Introduction
In recent years, there has been an increased awareness of childhood trauma and in particular childhood sexual abuse (CSA) and its adverse effects on individuals’ mental health in adulthood (Tonmyr et al. Reference Tonmyr, Jamieson, Mery and MacMillan2005; Afifi et al. Reference Afifi, Brownridge, Cox and Sareen2006; Zlotnicka et al. Reference Zlotnicka, Johnson, Kohn, Vicente, Rioseco and Saldivia2008). A recent population-based study in Ireland, reported rates of CSA involving physical contact, in 20% of girls and 16% of boys (McGee et al. Reference McGee, Garavan, Byrne, O'Higgins and Conroy2011). In the United Kingdom, when ‘severe maltreatment’ (including sexual, physical and emotional abuse and severe neglect) during childhood was examined by the National Society for the Protection of Cruelty to Children, 25% of young adults admitted to suffering severe maltreatment in childhood, with the authors, suggesting that this figure was probably an under-estimate (Radford et al. Reference Radford, Corral, Bradley, Fisher, Bassett, Howat and Collishaw2011). In adult mental health services, prevalence rates of CSA range from 22–85% and rates of physical abuse range from 17–87% of individuals with these rates somewhat higher in females (Read et al. Reference Read, Goodman, Morrison, Ross and Aderhold2004). To date, there are few prevalence studies evaluating the rates of other common forms of childhood maltreatment (i.e. emotional abuse, physical and emotional neglect) in individuals attending adult mental health services.
Childhood trauma and most frequently CSA or physical abuse in childhood has been associated with increased psychopathology and psychiatric disorder in adulthood. In particular, childhood trauma, including physical abuse, CSA and emotional abuse has been associated with increased rates of depression in adulthood (Goldberg, Reference Goldberg1994; Kinard, Reference Kinard1995; Ferguson & Dacey, Reference Ferguson and Dacey1997; Chapman et al. Reference Chapman, Anda, Felitti, Dube, Edwards and Whitfield2004). Similarly, childhood trauma and in particular CSA have been suggested as potential risk factors for the development of borderline personality disorder (PD; Park et al. 1994; Silk et al. Reference Silk, Lee, Hill and Lohr1995). More recent findings have, however, suggested that multiple types of childhood trauma (e.g. emotional, physical and sexual abuse) may be significant predictors of borderline PD (Huang et al. Reference Huang, Yang, Wu, Napolitano, Xi and Cui2012) with some studies suggesting that physical abuse or neglect, may potentially have a more significant aetiological role for the development of borderline PD than CSA (Widom et al. Reference Widom, Czaja and Paris2009).
Although many studies have investigated the putative association of one form of childhood trauma (often CSA or physical abuse) and mental health, fewer studies to date have evaluated a range of childhood traumas (including CSA, physical and emotional abuse, physical and emotional neglect) and their negative association with adult psychopathology. In relation to those studies that have been undertaken, childhood maltreatment of all forms have been associated with depression, suicide attempts, alcohol misuse and psycho-active substance misuse (Dube et al. Reference Dube, Felitti, Dong, Chapman, Giles and Anda2003; Chapman et al. Reference Chapman, Anda, Felitti, Dube, Edwards and Whitfield2004; Min et al. Reference Min, Farkas, Minnes and Singer2007; Wu et al. Reference Wu, Schairer, Dellor and Grella2010; Min et al. Reference Min, Minnes, Kim and Singer2013). Furthermore, increasing numbers of traumatic experiences (including parental separation, mental illness in the household and criminal household members in addition to childhood abuse or neglect) have been associated with poorer levels of psycho-social functioning (Wu et al. Reference Wu, Schairer, Dellor and Grella2010) and greater levels of psychological distress and recurrent depressive disorder (RDD; Chapman et al. Reference Chapman, Anda, Felitti, Dube, Edwards and Whitfield2004).
