Borderline Personality Disorder Features and History of Childhood Maltreatment in Mothers Involved With Child Protective Services

Abstract

This study examines the history of childhood maltreatment and Borderline Personality Disorder (BPD) symptoms in mothers whose children were removed from the home by Child Protective Services (CPS) to identify potential targets for future intervention efforts. Forty-one mothers of children removed from the home due to abuse and/or neglect and 58 community-control mothers without CPS involvement were assessed for history of childhood maltreatment, alcohol and drug use, and BPD features. CPS-involved mothers scored significantly higher on measures of childhood maltreatment history and BPD features than did control mothers. The highest BPD scores were associated with the most severe histories of mothers’ childhood maltreatment. In total, 50% of CPS-involved mothers reported elevated BPD features, compared with 15% of control mothers. Further, 19% of CPS-involved mothers had self-reported scores consistent with a BPD diagnosis, compared with 4% of control mothers. BPD features rather than maltreatment history per se predicted maternal involvement with CPS, controlling for alcohol and drug use predictors. The present data suggest that evidence-based treatments to address BPD symptoms may be indicated for some CPS-involved parents.

Keywords

abusive parents child maltreatment child welfare services/child protection psychopathology

Factors associated with parents’ own childhood maltreatment that negatively affect their subsequent parenting have not been sufficiently evaluated (Berlin, Appleyard, & Dodge, 2011; Bert, Guner, & Lanzi, 2009). Borderline personality disorder (BPD) is strongly associated with childhood maltreatment history (Herman, Perry, & van der Kolk, 1989; Widom, Czaja, & Paris, 2009; Zanarini et al., 1997) and may have particular relevance for understanding parenting difficulties, given the predominance of interpersonal problems observed in this disorder. This article examines the association between parent’s history of abuse and neglect, parental BPD, and involvement in CPS due to substantiated child maltreatment in their offspring.

Parental History of Childhood Maltreatment and Risk of Maltreatment in the Offspring

Parents who have experienced maltreatment during their own childhood have been shown to have higher incidence of punitive parenting styles and discipline practices (Dixon, Brown, & Hamilton-Giachritsis, 2005; Dixon, Hamilton-Giachritsis, & Brown, 2005), decreased responsiveness to their children, and stronger propensities for abusive behavior (Bert et al., 2009). Although many parents with childhood maltreatment histories do not become perpetrators, parental experience of maltreatment in childhood is a risk factor for engaging in abusive behaviors, and mechanisms of such transmission require further clarification (Berlin et al., 2011; Bert et al., 2009; Milner et al., 2010). Given the increased risk of abusive behavior in parents who have experienced childhood maltreatment, it is not surprising that such parents also have elevated rates of CPS involvement (e.g., English, Marshall, & Orme, 1999; Haapasalo & Aaltonene, 1999; Marshall & English, 1999; Rittner, 2002).

Childhood Maltreatment History and BPD Features

BPD is characterized by affective dysregulation, behavioral dyscontrol, suicidal ideations and attempts, nonsuicidal self-harm behaviors, and significant interpersonal difficulties (American Psychiatric Association [APA], 2000). The association between BPD and history of childhood maltreatment has been well documented, ranging from 60% to 92% in retrospective studies of adults with BPD (e.g., Gabbard, 1996; Herman et al., 1989; Stone, 1981; Zanarini et al., 1997). Further, this relationship has been demonstrated in studies examining rates of BPD in adults with histories of childhood maltreatment (Grover et al., 2007), in epidemiological research (Afifi et al., 2011), and in prospective studies of abused and neglected children (Johnson, Cohen, Chen, Kasen, & Brook, 2006; Widom et al., 2009). Several mechanisms by which early maltreatment may result in personality pathology have been described. Early trauma and adversity has a long-term impact on core personality domains, including affect regulation, ability to tolerate stress and anxiety, as well as the ability to form and maintain secure relationships and a stable sense of identity (Cicchetti & Rogosch, 2001; Herman et al., 1989). However, it has been noted that the history of childhood maltreatment is neither necessary nor sufficient for explaining the etiology of BPD, and that BPD is best understood as the product of Gene by Environment interaction (Paris, 1998).

BPD, Parenting Difficulties, and Risk of Maltreatment in the Offspring

Difficulties associated with borderline psychopathology include several potential challenges to effective parenting. For example, research has indicated that parents with BPD tend to rapidly oscillate between intrusive and rejecting contact with their children (Newman & Stevenson, 2005). In addition, children of parents with BPD are more frequently exposed to parental substance abuse, neglect, suicide attempts, and family conflict (Feldman et al., 1995; Moos & Moos, 1986).

Further, mothers struggling with borderline pathology may have difficulty maintaining an emotionally and physically secure environment. For example, individuals with BPD have high rates of marital distress and domestic violence and tend to have partners who also have personality disorders (PDs; Bouchard, Sabourin, Lussier, & Villeneuve, 2009; Whisman & Schonbrun, 2009). Thus, even if mothers with BPD do not directly maltreat their children, they may still engage in a lifestyle characterized by interpersonal difficulties and high-risk impulsive behaviors that increase the likelihood that their children will be at risk for maltreatment. In particular, BPD features are highly associated with alcohol and substance use (Bornovalova et al., 2005), and the rates of substance use among parents involved with CPS are typically found to be 70–80% (Young, Boles, & Otero, 2007). Given the interaction of negative environmental experiences with the accompanying genetic vulnerability, it is not surprising that children of parents with BPD have high rates of comorbid psychiatric diagnoses, including depression, substance abuse, and antisocial behaviors, as well as cognitive and interpersonal problems (Abela, Skitch, Auerbach, & Adams, 2005; Barnow, Spitzer, Grabe, Kessler, & Freyberger, 2006; Goldman, D’Angelo, & DeMaso, 1993; Herr, Hammen, & Brennan, 2008).

