Childhood adversity and personality disorders: Results from a nationally representative population-based study

Journal of Psychiatric Research 45 (2011) 814e822

Contents lists available at ScienceDirect

Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires

Childhood adversity and personality disorders: Results from a nationally representative population-based study Tracie O. Afifi a, b, *, Amber Mather b, Jonathon Boman b, William Fleisher b, Murray W. Enns b, a, Harriet MacMillan c, d, Jitender Sareen b, a, e a

Department of Community Health Sciences, University of Manitoba, S113 Medical Services Building, 750 Bannatyne Avenue, Winnipeg, Manitoba R3E 0W3, Canada Department of Psychiatry, University of Manitoba, Winnipeg, Manitoba, Canada Department of Psychiatry and Behavioural Neurosciences, McMaster University, Canada d Department of Pediatrics, McMaster University, Canada e Department of Psychology, University of Manitoba, Winnipeg, Manitoba, Canada b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 July 2010 Received in revised form 8 November 2010 Accepted 12 November 2010

Background: Although, a large population-based literature exists on the relationship between childhood adversity and Axis I mental disorders, research on the link between childhood adversity and Axis II personality disorders (PDs) relies mainly on clinical samples. The purpose of the current study was to examine the relationship between a range of childhood adversities and PDs in a nationally representative sample while adjusting for Axis I mental disorders. Methods: Data were from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC; n ¼ 34,653; data collection 2004e2005); a nationally representative sample of the United States population aged 20 years and older. Results: The results indicated that many types of childhood adversity were highly prevalent among individuals with PDs in the general population and childhood adversity was most consistently associated with schizotypal, antisocial, borderline, and narcissistic PDs. The most robust childhood adversity findings were for child abuse and neglect with cluster A and cluster B PDs after adjusting for all other types of childhood adversity, mood disorders, anxiety disorders, substance use disorders, other PD clusters, and sociodemographic variables (Odd Ratios ranging from 1.22 to 1.63). In these models, mood disorders, anxiety disorders, and substance use disorders also remained significantly associated with PD clusters (Odds Ratios ranging from 1.26 to 2.38). Conclusions: Further research is necessary to understand whether such exposure has a causal role in the association with PDs. In addition to preventing child maltreatment, it is important to determine ways to prevent impairment among those exposed to adversity, as this may reduce the development of PDs. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Child maltreatment Child abuse Neglect Exposure to intimate partner violence Personality disorders Psychiatric disorders

Exposure to childhood adversity is known to be associated with mental health impairment that can persist into adulthood. There are strong associations between adverse childhood experiences such as abuse, neglect, exposure to intimate partner violence, and parental divorce and suicidal behavior and adult Axis I mental disorders such as mood, anxiety, impulse control, and substance use disorders in representative population-based samples (Afifi et al., 2006, 2008, 2009, 2010; Bruffaerts et al., 2010; Enns et al., 2006; Kessler et al., 1997; MacMillan et al., 2001; Scott et al., 2010). Studies involving nationally representative samples have * Corresponding author. Department of Community Health Sciences, University of Manitoba, S113 Medical Services Building, 750 Bannatyne Avenue, Winnipeg, Manitoba R3E 0W3, Canada. Tel.: þ1 (204) 272 3138; fax: þ1 (204) 789 3905. E-mail address: t_afifi@umanitoba.ca (T.O. Afifi). 0022-3956/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jpsychires.2010.11.008

shown a relationship between traumatic events occurring in childhood and personality traits such as high neuroticism and openness to experiences (Allen and Lauterbach, 2007). Although, the relationship between childhood adversity and Axis I mental health conditions is well established, research on the link between childhood adversity and Axis II personality disorders (PDs) has focused mainly on clinical samples (Battle et al., 2004; Johnson et al., 2004; Rettew et al., 2003; Yen et al., 2002; Luntz and Widom, 1994; Zanarini et al., 1989, 1997, 2000, 2002; Bierer et al., 2003). PDs are generally persistent overtime, are often represented by patterns of behaviors and experiences that can negatively impact areas of cognition, affect, interpersonal functioning, and impulse control, and are frequently associated with impairment (American Psychiatric Association, 1994, 2000). Clinical studies

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

have indicated that the childhood experience of physical abuse, sexual abuse, emotional abuse, physical neglect, and emotional neglect are common among patients with PDs (Battle et al., 2004; Johnson et al., 2004; Rettew et al., 2003; Yen et al., 2002; Luntz and Widom, 1994; Zanarini et al., 1989, 1997, 2000, 2002; Bierer et al., 2003). Studies involving convenience (Gibb et al., 2001; Grover et al., 2007; Tyrka et al., 2009) and small community samples (Johnson et al., 1999, 2000, 2006) have also supported this association, but it remains unclear whether the relationship between childhood adversities and all PDs exists in representative general population samples. Another important limitation is the narrow examination of child adversity. To date, collectively, studies have looked at parenting behaviors and multiple types of child abuse and neglect. However, some studies have only examined child abuse (Yen et al., 2002; Gibb et al., 2001), neglect (Johnson et al., 2000), or a have combined child abuse and neglect together (Grover et al., 2007; Johnson et al., 1999; Luntz and Widom, 1994; Tyrka et al., 2009; Zanarini et al., 2000). Collapsing multiple types of child maltreatment is often necessary due to lack of statistical power based on small sample sizes. However, this approach precludes understanding the specific relationship between subtypes of maltreatment and impairment, such as PDs. The limited research involving community samples has all been based on a study of two New York State counties. The investigators combined multiple types of child maltreatment into child abuse and neglect categories (Johnson et al., 1999, 2000), only examined neglect (Johnson et al., 2000), and investigated parenting behaviors not including child abuse or neglect (Johnson et al., 2006). Another limitation of the current literature is the focus on only one or limited types of PDs. For example, there are numerous clinical studies showing a link between exposure to child sexual abuse and borderline personality disorder (Murray, 1993). Although this is an important association, less attention has been paid to other types of childhood adversity and PDs. An examination of a wider range of adverse childhood events with all PDs in a population-based sample would significantly extend the existing literature. Finally, only a few studies investigating childhood adversity and PDs have taken into account the effects of Axis I mental disorders on this relationship (Tyrka et al., 2009; Grover et al., 2007; Gibb et al., 2001). This is an important methodological consideration since Axis I mental disorders are highly comorbid with Axis II PDs (McGlashan et al., 2000; Lenzenweger, 2008). To our knowledge, this study is the first to examine the relationship between a wide range of adverse childhood experiences including child maltreatment and household dysfunction with all types of Axis II PDs in a nationally representative populationbased sample. It builds upon the existing literature, which is based on clinical and small community samples. Furthermore, we adjust for Axis I disorders, an important consideration, given the high prevalence of comorbidity between Axis I and Axis II disorders. 1. Methods 1.1. Survey Data were from the second wave of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) collected in 2004 to 2005 (n ¼ 34,653). The NESARC is a representative sample of the adult (20 years of age or older), civilian, non-institutionalized population of the United States; it included respondents living in households and assorted non-institutional group dwellings such as college quarters, group homes, and boarding houses. The response rate for Wave 2 was 86.7%. Interviews for both waves of the NESARC

