The long-term effects of child sexual abuse

CFCA Paper No. 11 – January 2013

The impact of child sexual abuse on mental health

Research has established a strong, albeit complex relationship between child sexual abuse and adverse mental health consequences for many victims (Fergusson & Mullen, 1999; Walsh, Fortier, & DiLillo, 2010). While much of the earlier research in this area used cross-sectional studies with clinical or convenience samples, more recent studies have increasingly used large random community samples, birth and twin cohorts. These more rigorous studies have arguably generated more reliable and generalisable findings, despite the assessment of child sexual abuse still being predominantly retrospective in design (Cutajar et al., 2010a, 2010b).

Noteworthy is a series of twin studies conducted over the last decade, which have consistently revealed a link between child sexual abuse and adverse mental health and related outcomes for survivors. Kendler et al. (2000), in an epidemiological and co-twin controlled analysis of 1,411 twin pairs, reported significant odds ratios for a range of psychiatric disorders in sexually abused women after controlling for family environment. The effects were strongest for drug and alcohol dependence and bulimia nervosa. Dinwiddie et al. (2000), in an Australian twin study with 5,995 twin pairs, also found significant odds ratios for child sexual abuse and major depression, panic disorder, and alcohol dependence. Similarly, Nelson et al. (2002) in another Australian study involving 1,991 twin pairs found that in twins where one had been sexually abused and the other not, the abused twins had significantly higher rates of major depression, attempted suicide, conduct disorder, alcohol dependence, nicotine dependence, social anxiety, rape as an adult, and divorce.

Negative mental health effects that have been consistently associated in the research with child sexual abuse include post-traumatic symptoms (Canton-Cortes & Canton, 2010; O'Leary & Gould, 2009; Ullman,Filipas, Townsend, & Starzynski, 2007); depression (Fergusson et al., 2008; Nelson et al., 2002); substance abuse (Lynskey & Fergusson, 1997; O'Leary & Gould, 2009); helplessness, negative attributions, aggressive behaviours and conduct problems; eating disorders (Jonas et al., 2011); and anxiety (Banyard, Williams, & Siegel, 2001; Nelson et al., 2002). More recently child sexual abuse has also been linked to psychotic disorders including schizophrenia and delusional disorder (Bendall, Jackson, Hulbert, & McGorry 2011; Lataster et al., 2006; Wurr & Partridge, 1996) as well as personality disorders (Cutajar, 2010b). Child sexual abuse involving penetration has, in particular, been identified as a risk factor for developing psychotic and schizophrenic syndromes (Cutajar et al., 2010a).

At the most serious extreme of mental health problems, the findings related to suicide ideation, suicide attempts and actual suicides are of particular concern, especially since the Victorian Parliamentary Inquiry into the Handling of Child Abuse by Religious and Other Organisations was instituted at least partly on the basis that 40 Victorian people allegedly abused by Catholic clergy had committed suicide in recent years.5 A number of studies indicate that sexual victimisation, both in childhood and beyond, is a significant risk factor for suicide attempts and for (accidental) fatal overdoses, among both men and women. This evidence comes from community and clinical samples, as well as epidemiological record-matching studies and several prospective longitudinal studies in various countries. Some earlier studies and reviews (Briere & Zaidi, 1989; Fondacaro & Butler, 1995) reported mixed findings, but other factors - such as co-existing child physical abuse, family dysfunction, depression, and the consequences of disclosing child sexual abuse - were often not considered. Some more recent and rigorous studies, however, have used large-scale data sources or longitudinal or follow-up designs, and reported significant links between child sexual abuse and later suicidal behaviour or ideation (Dube, Anda, & Whitefield 2005; Fergusson et al., 2008; Molnar, Berkman, & Buka, 2001).

In particular, the Christchurch longitudinal study in New Zealand (noted in Box 1) showed that exposure to childhood sexual abuse was related to "clear increases in the risks of later mental health problems" (Fergusson et al., 2008, p. 617). These included suicidality and depression, as well as anxiety disorders, conduct/anti-social personality disorder, and substance use. This association, from age 16 to 25 years, persisted after taking account of other adverse factors in childhood such as physical abuse, problematic parent-child attachment, and parental history of illicit drug use (Fergusson et al., 2000, 2008). There was no significant association between child sexual abuse and the family's socio-economic status. While physical abuse was also related to a range of mental health disorders including suicide attempts, but not suicide ideation, the long-term effects of child sexual abuse were generally larger than the long-term effects of physical abuse. Overall, after adjusting for a range of other factors, children exposed to sexual abuse involving attempted or completed sexual penetration had rates of mental health disorders, including suicidality, that were 2.4 times higher than those of children not so exposed. Estimates of the population attributable risk (PAR) suggested that the elimination of child sexual abuse within the Christchurch cohort would have reduced the overall rates of mental health disorder in adulthood by 13% (Fergusson et al., 2008, p. 617).