To examine the association between parents’ own history of maltreatment, BPD, and maltreatment in the offspring, we investigated families receiving child protective services (CPS). Research on parental BPD in this population is very limited, and we found only three relevant studies on the topic. In a study of 13 CPS-involved mothers, Stanley and Penhale (1999) reported that 46% were diagnosed with having a PD. However, in addition to the very small sample size, this study did not specify the type of PD diagnosis, and it relied on retrospective diagnostic reports of psychiatrists and general practitioners. A more systematic examination by Bools, Neale, and Meadow (1994) indicated that of the 19 CPS-involved mothers who specifically smothered or poisoned their children to fabricate illness, 89% met criteria for at least one PD diagnosis, including 42% histrionic, 26% borderline, 11% dependent, and 5% avoidant PD. A larger study by Laporte (2009) examined 100 CPS-involved mothers and 30 CPS-involved fathers with confirmed or suspected mental disorders and found that 48% of the mothers had a PD and 29% were diagnosed with BPD. Although these studies generally support the notion of elevated rates of BPD in CPS-involved families, their conclusions limited by the lack of a comparison group. In order to assess the relevance of BPD in CPS-involved parents, it is necessary to demonstrate that its incidence is higher than that found in demographically comparable individuals.

In order to help clarify the role of maltreatment history and BPD in parents of maltreated children, this study evaluated mothers with children removed from the home by CPS for substantiated abuse and/or neglect, as compared to matched mothers of children without maltreatment. By comparing parents based on the maltreatment status of their children, this research strategy has the benefit of potentially controlling for confounds, such as relationship status, education, employment, income, psychopathology, and treatment seeking status. Specifically, we hypothesized that (1) CPS-involved mothers would have higher rates of personal history of childhood maltreatment as compared to community-control mothers, (2) CPS-involved mothers would have greater BPD features as compared to community-control mothers, and (3) BPD features would significantly predict CPS involvement, even after controlling for history of maltreatment, alcohol and drug use, and demographic variables (education and income).

Method

Participants

The sample included the first three cohorts of CPS-involved and control mothers and children that participated in a study examining genetic and environmental factors associated with risk and resiliency in children (e.g., Kaufman et al., 2004; Kaufman et al., 2006). Of the 130 mothers of children that enrolled in that research, 99 were included in the present investigation. Thirty-one of the CPS-involved mothers were unreachable (i.e., lost custody of a child and their whereabouts was unknown).

Participants included 41 mothers of children who had been removed from the home by CPS from the Connecticut Department of Children and Families due to reports of abuse and/or neglect, and 58 community-control women with no history of CPS involvement. Families recruited for the CPS group met the following criteria: (1) a child was removed from parental care due to a substantiated report of abuse or neglect within 6 months of the study onset; and (2) Department of Children and Families was awarded 96-hr temporary custody of the children by the courts. Eligible families were informed about the study by their caseworker, and interested parents signed a form consenting for research staff to contact them about the study. Thirty-nine percent of mothers involved with CPS were European American, 39% were African American, 17% were Hispanic, and 5% were other. Relationship status was reported to be 43% married or living with a partner, 20.0% divorced or separated, 22.5% single, 15% with a partner but not living together, and 2.5% widowed.

Eligible community-control families met the following criteria: (1) reported annual household income of $25,000 or less; and (2) as reported by birth mother and children and verified by the Department of Children and Families computerized record system, no contact with CPS and no history of child abuse, neglect, and/or exposure to domestic violence, which was confirmed in a separate interview with parents and children. Control subjects were recruited through targeted mailings and newspaper advertisements, and prospective subjects were screened for study inclusion by telephone. Forty-three community-control mothers were European American, 30% were African American, 21% were Hispanic, and 5% were other. Fifty percent were married or living together with a partner, 19.2% divorced or separated, 17.3% single, and 13.5% with a partner but not living together.

All participants signed written consent. The Yale University Human Investigations Committee and the Department of Children and Families Institutional Review Board approved the present investigation.

Demographic data were collected by interview with age, education, and income recorded as continuous variables, while relationship status was coded as a dichotomous variable—“in a stable relationship,” such as married, living together, or stable partner but not living together, and “not in a relationship,” such as divorced, separated, single, widowed). Mothers involved with CPS and community-control mothers did not differ significantly in work status (56.1% and 62.1% working), age (M = 33.7, SD = 6.3 and M = 35.3, SD = 7.4, respectively), or relationship status (55.0% and 63.5% in a relationship, respectively). However, CPS-involved mothers had lower education level than community-control mothers (M = 11.0, SD = 1.8 and M = 13.1, SD = 1.8, respectively; t(95) = 5.361, p < .001), and lower annual income (M = $14,659, SD = 11,310 and M = $20,173, SD = 8,633, respectively; t(92) = 2.674, p < .01).

