Suicidal behaviour is a cause for concern among many western countries; in general, it is most common among young women. This research used qualitative methods to explore the narratives of 24 Cate Curtis, PhD, lectures in psychology at the University of Waikato, New Zealand. She is interested in female self-harming behaviour, including self-mutilation and suicidal behaviour; social factors implicated both in engaging in self-harm and in recovery, particularly the roles played by family and friends; and barriers to help-seeking behaviour such as stigma. She is also interested in the ways people diagnosable with mental illness make sense of their experiences of being “unwell” and their experiences as consumers of mental health services. Cate has also worked in a number of social service agencies as a youth and community worker.
Address correspondence to: Cate Curtis, PhD, Psychology Department, University of Waikato, Private Bag 3105, Hamilton, New Zealand (Email: [email protected]). The author wishes to thank the participants who candidly shared their experiences of suicidal behaviour and sexual abuse, and hopes that the opportunity to have their voices heard through this paper goes some way to repay their contribution. amine the meanings of events leading to and implicated in the recovery from suicidal behaviour. The research confirms sexual abuse as a common precursor to suicidal behaviour; several women asserted that they would not have attempted suicide if they did not have a sexual abuse history.
KEYWORDS. Sexual abuse, suicidal behaviour, adolescent mental health, intervention
Barriers to early death are increasingly strengthened through advances in medical science; we are more aware of the causes of premature death than ever before. Yet some young people continue to attempt (and in some cases succeed) to take their own lives. Internationally, adolescents and young adults are at greater risk of suicidal behaviour than other age groups (Gould et al., 1998; Romans, Martin, Anderson, Herbison, & Mullen, 1995), and while males complete suicide at higher rates than females, rates of suicidal behaviour in general are considerably higher for females (Ministry of Youth Affairs, Ministry of Health, & Te Puni Kokiri, 1998). Examinations of risk factors for suicidal behaviour have largely been quantitative in nature, seeking to determine correlations. Also, the majority of studies have been conducted with clinical populations.
The research discussed in this paper attempts to address these possible methodological issues through the use of qualitative methods with a community sample. The paper discusses the experiences of women who engaged in suicidal behaviour while under the age of 25 through their first-hand accounts. Of particular interest is the relationship between sexual abuse and subsequent suicidal behaviour, and how sexual abuse impacts upon help-seeking behaviour and the efficacy of interventions for suicidal behaviour. Adults who have been victims of sexual abuse as children or adolescents report significantly greater symptoms indicative of depression, anxiety, and self-abusive and suicidal behaviour.
In a 1992 study by Saunders, Villeponteaux, Lipovsky, Kilpatrick, and Veronen, abuse survivors were significantly more likely than others to meet diagnostic criteria for agoraphobia, panic disorder, obsessive-compulsive disorder, major depression, social phobia, and post-traumatic stress disorder (PTSD). Vajda and Steinbeck (2000) found that childhood sexual abuse is a stronger predictor of repeated suicidal behaviour than individual characteristics and other stressors, and Read, Agar, Barker-Collo, and Davies (2001) found that “Current suicidality was predicted better by child sexual abuse (experienced on average 20 years previously) than a current diagnosis of depression” (p. 367). Rodriguez-Srednicki (2001) reported increased rates of drug use, alcohol abuse, disordered eating, risky sex, dissociation, self-mutilation, and suicidality in a sample of 175 female college students who were survivors of childhood sexual abuse, as compared to 266 female college students with no reported history of abuse.
Likewise, elevated rates of depression, anxiety, low self-esteem, drug or alcohol abuse, suicide attempts, and psychiatric admission were found in McCauley and colleagues’ (1997) study of childhood physical and sexual abuse in American women. Similar findings are reported by Polusny and Follette (1995), Silverman, Reinherz, and Giacona (1996), Stepakoff (1998), Read et al. (2001), and Read, Agar, Argyle, and Aderhold (2003) have linked sexual abuse to hallucinations, delusions, and thought disorders.