815

were conducted face-to-face by trained lay interviewers. Further details of the NESARC have been published elsewhere (Ruan et al., 2008b; Grant et al., 2005). 1.2. Measures 1.2.1. Childhood adversity 1.2.1.1. Child maltreatment: abuse and neglect. Respondents’ experiences of a variety of adverse childhood events (events occurring before the age of 18) were assessed using questions based on those from the Adverse Childhood Experiences study (Dong et al., 2003; Dube et al., 2003). These questions were in turn a subset of the items from the Conflict Tactics Scale (Straus, 1979; Straus et al., 1996) and the Childhood Trauma Questionnaire (Bernstein et al., 1994). Respondents were asked to respond to all questions pertaining to abuse, neglect (except emotional neglect), and having a battered mother on a five-point scale (never, almost never, sometimes, fairly often, or very often). Emotional neglect questions employed an alternative five-point scale of never true, rarely true, sometimes true, often true, or very often true. All questions pertaining to general household dysfunction required yes/no responding (except questions regarding having a battered mother, as mentioned above). From the list of questions, several types of childhood adversity were coded. Physical abuse was defined as a response of “sometimes” or greater to either question when asked how often a parent or other adult living in the respondent’s home (1) pushed, grabbed, shoved, slapped, or hit the respondent; or (2) hit the respondent so hard it left marks or bruises, or caused an injury. Emotional abuse was identified as a response of “fairly often” or “very often” to any question when asked how often a parent or other adult living in the respondent’s home (1) swore at, insulted, or said hurtful things to the respondent; (2) threatened to hit or throw something at the respondent (but did not do it); or (3) acted in any other way that made the respondent afraid he/she would be physically hurt or injured. These definitions are consistent with child maltreatment definitions employed in the Adverse Childhood Experiences study (Dube et al., 2003; Dong et al., 2003). Sexual abuse was examined using a series of four questions (Wyatt, 1985). These questions were adapted for use in the AUDADIS-IV and were rated on the same five-point scale that was used for all other abuse and physical neglect questions. The questions examined the occurrence of sexual touching or fondling, attempted intercourse, or actual intercourse by any adult or other person when the respondent did not want the act to occur or was too young to understand what was happening. Any response other than “never” on any of the questions was taken to indicate sexual abuse. Physical neglect was defined as any response other than “never” on a series of four relevant questions. These questions explored respondents’ experiences of being left unsupervised when too young to care for themselves or going without needed clothing, school supplies, food, or medical treatment. Other studies using the Adverse Childhood Experiences Study have defined physical neglect differently than we have here (Dong et al., 2003; Dube et al., 2003); however, we were unable to follow the conventions outlined by these previous researchers because of the exclusion of one of the original physical neglect questions by the AUDADIS-IV (the original series included five questions examining physical neglect). To compensate for this discrepancy, an alternative definition of physical neglect was developed. Examination of the distribution of summed responses to all physical neglect questions in our dataset indicated a clear break in the distribution between those responding with “never” to all items versus those responding with “almost never” or higher to at least one item (74.4% of respondents

816

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

answered “never” to all questions). For this reason, the aforementioned definition of physical neglect was adopted. Emotional neglect was defined by five questions regarding whether the respondent felt a part of a close-knit family or whether anyone in the respondent’s family of origin made the respondent feel special, wanted the respondent to succeed, believed in the respondent, or provided strength and support. Consistent with previous research all five items were reverse-scored and summed; scores of 15 or greater were identified as emotional neglect (Dube et al., 2003; Dong et al., 2003). 1.2.1.2. Household dysfunction. To characterize the experience of having a battered mother, respondents were asked whether the respondent’s father, stepfather, foster/adoptive father, or mother’s boyfriend had ever done any of the following to the respondent’s mother, stepmother, foster/adoptive mother, or father’s girlfriend: (1) pushed, grabbed, slapped, or threw something at her; (2) kicked, bit, hit with a fist, or hit her with something hard; (3) repeatedly hit her for at least a few minutes; or (4) threatened to use or actually used a knife or gun on her. Any response of “sometimes” or greater for questions 1 or 2, or any response except “never” for questions 3 or 4, was defined as having a battered mother. Parental substance abuse was assessed with two questions regarding whether a parent or other adult living in the home had a problem with alcohol or drugs. A response of “yes” to either of these questions was defined as parental substance abuse. To characterize the remaining household dysfunction variables, respondents were asked to answer with either “yes” or “no” whether a parent or other adult in the home (1) went to jail or prison; (2) was treated or hospitalized for a mental illness; (3) attempted suicide; and/or (4) actually committed suicide. Responses of “yes” for any of these questions defined the corresponding general household dysfunction variable. Two variables were derived from each of these abuse, neglect, and general household dysfunction variables. An “any abuse or neglect” variable was created that indicated the presence of at least one type of abuse or neglect in the respondent’s childhood. A similar “any adverse childhood events” variable was created that identified respondents who experienced at least one type of adverse childhood events (abuse, neglect, and general household dysfunction). 1.2.2. Personality Disorder Diagnoses Diagnoses of PDs were made using the Alcohol Use Disorder and Associated Disabilities Interview Schedule-Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (AUDADIS-IV) (Grant et al., 2001; Ruan et al., 2008a). The AUDADIS-IV provides a fully structured interview protocol to assess various Axis I (mood, anxiety, substance use disorders) and Axis II (PDs) diagnoses. The reliability of the AUDADIS-IV for PDs has been assessed using test-retest methods and were determined to be good (Kappa ¼ 0.67e0.71; ICC ¼ 0.71e0.75; Alpha ¼ 0.75e0.83) (Grant et al., 2003; Ruan et al., 2008b) and equivalent or better than to reliabilities from patient samples using semi-structured personality interviews in short-term test-retest studies (Zimmerman, 1994). The authors from the reliability studies aptly concluded that the AUDADIS-IV diagnostic measures were reliable and useful research tools (Ruan et al., 2008b). The validity of the AUDADIS-IV for PDs using mental component summary, social functioning, and role emotional scores has also been assessed with linear regression analyses and were found to be highly significant (P < 0.01 to P < 0.001) (Grant et al., 2004). All 10 PDs were assessed in either Waves 1 or 2 of the NESARC. In the first wave, all but schizotypal, borderline, and narcissistic PDs