In Ireland, few studies to date have examined the association between childhood trauma and the prevalence of adult mental health difficulties. However, both physical abuse (Kelleher et al. Reference Kelleher, Harley, Lynch, Arseneault, Fitzpatrick and Cannon2008; Kelleher et al. Reference Kelleher, Keeley, Corcoran, Ramsey, Wasserman, Carli, Sarchiapone, Hoven, Wasserman and Canon2013) and CSA (Murphy et al. Reference Murphy, Houston, Shevlin and Adamson2013) have both been associated with higher rates of psychotic experiences in adolescents.
In this study, we wanted to assess the prevalence of childhood trauma including sexual, emotional and physical abuse and emotional and physical neglect among individuals attending a general adult mental health service. We also endeavoured to explore the presence of childhood trauma across a range of mental health disorders (both Axis I and Axis II disorders) in individuals attending inpatient and outpatient services and ascertain if there were any associations between childhood trauma and a range of clinical and socio-demographic factors.
Method
Participants
Individuals were recruited from both the inpatient and outpatient services of Roscommon Mental Health Services, Ireland. All consecutively admitted individuals to the Department of Psychiatry, Roscommon County Hospital between September 2010 and March 2011 (n=98) were invited to participate in this study. All consecutively attending individuals from two outpatient clinics (both scheduled weekly) attached to the Roscommon Mental Health Services between March 2011 and June 2011 (n=115) were also invited to participate (total n=213). Sixty-eight individuals in this latter group had previously been treated on one or more occasions as an inpatient in an adult mental health inpatient unit.
Demographic and clinical data were obtained from clinical interview with participants and clinical note review. All individuals were diagnosed according to the Diagnostic and Statistics Manual IV Revised (DSM-IV-TR) diagnostic criteria by their treating consultant psychiatrist. The presence of a PD(s) was evaluated by the Structured Clinical Interview for DSM-IV Axis II (SCID-II) PDs and was undertaken by trained clinicians (psychiatrists with considerable experience in psychiatry T.O., Z.N., B.H.).
Exclusion criteria included individuals <18 years of age, the presence of an intellectual disability (intelligence quotient <70), a diagnosis of dementia and the presence of acute psychosis. Individuals admitted to hospital with an acute psychotic episode or mania, were interviewed for the study before the end of their admission where possible, when their psychotic or manic episode had significantly resolved and thus if capable and agreeable to consent, they were included in the study.
Each participant provided written informed consent and the study was approved by the Roscommon County Hospital ethics committee. On completion of the assessment, psychotherapeutic support was offered if necessary.
Instruments
Childhood Trauma Questionnaire (CTQ)
The CTQ is a 28-item self-report retrospective inventory that measures childhood or adolescent trauma (Bernstein & Fink, Reference Bernstein and Fink1998). The CTQ has five subscales: emotional, physical and sexual abuse and emotional and physical neglect. A minimisation or denial scale is included. Each item is measured on a 5-point Likert scale ranging from ‘never true’ to ‘very often true’, and each subscale consists of five items. The CTQ items reflect common definitions of child abuse and neglect as found in the childhood trauma literature (Crouch & Milner, Reference Crouch and Milner1993; Malinosky-Rummell & Hansen, Reference Malinosky-Rummel and Hansen1993; Finkelhor, Reference Finkelhor1994; Knutson, Reference Knutson1995). The CTQ has demonstrated good internal consistency (r=0.63–0.95) and criterion-related validity (r=0.50–0.75).
Symptom Checklist-90-Revised (SCL-90-R)
The SCL-90-R is a 90-item self-report instrument that evaluates nine primary symptom dimensions [somatization (SOM), obsessive-compulsive (OC), interpersonal sensitivity (IS), depression (DEP), anxiety (ANX), hostility (HOS), phobic anxiety (PHOB), paranoid ideation (PAR), psychoticism (PSY)] and a global health score [Global Severity Index (GSI)], and is designed to provide an overview of a patient’s symptoms and their intensity at a specific point in time (Derogatis, Reference Derogatis1994). Each item consists of a 5-point rating scale ranging from ‘not at All’ to ‘extremely’.
Barratt Impulsiveness Scale (BIS-11)
The BIS-11 is a 30-item self-report questionnaire that evaluates three sub-scales of impulsiveness: attention, motor and non-planning impulsiveness (Patton et al. Reference Patton, Stanford and Barratt1995). Each item is measured on a 4-point Likert scale ranging from ‘rarely/never’ to ‘almost always/always’.