Sexual abuse has been linked to a number of negative psychological outcomes in addition to diagnosable disorders. Effects include trouble sleeping, nervousness, thoughts of hurting oneself, and learning difficulties. Women whose abusive experiences occurred within the family are at greater risk of disturbance than other women (Sedney & Brooks, 1984). Wagner and Linehan (1994) reported that not only are women who have been sexually abused more likely to engage in deliberate self-injury, their behaviour is also more likely to be lethal than that of women who did not report abuse. More recently, Gladstone,
Parker, Mitchell, and Malhi (2004) argued that depressed women with a history of childhood sexual abuse may require specifically tailored interventions. While a casual reading of the literature may suggest that most women who have been sexually abused go on to experience psychological problems, the Otago Women’s Health Study found that only one in five women who reported sexual abuse as a child developed a psychiatric disorder (Ministry of Health, 1998). Reviewing a number of studies, Goodyear-Smith (1993) argued that the other forms of abuse and family dysfunction that tend to occur alongside sexual abuse may in fact play at least as great a part in later depression and psychological problems. Boudewyn and Liem (1995) suggested that the longer the duration and the more frequent and severe the sexual abuse, the more depression and self-destructiveness is likely.
In a large random community study, Romans, Martin, and Mullen (1997) found that of their 252 participants, 26% of the participants reported sexual abuse before age 12 and 32% were sexually abused by the age of 16. Twenty-three (4.8%) of those interviewed reported a history of deliberate self-harm, and 22 of these 23 reported childhood sexual abuse. The one woman who self-harmed without a history of childhood sexual abuse reported sexual and physical assault as an adult. It should be noted that the vast majority of women who were sexually abused did not report self-harm. A clear “dose effect” was found: the more frequent and intrusive the childhood sexual abuse, the stronger the association with selfharm. This was also found in a study by Mullen, Martin, and Anderson (1996). Sexually abused participants who had self-harmed were more likely than other survivors of sexual abuse to report depression, anxiety disorders, eating disorders, and to drink alcohol in excess of the recommended guidelines.
They were also more likely to have experienced psychosocial disadvantage in their families of origin, such as low care/ high control relationships with their parents, parental discord, paternal depression or alcohol abuse, and physical abuse. Romans and colleagues’ study clearly demonstrates that although not all women who have been sexually abused go on to harm themselves, the majority of women who harm themselves have been sexually abused. Although exact figures cannot be obtained and various definitions1 are used in the research, sexual abuse in girls and young women is not uncommon, and it is generally accepted that females are far more likely to be victims of sexual abuse than males. For example, Saunders and colleagues (1992) reported that 10% of the women in their study in South Carolina had been raped during childhood, a further 15.6% had been molested, and another 12% had been the victims of non-contact sexual assault (such as indecent exposure).
When studied at age 18, 17% of females in a longitudinal study reported experiencing sexual abuse before age 16 (Fergusson, Lynskey, & Horwood, 1996). However, it has been suggested that survivors in this age group are inclined not to report the abuse they have suffered (Ministry of Health, 1998). A study of 3000 women aged 18-65 (Anderson, Martin, & Mullen, 1993) reported 32% being sexually abused before age 16. Using a somewhat narrower definition2 of sexual abuse than some, Muir (1993) found prevalence rates of 38% among women and 10% among men in her sample of university students. However, Romans, Martin, Anderson, Herbison, and Mullen (1995) argued that until recently most studies have involved atypical samples such as social agency clients and clinical inpatients, and criticisms of the link between childhood sexual abuse and later suicidality
have been based on these methodological issues; for example, clinical inpatients may not be representative of the wider population of survivors of sexual abuse. Disclosure of abuse appears to be difficult for many survivors of sexual abuse, particularly disclosure to police or others outside their circle of significant others. This may have implications for the efficacy of counselling and therapy. International studies reviewed by Muir (2001) suggested that 20-30% of survivors of childhood sexual abuse do not disclose until adulthood, and in over half the cases where disclosure occurred during childhood (usually to a parent or parent figure or friend; only 4% disclosed to a community figure, such as police, teachers, or social workers), no action was taken. The majority of participants in Muir’s (1993) study reported feeling scared, humiliated, guilty, and powerless at the time of the abuse. When asked how they felt immediately after disclosure, approximately one-third gave negative responses, such as guilt or shame. In some cases, the response of the person to whom they had disclosed was negative or unsupportive.
For some participants, negative responses to the question seem largely predicated on a lack of response by the person disclosed to, or little change to the situation. Another study conducted by Muir (2001) found that fear of the consequences frequently affected women’s ability to disclose. Anderson and colleagues (1993) reported that only 7% of sexually abused participants reported the abuse to police or social services, and Romans, Martin, and Mullen (1996) suggested that sexual abuse by a family member is much less likely to be reported to police or social services than if the abuser is outside the family. In many cases, fears were not unfounded: disclosures were often met with disbelief or rejection on the part of the confidant.