were assessed; in Wave 2, these three PDs were measured. Antisocial PD was assessed in both waves, and the diagnostic variable from Wave 2 is used in our analysis. Although PDs are subject to change over time, this disparity between the times of assessment of some of the PDs is not thought to be problematic since PDs are often persistent in nature for many individuals (American Psychiatric Association, 2000). In addition to analyses conducted on individual PD diagnoses, we also examined PD cluster variables, which included the presence of one or more PDs within each of cluster A (paranoid, schizoid, schizotypal), cluster B (antisocial, histrionic, borderline, narcissistic), and cluster C (avoidant, dependent, obsessivecompulsive). These derived variables were based on the DSM-IV classification of PDs into clusters determined by similarities in symptomatology (American Psychiatric Association, 2000). 1.2.3. Covariates Sociodemographic covariates included age (continuous), gender, household income (continuous), years of education (continuous), marital status (three categories: married/living common law, separated/divorced/widowed, and never married), and race/ethnicity (five categories: non-Hispanic White, nonHispanic Black, non-Hispanic American Indian/Alaska Native, nonHispanic Hawaiian/Pacific Islander, and Hispanic of any race). Three Axis I mental disorder variables were included in the models: (1) any lifetime mood disorder (depression, dysthymia, mania, or hypomania), (2) any lifetime anxiety disorder (panic disorder, agoraphobia, social phobia, specific phobia, generalized anxiety disorder, or post-traumatic stress disorder), and (3) any lifetime substance use disorder (abuse/dependence on alcohol, sedatives, tranquilizers, opioids, amphetamines, cannabis, cocaine, hallucinogens, inhalants/solvents, heroin, or other drugs). These mental health conditions were diagnosed using the AUDADIS-IV, as described above. To account for the effect of inter-cluster PD comorbidity, each analysis was adjusted for the two cluster variables that represented the PD clusters other than the cluster currently being examined. For example, in analyses of avoidant personality disorder, adjustments were made for any cluster A and any cluster B PDs. Intra-cluster PD comorbidity was not adjusted for because PDs tend not to present as distinct entities, and instead exhibit a high degree of overlap within clusters (Cox et al., 2007). Adjusting for within-cluster PDs may remove variability that is simply due to the common features of all PDs within a cluster, thereby perhaps negating effects that truly exist. 1.3. Statistical methods All analyses were conducted using the weight and stratification variables supplied with the Wave 2 NESARC data file. To account for the complex sampling design of the NESARC, Taylor series linearization was used as the variance estimation technique using SUDAAN software (Shah et al., 2004). In addition, due to the number of comparisons and the large sample size, a conservative 99% confidence interval was used to determine the statistical significance of the odds ratios. Crosstabs were calculated to determine the prevalence of childhood adversity among individuals who met criteria for PDs. For these analyses, the any general household dysfunction variable was utilized, instead of individually examining each general household dysfunction variable. The any general household dysfunction variable was used in these analyses to increase readability of the table; in all further analyses, the individual general household dysfunction variables were utilized. Logistic regression analyses were used to determine the association between each

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822 Table 1 Prevalence of childhood adversity in the general U.S. population. Type of adverse childhood event

N (%)

Abuse Physical Emotional Sexual

6294 (17.6) 2911 (8.1) 3854 (10.6)

Neglect Physical Emotional Any abuse or neglect Any general household dysfunction Any adverse childhood event

8561 3413 10524 14266 18010

(24.2) (9.4) (30.1) (40.3) (51.5)

N (%): number and percentage of respondents who experienced the given adverse childhood event. Ns are for the sample, whereas percentages are weighted to be representative of the US population. Any general household dysfunction: indicates whether a respondent has experienced at least one type of general household dysfunction (battered mother/ female caregiver, parent substance use problem, parental incarceration, parent mental illness, parent suicide attempt, or parent suicide completion). Any adverse childhood event: Indicates whether a respondent has experienced at least one type of abuse, neglect, or general household dysfunction.

adverse childhood event and PDs. These analyses were adjusted for sociodemographic variables, lifetime mood disorders, lifetime anxiety disorders, and lifetime substance use disorders, and out-ofcluster PDs. 2. Results Table 1 presents the prevalence of childhood adversity in the sample. In the entire sample, 30% experienced child abuse and/or neglect, 40% experienced household dysfunction, and 52% experienced any childhood adversity. Lifetime Axis I disorders were prevalent among those with PDs and any adverse childhood experiences. Among those with cluster A PDs, 65% had an anxiety disorder, 65% had a mood disorder, and 56% had a substance use disorder. The prevalence of anxiety disorders, mood disorders, substance use disorders was 55%, 57%, and 63%, respectively, among those with Cluster B PDs. Among those with cluster C PDs, 60% had an anxiety disorder, 59% had a mood disorder, and 53% had a substance use disorder. Among individuals experiencing any childhood adversity, the prevalence of anxiety disorders, mood disorders, and substance use disorders was 35%, 34%, and 44%, respectively. Table 2 presents the prevalence of each type of adverse childhood event among individuals with PDs. Childhood adversity was

817

highly prevalent among all types of PDs. The results from the logistic regression models examining the association between each type of adverse childhood event and all cluster A PDs are presented in Table 3. The findings indicate that several adverse childhood experiences were associated with increased odds of having cluster A PDs. More specifically, all types of abuse and neglect, having a battered mother, and parental substance use problems were associated with an increased likelihood of a cluster A PD. When examining the individual PDs, childhood adversity had the greatest link to schizotypal PD. Table 4 presents the results from the logistic regression models computing the relationship between childhood adversity and cluster B PDs. All adverse childhood events were significantly associated with increased odds of having a cluster B PD with the exception of completed parental suicide. When considering each cluster B PD individually, almost all types of child abuse, neglect, and household dysfunction were associated with an increased likelihood of having antisocial, borderline, and narcissistic PDs. Conversely, strong evidence for a relationship between childhood adversities and histrionic PD was not found. Table 5 presents the results from the logistic regression models examining the relationship between each adverse childhood event and cluster C PDs. The findings indicate that childhood adversity was not strongly associated with cluster C PDs. When looking at each cluster C PD individually, only emotional neglect was associated with avoidant PD, physical neglect with obsessive-compulsive PD, and no significant relationships were found between childhood adversity and dependent PD. Table 6 presents the results from the logistic regression models computing the relationships between child abuse and neglect, household dysfunction, mood disorders, anxiety disorders, substance use disorders, out-of-cluster PDs, and sociodemographic variables with each PD cluster. The findings indicate that several forms of child abuse and neglect remained associated with cluster A and B PDs when simultaneously accounting for the variance of all other covariates. Notably, the highest associations with PD clusters in these models were found for mental disorders including mood disorders, anxiety disorders, substance use disorders, and out-ofcluster PDs. 3. Discussion To our knowledge, this is the first study to examine the relationship between a wide range of childhood adversities and all Axis II PDs using a nationally representative sample controlling for Axis I

Table 2 Prevalence of childhood adversity among those with personality disorders. Personality disorder

Physical abuse N (%)

Emotional abuse N (%)

Sexual abuse N (%)

Physical neglect N (%)

Emotional neglect N (%)

Any general household dysfunction N (%)

Cluster A Paranoid Schizoid Schizotypal Cluster B Antisocial Histrionic Borderline Narcissistic Cluster C Avoidant Dependent Obsessive-Compulsive

1208 593 377 639 1767 532 224 957 875 983 273 52 830

786 405 234 433 1077 347 157 649 492 607 201 40 495

897 442 249 514 1209 308 166 770 572 699 234 46 562

1377 680 421 717 2004 588 268 1017 995 1174 334 71 993

664 341 238 324 853 270 118 487 363 528 208 44 402

1659 820 505 835 2382 653 315 1248 1164 1387 393 69 1157

(34.4) (35.0) (32.6) (40.9) (34.9) (41.9) (33.0) (40.5) (34.1) (29.0) (33.5) (34.0) (28.5)

(23.3) (24.9) (22.2) (28.0) (21.7) (27.1) (23.7) (28.6) (19.6) (18.0) (25.1) (28.0) (16.9)

(25.6) (25.4) (22.0) (33.4) (23.8) (23.8) (24.3) (33.6) (21.4) (21.2) (28.1) (31.3) (20.1)

(40.6) (40.8) (39.1) (47.9) (41.4) (48.4) (42.6) (45.2) (39.5) (35.9) (39.7) (45.6) (35.9)

(19.3) (20.1) (22.1) (20.2) (17.2) (20.6) (18.1) (21.0) (14.4) (15.7) (24.9) (27.0) (13.7)

(48.2) (49.0) (46.3) (54.5) (48.2) (52.1) (49.5) (55.0) (46.0) (42.2) (48.4) (52.0) (41.6)

N (%): number and percentage of respondents meeting criteria for the given personality disorder who experienced the given adverse childhood event. Ns are unweighted, percentages are weighted. Any general household dysfunction: indicates whether a respondent has experienced at least one type of general household dysfunction (battered mother/female caregiver, parent substance use problem, parental incarceration, parent mental illness, parent suicide attempt, or parent suicide completion).

