tell over time in different ways, having experienced a wide range of responses. Thus, it is important for professionals to inquire about reasons for withholding, all attempts to tell, responses following disclosure, and consequences of telling. Information should be gathered about family and other intimate relationships since certain family characteristics increase risk for lack of disclosure.

At the practice level, promoting primary prevention, such as awareness programs and parenting programs, and secondary prevention aimed at larger structures such as the media and justice system should be developed and sustained, as these influences can have profound effects on how children and youth perceive abuse, and anticipate responses to disclosure based on cultural messaging. The promotion of such programs also has the potential to reach parents and professionals to sensitize them to possible disclosures and preferable responses. More comprehensive work should also be focused at the community level and macro-systems with further research to assist in identifying and measuring these factors and their impact on disclosure, as well as responses to disclosure. It is important to keep in mind that the forces which influence children and youth in their disclosures are the same forces that permeate society and influence the responses of people to whom youth might disclose.