CPS records indicated that 65.7% of removed children were emotionally abused, 90.0% were neglected, 75.7% were physically abused, and 31.4% were sexually abused. Further, the biological mothers were an identified perpetrator of at least one type of maltreatment in 88.6% of the cases, with 54.3% of children emotionally abused, 84.3% neglected, 40.0% physically abused, and 0% sexually abused by the biological mother. CPS records did not classify the remaining cases within the above categories. Reasons for children removal from maternal care in these cases could have included other types of abuse and neglect, substance use, domestic violence, or failure to protect.

Assessments

The Childhood Trauma Questionnaire (CTQ; Bernstein et al., 2003) was used for the self-report of mother’s history of child maltreatment. CTQ is a brief screening questionnaire to assess the history of childhood maltreatment and includes five subscales, including emotional, physical and sexual abuse, and emotional and physical neglect. CTQ consists of 28 items on a 5-point Likert-type scale with higher scores indicating greater severity of traumatic experiences. Moderate-to-severe cutoff scores per maltreatment category are as follows: ≥13 for emotional abuse, ≥10 for physical abuse, ≥8 for sexual abuse, ≥15 for emotional neglect, and ≥10 for physical neglect. CTQ scale has good reliability (α =. 91) and validity and empirical confirmation for the 5-factor model structure (Paivio & Cramer, 2004; Scher, Stein, Asmundson, McCreary, & Forde, 2001; Wright et al., 2001). In the present sample, Cronbach’s α was as follows: CTQ total α = .94, physical neglect α = .64, physical abuse α = .87, emotional abuse α = .86, emotional neglect α = .88, and sexual abuse α = .95. The distributions of the history of maltreatment scale (total CTQ) differed from normality. To reduce the extreme skewness of the CTQ (skewness = 1.21, SE = .25; Kolmogorov–Smirnov Z = 1.39, p < .05), log₁₀ transformation was applied, which made it approximately normal (skewness = 0.43, SE = .25; Kolmogorov–Smirnov Z = 0.92, p = .37).

The Personality Assessment Inventory Borderline Features Scale (PAI-BOR; Morey, 1991) is a 24-item self-report scale that assesses features associated with BPD and has demonstrated good reliability and validity. Items are rated by respondents on a 4-point Likert-type scale (1 = very true, 2 = mainly true, 3 = slightly true, and 4 = false). The average internal consistency across normative, college, and clinical groups is α = .88 (Morey, 1991; Trull, 1995). Raw scores on PAI-BOR between 28 and 37 (T scores 60–90) indicate clinically significant elevated BPD features, raw scores between 38 and 59 (T scores 70–91) suggest BPD if all subscales are also elevated, and raw scores greater than 60 (T score of 92) are associated with personality functioning within the borderline range (Morey, 1991). PAI-BOR consists of four subscales, including affective instability (e.g., “My mood can shift quite suddenly”), identity problems (“I often wonder what I should do with my life”), negative relationships (e.g., “My relationships have been stormy”), and self-harm (e.g., “When I’m upset, I typically do something to hurt myself.”), with cutoff scores for moderate elevation of 7, 9, 9, and 6, respectively (T score of 60; Morey, 1991). In the present sample, Cronbach’s α was as follows: PAI-BOR total score (that was used for the analyses in this study) α = .88, affective instability α = .68, negative relationship α = .63, identity problems α = .72, and self-injury α = .73.

The Michigan Alcohol Screening Test (MAST; Selzer, 1971) is a 25-item self-report screening tool for long-term alcohol-related problems. A total score of 13 is used as a clinically significant cutoff (Ross, Gavin, & Skinner, 1990). The MAST is a widely used measure and has been found to be a reliable (α = .86) and valid measure that correlates well with Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition; DSM-IV) diagnostic criteria (Conley, 2001; Maisto, Connors, & Allen, 1995; Marion, Fuller, Johnson, Michels, & Diniz, 1996; Teitelbaum & Carey, 2000). In the present sample, Cronbach’s α for MAST was .86.

The Drug Abuse Screening Test (DAST; Skinner, 1982) is a 20-item drug screening tool to assess drug-related consequences of use and abuse for 12 general categories of drugs, not including alcohol. The total score reflects problems with generic “drug use” during the past 12 months. Scores range from 0 to 20, with higher scores reflecting greater drug use-related consequences. Generally, 6–10 is used as a cutoff for “moderate” problems, 11–15 as “substantial problems,” and 16–20 as “severe problems.” DAST has good reliability (α = .92) and validity (Cocco & Carey, 1994; Gavin, Ross, and Skinner, 1989; Skinner, 1982; Staley & El-Guebaly, 1990). In the present sample, Cronbach’s α for DAST was .96.

Alcohol use scale (total MAST; skewness = 3.09, SE = .25) and drug use scale (total DAST; skewness = 1.83, SE = .25) were not normally distributed and failed to normalize even with transformation. We tested the models with nontransformed MAST and DAST scores and, to examine whether the results would replicate, we also ran the regression models re-coding the MAST and DAST as dichotomous variables using their clinical cutoff scores (MAST ≥ 13 and DAST ≥ 6).