818

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

Table 3 Associations of childhood adversity with Cluster A personality disorders. Type of adverse childhood event

Cluster A personality disorder Any Cluster A

Paranoid

Schizoid

Schizotypal

Odds ratios (99% CI)

Odds ratios (99% CI)

Odds ratios (99% CI)

Odds ratios (99% CI)

Abuse Physical Emotional Sexual

1.39 (1.17e1.66) 1.71 (1.38e2.11) 1.44 (1.18e1.76)

1.26 (1.00e1.58) 1.52 (1.17e1.98) 1.05 (0.80e1.37)

1.18 (0.91e1.53) 1.35 (0.99e1.84) 0.99 (0.76e1.28)

1.62 (1.28e2.03) 1.76 (1.35e2.31) 2.05 (1.59e2.65)

Neglect Physical Emotional

1.29 (1.11e1.51) 1.50 (1.22e1.84)

1.15 (0.93e1.43) 1.31 (0.98e1.73)

1.13 (0.88e1.45) 1.68 (1.27e2.23)

1.61 (1.26e2.05) 1.35 (1.05e1.74)

General household dysfunction Battered mother/female caregiver Parent substance use problem Parent went to jail Parent mental illness Parent suicide attempt Parent completed suicide Any abuse or neglect Any adverse childhood event

1.27 1.21 1.26 1.04 1.17 0.85 1.40 1.54

1.24 1.15 1.18 1.02 1.05 0.92 1.14 1.29

1.21 1.13 1.22 1.13 0.99 1.22 1.16 1.32

1.33 1.42 1.48 1.09 1.22 0.95 2.01 2.28

(1.05e1.54) (1.04e1.41) (0.99e1.61) (0.78e1.38) (0.84e1.62) (0.50e1.47) (1.19e1.64) (1.31e1.81)

(0.97e1.60) (0.93e1.43) (0.84e1.65) (0.73e1.42) (0.67e1.66) (0.43e1.98) (0.93e1.41) (1.02e1.62)

(0.91e1.62) (0.87e1.45) (0.86e1.72) (0.80e1.62) (0.62e1.58) (0.54e2.74) (0.89e1.50) (1.01e1.74)

(1.03e1.70) (1.14e1.78) (1.09e2.00) (0.74e1.60) (0.81e1.84) (0.48e1.86) (1.61e2.52) (1.78e2.92)

Note: all odds ratios adjusted for age, gender, education, income, race/ethnicity, marital status, any cluster B PDs, any cluster C PDs, any Axis I lifetime mood disorders, any Axis I anxiety disorders, and any Axis I substance use disorders. Bold font indicates significant adjusted odds ratios (p < 0.01).

disorders and sociodemographic covariates. The findings show further evidence of the link between adverse experiences in childhood and mental health disorders in adulthood. First, the results indicate that many types of childhood adversity were highly prevalent among individuals with PDs in the general U.S. population. Second, childhood adversity was most strongly and consistently associated with clusters A and B PDs and specifically schizotypal, antisocial, borderline, and narcissistic PDs. Many of these findings remained significant even after simultaneously accounting for the variance of all types of child abuse and neglect, household dysfunction, mental disorders, and sociodemographic covariates. Third, childhood adversity in the form of household dysfunction including exposure to battering of a mother, parental substance use problems, parental incarceration, parental mental illness, and parental suicide attempts was associated with increased likelihood of PDs. Although all types of childhood adversities were highly prevalent among individuals with PDs in the current sample, this

prevalence was lower than estimates based on clinical samples (Battle et al., 2004; Bierer et al., 2003). For example, 29% of respondent with borderline PD reported experiencing emotional abuse in the current sample compared to 66% from a longitudinal clinical sample (Battle et al., 2004). Our findings suggest that childhood adversity may not be as prevalent in community samples of people with PDs compared to clinical samples, but adverse childhood events remain common among individuals with PD in the general population. Significant associations were found between childhood adversity and PDs from all three clusters. Significant relationships between child abuse and neglect and various PDs from clusters A, B, and C are consistent with previous research (Battle et al., 2004; Gibb et al., 2001; Grover et al., 2007; Johnson et al., 1999, 2000). However, our current findings indicate that childhood adversity more broadly defined as child abuse, neglect and household dysfunction was more robustly related to schizotypal PD and most PDs from cluster B. The relationship between childhood adversity

Table 4 Associations of childhood adversity with Cluster B personality disorders. Type of adverse childhood event

Cluster B personality disorder Any Cluster B

Antisocial

Histrionic

Borderline

Narcissistic

Odds Ratios (99% CI)

Odds Ratios (99% CI)

Odds Ratios (99% CI)

Odds Ratios (99% CI)

Odds Ratios (99% CI)

Abuse Physical Emotional Sexual

2.00 (1.77e2.27) 2.27 (1.92e2.68) 2.14 (1.83e2.51)

2.42 (1.97e2.98) 2.58 (1.95e3.40) 2.17 (1.63e2.89)

1.20 (0.90e1.60) 1.31 (0.92e1.86) 1.09 (0.76e1.58)

2.04 (1.70e2.45) 2.31 (1.87e2.87) 2.47 (2.05e2.97)

1.70 (1.45e1.98) 1.72 (1.39e2.12) 1.64 (1.34e2.00)

Neglect Physical Emotional

1.79 (1.55e2.07) 1.63 (1.38e1.94)

2.02 (1.60e2.54) 2.00 (1.54e2.60)

1.25 (0.91e1.70) 1.22 (0.81e1.84)

1.71 (1.45e2.03) 1.60 (1.25e2.04)

1.49 (1.26e1.77) 1.23 (0.99e1.54)

General household dysfunction Battered mother/female caregiver Parent substance use problem Parent went to jail Parent mental illness Parent suicide attempt Parent completed suicide Any abuse or neglect Any adverse childhood event

1.77 1.57 1.65 1.52 1.50 1.20 2.11 2.04

1.84 1.65 1.69 1.41 1.57 1.16 2.26 2.23

1.17 1.12 1.14 1.37 0.80 1.20 1.50 1.38

1.71 1.70 1.76 1.54 1.53 1.33 2.36 2.35

1.57 1.37 1.41 1.46 1.43 1.22 1.81 1.74

(1.49e2.11) (1.38e1.79) (1.36e2.00) (1.23e1.87) (1.14e1.95) (0.81e1.78) (1.86e2.41) (1.76e2.36)