Sheehan Disability Scale (SDS)
The SDS is a brief self-report tool that evaluates on a 10-point anchored visual analogue scale, functional impairment on three inter-related domains; work/school, social and family life (Rush, Reference Rush2000).
SCID-II PD
The SCID-II is a semi-structured diagnostic interview for assessing the 10 DSM-IV Axis II PDs, as well as depressive personality disorder and passive-aggressive PD (First et al. Reference First, Gibbon, Spitzer, Williams and Benjamin1997).
Data analysis
Statistical analysis was performed using the Statistical Package for Social Sciences 19.0 for Windows (SPSS Inc., USA). We utilised the student t-test or analysis of co-variance for parametric data and the χ 2-test for non-parametric categorical data where appropriate. We used the Pearson’s Product Moment Correlation (r) to determine the correlation between the various test variables (e.g. CTQ and SCL-90-R total and subscale scores). We utilised a p-value of <0.01 for determining significance for correlation analysis due to the number of variables being tested. Logistic regression was undertaken with the experience of childhood trauma or a form of childhood trauma as the dependent variable, and gender, age, diagnosis (Axis I disorder – schizophrenia, bipolar disorder, depression, anxiety disorder, alcohol or substance misuse) or borderline PD or antisocial PD), total BIS score, total SDS score and GSI from the SCL-90-R as the dependent variables.
Results
Of the 213 individuals invited to participate, 27 were excluded. Reasons for exclusion included intellectual disability (n=7), dementia (n=4), acute psychosis (n=14), and individuals been unattainable due to very brief inpatient admission stays of <24 hours (n=2). Of the remaining 186 people, 136 individuals (73%) agreed to participate in this study, including 60 inpatients and 76 outpatients. Fifty-six individuals in the outpatient group (74%) had previously had one or more admissions to an acute adult inpatient mental health unit. It was therefore decided to combine the inpatient and outpatient group for most comparisons. Demographic and clinical data were as follows: the mean age of participants was 45 years (s.d.=15), with 70 individuals (51.5%) being male. Forty-nine per cent of the sample was single followed by married (36%), and most individuals (75%) were unemployed. RDD (47%) was the most common Axis I disorder, followed by schizophrenia (21%) and bipolar affective disorder (14%). The most prevalent Axis II disorders was avoidant PD (34%), followed by passive aggressive PD (24%), paranoid PD (24%) and borderline PD (23%). The most common psychotropic medications prescribed were antipsychotics (76%), followed by antidepressants (74%).
One hundred and three individuals (76%) reported some form of childhood trauma, with no statistical difference in the frequency of trauma between men and women (Table 1). Emotional neglect was the most frequent type of childhood trauma reported (61%), followed by emotional abuse (47%), physical neglect (47%), physical abuse (28%) and sexual abuse (24%), with no significant difference in frequency in the prevalence of any of these childhood traumas between men and women (Table 1). For men who reported physical abuse and for women who reported sexual abuse (physical abuse=52%; sexual abuse=57%) over 50% of these individuals reported that the abuse suffered was in the ‘severe to extreme’ range. All individuals, who reported sexual abuse, also suffered at least one additional form of childhood trauma with 46% of individuals reporting that they suffered all four other forms of childhood trauma. The number of forms of childhood trauma reported by study participants was: none=30 (22.1%); 1=28 (20.6%); 2=27 (19.9%); 3=15 (11.0%); 4=18 (13.2%); 5=15 (11.0%).
