© 2007 Springer Publishing Company 577
Violence and Victims, Volume 22, Number 5, 2007
Family Environment and Adult
Attachment as Predictors of
Psychopathology and Personality
Dysfunction Among Inpatient
Shelley A. Riggs, PhD
University of North Texas
Gayla Sahl, PhD
Ellen Greenwald, MA
Heather Atkison, MA
Adrienne Paulson, MA
Texas Woman’s University
Colin A. Ross, MD
Timberlawn Psychiatric Hospital
The current study explored the role of early family environment and adult attachment style
in explaining long-term outcomes among child abuse survivors. Adult patients (N = 80)
in a trauma treatment program were assessed for clinical diagnosis and administered a
multiscale questionnaire. Hierarchical regression analyses were significant for dissociative
identity disorder (DID), substance abuse, anxiety disorder, posttraumatic stress, somatiza-
tion, and six personality disorder dimensions. Adult attachment styles were significant
predictors of most outcome variables. Of particular note was the strong contribution
of attachment avoidance to DID. Five family environment scales (Independence, Orga-
nization, Control, Conflict, Expressiveness) also contributed to various psychopathologi-
cal outcomes. Evidence emerged supporting a mediating role for attachment style in the
link between family independence and five personality disorder dimensions.
Keywords: child abuse; family environment; attachment style; psychopathology
dysfunction. Yet, not all abuse survivors suffer negative consequences, and those who do
demonstrate a broad variety of clinical symptoms and personality disturbance. Significant
heterogeneity within this population suggests that other factors, such as abuse characteristics
and pathogenic early home environments, contribute to diverse developmental trajectories
large body of evidence consistently indicates that adult survivors of physical,
sexual, or psychological abuse in childhood are more likely than adults without
a history of abuse to experience a wide range of psychological and personality
578 Riggs et al.
following abuse experiences. The research of Alexander and her colleagues (1992, 1993;
Alexander et al., 1998; Anderson & Alexander, 1996) suggests that adult attachment style
may mediate the association of child abuse with later distress, thus accounting for some of
the variability in outcomes. The current study extends past research to an inpatient popula-
tion and is the first to explore the combined contributions of early family characteristics
and adult attachment style to specific clinical syndromes and personality disturbance.
CHILDHOOD ABUSE, FAMILY BACKGROUND,
AND ADULT OUTCOMES
A variety of negative long-term consequences of child abuse have been reported, including
low self-esteem, interpersonal problems, and increased rates of depression, anxiety, sub -
stance abuse, somatization, and personality disorders (Jumper, 1995; Molnar, Buka, & Kessler,
2001; Mullen, Martin, Anderson, Romans, & Herbison, 1996; Polusny & Follette, 1995).
Although mixed, there is growing evidence that different types of abuse are associated
with specific symptomatology (Briere & Runtz, 1990; Engels, Moisan, & Harris, 1994;
Molnar et al., 2001). Moreover, compared to single forms of abuse, compound trauma
involving both sexual and physical abuse is specifically associated with depression, dis-
ordered eating, dissociation, posttraumatic stress symptoms, and psychiatric comorbidity
(Krupnick et al., 2004; see Higgins & McCabe, 2001, for review).
Because protection is the primary biological and evolutionary function of family care-
givers, childhood abuse may be especially traumatic when it occurs within the context
of family attachment relationships (Allen, 2001; Bowlby, 1980). However, while some
studies support the expectation that a closer family relationship to the perpetrator is asso-
ciated with increased levels of maladjustment and symptomatology, other studies report
no differences in outcome between survivors of intrafamilial or extrafamilial abuse (see
Tyler, 2002, for review). The literature also indicates that parental psychopathology and
parental substance abuse are associated with increased risk for later psychological dif-
ficulties (Benedict & Zautra, 1993; Trull, 2001; Yama, Tovey, Fogas, & Morris, 1995).
These parental problems, in turn, increase the risk of dysfunctional family environments
(Ellis & Zucker, 1997; Mullen et al., 1996) and may influence outcomes indirectly through
associations with disturbed parent–child interactions and increased levels of conflict and
disorganization within the family (Cummings & Davies, 1994; Rutter & Quinton, 1984).
Families with a history of physical or sexual abuse are generally characterized as less
supportive, less organized, and more isolated. They may also demonstrate low levels of
independence and high levels of control (Justice & Calvert, 1990; Moos & Moos, 2002).
Low family cohesion and high conflict and/or control has been linked to depression, anxi-
ety, and posttraumatic stress symptoms in college and community samples of child abuse
survivors (Kamsner & McCabe, 2000; Myerson, Long, Miranda, & Marx, 2002; Yama,
Tovey, & Fogas, 1993). In samples unselected for child abuse, dependent personality
disorder was related to high levels of family control and low levels of expressiveness and
independence; histrionic personality disorder was associated with low cohesion and high
control (Baker, Capron, & Azorlosa, 1996; Head, Baker, & Williamson, 1991). High fam-
ily conflict and disorganization may characterize the early family experiences of patients
diagnosed with borderline personality disorder (Ludolph et al., 1990). Some researchers
have suggested that family background factors are better predictors of psychological out-
comes than abuse-specific variables (Higgins & McCabe, 2003; Mullen et al., 1996; Nash,
Family, Attachment, and Psychopathology 579
Hulsey, Sexton, Harralson, & Lambert, 1993); however, other research indicates no strong
pattern of associations (Harter & Vanecek, 2000; Polusny & Follette, 1995). More research
is needed to clarify the role of family environment in the development of specific Axis I
and II disorders among adults with histories of childhood abuse.
Although scarce, evidence is emerging to support theoretical links between adult attachment
security and healthy early family environments (Diehl, Elnick, Bourbeau, & Labouvie-Vief,
1998; Mikulincer & Florian, 1999; Pfaller, Kiselica, & Gerstein, 1998). Adult romantic
attachment style is presumed to be an outgrowth of early attachment organization, and
research has shown that adult attachment insecurity is related to memories of unavailable and
nonsupportive parenting (Brennan & Shaver, 1998; Mickelson, Kessler, & Shaver, 1997).