A recent Australian study using quite a different methodology focused on completed suicides and fatal drug overdoses. This study did not rely on self-report data but was not able to take account of other contributory factors either early in life or closer to the fatality. In this study, Cutajar et al. (2010b) linked the forensic medical records of over 2,500 victims of child sexual assault in Victoria over a 30-year period with the coronial data for a 44-year-span follow-up. They concluded that "child sexual abuse victims are at increased risk of suicide and accidental fatal drug overdose" but that "it is not possible to reliably attribute the association entirely to the experience of CSA" (p.186), given the non-random nature of the children who come to the attention of child protection services and the police as a result of child sexual abuse allegations.

Table 1 provides a summary of the findings from a number of studies on the impact of child sexual abuse on mental health and psychiatric disorder, including suicide.

Importantly, not all victims of child sexual abuse develop mental health or adjustment difficulties in adulthood. Lynskey and Fergusson (1997), for example, reported that one-quarter of those exposed to child sexual abuse in their cohort study, did not meet the criteria for any psychiatric diagnoses or adjustment difficulties in early adulthood. However, it is important to be alert to sleeper effects with problems possibly emerging at later stages in life or triggered by significant life events.

Walsh et al. (2010) have characterised child sexual abuse as a "non-specific risk factor" (p. 2) for adjustment difficulties, since up to 25% of victims experienced no direct psychological problems in childhood and up to 40% of victims exhibited no clear symptomatology in adulthood. Green et al. (2010) also found that there was little specificity for a range of childhood adversities, including sexual abuse and maladaptive family functioning, being associated with various psychiatric disorders in a large-scale community survey. Further, while there was a cumulative impact, this was not a straight additive effect and it also declined with age.

Although a robust body of research demonstrates the link between child sexual abuse and mental health problems, it is important to note that some studies fail to control adequately for potentially confounding variables. As outlined earlier, these include other childhood adversities, such as other forms of abuse, family functioning and socio-demographic factors. The picture is complex, however, for two reasons. First, there is evidence (see following section) that children who have already been victimised in various ways are more likely to be re-victimised sexually or physically both as adolescents and adults. Second, recent large-scale studies in the US have found evidence of a stress sensitisation effect - that is, being exposed to a range of childhood adversities including sexual abuse exacerbates the impact of stressful life events in adulthood (Kendler et al., 2004; Espejo et al., 2006). McLaughlin et al. (2010) found, for example, that both men and women with such adversities in childhood were more likely to have psychiatric disorders when exposed to stressful life events in adulthood than those without such early adversities. Finkelhor, Ormrod, and Turner (2007) also found a similar effect within childhood, with children revealing elevated risks of trauma symptoms if they had been subjected to several kinds of victimisation within the past year. Maker, Kemmelmeier, and Peterson (2001) highlighted that victims of child sexual abuse are at greater risk of adult sexual assault and that the negative psychological outcomes attributed to child sexual abuse may in fact be more strongly associated with sexual assault in adulthood "as measures of psychological functioning may be more sensitive to the effects of recent sexual trauma than the impact of more distal child abuse" (p. 353). Importantly, more research is needed to examine the extent to which interventions like counselling may improve the outcomes for survivors and mediate some of the potentially negative consequences.

Studies that have specifically examined the long-term mental health outcomes for male survivors of child sexual abuse are limited. Overall, research findings have indicated that women survivors either experience more severe problems following child sexual abuse (Ryan, Kilmer, Cauce, Watanabe, & Hoyt, 2000) compared with men, or that their experiences are largely comparable (Boudewyn & Huser Liem, 1995; Roesler & McKenzie, 1994). However, some research findings suggest that male victims of child sexual abuse may experience different and, in some respects, more adverse mental health outcomes than female victims. For example, J. Hunter (1991) found that male victims were more likely than women to experience anxiety, rumination and worry. Gold et al. (1999) found that relative to their respective normative samples, male survivors drawn from a clinical sample demonstrated greater symptomatology compared with women survivors on measures of interpersonal sensitivity, depression, anxiety and phobic anxiety. The picture, however, may be more complex than the findings using various measures and diagnoses indicate. For example, Hillberg et al. (2011) concluded that while a series of meta-analyses have failed to demonstrate significant gender differences on mental health difficulties, there is empirical evidence of gender differences at least in victims' perceived mental health consequences. This finding is consistent with research that suggests that male survivors of child sexual abuse are more susceptible to internalising effects, while women are more likely to experience externalising effects (Dorahy & Clearwater, 2012; Romano & De Luca, 2001). This contrasts with findings from research in other areas indicating that men are more likely to externalise their problems. The difference may be related to gender norms that make it difficult for men to discuss sexual abuse, and possibly even to a cultural bias that sees women's, but not men's, promiscuity as an "externalising" problem.