Table 1 Childhood trauma severity in both genders
No difference in age, marital status, employment status, education level achieved and family structure was ascertained between individuals who had suffered childhood trauma compared with those who did not (Table 2), however, individuals who suffered childhood trauma had more inpatient admissions to adult mental health units (t=5.36, p<0.001). Individuals who suffered childhood trauma, had no difference in the rate or type of Axis I disorders, but had higher rates of several PDs including passive aggressive, paranoid, borderline and antisocial PD (p<0.01) when compared with those who did not report trauma. The prevalence of the five childhood maltreatment categories for both Axis I and Axis II disorders at different degrees of severity are further presented in Tables 3 and 4. Axis I psychiatric disorders were divided into five groups: schizophrenia, bipolar disorder, RDD, anxiety disorders and alcohol/substance abuse or dependence. The highest rate of childhood trauma was reported in individuals with alcohol and/or substance abuse or dependence (92%), however, this was not significantly different to individuals with other Axis I disorders (χ 2=5.531, df=4, p=0.237). Individuals with a primary diagnosis of anxiety disorders had lower levels of childhood trauma (50%), and none reported sexual abuse, however, only eight individuals were included in this group (Table 3). Of the 85 individuals diagnosed with PD, 73 (87%) reported childhood trauma, with particularly high rates in individuals diagnosed with antisocial PD (100%), passive aggressive PD (97%), paranoid PD (97%) and borderline PD (94%). The highest rate of CSA was ascertained in individuals with antisocial PD (62%), and significantly higher rates of CSA were demonstrated for individuals with antisocial PD compared with those without antisocial PD. No other PD was significantly associated with CSA. Several PDs were associated with physical abuse, physical neglect, emotional abuse and emotional neglect (Table 4). Many individuals who fulfilled diagnostic criteria for one PD also fulfilled diagnostic criteria for one or more other PDs (60.3%), making interpretations regarding individual PDs more difficult.
Table 2 Comparing clinical factors in individuals who did and did not suffer childhood trauma
BIS, Barratt Impulsiveness Scale; SDS, Sheehan Disability Scale; SCL-90-R, The Symptom Checklist-90-R; SOM, somatization; OC, obsessive-compulsive; IS, interpersonal sensitivity; DEP, depression; ANX, anxiety; HOS, hostility; PHOB, phobic anxiety; PAR, paranoid ideation; PSY, psychoticism; GSI, Global Severity Index.
Data are number (percentage) or mean (±s.d.), when appropriate. Abuse suffered indicates patients with childhood trauma history (includes sexual, physical, emotional abuse and emotional, physical neglect).
*p-values by t-tests for independent samples or χ 2. (In the analysis, both tests, t-tests for independent samples and U-Mann–Whitney, were conducted and they revealed equal results. Statistically stronger tests were selected).
Table 3 Childhood trauma in individuals with Axis I disorders
One individual, with hypochondriacal disorder, one with organic induced personality change and one with anorexia nervosa are not included above.
Table 4 Childhood trauma in individuals with Axis II disorders
PD, personality disorder.
We excluded individuals with schizotypal PD (n=6), schizoid PD (n=4) and histrionic PD (n=2) from all statistical analysis due to the low numbers.
Associations between the types of trauma and each PD are presented with values of *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001 (χ 2 analysis was utilised). Due to low numbers, subcategories of each type of childhood trauma and their association with each personality disorder were not compared.
Insufficient numbers were present to investigate if sub-categories of childhood trauma were associated with the presence or absence of Axis I and II disorders. Sufficient data was available (n>15) to examine individuals with dual diagnoses for individuals diagnosed with RDD and co-morbid avoidant, obsessive-compulsive, passive-aggressive and paranoid, PD. Individuals with RDD who also had co-morbid obsessive-compulsive PD, had higher rates of CSA (75%; χ 2=4.233; p=0.040), physical abuse (83.3%; χ 2=5.440; p=0.02) and emotional neglect (68.4%; χ 2=3.859; p=0.049) compared with those with RDD alone.
Individuals who were inpatients at the time of data collection reported higher rates compared with outpatients, of CSA (60.6% v. 39.4%; χ 2=4.693; p=0.03), emotional abuse (57.1% v. 42.9%; χ 2=6.403; p=0.011), physical abuse (57.9% v. 42.1%; χ 2=3.948; p=0.047) and physical neglect (54.7% v. 45.3%; χ 2=5.330; p=0.021). When we compared all individuals who were previously admitted to an inpatient adult mental health unit (n=114) to individuals only ever treated as outpatients (n=22), the only significant finding remaining was a higher rates of childhood emotional neglect in the inpatient group (89.2% v. 10.8%; χ 2=4.062; p=0.044).