Researchers have described adult romantic attachment as comprising two polar dimensions
of anxiety and avoidance (Brennan, Clark, & Shaver, 1998), which form four quadrants or
categories of attachment. According to Bartholomew and Horowitz (1991), secure individu-
als show a balance between a healthy connection to others and self-reliance; they tend to
have a positive self-image (low anxiety) and demonstrate trust and open communication in
relationships (low avoidance). In contrast, preoccupied adults question their self-worth and
fear abandonment (high anxiety), whereas dismissing–avoidant adults distrust others and
minimize or shun interpersonal intimacy (high avoidance). Fearful–avoidant adults experi-
ence high levels of both attachment anxiety and avoidance, which may contribute to ongo-
ing struggles regarding approach/avoidance behaviors in personal relationships (Simpson &
Rholes, 2002), a conflict frequently observed in clinical work with abuse survivors.
Insecure models of attachment, while not synonymous with psychopathology, create
a risk for the development of psychological problems. Preoccupied adult attachment style
is characterized by hyperactivating strategies of coping and emotional regulation (Miku-
lincer & Shaver, 2003) and has been linked to low self-control and tolerance, interpersonal
dependence/reliance, as well as histrionic, dependent, and borderline personality traits
(Allen, Coyne, & Huntoon, 1998; Brennan & Shaver, 1998; Diehl et al., 1998; Onishi,
Gjerde, & Block, 2001). In contrast, dismissing–avoidant attachment is characterized by
deactivating strategies (Mikulincer & Shaver, 2003) and in nonclinical samples has been
associated with substance abuse, somatization, and repressive tendencies (Mickelson
et al., 1997; Mikulincer, Florian, & Weller, 1993; Onishi et al., 2001), with possible links
suggested to dissociation and narcissistic, schizoid, antisocial, paranoid, and obsessive–
compulsive personality traits (e.g., Blatt & Levy, 2003). Findings in nonclinical samples
also have documented significant relationships between fearful–avoidant attachment and
depression, somatic anxiety, substance abuse, dissociation, and paranoid, schizoid, schizo-
typal, avoidant, self-defeating, borderline, narcissistic, and obsessive–compulsive person-
ality traits (Anderson & Alexander, 1996; Brennan & Shaver, 1998; Brennan, Shaver, &
Tobey, 1991). Among inpatients, Allen et al. (1998) reported that high attachment anxiety
and an inability to depend (i.e., fearful avoidance) were related to paranoid, schizotypal,
and borderline personality.
Recently, researchers have become interested in the role of the attachment system in
long-term outcomes of abuse survivors. Retrospective reports of maltreatment in childhood
are linked to insecure romantic attachment among undergraduates (Gauthier, Stollak,
Messe, & Aronoff, 1996; Roche, Runtz, & Hunter, 1999; Styron & Janoff-Bulman, 1997).
580 Riggs et al.
For example, Roche et al. (1999) reported that a negative model of other/attachment
avoidance was related to depression and trauma symptomatology (e.g., defensive avoid-
ance, dissociation). Negative view of self/attachment anxiety has been linked to symptoms
of depression, anxiety, and traumatic stress in college and community samples (Muller,
Lemieux, & Sicoli, 2001; Roche et al.). Among female incest victims, preoccupied attach-
ment was associated with dependent personality traits, whereas fearful–avoidant attach ment
was associated with borderline and avoidant personality traits (Alexander, 1993; Alexander
et al., 1998). Anderson and Alexander (1996) reported that fearful–avoidant attachment
was related to high levels of dissociation and was much more likely to characterize a small
subsample (n = 8) of women identified as having dissociative identity disorder (DID).
Moreover, several researchers have reported recently that internal models of attachment
can both mediate and moderate the impact of childhood abuse and long-term psychologi-
cal outcomes, including dissociation, depression, and traumatic stress (Roche et al., 1999;
Shapiro & Levendosky, 1999; Wekerle & Wolfe, 1998).
THE CURRENT STUDY
Although family systems and attachment theories clearly share similar concepts regarding
development (Hill, Fonagy, Safier, & Sargent, 2003; Marvin & Stewart, 1990), little effort
has been made to integrate these two bodies of literature, and systematic research explor-
ing the interrelationships among family and attachment constructs is lacking. Further, the
bulk of studies examining either of these constructs in relation to psychological outcomes
among abuse survivors have used college or community samples (e.g., Anderson & Al-
exander, 1996; Muller et al., 2001; Roche et al., 1999), with rare exceptions using outpa-
tient clinical samples (Gold, Hyman, & Andres-Hyman, 2004; Nash et al., 1993). However,
college samples, and to a lesser extent community samples, generally show small effect
sizes for the long-term consequences of child abuse and tend to report shorter, less violent,
and less invasive abuse experiences than survivors in clinical samples (Jumper, 1995). The
current study used a geographically diverse sample of inpatients admitted to a specialized
treatment program for survivors of severe trauma. Because the sample was characterized
by a wide range of Axis I and Axis II psychopathology, it was possible to explore the
relative contributions of family environment and adult attachment style in the prediction
of specific types of disorders and personality dysfunction, beyond what is accounted for
by multitype abuse, parental psychopathology, and parental substance abuse. Notably, due
to the nature of the trauma program, a high number of participants were diagnosed with
DID, which offered a unique opportunity to study a disorder thought to originate in the
context of extreme child abuse but rarely seen in practice. A second goal of the study was
to address the suggestion that adult romantic attachment style mediates the association
between early family experiences and later outcomes among child abuse survivors.
Six family environment qualities were examined: five identified by Reichertz and
Frankel (1990) as characteristic of conflicted families (i.e., high conflict and high control, as
well as low cohesion, expressiveness, and independence), and a sixth scale (low organiza-
tion) reported to be associated with a history of child sexual abuse (Moos & Moos, 2002).
Although research examining early family environment and adult romantic attachment in
relation to specific psychological disorders among trauma survivors is relatively scarce,
based on the existing theory and literature previously reviewed, two general hypotheses
were tested. First, we expected family environment and adult attachment style to explain
Family, Attachment, and Psychopathology 581
a significant amount of variance in the full regression models for psychopathology out-
comes, beyond what can be accounted for by multitype or intrafamilial abuse, and parent
mental disorder or substance abuse. Second, we predicted that romantic attachment style
would mediate the relationship between some family environment variables (e.g., cohesion
and independence, which seem to best represent the polar dimensions of attachment and
self-reliance) and psychopathology. More specific hypotheses included:
1. Nonoptimal family environments (i.e., high levels of conflict and control; low
levels of cohesion, expressiveness, independence, and organization) will predict
greater psychopathology. For example: (a) Low cohesion and high conflict or
control will predict depression, anxiety, and posttraumatic stress symptoms, (b)
dependent personality will be related to high family control, and low expressive-
ness and independence, and (c) borderline personality will be related to high fam-
ily conflict and low organization.