Similarly, Myer (1985) reported that of 43 mothers who attended a programme for mothers of father-daughter incest victims, only 56% protected their daughters, with 9% taking no action and the remaining 35% rejecting their daughters in favour of their partners (the perpetrators of abuse). Members of the latter two groups either denied the abuse took place, or blamed their daughters, claiming, for instance, that their daughters were seductive, provocative, or pathological liars. Denial of abuse during childhood is often particularly disempowering and engenders a sense of betrayal and may result in the abuse continuing. Withholding disclosure may be a way to retain control over one’s memories and emotions; as Muir (2001) discussed, control (or the lack of it) is frequently an important issue for survivors of abuse. Control may also be maintained through selective disclosure, or choosing confidants that maximise confidence about disclosure.
It seems possible that disclosure may impact mental health and potential suicidality, depending on how the disclosure is dealt with. If abuse is disclosed when it first occurs and is appropriately dealt with, the abuse survivor will have the opportunity to take whatever steps she feels necessary to re-establish her emotional equilibrium. On the other hand, if she is unable to disclose the abuse, or it is not dealt with appropriately, she may be at greater risk of feelings of guilt, anxiety, low self-esteem, and depression and the abuse may continue. These emotional responses could, in turn, lead to suicidality. Holguin and Hansen (2003) suggested that in addition to the impact of the abuse itself, the consequences of being labelled as sexually abused may have detrimental effects. They argued that the combination of negative expectations and biases may create a self-fulfilling prophecy, but acknowledge that further research into this area is required.
Additionally, the mother’s response to a disclosure of sexual abuse is central to her daughter’s recovery (Candib, 1999). Furthermore, perpetrators of sexual abuse are rarely identified and even more rarely punished (Candib, 1999); if a girl’s mother does not support her, she may well receive no support at all. A link between delayed disclosure of abuse or inadequate response to disclosure and subsequent suicidal behaviour does not appear to have been researched. However, given what is known about increased likelihood of suicidal behaviour among survivors of sexual abuse, it seems plausible that the addition of a lack of support to deal with the abuse may exacerbate suicidality. In summary, the literature reviewed suggested that sexual abuse leads to an increased likelihood of depression, anxiety, trauma, and substance abuse, all of which have been associated with suicidal behaviour.
While it would be incorrect to say that the majority of survivors of sexual abuse engage in suicidal behaviour, there is no doubt that the risk is increased, and Romans and colleagues’ findings (Romans et al., 1995, 1997) suggested that the majority of young women who attempt suicide have been sexually abused. Candib’s (1999) and Muir’s (2001) findings that those who do not disclose abuse or whose disclosure does not result in appropriate responses are at increased risk of distress suggests that the likelihood of subsequent suicidal behaviour may be raised in these groups; however, there appears to be little research on this topic. This research sought to explore the perceptions of women who had engaged in suicidal behaviour. While the literature discussed above clearly points to a link between childhood sexual abuse and subsequent suicidal behaviour, the author was concerned to avoid assumptions about linkages between the two.
The purpose of the research was to determine how the women concerned (i.e., the survivors of suicidal behaviour) saw their behaviour, what they considered to be the factors that led to their suicidality, how they made sense of their actions, and how they moved beyond suicidal behaviour. Therefore few restrictions were placed on criteria for inclusion in the study and a research method was chosen that minimises the impact of the preconceived ideas of the researcher. Experience of sexual abuse was not a necessary criterion for inclusion in the study. Rather, this was a study of female suicidal behaviour that revealed a high prevalence of sexual abuse in participants’ histories. METHODS The population of interest in this research was women who engaged in non-fatal suicidal behaviour while under age 25. As noted by Gould and colleagues (1998), clinical samples demonstrate higher rates of co-morbidity than community samples, leading to an increased risk of sample bias.
Due to this potential bias, and because a large number of young female self-injurers do not come to the attention of a mental health professional, a qualitative approach that included a non-clinical population was used. Eligibility for participation in the research included prior engagement in suicidal behaviour, while under age 25, cessation of suicidal behaviour for at least one year, and self-assessed as having recovered from suicidality. The latter two criterion were included both to ensure a degree of safety from distress that might have occured through taking part in the research and to increase the likelihood of participants having had some time to reflect on the cessation of their suicidal behaviour.