Procedures

Participants underwent interviews at their current place of residence in two interview sessions, including additional interview and self-report measures that were not used in the present study (see details in Kaufman et al., 2004 and 2006). The first session began with discussion of informed consent, including responding to any questions, before parents signed written consent forms. Parents received $25 as compensation for their participation at the end of the first interview and $15 for their participation after completing the second interview.

Data Analysis

Prior to analyses, all variables were examined for the accuracy of data entry, and adherence of distributions to the assumptions of multivariate analyses was checked using SPSS 17.0 for Windows. Hypotheses 1 and 2 regarding group differences on history of childhood maltreatment and BPD features, respectfully, were examined using χ² and analysis of covariance. Hierarchical logistic regression analysis was performed to test Hypothesis 3 involving prediction of CPS-involved versus community-control group status from BPD features, controlling for the history of maltreatment, alcohol and drug use, and certain demographic variables (i.e., education level and annual income). These demographic variables were included in the analyses as possible confounding variables, as these characteristics differed significantly between groups despite our efforts to match groups. We utilized Holm-Bonferroni corrections to adjust the α level downward to prevent chance capitalization in multiple comparisons. The Holm-Bonferroni sequential approach allows for increased power while controlling for Type I error (Abdi, 2010).

Results

History of Childhood Maltreatment in CPS-Involved and Community-Control Mothers

As compared to community-control mothers (M = 39.40, SD = 13.40), CPS-involved mothers reported significantly higher maltreatment scores (M = 54.03, SD = 23.63), F(1,83) = 10.80, p < .001. CPS mothers, as compared to community-control mothers, were more likely to report maltreatment at the moderate-to-severe levels. Specifically, CPS-involved mothers, as compared to community-control mothers, reported more severe levels of emotional abuse (44.7% vs. 16.7%, χ²(1) = 8.7, p < .01), physical abuse (44.7% vs. 20.4%, χ²(1) = 6.3, p < .05), sexual abuse (50.0% vs. 25.0%, χ²(1) = 5.6, p < .05), emotional neglect (36.8% vs. 18.5%, χ²(1) = 3.9, p < .05), and physical neglect (39.5% vs. 7.4%, χ²(1) = 13.9, p < .001).

BPD Features in CPS-involved mothers/community-control mothers

As predicted, mothers involved with CPS were more likely to have clinically elevated BPD features (50%), as compared with control mothers (14.5%), χ²(1) = 13.7, p < .001, and 19.4% of the CPS mothers had self-reported scores consistent with a BPD diagnosis, as compared with 3.6% of the control mothers χ²(1) = 6.1, p < .02. When controlling for SES covariates (annual income and education), CPS mothers scored significantly higher on BPD features (M = 29.19, SD = 13.27), as compared with control mothers (M = 19.74, SD = 8.24), F(1,78) = 14.40, p < .000.

Predicting Group Status in CPS-Involved and community-control Mothers

Preliminary analyses indicated that correlation coefficients between predictors were low to moderate (from r = .04 to r = .58), except for substance use variables (r = .72). Variance inflation factors (VIFs) were calculated before inclusion of substance use variables in the analyses. VIFs for drug and alcohol use in the tested model were ≤5 (VIF for alcohol use = 1.85; VIF for drug use = 2.27), indicating that multicollinearity did not inflate standard errors. Therefore, both variables were left in the regression model.

BPD features, history of maltreatment, and alcohol and drug use significantly predicted group status when tested separately using logistic regression analyses (see Table 1). To test the unique predictive significance of the two variables of interest (the history of maltreatment and BPD features) on group status (CPS vs. control), we performed hierarchical regression that included history of maltreatment, the alcohol and drug use predictors, and SES covariates (education level and annual income) in Step 1 and BPD feature scores in Step 2. The Hosmer-Lemeshow statistic of 0.64 implied a good fit of Model 1 that included variables entered in Step 1, with the model explaining 45% of variance and classifying 77.2% of the cases correctly. Results indicated that while maltreatment history and drug and alcohol use predicted group status in separate regression analyses, when entered together, only drug use remained a significant predictor (see Table 2). We would also like to note that education level was a significant predictor of group status.

Table 1.

Separate Logistic Regressions Predicting Group Status (CPS-Involved vs. Control Mothers) From Personality Assessment Inventory Borderline Features Total Scores, Maltreatment History, and Alcohol and Substance Use Measures

Predictor	B	SE B	e^B	χ² (df = 1)	Nagelkerke R ²
Maltreatment history	5.305^***	1.470	201.430	15.732***	.212
Drug use	.177^**	.053	1.193	18.891***	.225
Alcohol use	.109^*	.052	1.115	5.1718*	.079
BPD features	.082***	.023	1.085	17.327***	.229

Note. After adjustment of the alpha levels for the total number of comparisons, using Holm-Bonferroni correction, alcohol use was no longer a significant predictor of group status.

p < .05. **p < .01. ***p < .001.

Table 2.