2. Adult attachment style will predict diagnostic category and personality dimen-
sions. Specifically: (a) Attachment anxiety will predict bipolar, anxiety, and
dependent personality disorders, (b) attachment avoidance will predict substance
abuse, DID, somatoform disorder, and schizoid, antisocial, and compulsive person-
ality, and (c) high attachment anxiety and high attachment avoidance will predict
major depression and posttraumatic stress disorder (PTSD), and avoidant, border-
line, schizotypal, and paranoid personality, but pathologically low levels of histri-
onic and narcissistic personality representing low sociability and low self-esteem
(Millon, Davis, & Millon, 1997).
Participants were 80 patients (74 females, 6 males) with a mean age of 36.56 (range = 18
to 66) recruited over a 7-month period from a specialized hospital treatment program for
trauma-related disorders. Patients were excluded from the study if they were psychotic,
did not speak fluent English, or were judged to be too unstable by the attending physician.
Ethnicity was predominantly White (81.3%; n = 65), but also included 3 Hispanics, 3 Native
Americans, 1 African American, 1 Asian American, 5 multiracial backgrounds, and 2 report-
ing “other” ethnicity. Educational attainment was fairly high, with 43.8% of participants
reporting a college degree, 47.5% some college or technical school, and 8.8% a high school
degree or less. Marital status was fairly evenly distributed among single (38.8%), married
(33.8%), and divorced (27.5%) individuals. Similarly, participants reported varied annual
family income, with 22.5% reporting below $15,000, 21.3% reporting $15,000–$30,000,
16.3% reporting $30,000–$45,000, 16.3% reporting $45,0000–$75,000, and 22.5% reporting
more than $75,000. About two thirds (66.3%) of the participants claimed heterosexual orien-
tation, 20% claimed bisexual orientation, and 7.5% claimed gay/lesbian sexual orientation.
All procedures were reviewed and approved by the hospital staff and medical director, as
well as the Institutional Review Board of the Principal Investigator's educational institu-
tion. Prior to recruitment during hospital intake interviews, patients were independently
582 Riggs et al.
assessed by their attending physicians to determine a diagnosis according to the Diagnostic
and Statistical Manual of Mental Disorders-Fourth Edition–Text Revision (DSM-IV-TR;
Amer ican Psychiatric Association [APA], 2000) and to determine their appropriateness for
participation in the study based on clinical status. Eligible patients interested in participat-
ing scheduled an appointment with a research assistant. After the study was explained in
depth, the patient signed a consent form and a release of information form for diagnostic
records. Doctoral-level counseling psychology students provided specific verbal and writ-
ten instructions as they individually administered study instruments. Questionnaires were
completed that day if possible, or returned within 48 hours, at which time participants
could ask questions or discuss any concerns about the study.
A background questionnaire was developed to assess age, ethnicity, income, marital status,
education level, employment status, and sexual orientation. Participants also responded
yes/no to items regarding family history of psychological difficulties (i.e., alcoholism, drug
abuse, suicide attempt, psychiatric hospitalization, diagnosed mental disorder); for each
“yes” response, participants identified the relevant family member and specific disorder. In
addition, five self-report instruments measured family environment, adult attachment style,
and psychopathological symptoms.
Experiences in Close Relationships Scale (ECR; Brennan et al., 1998). The ECR is
the latest benchmark self-report measuring adult romantic attachment style. Using a large
number of items chosen from existing measures, the authors identified two major factors.
Based on the highest absolute-value correlations with one of the two major factors, a
36-item instrument was created with two 18-item scales: attachment anxiety and attach-
ment avoidance. Each item is rated on a 7-point Likert scale ranging from “not at all like
me” to “very much like me.” The scales were almost uncorrelated (r = .11) and demon-
strated coefficient alphas above .90. Item-total correlations ranged from .50 to .73. Crowell,
Fraley, and Shaver (1999) reported that the two scales demonstrated internal consistency
and test-retest reliability and had high construct, predictive, and discriminant validity.
Family Environment Scale (FES; Moos & Moos, 2002). The FES is a self-report mea-
sure consisting of 90 true/false items designed to assess early family environment. The FES
has been used extensively with college, community, and clinical populations, with norms for
both normal and distressed families. Of the 10 subscales, the current study used the Cohesion
(mutual commitment, support), Expressiveness (encouragement of emotional expression),
Conflict (open anger/conflict), Independence (assertiveness, self-sufficiency, independent
decision-making), Organization (structured activities, responsibilities), and Control (set
rules, procedures) scales. Content and construct validity, test-retest reliability, and good item
total correlations have been reported. Cronbach’s alphas range from .61 to .82. Standard
scores below 40 are generally considered low, and scores above 60 are considered elevated.
Millon Multiaxial Clinical Inventory-III (MCMI-III; Millon, 1994). The MCMI-III
is the latest revision of the MCMI, a widely used 175-item, true/false instrument designed
to assess both personality disorders and symptom syndromes in clinical populations. The
22 scales are based on Millon’s (1990; Millon & Davis, 1996; Millon, Davis, Millon,
Escovar, & Meagher, 2000) evolutionary theory of personality and psychopathology, as
well as DSM-IV (APA, 1994) diagnostic criteria. The present study used 10 personality
disorder scales (Schizoid, Avoidant, Dependent, Histrionic, Narcissistic, Antisocial,
Compulsive, Schizotypal, Borderline, Paranoid) and 8 clinical syndrome scales (Anxiety,
Family, Attachment, and Psychopathology 583
Somatoform, Bipolar/Manic, Dysthymia, Alcohol Dependence, Drug Dependence, Post-
traumatic Stress, Major Depression). Raw scores were transformed into base rate (BR)
scores to allow comparison to a normative group of patients and to reflect the nonnormal
distribution and actual prevalence of the disorder among patient populations. A BR score
of 85 or above indicates the prominence or likely presence of a particular disorder (Millon
et al., 1997). Extensive research supports the validity and reliability of earlier versions
of the MCMI (Choca & Van Denburg, 1997; Craig, 1993). Internal consistency for the
clinical scales ranged from .66 to .90, and test-retest reliabilities ranged from .82 to .96.