A small number of recent studies on clergy-perpetrated sexual abuse also indicates that boys may be particularly susceptible to abuse of this type and to the effects that play out in adulthood. A large-scale study on abuse allegations in the Catholic Church in the US and a smaller study in Australia on allegations against Anglican clergy found that the majority of these allegations involved male victims. In the US study by the John Jay College Research Team (2004), 81% of the victims were male, and 40% of all victims were males aged 11-14 years.6 In the Australian study, 75% of the 180 victims in 191 complaints were male (Parkinson, Oates, & Jayakody, 2010). The average time from the alleged abuse to making a complaint was 25 years for males, and 18 years for females. Neither of these studies was designed to look at the impact of the abuse on the victims, and as Fogler et al. (2008) pointed out, "our knowledge of the effects of CPSA [clergy-perpetrated sexual abuse] is still in its infancy" (p. 349).

There are indications, however, that sexual abuse by clergy and other powerful authority figures may have particularly devastating effects. Brady (2008) drew strong parallels here with the features of abuse within the family that are deemed particularly damaging and difficult for children to deal with. These include the fact that:

the families of many victims were closely allied with the life of their church - a spiritual family; the abuse tended to occur over an extended period of time, similar to many cases of incest; adults frequently did not believe reports of abuse when alerted to it, which often also occurs in cases of incest; church leaders tried to silence victims to avoid scandal, also a repeated theme in incest; and many victims did not disclose the abuse until adulthood, again similar to many cases of incest. (Doyle, 2003, as cited in Brady, 2008, p. 360)

In the same special issue of the Journal of Child Sexual Abuse, which was concerned with the trauma of clergy sexual abuse, Fogler et al. (2008) drew together the literature and provided some theoretical foundations for their conclusion that clergy-perpetrated sexual abuse "can catastrophically alter the trajectory of psychosocial, sexual, and spiritual development" (p. 330). Fogler et al. attributed the damaging impact of sexual abuse by clergy, which commonly occurs around the ages of 11-14 years, to the way in which it undermines the victims' trust, sense of self, sexual identity, and social and cognitive development.

As the body of research on the mental health consequences of child sexual abuse continues to grow, more sophisticated and focused research is needed to tease out possible gender differences as well as the influence of potential mediating factors on the mental health outcomes for victims of child sexual abuse.

Table 1. Studies reporting on mental health disorders and suicidality
Authors Sample/type of study Findings Gender differences
Australia
Cutajar et al. (2010b) Data linkage for cohort of 2,759 victims of child sexual abuse in forensic medical records 1964-1995 with coronial records up to 44 years later. Significantly higher rate of suicide or accidental fatal overdose among child sexual abuse victims than in general population. Female sexual abuse victims had 40 times higher risk of suicide, 88 times higher for fatal overdose; for males, 14 times and 38 times respectively.
Martin, Bergin, Richardson, Roeger, & Allison (2004) Cross-sectional community survey with 2,485 adolescents at 27 SA schools. Strong association between sexual abuse and suicidal ideation and behaviour (plans, threats and attempts), especially for boys:

 

  • 10-fold increased risk for suicidal plans and threats compared with non-abused peers;
  • 15-fold increase for attempted suicide; and
  • 3-fold increase for girls that was mediated by distress, hopelessness and family functioning.
Prevalence of self-reported child sexual abuse (undefined) was 5% for girls and 2% for boys; stronger association between sexual abuse and suicidality among males.
Nelson et al. (2002) Co-twin: Examined 1,991 same-sex pairs of twins (1,159 female and 832 male pairs). The twin reporting child sexual abuse had significantly greater risk for all 8 adverse outcomes (major depression, suicide attempt, conduct disorder, alcohol dependence, nicotine dependence, social anxiety, rape after the age of 18 years, and divorce) than their non-abused twin.

 

Increased risks associated with child sexual abuse involving intercourse.