However, it is acknowledged that the decision that suicidal behaviour should have ceased at least one year prior to participation is somewhat arbitrary. Personal experience of sexual abuse was not a criterion for participation. Participants were recruited through presentations made to third-year and graduate psychology classes, through items placed in magazines and newspapers, via the e-mail lists of relevant information networks, and through letters and information sheets sent to relevant community organisations such as women’s groups and community support groups. Potential participants were invited to contact the researcher. An initial recruitment discussion took place during which the purpose of the study was discussed, along with eligibility criteria. Participants were asked if they had questions and then offered a written information sheet.
Following this, eligible participants were invited to take some time to consider whether they wished to proceed and to contact the researcher again if they did. Most participants decided immediately that they wished to proceed and made an appointment for an interview. Two possible participants were not heard from subsequent to the recruitment discussion, and one cancelled her appointment. Participants A total of 24 women took part in the research. The participants formed a diverse group; efforts were made to recruit participants from a variety of backgrounds, in an attempt to obtain a sample as representative of the population of interest as possible, given the sample size. The youngest participant was aged 21 at the time of the interview, while the oldest was 46. The average age was 29.6 years.
All participants were born in New Zealand and of European descent, although two also were of Maori ethnicity (the indigenous people of New Zealand) and one was part-Asian. Most (n = 23) of the participants spent their childhood living with at least one biological parent, with the remaining participant had been adopted. However, only 11 participants reported that their biological parents were cohabiting at the time of the participant’s first suicide attempt. In eight cases, the parents separated at some point of the participant’s childhood or adolescence; in three cases one parent had died, and in one case both parents had died. Eleven participants were living in cities at the time of their first suicide attempt, nine in towns, and four in rural areas. Socio-economic status and education levels were mixed, possibly as a result of the recruitment process; thirteen participants had completed some university courses.
Interview and Procedures An open-ended, semi-structured method of interviewing was chosen in recognition that an attempt to fit the participants’ varied experiences into a “one size suits all” structure would risk losing the subtleties of their interpretations. This method facilitates access to information the researcher could not have considered (Burns, 1994). In line with the narrative approach, once the preliminaries to the interview had been conducted (discussion about consent, recording of the interview, making the participant comfortable, discussion of the topic, etc.), the participants were encouraged to tell their “story,” beginning with the background to becoming suicidal. Participants were asked in general terms how or why they became suicidal. They were not prompted by having possible risk factors suggested, such as sexual abuse. During this stage, the researcher’s role was solely one of encouraging the process of story-telling.
The second stage was one of seeking clarification and elaboration as required. Interviews lasted an average of two hours and all except three were conducted face to face, with two others being conducted by telephone and the remaining one a combination of telephone and electronic mail. Face-to-face interviews were conducted at the place of choice of the participant (in one case, at the participant’s workplace; in another, at the offices of a participant’s counsellor; and the remainder evenly split between the researcher’s university office and the participants’ homes). All face-to-face and telephone interviews were audio-taped. When transcribing was complete (within two weeks), a copy of the transcription was given to each participant to check for accuracy. No participants requested changes be made other than adding or correcting some details.
Thematic data analysis was performed utilising the QSR Nud*ist qualitative data analysis software package. A suitable coding structure was developed through this process, with branches for risk factors, other self-harming behaviours, interventions/therapies, and factors in cessation. Results A range of both proximal and distal factors were discussed by participants as contributing to their suicidal behaviour. Although suicidal behaviour was often triggered by an immediately preceding event, it was clear from the participants’ narratives that suicidal behaviour occurred against a background of long-term disturbance and dysfunction.
All participants spoke about combinations of factors, and, with the exclusion of two women who considered that their suicidal behaviour was primarily due to biological causes (Kate and Lucy3), all the participants spoke of issues within their family being key contributors. While the divorce of parents does not seem particularly common (eight out of 24 participants), most of the others spoke of a large amount of parental conflict that did not result in divorce. Additionally, there were five parental deaths, two of which were suicides, and a number of other issues as discussed above. Almost all participants spoke of some level of physical or emotional abuse within the family that was sufficiently severe to be considered a cause of their suicidal behaviour.