Answer forms were mailed to the publisher for computer scoring. Because missing data
constituted a low proportion (10%), mean substitution was used to replace missing values
to maximize power for MCMI analyses.
Dissociative Experiences Survey (DES; Bernstein & Putnam, 1986). The DES is a 28-
item self-report designed to measure the frequency of dissociative experiences in clinical
populations. Using 10% increments ranging from 0% (never) to 100% (always), respon-
dents circle the degree to which a particular experience applies. The DES has demonstrated
good validity and reliability, and good overall psychometric properties (Carlson & Putnam,
1993; Carlson et al., 1993; van IJzendoorn & Schuengel, 1996), especially in the ability to
discriminate DID from other diagnostic groups. Carlson et al. (1993) reported that a cutoff
score of 30 optimally maximized the accuracy of predicting a DID diagnosis.
Dissociative Disorders Interview Schedule (DDIS; Ross et al., 1989). The DDIS is a
131-item structured interview that assesses the presence of DSM-IV (APA, 1994) disorders,
including somatization disorder, major depressive disorder, borderline personality disor-
der, and five dissociative disorders. The interview also includes questions pertaining to
details of childhood abuse (e.g., type, severity, onset, duration, frequency, perpetrators),
substance abuse, schizophrenia, extrasensory/paranormal experiences, and general psychi-
atric history. The DDIS showed excellent diagnostic concordance for DID and dissociative
disorder not otherwise specified (DDNOS) with the DES-T (κ = .81), SCID-D (κ = .74),
and clinician diagnosis based on clinical interview (κ = .71) (Ross, Duffy, & Ellason,
2002). For this study, a self-report version was administered.
Descriptive statistics for MCMI-III, DES, and FES scales are presented in Table 1. MCMI
descriptive data are comparable to similar samples reported in the literature (Allen et al., 1998;
Ellason, Ross, & Fuch, 1995). Although MCMI debasement and disclosure scale elevations
may be indicative of “faking bad,” these scale elevations in an inpatient psychiatric sample
are more likely to indicate severe emotional distress and psychopathology (Allen et al.,
1998; Wetzler & Marlowe, 1990). The DES mean score of 36 is higher than the cutoff score
of 30 for identifying severely dissociative features and comparable to previously reported
DES means for PTSD and DID/DDNOS populations (Bernstein & Putnam, 1986; Carlson
et al., 1993; Coons, Bowman, Pellow, & Schneider, 1989; Frischholz et al., 1990; Ross
et al., 1989). FES descriptive data show that this sample was generally characterized by high
levels of conflict and control and low levels of cohesion, expressiveness, and independence.
These findings reflect the expected profile of conflicted families (Reichertz Frankel, 1990)
and families characterized by physical and/or sexual abuse (Justice & Calvert, 1990).
584 Riggs et al.
To increase diagnostic certainty, the assignment of clinical diagnoses considered several
sources of information, including clinician diagnosis, MCMI-III scale scores 85 or greater,
DDIS indices, and a DES cutoff score of 30 or greater. To receive a particular diagnosis,
two or more data sources were required to indicate the presence of each disorder. For
example, the participant was assigned a DID diagnosis if the clinician’s diagnosis was DID
and either the DDIS or DES score (>30) supported this diagnosis. Using this multisource
method, 87.5% of participants were diagnosed with depression, 17.5% with bipolar dis-
order, 48.8% with any anxiety-related disorder other than PTSD, 8.8% with somatoform
disorder, 13.8% with PTSD, 16.3% with substance abuse, 55% with DID, and 21.3%
with borderline personality disorder. The mean number of comorbid diagnoses was 3.09
(SD = 1.43). Of the 10 DSM-IV-TR personality disorders, only borderline personality dis-
order (BPD) could be assessed using the criteria of two or more instruments, so subsequent
TABLE 1. Descriptive Statistics for MCMI-III, DES, and FES
MCMI personality dimensions
Note. Clinical significance for MCMI > 85. Clinical significance for DES > 30;
for FES, low < 40 or high > 60.
Family, Attachment, and Psychopathology 585
analyses utilized MCMI personality dimensions to operationalize the other nine DSM-IV-
TR personality disorders (the borderline personality scale was not considered separately
from its contribution to the diagnostic category). In addition, to offset the inability to run
analyses with PTSD and somatoform diagnostic categories due to low frequencies, these
MCMI syndrome scales were added as outcome variables.
The sample was characterized by extreme traumatic experiences and parental dysfunc-
tion. Parental psychopathology (e.g., depression, anxiety) was reported by 37% of the
participants and parental alcohol and/or drug abuse was reported by 55%. Over 91% of the
sample reported a history of child sexual abuse, with 75% reporting both sexual and physi-
cal abuse, and 7.5% reporting neither sexual nor physical abuse but a significant degree
of psychological abuse (e.g., harsh rejection) or neglect by caregivers. Of those reporting
sexual abuse, 66% (n = 48) indicated the perpetrator was in the immediate family (i.e.,
parent, stepparent, or sibling) and 24% reported one perpetrator, 28% reported two, and
47% reported three or more perpetrators. Types of sexual abuse ranged from intercourse
with penetration (78%), oral sex (73%), anal sex (44%), to pornographic photography/film
(31%). Chronic sexual abuse was the norm with only 11% reporting fewer than 6 separate
incidents and 58% reporting more than 50 separate incidents before the age of 18.