Hierarchical Logistic Regression Model Predicting Group Status (CPS-Involved vs. Control Mothers) From Personality Assessment Inventory Borderline Features Total Scores, Controlling for Maltreatment History, Alcohol and Substance Use Measures and SES Variable (N = 79)

	Model 1			Model 2
Predictor	B	SE B	e^B	B	SE B	e^B
Education	−.953^*	.378	.386	−1.083**	.417	.339
Income	.000	.000	1.000	.000	.000	1.000
Maltreatment history	3.443	1.844	31.278	2.604	1.970	13.518
Alcohol use	−.051	.081	.950	−.091	.084	.913
Drug use	.154*	.077	1.167	.122	.072	1.130
BPD features				.079*	.038	1.083
Constant	−2.782	3.138	.062	−2.933	3.212	.053
χ ²	31.973***			37.087***
df	5			6
Nagelkerke R ²	.449			.506

*p < .05. **p < .01. ***p < .001.

Model 2 included all variables of interests and SES covariates entered at Step 2. The Hosmer-Lemeshow statistic of 0.60 implied a good fit of the model, explaining 51% of variance, and classifying 79.7% of the cases correctly. Results indicated that BPD features significantly improved prediction of the group status when all other covariates were accounted for. Maltreatment history and alcohol and drug use were not significant predictors, while education level remained significant (see Table 2). Testing both models with alcohol and drug use scales dichotomized at the clinical cutoff point yielded the same pattern of results.

Discussion

The main aims of this study were to evaluate the history of childhood maltreatment and the prevalence of BPD in mothers of children receiving CPS, as compared to community-control mothers, and to examine the relative predictive significance of borderline pathology and parental history of maltreatment on child maltreatment status, as indicated by CPS involvement. Consistent with study hypotheses, CPS-involved mothers had higher rates of childhood maltreatment relative to control mothers. In addition, 50% of the CPS mothers reported clinically significant BPD features, including nearly one in five reporting symptoms consistent with a BPD diagnosis. These findings were consistent with prior research showing that about 29% of mothers involved with CPS meet the BPD diagnosis (Bools, Neale, & Meadow, 1994; Laporte, 2009) and, consistent with study expectations, were significantly higher than those observed in the community-control mothers in the present study. Only about 3.6% of the community-control mothers reported symptoms consistent with a BPD diagnosis, which is within the expected range of 2% to 5.9% of the general population (Grant et al., 2008; Paris, 1999).

History of childhood maltreatment predicted BPD features and diagnostic status in both CPS-involved and control mothers. CPS-involved mothers reported significantly higher rates of childhood maltreatment history than the community controls, which is generally consistent with the findings from the prior research comparing mothers with and without CPS involvement (e.g., Haapasalo & Aaaltonen, 1999). However, as hypothesized, results of this study indicated that BPD features rather than maltreatment history or substance use, uniquely contributed to predicting maternal involvement with CPS. Although the history of maltreatment and alcohol and drug use predicted group status when tested separately, BPD features predicted group status above and beyond these factors. This finding provides a potential direction for developing models of intervention. Caregivers' history of maltreatment is not a modifiable risk factor; however, a growing body of treatment outcome research indicates that trauma-related psychopathology including BPD features is amenable to treatment.

It is noted that current services for parents involved with CPS are typically limited to parenting skills training, child development–oriented programs, home visiting activities (e.g., prenatal programs), and assistance with meeting basic needs (Chaffin, Bonner, & Hill, 2001). Treatment does not typically include services that address parental emotional and behavioral problems, particularly the interpersonal difficulties identified in this study, which might explain the limited treatment benefits of these programs as indicated by high recidivism (Casanueva, Martin, Runyan, Barth, & Bradley, 2008; Chaffin et al., 2001). Although there have been recent efforts to improve access to effective substance use interventions for CPS-involved parents (Oliveros & Kaufman, 2011), it is noted that the alcohol and drug difficulties did not predict CPS-involvement beyond BPD features and education. Based on our findings, the underlying common phenomenology between these disorders, (e.g., emotion dysregulation and impulsivity) may represent a more efficient target for interventions for CPS-involved parents.

Several interventions have demonstrated efficacy for treating BPD in randomized controlled trials and might be considered for intervening with mothers with histories of maltreatment and BPD features. These include dialectical behavior therapy (Linehan et al., 2006), transference-focused psychotherapy (TFP; Clarkin, Yeomans, & Kernberg, 1999), mentalization-based treatment (MBT; Fonagy & Bateman, 2007), systems training for emotional predictability and problem solving (STEPPS; Blum et al., 2008), dynamic deconstructive psychotherapy (DDP; Gregory et al., 2008), and schema-focused therapy (SFT; Young, Klosko, & Weishaar, 2003). Insight-oriented therapies (e.g., TFP, MBT, DDP, SFT) may have an advantage for exploring the relational consequences of abuse experiences, while the more behavioral interventions (e.g., DBT, STEPPS) may have an advantage in improving the behavioral control necessary for effective parenting. However, a recent meta-analysis of DBT efficacy in BPD determined that the largest effects from treatment were found in the area of social adjustment (ES 1.09; Kroger & Kosfelder, 2007), suggesting that DBT might also effectively address relational difficulties. Although all the above mentioned treatments have empirical support, there is no evidence that one psychotherapy is superior to another (Leichsenring, Leibing, Kruse, New, & Leweke, 2011). Further, to date, none of these interventions have been systematically examined in a child welfare population.