(1.40e2.43) (1.32e2.05) (1.25e2.27) (0.99e1.84) (1.05e2.36) (0.60e2.24) (1.80e2.83) (1.73e2.87)

(0.83e1.66) (0.83e1.51) (0.74e1.76) (0.84e2.23) (0.41e1.59) (0.46e3.12) (1.11e2.02) (1.00e1.90)

(1.42e2.06) (1.44e2.01) (1.36e2.27) (1.18e2.01) (1.08e2.16) (0.74e2.39) (1.99e2.81) (1.92e2.86)

(1.30e1.89) (1.15e1.64) (1.09e1.83) (1.12e1.91) (1.06e1.93) (0.75e1.99) (1.55e2.12) (1.47e2.07)

Note: all odds ratios adjusted for age, gender, education, income, race/ethnicity, marital status, any cluster A PDs, any cluster C PDs, any Axis I lifetime mood disorders, any Axis I anxiety disorders, and any Axis I substance use disorders. Bold font indicates significant adjusted odds ratios (p < 0.01).

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

819

Table 5 Associations of childhood adversity with Cluster C personality disorders. Type of adverse childhood event

Cluster C personality disorder Any Cluster C

Avoidant

Dependent

Obsessive-compulsive

Odds ratios (99% CI)

Odds ratios (99% CI)

Odds ratios (99% CI)

Odds ratios (99% CI)

Abuse Physical Emotional Sexual

1.16 (0.98e1.37) 1.19 (0.99e1.43) 1.16 (0.95e1.42)

1.16 (0.98e1.37) 1.22 (0.89e1.66) 1.13 (0.86e1.48)

0.81 (0.44e1.47) 0.93 (0.47e1.84) 0.89 (0.45e1.74)

1.15 (0.95e1.39) 1.11 (0.91e1.36) 1.11 (0.90e1.37)

Neglect Physical Emotional

1.17 (1.01e1.36) 1.15 (0.96e1.39)

1.02 (0.78e1.34) 1.60 (1.18e2.16)

1.03 (0.61e1.75) 1.31 (0.75e2.30)

1.20 (1.04e1.40) 0.99 (0.79e1.23)

General household dysfunction Battered mother/female caregiver Parent substance use problem Parent went to jail Parent mental illness Parent suicide attempt Parent completed suicide Any abuse or neglect Any adverse childhood event

1.09 1.08 0.93 1.14 1.07 1.05 1.22 1.17

1.05 1.08 0.95 1.15 0.91 0.83 1.18 1.17

0.81 0.64 0.97 1.40 1.06 2.02 0.78 0.98

1.06 1.10 0.88 1.11 1.04 1.18 1.21 1.16

(0.90e1.32) (0.92e1.28) (0.72e1.20) (0.90e1.43) (0.81e1.41) (0.70e1.58) (1.06e1.40) (1.00e1.37)

(0.78e1.43) (0.80e1.45) (0.65e1.39) (0.81e1.63) (0.58e1.43) (0.35e1.95) (0.91e1.53) (0.86e1.59)

(0.38e1.73) (0.34e1.22) (0.42e2.25) (0.63e3.12) (0.42e2.66) (0.62e6.53) (0.43e1.40) (0.51e1.86)

(0.86e1.30) (0.92e1.30) (0.67e1.15) (0.87e1.43) (0.78e1.39) (0.79e1.79) (1.04e1.41) (0.99e1.37)

Note: all odds ratios adjusted for age, gender, education, income, race/ethnicity, marital status, any cluster A PDs, any cluster B PDs, any Axis I lifetime mood disorders, any Axis I anxiety disorders, and any Axis I substance use disorders. Bold font indicates significant adjusted odds ratios (p < 0.01).

and schizotypal PD is consistent with previous research that found an association between childhood adversity and schizotypal symptoms (Steel et al., 2009; Berenbaum et al., 2003, 2008). This connection may be partly explained through the shared variance between childhood adversity, schizotypal symptoms, and dissociative tendencies (Irwin, 2001). Almost all forms of abuse, neglect,

and household dysfunction were associated with increased odds of having schizotypal PD, antisocial PD, borderline PD, and narcissistic PD. This observation is in keeping with the clinical impression of a particularly strong association between cluster B PDs and childhood adversities. Cluster B PDs are characterized by dramatic, emotional, and erratic behavior (American Psychiatric Association,

Table 6 Associations of childhood adversity, household dysfunction, and sociodemographic covariates with personality disorders. Cluster A

Cluster B

Cluster C

Adjusted Odds ratios (99% CI)

Adjusted Odds ratios (99% CI)

Adjusted Odds ratios (99% CI)

Child abuse and neglect Physical abuse Emotional abuse Sexual abuse Physical neglect Emotional neglect

1.08 1.33 1.22 1.09 1.26

(0.86e1.35) (1.01e1.74) (1.00e1.50) (0.92e1.30) (0.99e1.59)

1.42 1.29 1.63 1.34 1.09

(1.22e1.65) (1.03e1.61) (1.37e1.93) (1.14e1.57) (0.89e1.33)

1.09 1.08 1.10 1.12 1.04

(0.87e1.36) (0.85e1.37) (0.89e1.36) (0.95e1.32) (0.84e1.30)

General household dysfunction Battered mother/female caregiver Parent substance use problem Parent went to jail Parent mental illness Parent suicide attempt Parent completed suicide

0.98 1.04 1.11 0.90 1.15 0.71

(0.77e1.25) (0.85e1.28) (0.84e1.46) (0.64e1.28) (0.74e1.79) (0.37e1.35)

1.02 1.19 1.11 1.23 0.99 0.86

(0.82e1.27) (1.03e1.38) (0.88e1.41) (0.93e1.62) (0.66e1.48) (0.49e1.50)

0.96 1.05 0.81 1.15 0.93 1.12

(0.75e1.21) (0.87e1.26) (0.61e1.08) (0.88e1.52) (0.64e1.37) (0.67e1.88)

Mental disorders Mood disorders Anxiety disorders Substance use disorders Cluster A Cluster B Cluster C

2.25 (1.90e2.65) 2.38 (1.98e2.86) 1.17 (0.99e1.37) Not included 4.94 (4.16e5.87) 5.66 (4.72e6.79)

2.20 (1.91e2.53) 1.83 (1.60e2.08) 2.05 (1.79e2.34) 4.84 (4.08e5.74) Not included 1.61 (1.35e1.93)

2.17 (1.82e2.58) 2.05 (1.76e2.40) 1.26 (1.07e1.48) 5.44 (4.52e6.55) 1.58 (1.32e1.88) Not included

Sociodemographic covariates Age (continuous) Gender (female reference) Household income (continuous) Education (continuous) Marital Status (married/common law) Widowed/separated/divorced Never married Ethnicity (White reference) Black American Indian/Alaska Native Hawaiian/Pacific Islander Hispanic

0.99 1.19 0.98 0.94 1.00 1.34 1.30 1.00 2.09 1.55 1.00 1.41

0.98 2.00 0.98 0.98 1.00 1.30 1.19 1.00 1.71 1.38 1.10 1.10

1.00 1.05 1.00 1.05 1.00 0.77 0.82 1.00 0.80 0.72 0.76 0.82

(0.99e1.00) (1.01e1.41) (0.96e0.99) (0.91e0.98) (e) (1.12e1.59) (1.08e1.57) (e) (1.71e2.56) (0.99e2.42) (0.61e1.62) (1.11e1.79)

All independent variables simultaneously entered into each PD cluster model.