Significant higher scores were demonstrated on the BIS-11 total (t=1.605, p=0.022) and motor impulsivity scales (t=1.519, p=0.044), and on the SDS total (t=1.981, p=0.056) and SDS family/home responsibilities scales (t=2.215, p=0.029) for individuals who suffered childhood trauma compared with those who did not (Table 2). Both the Global Symptom Index (GSI) and nine symptom domains of the SCL-90-R (SOM, OC, IS, DEP, ANX, HOS, PHOB, PAR, PSY), revealed higher scores in individuals who suffered childhood trauma (p<0.01). These findings were replicated for all five forms of childhood trauma (p<0.01) apart from CSA in the SOM, DEP and ANX domains, which failed to reach statistical significance.
Logistic regression demonstrated that total BIS scores were predictive of individuals having suffered physical abuse in childhood (B=0.085, p=0.008, OR=1.09, 95% CI 1.03–1.16). SCL-90-R GSI scores were predictive of individuals having suffered physical neglect in childhood (B=1.03, p=0.002, OR=2.81, 95% CI 1.48–5.3), with no evidence for age, gender, type of Axis I disorder and total SDS score predicting childhood trauma. Evaluating Axis II disorders using this model (substituting Axis I D/O with borderline PD or antisocial PD in the model); we found borderline PD was predictive of having suffered any form of childhood trauma (B=1.14, p=0.045, OR=3.13, 95% CI 1.03–9.55); having suffered emotional abuse (B=1.57, p=0.006, OR 4.79, 95% CI 1.58, 14.50), physical abuse (B=1.58, p=0.004, OR=4.87, 95% CI 1.65–14.39) and emotional neglect (B=1.41, p=0.025, OR 4.08, 95% CI 1.20–13.90). Antisocial PD was predictive of having suffered CSA (B=1.614, p=0.021, OR=5.02, 95% CI 1.28–19.71), physical abuse (B=2.11, p=0.015, OR=8.26, 95% CI 1.51–45.07) and emotional abuse (B=2.19, p=0.047, OR=8.92, 95% CI 1.03–76.87) in childhood.
The number of reported types of childhood traumas was significantly associated with both perceived level of disability (F=4.28, df=5, p=0.001) and GSI (F=9.40, df=5, p<0.0001), when we classified individuals by the number of trauma categories (0–5) they experienced and examined if a ‘dose–response relationship’ existed with their present symptomatology as measured by the SDS perceived level of disability and SCL-90-R global severity index (GSI).
The correlation between experiencing one form of childhood trauma and another were medium to large in magnitude and are detailed in Table 5. The highest correlation notes was found between emotional abuse and physical abuse (r=0.65, p<0.0001) and the lowest correlation were found between emotional neglect and CSA (r=0.382, p<0.0001). Correlations between the SCL-90-R, BIS-11 (not the non-planning subscale) and SDS were also medium to large in magnitude, with the highest correlation noted between the SCL-90-R GSI scale and the SDS total score scale (r=0.579, p<0.001).
Table 5 Correlation between childhood trauma and psychometric measures
CTQ, Childhood Trauma Questionnaire; SDS, Sheehan Disability Scale; BIS, Barratt Impulsiveness Scale; SCL-90-R, The Symptom Checklist-90-R; SOM, somatization; OC, obsessivecompulsive; IS, interpersonal sensitivity; DEP, depression; ANX, anxiety; HOS, hostility; PHOB, phobic anxiety; PAR, paranoid ideation; PSY, psychoticism; GSI, Global Severity Index.
Correlation (r) is significant at the level: *p<0.01; **p<0.001; ***p<0.0001.
Discussion
This study demonstrated high levels of childhood trauma in both male and female (approximately three-quarters of the sample) attendees of adult mental health services (inpatient or outpatient), with emotional neglect and abuse the most common types of traumas experienced. There was a ‘dose–response’ relationship between the number of types of childhood trauma experienced and the severity of subjective psychopathology. All forms of childhood trauma were associated with higher levels of symptom severity, independent of symptom dimension or psychiatric diagnosis (except for a lower incidence of childhood trauma in individuals with anxiety symptoms), in individuals attending adult mental health services.