To assess the need to control for variables previously linked to adult outcomes, dummy
variables were created for reported history of parent mental disorder, parent substance
abuse, intrafamilial abuse, and multitype (vs. single type) abuse. Pearson chi-square tests
compared each dummy variable with each diagnostic category. With the Bonferroni cor-
rection rate established at .008, results were significant for multitype abuse and DID, χ2
(1, 79) = 8.75, p < .003, as well as anxiety disorder, χ2 (1, 79) = 8.02, p < .005. DID was
also marginally significantly associated with reported parent substance abuse, χ2 (1, 79) =
5.28, p < .02. Next, multivariate analyses of variance (MANOVAs) were conducted to
assess the relationship between these dummy variables and MCMI-III scales representing
nine DSM-IV-TR personality disorders. Because MANOVAs control for the possibility of
inflated overall type I error rate and also incorporate correlations among variables into the
test statistic (Stevens, 1996), it was determined that a .05 alpha was sufficient to deter-
mine significance. Wilks’s Lambda multivariate tests showed significant associations for
parental substance abuse, F(1, 79) = 2.26, p < .03, and multitype abuse, F(1, 79) = 2.43,
p < .02, and a marginal association with parental psychopathology, F(1, 80) = 1.90, p <
.06, but not intrafamilial abuse, F(1, 73) = .90, p < .54. Similar findings emerged for the
posthoc MANOVA with the two MCMI Axis I syndromes (i.e., PTSD, somatoform). As a
result of these findings, intrafamilial abuse was dropped from consideration, but parental
psychopathology, parental substance abuse, and/or multitype abuse where significant were
included in the full regression analyses.
Analyses to determine if age, gender, or ethnicity were related to key variables were
nonsignificant. However, higher income was associated with MCMI-III Antisocial,
F(2, 78) = 5.57, p < .005, and Borderline, F(2, 78) = 3.04, p < .05, scales. In addition,
divorced participants reported higher attachment anxiety than single participants, F(2, 79) =
3.06, p < .05, and college graduates reported higher attachment avoidance than participants
with less than a high school education, F(2, 79) = 4.30, p < .02.
As part of the determination for mediation (Baron & Kenny, 1986), three sets of initial
simultaneous regressions were run to establish simple direct associations among family
586 Riggs et al.
environment, attachment style, and psychopathology outcome variables. Results of the
initial regressions indicated that mediation effects of attachment anxiety and avoidance
could be tested for only associations between FES Independence and five personality
dimensions (schizoid, avoidant, dependent, histrionic, narcissistic). Subsequently, dummy
variables and any predictors associated with outcome variables in preliminary analyses
were included in the full regression models to test the primary hypothesis that family envi-
ronment and adult attachment significantly contribute to the prediction of five diagnostic
categories (bipolar, anxiety, substance abuse, DID, and BPD) and 10 MCMI-III scales
(schizoid, avoidant, dependent, histrionic, narcissistic, compulsive, schizotypal, paranoid,
somatoform, PTSD). Dummy variables were entered in the first block, followed by FES
scale(s) in the second block and adult attachment style scale(s) in the third and last step of
the model. To protect against the possibility of Type I error in the full regression models,
a more conservative .001 alpha level was used for the model chi-square or final F value,
though subsequent interpretation of betas and t-values used the standard .05 significance
level. Tolerance and variance inflation factor (VIF) values indicated that multicollinearity
was not a problem in this sample. Results of all full regressions are presented in Tables 2
and 3, but only those meeting the .001 level are reported and discussed in text to highlight
As shown in Table 2, three of the full logistic regression models were significant at the .001
level or better (DID, substance abuse, anxiety), and three were only marginally significant
at the .01 or .05 levels (depression, bipolar, BPD). As predicted, attachment avoidance sig-
nificantly contributed to DID, raising the odds of this diagnosis by about 84%. Similarly,
higher family control significantly increased the odds of a DID diagnosis by a factor of 1.08.
Although multitype abuse significantly predicted DID in the first step (OR = .17, 95% CI =
.05–.61, p < .006) and maintained significance in the second step after adding family control
(OR = .20, 95% CI = .05–.75, p < .02), it was not a significant contributor to DID in the final
model, which accounted for 43% of the variance according to the Nagelkerke R2.
Attachment anxiety significantly increased the odds of a substance abuse diagnosis by a
factor of 2.66, and the full regression model accounted for 37% of the variance. Although
family expressiveness was initially a significant predictor of substance abuse (OR = 1.06,
95% CI = 1.00–1.12, p < .04), it dropped to nonsignificance when attachment anxiety was
added. In contrast, family organization went from a trend in the first step (OR = .96, 95%
CI = .91–1.00, p < .06) to significance in the final model, indicating that higher family
organization decreased the odds of a substance abuse diagnosis. Contrary to predictions,
attachment anxiety was not associated with anxiety disorder; however, multitype abuse and
family control significantly decreased the odds of an anxiety diagnosis in a full model
accounting for 22% of the variance.
The full regression models for most MCMI-III scales were significant, except for two
personality dimensions (compulsive, paranoid) that did not reach the .001 significance
level (see Table 3). All independent variables significantly contributed to the prediction
of schizoid personality when first entered into the equation; however, in the final model
only the contributions of parent substance abuse and attachment avoidance remained
significant, together accounting for 26% of the variance. Although higher levels of FES
independence were associated with fewer schizoid personality traits in the second step, its
unique contribution dropped to nonsignificance after attachment avoidance was included.
Likewise, all predictor variables significantly contributed to MCMI avoidant personality
score initially, but only parent substance abuse and both attachment scales remained signif-
icant, explaining 30% of the variance. Similarly, for histrionic and narcissistic personality
TABLE 2. Full Logistic Regression Models for Diagnostic CategoriesCriterion
Block / Model
N = 78
1. Multitype abuse
Parent substance abuse
2. FES control 3. ECR avoidance
abuse N = 80
1. FES expressiveness
2. ECR anxiety
N = 79
1. Multitype abuse2. FES control
N = 80
1. ECR avoidance
N = 80
1. FES conflict2. ECR anxiety
N = 80
1. FES organization2. ECR anxiety
Note. OR = odds ratio; CI = confidence interval.
†p < .10. *p < .05. **p < .01. ***p < .001.