(0.98e0.98) (1.73e2.32) (0.97e1.00) (0.95e1.01) (e) (1.13e1.50) (1.00e1.41) (e) (1.44e2.03) (0.97e1.98) (0.67e1.81) (0.89e1.36)

(1.00e1.01) (0.91e1.20) (0.98e1.02) (1.01e1.08) (e) (0.64e0.92) (0.67e0.99) (e) (0.67e0.97) (0.44e1.18) (0.52e1.11) (0.66e1.02)

820

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

1994, 2000). Theory suggests that personality develops from emotions and emotion-related experiences beginning in early childhood (Cohen, 2008). Adverse childhood experiences may, in part, shape personality and have the potential to negatively influence the development of personality traits, PD symptoms, and PDs. Another interesting finding from this research was that many adverse childhood events were not strongly related to cluster C disorders. It may be that the nature of the psychopathology associated with cluster C PDs (i.e. the anxious, fearful cluster) is more closely overlapping with the “distress” Axis I disorders (i.e. mood, anxiety). If etiological factors (or “symptoms” used in making the diagnoses) are more closely overlapping, then controlling for such Axis I disorders may more completely account for any association that might have been seen between cluster C and childhood adversities. In the fully adjusted models presented in Table 6, Axis I disorders including mood disorders, anxiety disorders, and substance use disorders remained significantly associated with cluster A, cluster B, and cluster C PDs with the exception of substance use disorders with cluster A PDs. It is noted that mood disorders, anxiety disorders, and substance use disorders were among the largest odds ratios associated with cluster A, B, and C PDs meaning that a history of Axis I mental disorders remains a strong predictor of PDs. This is in keeping with the high prevalence of comorbid Axis I and Axis II mental disorders found in this study. These highly comorbid relationships between Axis I mental disorders and Axis II PDs have treatment implications; poorer clinical outcomes may result for some individuals presenting with this comorbidity (Reich, 2007). Although child abuse and neglect significantly increases the likelihood of cluster A and cluster B PDs, Axis I disorders are also important correlates. Of particular note, most individuals who experience childhood adversity do not develop PDs; it is important in developing approaches to reduce impairment to understand the mediators and moderators of this association. As outlined by Bornovalova et al. (2009), there is a lack of information about the longitudinal trajectories of PDs; to understand the development of PDs, it is essential to measure the onset and course of PDs with repeated assessments from youth and adulthood (indeed we would argue, beginning in childhood with measurement of environmental adversity). The investigation by Kim et al. (2009) is one of the few longitudinal studies to examine the influence of maltreatment on personality processes and subsequent adjustment in a sample of children. Ideally, such a sample would be followed through to young adulthood. In addition to the need for longitudinal follow-up, future research should model the genetic and environmental effects on personality traits, and include gene-environment interactions (Bornovalova et al., 2009). This might help explain why some individuals who experience adversity during childhood do not develop PDs (Paris, 1997, 1998). Perhaps some individuals, based on their genetic make-up, or other factors that buffer the environmental adversity, such as experiences of nurturing parenting are protected from the negative impact of child maltreatment. Many domains of personality are highly heritable (Jang et al., 1996). It may be that a PD is a result of a specific genotypes interacting with the adverse environmental factors that leads to expression of dysfunctional personality traits, PD symptoms, or PDs. Similarly, not all individuals with PDs have a history of childhood adversity. Clearly, there are multiple pathways that lead to the development of PDs. Investigation of pathway models using behavioral-genetics and molecular study designs including measures of resiliency are necessary in determining the factors that influence the development of PDs, conditions which are associated with major morbidity and some mortality.

Estimated societal costs associated with child abuse (World Health Organization, 2006) and PDs (van Asselt et al., 2007) are substantial. Although some progress has been made in preventing child maltreatment (MacMillan et al., 2009), it is also important to determine ways of reducing impairment, as well as recurrence among those who have suffered maltreatment in childhood. For example, among sexually abused children who experience PTSD symptoms, there is evidence that trauma-focused cognitivebehavior therapy (TF-CBT) can reduce PTSD as well as anxiety and depression (Macdonald et al., 2006). However, given that most follow-ups after TF-CBT do not extend beyond 12 months (Stallard, 2006), the long-term effects of such treatment are unknown. Do interventions shown to be effective in reducing mental health problems following maltreatment in childhood lead to better outcomes in adulthood such as reduced risk of PDs? This is a critical area of research that requires long-term follow-up of patients. Although it is not ethical to withhold an effective treatment from one group, it is possible to compare treatments (for example, usual care and enhanced treatment) and follow patients long-term. We are not aware of any trials following children to adulthood after treatment for conditions related to child maltreatment; however, Olds and colleagues are following participants in each of three prevention trials of home visitation to determine the long-term outcome of this intervention (Olds et al., 2007). Prevention strategies that are effective in reducing adverse childhood events may also help to reduce PDs in the general population, but this is currently unknown. Furthermore, programs that address one type of maltreatment or related symptoms, such as PTSD, cannot be assumed to generalize to other types of maltreatment or impairment. Cohen and colleagues are currently evaluating whether TFCBT is effective in reducing symptoms among children exposed to domestic violence. Such programs have the potential to prevent a wide range of mental health problems in adulthood, but this is yet to be determined. Preventing child maltreatment is not a simple task. Evidence from prevention and intervention research should be replicated in other samples. It is also important for clinicians and researchers to be aware of the types of household dysfunction that are related of PDs. Clinicians need to consider the broader range of household dysfunction when inquiring about child abuse and neglect. Also, researchers could include these and other household dysfunction variables to broaden the examination of childhood adversity in future research on PDs. Limitations of the current study should be considered. First, the cross-sectional design precludes determining any causal inferences in the relationship between childhood adversity and PDs. Second, data on childhood adversity were collected retrospectively, which may introduce some sampling error due to recall and reporting bias. For example, it is possible that individuals with PDs might be more likely to subjectively recall an experience as abusive or traumatic. However, there is evidence that supports the validity of accurate recall of adverse childhood events (Hardt and Rutter, 2004). Additionally, although several items were used to measure child abuse and neglect, the assessment of other family violence was limited to violence against a mother or female caregiver. Although our study included a wide range of adverse childhood events this is not an exhaustive list; other types of adverse childhood should be included in future research. Third, although all mental disorder diagnoses were made by a reliable structured interview conducted by trained lay interviews, this assessment approach may not match the accuracy of an experienced clinician. Structured clinical interviews for DSM based diagnoses would be ideal, but is not possible in nationally representative epidemiologic surveys due to expense. However, the assessment of Axis I and Axis II PDs included in the NESARC using the AUDADIS-IV provides