The results of this study should be considered in light of its limitations. Our sample was derived from a larger investigation on the Gene and Environment interaction of depression in children and only assessed the mothers; therefore, the present results do not speak to risk factors for CPS involvement in fathers. Although all of the mothers of participating control children were included in this study, we were unable to reach some of the mothers of children receiving CPS. Although it was not possible to evaluate whether participating and not participating CPS mothers differed on the variables of interested, it is not likely that mothers whose whereabouts was unknown would have less severe personal history of maltreatment and trauma-related psychopathology. Further, despite our efforts to match groups on the socioeconomic status, there were significant differences between control mothers and mothers involved with CPS in education level and annual income, consistent with past findings on parents at risk for perpetrating maltreatment (Lohman, Pittman, Coley, & Chase-Lansdale, 2004). However, we statistically controlled for these confounds.

Further, the study design was cross sectional, which does not usually allow for assessing time order of exposure and effect. At the same time, we can safely assume that maternal history of childhood maltreatment occurred before maternal involvement with CPS. Yet, this technique is weaker in establishing time order than through observations using longitudinal designs and does not allow for meditational analyses of causal pathways. While a strength of the study involved examining the role of BPD features in predicting CPS involvement in the context of other potentially related risk factors (i.e., drug and alcohol use), it is noted that the present investigation did not include additional diagnostic categories that could also affect the relationship of maltreatment history and CPS involvement, such as depression, posttraumatic stress disorder, or other PD diagnoses such as antisocial personality disorder. It is recommended that such constructs be included in replications of the present findings. Further, the alcohol and drug use data were not normally distributed, thus potentially affecting results. Finally, all measures were also self-report and, thus, subject to potential bias. Self-report measures were retrospective and also were taken after children were removed, thus the stress of the investigation, removal, court proceedings, and other associated issues may have influences maternal symptom report.

The findings of this study suggest that BPD features may be associated with negative outcomes for mothers. Understanding the mechanisms by which this risk is increased may inform more targeted treatment interventions in at-risk groups. Further research may examine such factors, as emotion dysregulation, identity instability, and dysfunctional interpersonal processes, as possible moderating or mediating variables.

Footnotes

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abdi

(2010). Holm’s sequential Benferroni procedure. In Salkind

(Ed.), Encyclopedia of Research Design (pp. 1–8). Thousand Oaks, CA: Sage.

Abela

J. R. Z.

Skitch

S. A.

Auerbach

R. P.

Adams

(2005). The impact of parental borderline personality disorder on vulnerability to depression of children of affectively ill parents. Journal of Personality Disorders, 19, 68–83.

Afifi

T. O.

Mather

Boman

Fleisher

Enns

M. W.

Macmillan

Sareen

(2011). Childhood Adversity and Personality Disorders: Results from a nationally representative population-based study. Journal of Psychiatric Research, 45, 814–822.

American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (4th ed., Text Revised). Washington, DC: Author.

Barnow

Spitzer

Grabe

H. J.

Kessler

Freyberger

H. J.

(2006). Individual characteristics, familial experience, and psychopathology in children of mothers with borderline personality disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 45, 965–972.

Berlin

L. J.

Appleyard

Dodge

K. A.

(2011). Intergenerational continuity in child maltreatment: Mediating mechanisms and implications for prevention. Child Development, 82, 162–176.

Bernstein

D. P.

Stein

J. A.

Newcomb

M. D.

Walker

Pogge

Ahluvalia

Zule

. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse and Neglect, 27, 169–190.

Bert

S. C.

Guner

B. M.

Lanzi

R. G.

(2009). The influence of maternal history of abuse on parenting knowledge and behavior. Family Relations, 58, 176–187.

Blum

Pfohl

St.John

Pfohl

Stuart

McCormick

Black

D. W.

(2008). Systems training for emotional predictability and problem solving (STEPPS) for outpatients with borderline personality disorder: A randomized controlled trial and 1-year follow-up. American Journal of Psychiatry, 165, 468–478.

10.

Bools

Neale

Meadow

(1994). Munchausen syndrome by proxy: A study of psychopathology. Child Abuse & Neglect, 18, 773–788.

11.

Bornovalova

Lejuez

Daughters

Rosenthal

Lynch

(2005). Impulsivity as a common process across borderline personality and substance use disorders. Clinical Psychology Review, 25, 790–812.

12.

Bouchard

Sabourin

Lussier

Villeneuve

(2009). Relationship quality and stability in cuples when one partner sufferes from Borderline Personality Disorder. Journal of Marital and Family Therapy, 35, 446–455.

13.

Casanueva

Martin

S. L.

Runyan

D. K.

Barth

R. P.

Bradley

R. H.

(2008). Parenting services for mothers involved with child protective services: Do they change maternal parenting and spanking behaviors with young children? Children and Youth Services Review, 30, 861–878.

14.

Chaffin

Bonner

B. L.

Hill

R. F.

(2001). Family preservation and family support programs: Child maltreatment outcomes across client risk levels and program types. Child Abuse and Neglect, 25, 1269–1289.

15.

Cicchetti

Rogosch

(2001). Diverse patterns of neuroendocrine activity in maltreated children. Development and Psychopathology, 13, 677–693.

16.

Clarkin

J. F.

Yeomans

F. E.

Kernberg

O. F.

(1999). Psychotherapy for borderline personality. New York, NY: Wiley.

17.

Cocco

K. M.

Carey

K. B.

(1994), Psychometric properties of the Drug Abuse Screening Test in psychiatric outpatients. Psychological Assessment, 10, 408–414.

18.