In this study, over half the cohort suffered emotional neglect, half suffered emotional abuse or physical neglect and a quarter suffered CSA or physical abuse in childhood. These rates of CSA and physical abuse are consistent with previous reports (Chu & Dill, Reference Chu and Dill1990); however, we report somewhat higher rates of physical neglect and emotional abuse compared with other studies of similar populations (Gould et al. Reference Gould, Stevens, Ward, Carlin, Sowell and Gustafson1994; Simon et al. Reference Simon, Herlands, Marks, Mancini, Letamendi, Li, Pollack, Van Ameringen and Stein2009). More than half of the participants in this study experienced more than one type of childhood trauma, with significant correlations demonstrated between experiencing different forms of childhood trauma. This is consistent with other studies demonstrating that individuals who experience one form of childhood trauma are very likely to suffer other forms of childhood trauma (Dong et al. Reference Dong, Anda, Felitti, Dube, Williamson, Thompson, Loo and Giles2004; Finkelhor et al. Reference Finkelhor, Ormrod and Turner2007). The prevalence rates of individuals suffering multiple forms of childhood trauma is somewhat higher than some other studies (Swett & Halpert, Reference Swett and Halpert1993; Edwards et al. Reference Edwards, Holden, Felitti and Anda2003), which may be due to the use of a very sensitive questionnaire (CTQ). Consequently, examination of particular types of abuse in isolation may not be the optimum method of examining the association between childhood trauma and mental health problems in adulthood as our research indicates that there may be a complex interplay of childhood traumas in many individuals.
In addition, we demonstrated a ‘dose–response’ relationship between the number of forms of traumatic experiences suffered and levels of disability, symptom severity and number of inpatient psychiatric admissions. These findings are consistent with previous research showing increasing exposure to different childhood traumas associated with greater adverse health outcomes including greater levels of depression, alcohol dependence, psycho-active substance misuse, medical problems, aggression and lower levels of quality of life (Chapman et al. Reference Chapman, Anda, Felitti, Dube, Edwards and Whitfield2004; Anda et al. Reference Anda, Felitti, Bremner, Walker, Whitfield, Perry, Dube and Giles2006; Wu et al. Reference Wu, Schairer, Dellor and Grella2010).
Contrary to other studies, where CSA and physical abuse has been demonstrated to be more prevalent in women (Springer et al. Reference Springer, Sheridan, Kuo and Carnes2003), men in our study reported slightly higher levels (albeit not statistically significant) of CSA, physical abuse and physical and emotional neglect compared with females. A possible tentative explanation for these high rates of childhood trauma and CSA might be the history of abuse by clergy and institutions that occurred in Ireland up to ∼1990. Additionally, it has been postulated that male childhood abuse survivors previously did not report childhood trauma and that as a coping strategy; they denied the impact of CSA on their lives (Holmes et al. Reference Holmes, Offen and Waller1997). However, with enhanced awareness of childhood abuse and greater counselling services now available in Ireland, more men have potentially been able to disclose their experiences.
Previous studies have reported that childhood trauma, and in particular CSA, have been associated with an increased prevalence of a variety of symptoms of mental illness and mental health disorders (Saunders et al. Reference Saunders, Villeponteaux, Lipovsky, Kilpatrick and Veronen1992), rather than any specific symptom or Axis I disorder (Kendall-Tackett et al. Reference Kendall-Tackett, Williams and Finkelhor1993). Similarly, in this study, except for anxiety disorders (where we had low numbers to investigate accurately), high rates of childhood traumas were reported by individuals who were diagnosed with a range of Axis I disorders. Individuals who experienced a history of childhood maltreatment reported significantly more frequent lifetime inpatient psychiatric admissions. This finding is consistent with previous literature (Finestone et al. Reference Finestone, Stenn, Davies, Stalker, Fry and Koumanis2000; Chartier et al. Reference Chartier, Walker and Naimark2007) and suggests that individuals who have experienced childhood maltreatment have higher health care utilisation. Recent research has also linked childhood trauma and the development of psychosis (Van Winkel et al. Reference Van Winkel, Van Nierop, Myin-Germeys and Van Os2013), and although three times as many patients with schizophrenia reported childhood trauma compared with those who did not, the rate of childhood trauma was similar across Axis I disorders.