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

a unique opportunity to study PDs in the general population. Finally, the existence of numerous Axis I disorders were included as covariates. However, not all Axis I disorders were assessed in the data (e.g. psychotic disorders or obsessive-compulsive disorder), which is an important limitation. In conclusion, the present findings suggest that childhood abuse, neglect and household dysfunction are related to PDs in the general U.S. population. Due to the trauma of childhood adversity and impairment related to PDs, the present study offers important policy implications. Reducing childhood adversity may help to reduce PDs in the general population. These disorders are associated with a huge burden of suffering and determining approaches to reduce them should be a priority. Acknowledgements The authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The authors would like to thank Christine Henriksen for editing the manuscript. Funding Preparation of this article was supported by a Canadian Institutes of Health Research (CIHR) Postdoctoral Fellowship award to the first author. Contributors Afifi conducted literature searches, designed the analysis, wrote sections of the manuscript. Mather designed the analysis, conducted the statistical analysis, wrote sections of the manuscript, edited and revised the manuscript. Boman designed the analysis, wrote sections of the manuscript, edited and revised the manuscript. Fleisher designed the analysis, wrote sections of the manuscript, edited and revised the manuscript. Enns designed the analysis, wrote sections of the manuscript, edited and revised the manuscript. MacMillan designed the analysis, wrote sections of the manuscript, edited and revised the manuscript. Sareen designed the analysis, wrote sections of the manuscript, edited and revised the manuscript. All authors contributed to and have approved the final manuscript. Conflict of interest No conflicts of interest to declare. References Afifi TO, Boman J, Fleisher W, Sareen J. The relationship between child abuse, parental divorce, and lifetime mental disorders and suicidality in a nationally representative adult sample. Child Abuse & Neglect 2009;33:139e47. Afifi TO, Brownridge DA, Cox BJ, Sareen J. Physical punishment, childhood abuse, and psychiatric disorders. Child Abuse & Neglect 2006;30:1093e103. Afifi TO, Brownridge DA, MacMillan H, Sareen J. The relationship of gambling to intimate partner violence and child maltreatment in a nationally representative sample. Journal of Psychiatric Research 2010;44:331e7. Afifi TO, Enns MW, Cox BJ, Asmundson GJG, Stein MB, Sareen J. Population attributable fractions of psychiatric disorders and suicidal ideation and attempts associated with adverse childhood events in the general population. American Journal of Public Health 2008;98:946e52. Allen B, Lauterbach D. Personality characteristics of adult survivors of childhood trauma. Journal of Traumatic Stress 2007;20:587e95.

821

American Psychiatric Association. Diagnostic & statistical manual for mental disorders (DSM). Washington, DC: American Psychiatric Press, Inc; 1994. American Psychiatric Association. Diagnostic criteria from DSM-IV-TR. Washington, D.C.: American Psychiatric Association; 2000. Battle CL, Shea T, Johnson DM, Yen S, Zlotnick C, Zanarini MC, et al. Childhood maltreatment associated with adult personality disorders: findings from the collaborative longitudinal personality disorders study. Journal of Personality Disorders 2004;18:193e211. Berenbaum H, Thompson RJ, Milanek ME, Bredemeier K. Psychological trauma and schizotypal personality disorder. Journal of Abnormal Psychology 2008;117:502e19. Berenbaum H, Valera EM, Kerns JG. Psychological trauma and schizotypal symptoms. Schizophrenia Bulletin 2003;29:143e52. Bernstein DP, Fink L, Handelsman L, Foote J, Lovejoy M, Wenzel K, et al. Initial reliability and validity of a new retrospective measure of child abuse and neglect. American Journal of Psychiatry 1994;151:1132e6. Bierer LM, Yehuda R, Schmeidler J, Mitropoulou V, New AS, Silverman JM, et al. Abuse and neglect in childhood: relationship to personality disorder diagnoses. CNS Spectrum 2003;8:737e54. Bornovalova MA, Hicks BM, Iacono WG, McGue M. Stability, change, and heritability of borderline personality disorder traits from adolescence to adulthood: a longitudinal twin study. Development and Psychopathology 2009;21:1335e53. Bruffaerts R, Demyttenaere K, Borges G, Haro JM, Chiu WT, Hwang I, et al. Childhood adversities as risk factors for onset and persistence of suicidal behaviour. British Journal of Psychiatry 2010;197:20e7. Cohen P. Child development and personality disorder. Psychiatric Clinics of North America 2008;31:477e93. Cox BJ, Sareen J, Enns MW, Clara I, Grant BF. The fundamental structure of Axis II personality disorders assessed in the National epidemiologic survey on alcohol and related conditions. Journal of Clinical Psychiatry 2007;68:1913e20. Dong M, Anda RF, Dube SR, Giles WH, Felitti VJ. The relationship of exposure to childhood sexual abuse to other forms of abuse, neglect, and household dysfunction during childhood. Child Abuse & Neglect 2003;27:625e39. Dube SR, Felitti VJ, Dong M, Chapman DP, Giles WH, Anda RF. Childhood abuse, neglect, and household dysfunction and the risk of illicit drug use: the adverse childhood experiences study. Pediatrics 2003;111:564e72. Enns MW, Cox BJ, Afifi TO, de Graaf R, ten Have M, Sareen J. Childhood adversities and risk for suicidal ideation and attempts: a longitudinal population-based study. Psychological Medicine 2006;32:1769e78. Gibb BE, Wheeler R, Alloy LB, Abramson LY. Emotional, physical, and sexual maltreatment in childhood versus adolescence and personality dysfunction in young adulthood. Journal of Personality Disorders 2001;15:505e11. Grant BF, Dawson DA, Hasin DS. The alcohol use disorder and associated disabilities interview schedule-DSM-IV version. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2001. Grant BF, Dawson DA, Stinson FS, Chou PS, Kay W, Pickering RP. The alcohol use disorder and associated disabilities interview schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression, and psychiatric diagnostic modules in a general population sample. Drug and Alcohol Dependence 2003;71:7e16. Grant BF, Hasin DS, Stinson FS, Dawson DA, Chou SP, Ruan WJ. Co-occurrence of DSM-IV personality disorders in the US: results from the National epidemiologic survey on alcohol and related conditions. Comprehensive Psychiatry 2005;46:1e5. Grant BF, Stinson FS, Dawson DA, Chou SP, Ruan WJ, Pickering RP. Co-occurrence of 12-month alcohol and drug use disorders and personality disorders in the United States: results from the national epidemiologic survey on alcohol and related conditions. Archives of General Psychiatry 2004;61:361e8. Grover KE, Carpenter LL, Price LH, Gagne GG, Mello AF, Mello MF, et al. The relationship between childhood abuse and adult personality disorder symptoms. Journal of Personality Disorders 2007;21:442e7. Hardt J, Rutter M. Validity of adult retrospective reports of adverse childhood experiences: review of the evidence. Journal of Child Psychology and Psychiatry and Allied Disciplines 2004;45:260e73. Irwin HJ. The relationship between dissociative tendencies and schizotypy: an artifact of childhood trauma? Journal of Clinical Psychology 2001;57:331e42. Jang KL, Livesley WJ, Vernon PA, Jackson DN. Heritability of personality disorder traits: A twin study. Acta Psychiatrica Scandinavica 1996;94:438e44. Johnson DM, Shehan TC, Chard KM. Personality disorders, coping strategies, and posttraumatic stress disorder in women with histories of childhood sexual abuse. Journal of Child Sexual Abuse 2004;12:19e39. Johnson JG, Cohen P, Brown J, Smailes EM, Bernstein DP. Childhood maltreatment increases risk for personality disorders during early adulthood. Archives of General Psychiatry 1999;56:600e8. Johnson JG, Cohen P, Chen H, Kasen S, Brook JS. Parenting behaviors associated with risk for offspring personality disorder during adulthood. Archives of General Psychiatry 2006;63:579e87. Johnson JG, Smailes EM, Cohen P, Brown J, Bernstein DP. Association between four types of childhood neglect and personality disorder symptoms during adolescence and early adulthood: findings of a community-based longitudinal study. Journal of Personality Disorders 2000;14:171e87. Kessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychological Medicine 1997;27:1101e19. Kim J, Cicchetti D, Rogosch FA, Manly JT. Child maltreatment and trajectories of personality and behavioral functioning: implications for the development of personality disorder. Development and Psychopathology 2009;21:889e912.