TABLE 3. Full Regression Models for MCMI–III Personality and Syndrome ScalesCriterion
1. Multitype abuse
Parent substance abuse Parent mental disorder
2. FES independence3. ECR avoidance
3, 74 4, 735, 72
1. Parent substance abuse
Parent mental disorder
2. FES independence3. ECR avoidance
2, 78 3, 755, 73
–.24* –.30** –.31**–.34**–.26**
1. Parent substance abuse
Parent mental disorder
2. FES independence
FES conflict FES organization
3. ECR avoidance
2, 765, 73 7, 71
–.13 –.25*–.35** –.25* –.22*
1. Multitype abuse
Parent substance abuse Parent mental disorder
2. FES independence3. ECR avoidance
3, 744, 736, 71
1. Multitype abuse
Parent substance abuse Parent mental disorder
1. FES independence2. ECR anxiety
1, 782, 77
1. Multitype abuse
Parent substance abuse
2. FES expressiveness3. ECR avoidance
2, 753, 745, 72
1. Multitype abuse
Parent mental disorder
2. FES conflict3. ECR avoidance
2, 78 3, 755, 73
1. Multitype abuse
Parent substance abuse Parent mental disorder
2. ECR avoidance
3, 744, 73
1. Multitype abuse 2. FES independence3. ECR avoidance
1, 772, 764, 74
.15 .17 .26
Note. The “Initial β” column represents the standardized beta weight value and t significance levels for predictors when first entered into the
equation. “Final β” indicates the standardized beta weight value and t significance levels for predictors in the final step of the model.
†p < .10. *p < .05. **p < .01. ***p < .001.
2. FES independence 3. ECR avoidance
4, 73 6, 71
590 Riggs et al.
dimensions, FES independence was no longer a significant contributor after adding the
two ECR scales, which together accounted for an additional 8% and 12% of the variance,
respectively. Patients reporting high levels of attachment anxiety and avoidance endorsed
fewer histrionic and narcissistic traits. Given results of initial regressions, these findings
suggest that attachment style fully mediates the relationship between family independence
and MCMI schizoid, avoidant, histrionic, and narcissistic personality scales, after account-
ing for multitype abuse, parent mental illness and parent substance abuse.
However, FES scales remained significant in the final regression models for the MCMI
dependent and schizotypal personality scales. The dependent personality dimension was
significantly and negatively related to the FES independence, conflict and organization
scales in the second step, which accounted for 20% of the variance. In the final step of
the full model, higher ECR anxiety scores explained an additional 7% of the variance. For
schizotypal personality, the final regression indicated that high schizotypal scores were
significantly associated with low levels of FES expressiveness, more multitype abuse, and
high levels of attachment anxiety, with the full model explaining 21% of the variance.
In contrast to the previous findings, FES variables did not demonstrate unique contri-
butions to the prediction of MCMI somatoform, or PTSD scales. The somatoform scale
was significantly associated with parent substance abuse and attachment avoidance, which
together explained 23% of the variance. In addition, both ECR scales, along with multitype
abuse, significantly predicted the PTSD syndrome scale, with the final model accounting
for 26% of the variance.
Results of the present study generally support the view that early family environment and
adult attachment style contribute to psychological outcomes among adult trauma survivors
beyond what can be explained by multitype abuse, parent mental disorder, and parent sub-
stance abuse. In addition, evidence emerged that bolstered the suggestion that adult attach-
ment style mediates the relationship between some types of early family environments
and later psychopathology. After briefly addressing results related to the control variables,
significant findings (< .001) for the full regression models predicting diagnostic categories
and MCMI-III scales will be discussed.
Preliminary analyses indicated that multitype abuse, parent mental disorder, and parent
substance abuse were rarely associated with diagnostic category and frequently were
reduced to nonsignificance in the final regression models for MCMI-III scales. For ex-
ample, parent mental disorder significantly predicted the criterion in the first two steps
of five regression analyses (schizoid, avoidant, dependent, histrionic, somatoform), but
in each case it dropped to nonsignificance after the attachment scales were added in the
final step. Similarly, multitype abuse dropped to nonsignificance in four of six regressions,
retaining significance only for schizotypal personality and PTSD. These results suggest
that multitype abuse and parent mental disorder often influence personality functioning
only indirectly through associations with adult attachment style.
Similarly, parent substance abuse was not related to a participant diagnosis of sub-
stance abuse as expected. This finding may be due to sample characteristics. Specifically,
Family, Attachment, and Psychopathology 591
although not screened for substance abuse, participants were drawn from a specialized
trauma treatment program rather than a substance abuse program. It is possible that the
more severe cases of substance abuse showing the typical intergenerational pattern might
have been referred to a program specifically targeting chemical dependency. In addition,
the current sample contained a preponderance of women, who generally are less likely
than men to abuse substances (APA, 2000). In contrast, current findings imply that par-
ent substance abuse has a direct impact on some personality dimensions, independent of
family environment or attachment style. However, counter to hypotheses and previous
suggestions (Cummings & Davies, 1994; Rutter & Quinton, 1984), participants reporting
parent substance abuse and/or mental disorder scored lower than those without such a
history on the schizoid, avoidant, and somatoform scales and higher on the histrionic scale.
According to Jacob and Leonard (1994), social learning theory would predict that children
are most likely to model parent behaviors, such as alcohol abuse, when they admire,
respect, or identify with the parent in some way. In this trauma sample where parent–child
relations were marked by extreme parental abuse or neglect, the optimal conditions that
promote parental modeling may not be present (e.g., Johnson & Pandina, 1991; Russell,
1990; Stein, Burden, & Nyamathi, 2002). Although there are clearly cases when adults do
not admire parents and still unconsciously exhibit the same behaviors, it is plausible that
some survivors might attribute abusive or neglectful behavior by parents to a mental ill-
ness or chemical dependency that is beyond parents’ control. As a result, these individuals
might be able to break the typical intergenerational pattern by preserving or developing
in therapy more positive internal representations of self and other, which would increase
well-being and decrease the risk of psychological dysfunction. However, current results
are only suggestive, and additional research is needed with psychiatric samples to directly
test these inferences.
Full regressions were significant at the required .001 level for 3 of the 6 diagnostic cat-
egories: DID, substance abuse, and anxiety disorder. Consistent with research suggesting
that sexually abused children are often rigidly controlled and isolated by their perpetrators
(Justice & Calvert, 1990; Moos & Moos, 2002; Yama et al., 1993), the odds of a DID
diagnosis rose by 8% with increases in family control. Most striking, however, was the
84% gain in the odds of DID when attachment avoidance increased. Together family con-
trol and attachment avoidance accounted for a robust 43% of the variance. Dissociation
often occurs when individuals feel they have no control over their physical situation and
thus resort to passive avoidance by dissociating mentally to separate themselves from the
aversive experience (Nijenhuis, Vanderlinden, & Spinhoven, 1998). While dissociation
may be temporarily adaptive during abuse experiences, repeated use of this defense may
produce an enduring predisposition to avoid stressful emotions and close relationships,
which may be perceived as threatening. Current findings extend previous research showing
associations of infant avoidant attachment and adult fearful–avoidant attachment to disso-
ciative tendencies in adolescence or adulthood (Carlson, 1998; Ogawa, Sroufe, Weinfield,
Carlson, & Egeland, 1997) and reinforce the only prior report linking adult attachment
style to DID in eight incest survivors (Anderson & Alexander, 1996).