Prevalence of child sexual assault of 17% for women and 5% for men; significantly increased risk for suicide among both women and men, after taking account of family background.
Plunkett et al. (2001) Prospective 9-year follow-up of 183 male and female sexually abused children. The observed suicide rate in sexually abused children was 10.7-13.0 times that of the Australian national rate. 24% females and 9% of males had attempted suicide by 9-year follow-up.
New Zealand
Fergusson et al. (1996)

 

Fergusson, Beautrais & Horwood (2003)

Fergusson et al. (2008)

Prospective longitudinal cohort study of 1,265 children born in 4-month period in mid 1977, followed regularly to age 25 years in this New Zealand study (Christchurch Health and Development Study). 25 year-olds who experienced attempted or completed sexual penetration as children had rates of mental health disorder (including suicide ideation and attempts, depression and anxiety, substance dependence) that were 2.4 times higher than those not exposed to child sexual abuse; this effect remained significant after taking into account various measures of family functioning and socio-economic status. No gender difference found.
Martin, Anderson, Romans, & Herbison (1993) Random, stratified community sample of 1,376 adult women. Significant associations found between child sexual abuse and higher levels of psychopathology, with higher rates of substance abuse and suicidal behaviour, after controlling for family dysfunction; more severe the abuse, the higher the level of psychopathology. Female sample only.
Scott et al. (2010) Retrospective nationally representative cohort study of 2,144 16-27 year-olds from a mental health survey; 221 were identified as having records on a national child protection agency database. After adjusting for demographic and socio-economic correlates, child protection agency history was associated with several individual mental disorders, mental disorder co-morbidity, and all mental disorder groups, both 12-month and lifetime. Adjusted for sex, as well as age, ethnicity, maternal education, respondent education, and current household income.
USA
Briere & Elliott (2003) Random: Geographically stratified, general population sample of 1,442 adults. Child sexual abuse was associated with a range of trauma symptoms including depression, anxiety, anger, intrusive experiences and sexual concerns after controlling for age, sex, race and income and history of physical abuse. 14% of males and 32% of females reported child sexual abuse.
Brown et al. (1999) Prospective: A cohort of 776 randomly selected children, followed for 17 years. Compared with physical abuse and neglect, child sexual abuse was found to carry the greatest risk for depression and suicide, independent of demographic, parent and child characteristics. Gender and age were taken into account in the analyses but no differences were reported.
Kendler et al. (2000) Twin study in which one twin had been sexually abused, drawn from a sample of 1,411 adult female twins. The twin reporting child sexual abuse was consistently at higher risk for lifetime psychiatric and substance use disorders compared with their non-abused co-twin; as severity of the abuse increased, so did the odds ratios. Female sample only.
Molnar, Berkman et al. (2001) Nationally representative sample of 5,877 Americans aged 15 to 54 years. Among those sexually abused as children, odds of suicide attempts were 2-4 times higher among women and 4-11 times higher among men, compared with those not abused, after controlling for other adversities. Higher odds suicide for males than females.
Trickett, Noll, & Putnam (2011) 84 females (6-16 years old) with Child-Protection-Service-substantiated sexual abuse, including genital contact and/or penetration by a family member and a demographically similar comparison group ( n = 82); children and older caregivers for key participants included. Sexually abused women at follow-up aged 25 more likely to engage in self-mutilation, risky sexual activity, abuse drugs and alcohol, experience more lifetime traumas, PTSD, fail to complete high school, and qualify for at least one DSM diagnosis.

 

Potent "sleeper effects" emerge over longer developmental time spans than previously documented, including increasing obesity and high rates of intimate partner abuse in early adulthood.

Female sample only.
Various - meta-analyses
Neuman, Houskamp, pollock, & Briere (1996) Meta-analysis of 38 studies involving adult women. Significant associations between sexual abuse and a number of measures of psychological adjustment - anxiety, anger, depression, suicidality, self-mutilation, sexual problems, substance abuse, impairment of self-concept, interpersonal problems, obsessions and compulsions, dissociation, post-traumatic stress responses, and somatisation as well as re-victimisation. Females examined only.
Paolucci, Genuis, & Violato (2001) Meta-analysis of 37 studies published between 1981 and 1995 involving 25,367 people. Strong effect sizes before and after taking account of various factors, with average unweighted and weighted d's for each of the outcome variables: for PTSD .50 and .40; for depression .63 and .44; for suicide .64 and .44; for sexual promiscuity .59 and .29; for victim-perpetrator cycle .41 and .16; and for academic performance .24 and .19. Factors taken into account included gender, socioeconomic status, type of abuse, age when abused, relationship to perpetrator, and number of abuse incidents.

Footnotes

5 See Inquiry into the Handling of Child Abuse by Religious and Other Organisations <www.parliament.vic.gov.au/fcdc/inquiry/340>.

6 This study identified 10,667 claims of child sexual abuse by Catholic priests between 1950 and 2002; the claims were made against 4,392 priests, comprising about 4% of all Catholic clerics in the United States.