822

T.O. Afifi et al. / Journal of Psychiatric Research 45 (2011) 814e822

Lenzenweger MF. Epidemiology of personality disorders. Psychiatric Clinics of North America 2008;31:395e403. Luntz BK, Widom CS. Antisocial personality disorder in abused and neglected children grown up. American Journal of Psychiatry 1994;151:670e4. Macdonald GM, Higgins JP, Ramchandani P. Cognitive-behavioural interventions for children who have been sexually abused. Cochrane Database of Systematic Reviews 2006;4:CD001930. MacMillan HL, Fleming JE, Streiner DL, Lin E, Boyle MH, Jamieson E, et al. Childhood abuse and lifetime psychopathology in a community sample. American Journal of Psychiatry 2001;158:1878e83. MacMillan HL, Wathen CN, Barlow J, Fergusson DM, Leventhal JM, Taussig HN. Interventions to prevent child maltreatment and associated impairment. Lancet 2009;373:250e66. McGlashan TH, Grilo CM, Skodol AE, Gunderson JG, Shea TM, Morey LC, et al. The collaborative longitudinal personality disorders study: baseline axis I/II and II/II diagnositic co-occurrence. Acta Psychiatrica Scandinavica 2000;102:256e64. Murray JB. Relationship of childhood sexual abuse to borderline personality disorder, posttraumatic stress disorder, and multiple personality disorder. Journal of Psychology 1993;127:657e76. Olds DL, Sadler L, Kitzman H. Programs for parents of infants and toddlers: recent evidence from randomized trials. Journal of Child Psychology and Psychiatry and Allied Disciplines 2007;48:355e91. Paris J. Childhood trauma as an etiological factor in the personality disorders. Journal of Personality Disorders 1997;11:34e49. Paris J. Childhood trauma causes personality disorders? Canadian Journal of Psychiatry 1998;43:148e53. Reich J. State and trait in personality disorders. Annals of Clnical Psychiatry 2007;19:37e44. Rettew DC, Zanarini MC, Yen S, Grilo CM, Skodol AE, Shea T, et al. Childhood antecedents of avoidant personality disorder: A retrospective study. Journal of the American Academy of Child and Adolescent Psychiatry 2003;42:1122e30. Ruan WJ, Goldstein RB, Chou SP, Smith SM, Saha TD, Pickering RP, et al. The alcohol use disorder and associated disabilities interview schedule-IV (AUDADIS-IV): reliability of new psychiatric diagnostic modules and risk factors in a general population sample. Drug Alcohol Dependence 2008a;92:27e36. Ruan WJ, Goldsten RB, Chou SP, Smith SM, Saha TD, Pickering RP, et al. The Alcohol use disorder and associated disabilities interview schedule-IV (AUDADIS-IV): Reliability of new psychiatric diagnostic modules and risk factors in a general population sample. Drug and Alcohol Dependence 2008b;92:27e36. Scott KM, Smith DR, Ellis PM. Prospectively ascertained child maltreatment and its association with DSM-IV mental disorders in young adults. Archives of General Psychiatry 2010;67:712e9.

Shah BV, Barnwell BG, Bieler GS. SUDAAN user’s manual: release 9.0. Triangle Park, NC: Research Triangle Institute; 2004. Stallard P. Psychological interventions for post-traumatic reactions in children and young people: a review of randomised controlled trials. Clinical Psychology Review 2006;26:895e911. Steel C, Marillier S, Fearson P, Ruddle A. Childhood abuse and schizotypal personality. Social Psychiatry and Psychiatric Epidemiology 2009;44:917e23. Straus MA. Measuring intrafamily conflict and violence: the Conflict Tactics (CT) Scales. Journal of Marriage and the Family 1979;41:75e88. Straus MA, Hamby SL, Boney-McCoy S, Sugarman DB. The revised Conflict Tactics Scales (CTS2): development and preliminary psychometric data. Journal of Family Issues 1996;17:283e316. Tyrka AR, Wyche MC, Kelly MM, Price LH, Carpenter LL. Child maltreatment and adult personality disorder symptoms: influence of maltreatment type. Psychiatry Research 2009;165:281e7. van Asselt AD, Dirksen CD, Arntz A, Severens JL. The cost of borderline personality disorder: societal cost of illness in BPD-patients. European Psychiatry 2007;22:354e61. World Health Organization. Preventing child maltreatment: A guide to taking action and generating evidence. WHO; 2006. Wyatt GE. The sexual abuse of Afro-American and white American women in childhood. Child Abuse & Neglect 1985;9:507e19. Yen S, Shea T, Battle CL, Johnson DM, Zlotnick C, Dolan-Sewell R, et al. Traumatic exposure and posttraumatic stress disorder in borderline, schizotypal, avoidant, and obessive-compulsive personality disorders: findings from the collaborative longitudinal personality disorders study. Journal of Nervous and Mental Disease 2002;190:510e8. Zanarini MC, Frankenburg FR, Reich DB, Marino MF, Lewis RE, Williams AA, et al. Biparental failure in the childhood experiences of borderline patients. Journal of Personality Disorders 2000;14:264e73. Zanarini MC, Gunderson JG, Marino MF, Schwartz EO, Frankenburg FR. Childhood experiences of borderline patients. Comprehensive Psychiatry 1989;30:18e25. Zanarini MC, Williams AA, Lewis RE, Reich RB, Vera SC, Marino MF, et al. Reported pathological childhood experiences associated with the development of borderline personality disorder. American Journal of Psychiatry 1997;154:1101e6. Zanarini MC, Yong L, Frankenburg FR, Hennen J, Reich DB, Marino MF, et al. Severity of reported childhood sexual abuse and its relationship to severity of borderline psychopathology and psychosocial impairment among borderline inpatients. Journal of Nervous and Mental Disease 2002;190:381e7. Zimmerman M. Diagnosing personality disorders: A review of issues and research methods. Archives of General Psychiatry 1994;51:225e45.