In keeping with previous findings (Grover et al. Reference Grover, Carpenter, Price, Gagne, Mello, Mello and Tyrka2007), childhood trauma was a very frequent finding (87%) in individuals who fulfilled diagnostic criteria for a PD, with particularly high rates noted in individuals with passive-aggressive, paranoid, borderline and antisocial PDs. Our findings of childhood trauma in individuals with PDs are somewhat higher than those of other studies (Zimmerman et al. Reference Zimmerman, Rothschild and Chelminski2005), however, our increased rates may be in some part be due to our patient cohort including both inpatients and outpatients unlike other cohorts, which were predominantly outpatient in nature (Zimmerman et al. Reference Zimmerman, Rothschild and Chelminski2005).
CSA has previously been suggested as a significant factor in the aetiology of borderline PD (Byrne et al. Reference Byrne, Velamoor, Cernovsky, Cortese and Losztyn1990; Paris, et al. Reference Paris, Zweig-Frank and Guzder1994; Silk et al. Reference Silk, Lee, Hill and Lohr1995; Khalily & Hallahan, Reference Khalily and Hallahan2010). Thirty-five per cent of individuals with borderline PD in this study previously experienced CSA, however, due to the cross-sectional nature of this study, it is not possible to infer if CSA was either causative or a contributory factor for the development of borderline PD. Indeed, other forms of childhood trauma were more prevalent in this group than CSA. It has previously been suggested that multiple types of childhood trauma (e.g. emotional, physical and sexual abuse) are significant predictors of borderline PD (Huang et al. Reference Huang, Yang, Wu, Napolitano, Xi and Cui2012) with reports that experiencing physical abuse and neglect, but not CSA are risk factors for the development of borderline PD (Widom et al. Reference Widom, Czaja and Paris2009). Taken together, these findings suggest that the aetiology of borderline PD is very complex and involves, in addition to possibly genetic and biological factors, several psychosocial factors including perhaps multiple childhood trauma experiences and not just CSA alone (Joyce et al. Reference Joyce, McKenzie, Luty, Mulder, Carter, Sullivan and Cloninger2003; Huang et al. Reference Huang, Yang, Wu, Napolitano, Xi and Cui2012). There are a number of therapeutic strategies suggested that may potentially be beneficial in individuals with borderline PD who have a history of childhood trauma including schema focused therapy (Khalily & Hallahan, Reference Khalily and Hallahan2010). Having an awareness of the various forms of childhood trauma suffered by individuals, may increase the therapist’s ability to engage in such therapy in this cohort of individuals.
The only PD that was significantly more associated (compared with individuals with other PDs) with the experience of CSA in this study was antisocial PD. However, as with borderline PD, high rates of all childhood traumas in all categories were present, with all 13 individuals experiencing at least one form of childhood trauma. Our results finding an association between antisocial PD and CSA are consistent with some previous research (Luntz & Widom, Reference Luntz and Widom1994; Bergen et al. Reference Bergen, Martin, Richardson, Allison and Roeger2004). However, other studies suggest that other forms of childhood trauma such as physical abuse but not CSA are aetiologically associated with antisocial PD (Bernstein et al. Reference Bernstein, Stein and Handelsman1998). As with borderline PD, the aetiology appears to be complex, although it is possible that CSA has a specific association in addition to other psychosocial factors in the aetiology of antisocial PD. As there were only 13 individuals fulfilling criteria for antisocial PD, caution is required in these interpretations. As in other studies (Fossati et al. Reference Fossati, Maffei, Bagnato, Battaglia, Donati and Donini2000), due to the high rates of co-morbidities between Axis II diagnoses, separating out associations between specific childhood traumas and specific PDs is difficult. However, this study indicates a strong association between particular types of PDs and childhood trauma.
Our findings confirm previous reports of an association between childhood maltreatment and impulsivity (Brodsky et al. Reference Brodsky, Oquendo, Ellis, Haas, Malone and Mann2001; Roy, Reference Roy2005). However, due to the cross-sectional design of this study, it is not possible to infer any causality between childhood maltreatment and impulsivity. It is possible, for example, that being impulsive could be a consequence of experiencing childhood trauma or alternatively being impulsive may make certain individuals more susceptible to experiencing trauma.