Conley

T. B.

(2001). Construct validity of the MAST and AUDIT with multiple offender drunk drivers. Journal of Substance Abuse Treatment, 20, 287–295.

19.

Dixon

Brown

Hamilton-Giachritsis

(2005). Risk factors of parents abused as children: A mediational analysis of the intergenerational continuity of child maltreatment (Part I) Journal of Child Psychology and Psychiatry, 46, 47–57.

20.

Dixon

Hamilton-Giachritsis

Brown

(2005). Attributions and behaviors of parents abused as children: A mediational analysis of the intergenerational continuity of child maltreatment (Part II) Journal of Child Psychology and Psychiatry, 46, 58–68.

21.

English

D. J.

Marshall

D. B.

Orme

(1999). Characteristics of repeated referrals to child protective services in Washington State. Child Maltreatment, 4, 297–307.

22.

Feldman

R. B.

Zelkowitz

Weiss

Heyman

Vogel

Paris

(1995). A comparison of the families of borderline personality disorder mothers and the families of other personality disorder mothers. Comprehensive Psychiatry, 36, 157–163.

23.

Fonagy

Bateman

A. W

(2007). Mentalizing and borderline personality disorder. Journal of Mental Health, 16, 83–101.

24.

Gabbard

G. O.

(1996). Psychotherapeutic strategies for borderline personality disorders. In The Hatherleigh guide to psychiatric disorders (pp. 119–135). New York, NY: Hatherleigh Press.

25.

Gavin

D. R.

Ross

H. E.

Skinner

H. A.

(1989). Diagnostic validity of the drug abuse screening test in the assessment of DSM-III drug disorders. British Journal of Addiction, 84, 301–307.

26.

Goldman

S. J.

D’Angelo

E. J.

DeMaso

M. D.

(1993). Psychopathology in the families of children and adolescents with borderline personality disorder. American Journal of Psychiatry, 150, 1832–1835.

27.

Grant

B. F.

Chou

S. P.

Goldstein

R. B.

Huang

Stinson

F. S.

Saha

T. D.

Ruan

W. J.

(2008). Prevalence, correlates, disability, and comorbidity of DSM-IV borderline personality disorder: Results from the Wave 2 National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry, 69, 533–545.

28.

Gregory

R. J.

Chlebowski

Kang

Remen

A. L.

Soderberg

M. G.

Stepkovitch

Virk

(2008). A controlled trial of psychodynamic psychotherapy for co-occurring borderline personality disorder and alcohol use disorder. Psychotherapy: Theory, Research, Practice, Training, 45, 28–41.

29.

Grover

K. E.

Carpenter

L. L.

Price

L. H.

Gagne

G. G.

Mello

A. F.

Mello

M. F.

Tyrka

A. R.

(2007). The relationship between childhood abuse and adult personality disorder symptoms. Journal of Personality Disorders, 21, 442–447.

30.

Haapasalo

Aaltonene

(1999). Mothers’ abusive childhood predicts child abuse. Child Abuse Review, 8, 231–250.

31.

Herman

J. L.

Perry

J. C.

van der Kolk

B. A.

(1989). Childhood trauma in borderline personality disorder. American Journal of Psychiatry, 146, 490–495.

32.

Herr

N. R.

Hammen

Brennan

P. A.

(2008). Maternal borderline personality disorder symptoms and adolescent psychosocial functioning. Journal of Personality Disorders, 22, 451–465.

33.

Johnson

J. G.

Cohen

Chen

Kasen

Brook

J. S.

(2006). Parenting behaviors associated with risk for offspring personality disorders during adulthood. Archives of General Psychiatry, 63, 579–587.

34.

Kaufman

Yang

B. Z.

Douglas-Palumberi

Houshyar

Lipschitz

Krystal

Gelerner

(2004). Social supports and serotonin transporter gene moderate depression in maltreated children. Proceedings of the National Academy of Science USA, 101, 17316–17321.

35.

Kaufman

Yang

B. Z.

Douglas-Palumberi

Grasso

Lipschitz

Houshyar

Gelerner

(2006). Brain-derived neurotropic factor-5-HTTLPR gene interactions and environmental modifiers of depression in children. Biological Psychiatry, 59, 673–680.

36.

Kroger

Kosfelder

(2007). Eine Meta-Analyse zur Wirksamkeit der Dialektisch Behavioralen Therapie bei Borderline-Personlichkeitsstorungen. Zeitschrift fur Klinische Psychologie und Psychotherapie: Forschung und Praxis, 36, 11–17.

37.

Laporte

(August 2009). Children involved with youth protection services: The additional adversity of maternal BPD. Paper presented at the 11th International Congress of the International Society for the Study of Personality Disorders, New York, NY.

38.

Leichsenring

Leibing

Kruse

New

A. S.

Leweke

(2011). Borderline personality disorder. The Lancet, 377, 74–84.

39.

Linehan

M. M.

Comtois

K. A.

Murray

A. M.

Brown

M. Z.

Gallop

R. J.

Heard

H. L.

Lindenboim

(2006). Randomized controlled trial and follow-up of dialectical behavior therapy vs therapy by experts for suicidal behaviors and borderline personality disorder. Archives of General Psychiatry, 63, 757–766.

40.