Regression results for substance abuse were also highly significant. Family organiza-
tion decreased the odds of this diagnosis, which suggests that the early provision of a
clear family structure may deter later use of alcohol or drugs. Contrary to predictions
592 Riggs et al.
based on previous findings (Mickelson et al., 1997), however, it was not attachment
avoidance, but instead attachment anxiety that increased the odds of this diagnosis by a
factor of 2.66. Much of the recent literature suggests that males and females have different
reasons for abusing substances; males use drugs and/or alcohol as a mechanism to avoid
thinking about problems or enhance positive emotions and socialization, whereas females
use psychoactive substances to alleviate internalized distress or increase self-confidence
(Boys, Marsden, & Strang, 2001; Chassin, Pitts, DeLucia, & Todd, 1999; Lillehoj, Trudeau,
Spoth, & Wickrama, 2004). Because this sample was predominantly women, current
findings suggest that attachment anxiety might be more likely than attachment avoidance
to contribute to substance use among female inpatients.
In contrast to Muller et al.’s (2001) findings, attachment anxiety was not related to
anxiety disorder. Different measures and sample composition may account for this finding.
Unlike the current study’s use of inpatient diagnostic category, Muller et al. used the Beck
Anxiety Inventory (Beck & Steer, 1990) in a community sample. Even though one major
incident of abuse in childhood met study criteria for abuse, rates of child abuse were sub-
stantially lower than the current sample, so it is possible that the severity of abuse and
psychological disturbance may have produced a restricted range that obscured the rela-
tionship between attachment style and anxiety disorder in this study. Alternatively, the
grouping of all anxiety disorders other than PTSD may have compromised the results and
contributed to the unexpected finding that multitype abuse significantly decreased the odds
of an anxiety diagnosis. Although physical violence may somehow reduce anxiety associ-
ated with concurrent sexual abuse or neglect, current findings suggest that multitype abuse
experiences contribute to anxious symptomatology more characteristic of PTSD than other
anxiety disorders. Finally, while low levels of family control might suggest other family
factors obviating the need for control (e.g., clear rules), according to Moos and Moos
(2002), very low levels of control can reflect an underorganized system lacking rules and
standard procedures, which may stimulate a generalized anxiety regarding consequences
The full regression models were significant at the required .001 level for 8 of 10 MCMI
scales (schizoid, avoidant, dependent, histrionic, narcissistic, schizotypal, PTSD, somato-
form). A pattern of findings emerged partially supporting the hypothesis that adult attach-
ment style would mediate the relationship between family environment and personality
dysfunction. Following preliminary tests of direct associations, in full regression models
for 4 of 5 personality dimensions, family independence was a significant predictor at initial
entry but dropped to nonsignificance after the attachment predictor(s) were entered. From
a theoretical standpoint, the family’s fostering of independence and attachment represent
two complementary behavioral systems, which should be balanced for optimal develop-
ment (Bowlby, 1988). Consequently, family environments that inappropriately discourage
or conversely insist on independence can deleteriously influence the emergent attachment
style, which in turn may influence personality development.
Patients reporting high levels of both adult attachment anxiety and attachment avoid-
ance scored significantly higher on the PTSD and avoidant personality scales. The nega-
tively skewed self–other perspective represented by high levels of both attachment anxiety
and avoidance is consistent with descriptions of trauma survivors, who tend to view the
self as less worthy and the world as more malevolent (Janoff-Bulman, 1992). Similarly,
Family, Attachment, and Psychopathology 593
a negative sense of self and social inhibition characterize avoidant personality disorder.
Expectations formed in childhood for others to be hurtful and the self to be inadequate may
engender contradictory behavioral tendencies theoretically consonant with the irresolvable
approach–avoidance dilemma described by Main and Hesse (1990) in relation to disor-
ganized infants. The inability of individuals with high attachment anxiety and avoidance
to resolve this paradox could conceivably produce disorganization (Simpson & Rholes,
2002), increasing the risk for PTSD and personality disturbance.
In contrast to results for other personality disorder scales, high FES independence
scores were associated with high scores on the histrionic and narcissistic personality
dimensions. Conversely, the two attachment scales were negatively related to these two
personality scales. Because this sample of severely traumatized adults was characterized
by extremely low mean scores on these two personality dimensions, the finding that greater
security (i.e., low attachment anxiety and low attachment avoidance) is associated with
higher histrionic and narcissistic features is consistent with Millon et al.’s (1997) report
that moderate levels of these two personality dimensions represent healthier functioning
in sociability and self-esteem. Similarly, a greater emphasis on independence in families
of child abuse survivors would predict better functioning.
Results suggested that low family independence is uniquely salient in the etiology of
dependent personality disorder. This finding fits with prior theory and research describing
families of individuals with dependent personality disorder as enmeshed, overprotective,
discouraging of autonomy, and indulgent (Millon et al., 2000). In addition, dependent
personality was associated with low family organization and conflict, which may reflect
DSM-IV-TR (APA, 2000) criteria involving a need for others to assume responsibility and
reluctance to express disagreement. Low family organization indicates a lack of structured
responsibilities entrusted to children, who then may fail to acquire appropriate compe-
tencies or the psychological maturity to form personal opinions differing from parent
decisions, leading to low conflict levels. The tendency of individuals with dependent per-
sonality disorder to idealize and/or try to please attachment figures and to devalue the self
extends to romantic love (Millon et al., 2000), which may account for present associations
with romantic attachment anxiety and is consistent with previous links of preoccupied
attachment to dependent personality disorder (Brennan & Shaver, 1998).