An important component of this study was the examination of multiple forms of childhood trauma and the assessment of varying degrees of severity of these traumas rather than examining only the presence or absence of CSA or physical abuse. Individuals disclose more experiences, when they have more opportunities to report, and when given multiple cues about the type of experiences researchers are investigating, compared with one single question (Finkelhor, Reference Finkelhor1986). Consequently, we believe that the risk of under-reporting all types of childhood trauma, despite the sensitive nature of this research topic is reduced in this study. No one requested psychotherapeutic support after completion of the study. It is, however, possible that some who refused to participate (27% of the entire sample), decided to do so due to the sensitive nature of this topic. There is also a risk of recall bias regarding the retrospective reports of childhood trauma; however, where psychometric instruments are of high quality the risk of recall bias is significantly reduced (Bernstein et al. Reference Bernstein, Fink, Handelsman, Foote, Lovejoy, Wenzel, Sapareto and Ruggiero1994; Hardt & Rutter, Reference Hardt and Rutter2004). Furthermore, recent evidence suggests that memories of specific childhood experiences are predominantly accurate (Lindsay et al. Reference Lindsay, Pitzer and Fingerman2010). Consequently, we believe that the utilisation of a sensitive instrument such as the CTQ enables researchers or clinicans to attain more accurate findings in relation to individuals suffering childhood trauma.
Limitations
This study has a number of limitations. First, the sample size for individuals with anxiety disorders, alcohol or substance misuse/dependence and some Axis II disorders is modest, and a larger cohort of individuals would be required to examine accurately the association of childhood trauma and these disorders. Second, our data in relation to the 50 individuals who declined to participate is limited, and consequently they may have differed on certain clinical (presence of a PD) or demographic data compared with those individuals who participated in the study. Third, sociodemographic variables were not significantly associated with the occurrence of childhood trauma or the presence of Axis I or II disorders, however, an in-depth evaluation of other important sociodemographic factors including childhood adversities (Reigstad et al. Reference Reigstad, Jorgensen and Wichstrom2006), as childhood trauma should not be considered as an isolated experience (Horwitz et al. Reference Horwitz, Widom, McLaughlin and White2001). Fourth, as this is a cross-sectional study, no causality can be inferred between suffering childhood maltreatment and developing an Axis I or II disorder. Most individuals diagnosed via the SCID-II interview, did not have a prior diagnosis of a PD in their clinical notes, particularly those fulfilling diagnostic criteria for a passive-aggressive or avoidant PD, consequently despite fulfilling criteria for these conditions, difficulties relating to the presence of these PDs may not have been significant. Finally, post-traumatic stress disorder appears under-reported compared with previous studies (Fitzpatrick et al. Reference Fitzpatrick, Carr, Dooley, Flanagan-Howard, Flanagan, Tierney, White, Daly, Shevlin and Egan2010) and this may relate to it either not being the primary diagnosis, not being diagnosed as no formal psychometric instrument (i.e. SCID-1) was utilised or because only the current Axis I diagnosis was requested off their treating consultant psychiatrist. Thus, individuals may previously have fulfilled diagnostic criteria for PTSD, but no longer satisfied these criteria at study entry.
Conclusion
Childhood trauma was extremely prevalent among both men and women attending an Irish adult mental health service, who suffered from a wide range of Axis I and II disorders. Childhood trauma was associated with significant psychopathology and distress with a ‘dose–response’ relationship noted between the number of forms of childhood trauma experienced and the level of distress and symptom severity found. Thus, examining the prevalence of particular types of abuse in isolation may not be the optimum method of evaluating the association between childhood trauma and mental health problems in adulthood due to the complex interplay of childhood traumas in many individuals. Consequently, utilising a sensitive instrument such as the CTQ can inform the clinician of the various forms of trauma and degree of severity of such trauma experienced by their patient.
Acknowledgement
The authors would like to thank the administration staff at the Department of Psychiatry, Roscommon County Hospital for the clerical support provided during this study.