Lohman

B. J.

Pittman

L. D.

Coley

R. L.

Chase-Lansdale

P. L.

(2004). Welfare history, sanctions, and developmental outcomes among low-income children and youth. Social Service Review, 78, 41–73.

41.

Maisto

S. A.

Connors

G. J.

Allen

J. P.

(1995). Contrasting self-report screens for alcohol problems: A review. Alcoholism: Clinical and Experimental Research, 19, 1510–1516.

42.

Marion

L. N.

Fuller

S. G.

Johnson

N. P.

Michels

P. J.

Diniz

(1996). Drinking problems in nursing students. Journal of Nursing Education, 35, 196–203.

43.

Marshall

D. B.

English

D. J.

(1999). Survival analysis of risk factors for recidivism in child abuse and neglect. Child Maltreatment, 4, 287–296.

44.

Milner

J. S.

Thomsen

C. J.

Crouch

J. L.

Rabenhorst

M. M.

Martens

P. M.

Dyslin

C. W.

Merrill

L. L.

(2010). Do trauma symptoms mediate or moderate the association between a childhood history of physical abuse and adult child abuse risk? Child Abuse & Neglect, 34, 332–344.

45.

Moos

(1986). The Family Environment scale manual (2nd ed.). Palo Alto, CA: Consulting Psychologists Press.

46.

Morey

(1991). The personality assessment inventory professional manual. Odessa, FL: Psychological Assessment Resources.

47.

Newman

Stevenson

(2005). Parenting and Borderline Personality Disorder: Ghosts in the nursery. Clinical Child Psychology and Psychiatry, 10, 385–394.

48.

Oliveros

Kaufman

(2011). Addressing substance abuse treatment needs of parents involved with the child welfare system. Child Welfare, 90, 25–41.

49.

Paivio

S. C.

Cramer

K. M.

(2004). Factor structure and reliability of the Childhood Trauma Questionnaire in a Canadian undergraduate student sample. Child Abuse and Neglect, 28, 889–904.

50.

Paris

(1998). Does childhood trauma cause personality disorder in adults? Canadian Journal of Psychiatry, 43, 148–153.

51.

Paris

(1999). Borderline personality disorder. In Millon

Blaney

Davis

(Eds.), Oxford textbook of psychopathology (pp. 628–652). New York, NY: Oxford University Press.

52.

Rittner

(2002). The use of risk assessment instruments in child protective services case planning and closures. Children and Youth Services Review, 24, 189–207.

53.

Ross

H. E.

Gavin

D. R.

Skinner

H. A.

(1990). Diagnostic validity of the MAST and the Alcohol Dependence Scale in the assessment of DSM-III alcohol disorders. Journal of Studies on Alcohol, 51, 506–513.

54.

Scher

C. D.

Stein

M. B.

Asmundson

G. J.

McCreary

D. R.

Forde

D. R.

(2001). The childhood trauma questionnaire in a community sample: psychometric properties and normative data. Journal of Trauma and Stress, 14, 843–857.

55.

Selzer

M. L.

(1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127, 1653–1658.

56.

Skinner

H. A.

(1982). The drug abuse screening test. Addictive Behaviors, 7, 363–371.

57.

Staley

El-Guebaly

(1990). Psychometric properties of the drug abuse screening test in a psychiatric patient population. Addictive Behaviors, 15, 257–264.

58.

Stanley

Penhale

(1999). The mental health problems of mothers experiencing the child protection system: Identifying needs and appropriate responses. Child Abuse Review 8, 34–45.

59.

Stone

M. H.

(1981). Psychiatrically ill relatives of borderline patients: A family study. Psychiatric Quarterly, 58, 71–83.

60.

Teitelbaum

L. M.

Carey

K. B.

(2000). Temporal stability of alcohol screening measures in psychiatric setting. Psychology of Addictive Behaviors, 14, 401–404.

61.

Trull

T. J.

(1995). Borderline personality disorder features in nonclinical young adults: I. Identification and validation. Psychological Assessment, 7, 33–41.

62.

Whisman

M.A.

Schonbrun

Y.C.

(2009). Social consequences of Borderline Personality Disorder symptoms in a population-based survey: Marital distress, marital violence, and marital disruption. Journal of Personality Disorders, 23, 410–415.

63.

Widom

C. S.

Czaja

S. J.

Paris

(2009). A prospective investigation of borderline personality disorder in abused and neglected children followed up into adulthood. Journal of Personality Disorders, 23, 433–446.

64.

Wright

K. D.

Asmundson

G. J.

McCreary

D. R.

Scher

Hami

Stein

M. B.

(2001). Factorial validity of the childhood trauma questionnaire in men and women. Depression and Anxiety, 13, 179–183.

65.

Young

N. K.

Boles

S. M.

Otero

(2007). Parental substance use disorders and child maltreatment: Overlap, gaps, and opportunities. Child Maltreatment, 12, 137–149.

66.

Young

J. E.

Klosko

J. S.

Weishaar

M. E.

(2003). Schema therapy: A practitioner’s guide. New York, NY: Guilford.

67.

Zanarini

M. C.

Williams

A. A.

Lewis

R. E.

Reich

R. B.

Vera

S. C.

Marino

M. F.

Frankenburg

F. R.

(1997). Reported pathological childhood experiences associated with the development of borderline personality disorder. The American Journal of Psychiatry, 154, 1101–1106.