Schizotypal personality style was predicted by multitype abuse, low family expressive-
ness, and high attachment anxiety. In contrast to the term “expressed emotion,” which in
the schizophrenia literature refers to critical and/or emotional overinvolvement (Asarnow,
Tompson, Hamilton, Goldstein, & Guthrie, 1994; Miklowitz, 2004), the developers of
the FES define expressiveness as “the extent to which family members are encouraged
to express their feelings directly” (Moos & Moos, 2002, p. 1). Low levels of expres-
siveness would be more closely related to parental neglect and/or rejection than critical
overinvolvement, and thus current results are in line with previous research indicating that
childhood neglect may be a precursor to schizotypal personality (Berenbaum, Valera, &
Kerns, 2003; Johnson, Smailes, Cohen, Brown, & Bernstein, 2000; Torgesen & Alnaes,
1992). Additionally, consistent with predictions based on the DSM-IV-TR (APA, 2000)
criterion of social anxiety, high levels of attachment anxiety were associated with schizo-
typal personality. However, contrary to expectations based on descriptions of schizotypal
persons as paranoid and lacking close relationships, attachment avoidance did not predict
schizotypal personality in the final regression model. In this predominantly female sample,
this finding may reflect previously reported gender differences in schizotypy that greater
social anxiety is found among women and higher negative symptoms (withdrawal, no
594 Riggs et al.
friends) are found among men (Fossati, Raine, Carretta, Leonardi, & Maffei, 2003; Miller &
Burns, 1995). Social avoidance may be relevant to schizotypy only among men or alternatively
in larger social contexts consisting of people outside the established attachment network.
Therefore, avoidant behaviors would not necessarily manifest in close romantic relationships.
Consistent with predictions and previous findings (Brennan & Shaver, 1998), results
indicated that attachment avoidance significantly predicted schizoid traits. Although mul-
titype abuse, parent mental disorder, and low family independence initially were unique
predictors, all three variables dropped to nonsignificance after attachment avoidance was
included in the model. Only parent substance abuse continued to significantly predict
schizoid personality in the final model with attachment avoidance. Results indicated
that somatoform syndrome was also associated with attachment avoidance. In clinical
reports, somatic complaints have been linked frequently with the reluctance to experi-
ence or express emotion (Brewin, Dalgleish, & Joseph, 1996; Zerbe, 1999) and child-
hood trauma (Heim, Ehlert, Hanker, & Hellhammer, 1998; Thakkar & McCanne, 2000).
Because physical illness and inhibition of negative emotions may increase nurturance and
diminish aversive behaviors in perpetrators, somatic expressions of distress may develop
in an attempt to adapt to an abusive and/or neglectful home environment. This tendency to
express negative emotion in bodily functions may persist into adult romantic attachments
as a way to gain attention or avert partner abuse, which is more commonly reported by
child abuse survivors than nonabused controls (Acierno, Resnick, Kilpatrick, Saunders, &
Best, 1999; Cloitre, Scarvalone, & Difede, 1997).
From a clinical standpoint, current results have a number of interesting implications for
intervention. First, an examination of family experiences may help identify at-risk children
and early prevention efforts. Clinicians might effectively target family environment char-
acteristics in family therapy with at-risk or abused children and their families, specifically
working to improve the family’s organization, control, and expressiveness. In particular,
given current findings suggesting the importance of family independence and attachment
relationships, family therapy can be designed to help parents establish an optimal balance
between the encouragement of autonomy and maintenance of family connection, which
will foster secure attachment among family members.
On the other hand, clinicians working with adult abuse survivors might find an explora-
tion of current romantic attachment style useful in ameliorating the long-term impact of
abuse. Attachment theory is easy to explain and provides a solid framework that makes
intuitive sense to help survivors understand the effects of attachment trauma and the role of
close relationships in coping responses and emotional regulation strategies (Allen, 2001). In
addition to psycho-education and the provision of a secure therapeutic alliance, interventions
designed to promote more adaptive interpersonal skills and foster a sense of security in cur-
rent attachment relationships with partners or other attachment figures (e.g., couples therapy)
may counteract earlier adverse attachment experiences, decreasing psychological distress
and improving overall functioning. Due to the risk of revictimization among individuals with
a history of childhood abuse (Acierno et al., 1999; Cloitre et al., 1997), a therapeutic focus on
the potential re-enactment of early attachment trauma in current romantic relationships also
may be beneficial in treating psychopathology among adults with a history of child abuse.
Allen (2001) provides specific recommendations for clinicians on the application of attach-
ment theory to intervention approaches and techniques with adult trauma survivors.
Family, Attachment, and Psychopathology 595
LIMITATIONS AND CONCLUSIONS
Overall, this study shed new light on the role of family environment and adult romantic
attachment in personality functioning and psychopathology among child abuse survivors.
Current results should be interpreted in light of the study’s strengths and limitations.
Multiple diagnostic measures raise confidence in the accuracy of participant diagnosis for
logistic regressions, although comorbidity must be taken into account as representative of
a psychiatric sample. However, the possibility of inflated findings for the MCMI analyses
due to common method variance in using self-reports for predictors and criterion must be
considered. The use of a psychiatric trauma sample extends previous research with com-
munity or college samples and offers a unique opportunity to explore attachment processes
in relation to specific forms of psychopathology, most notably DID. By the same token,
however, the sample composition and size may have narrowed the range of possible out-
comes and obscured potential associations. For example, although there is some evidence
that a history of child sexual abuse is more common among homosexual adults (Beitchman
et al., 1992; Hughes, Johnson, & Wilsnack, 2001), the high proportion of homosexual
and bisexual participants in this study may indicate a sampling bias that possibly affected
results. Additionally, retrospective reports of childhood trauma and early family environ-
ment may have introduced subjective biases reflected in distorted memories or lack of self-
disclosure. In particular, we had no corroborating evidence of childhood sexual or physical
abuse. Finally, the correlational nature of this study precludes an examination of temporal
and causal relationships between adult attachment and psychopathology.
Clearly, more research using larger samples with psychiatric and control groups is
needed to clarify how these constructs relate to one another and differentially relate to per-
sonality and psychopathology. Longitudinal research following at-risk children into adult-
hood would better address questions related to specific developmental trajectories arising
from particular abuse experiences and dysfunctional family environments. Despite the
need for continued investigation, results of the current study enhance our understanding of
family and relationship factors that contribute to different psychopathological outcomes.
In particular, while early family environments and experiences with parents are important,
this study’s findings regarding the mediation effects of adult attachment style highlight the
importance of considering patients’ current approaches to adult relationships. Clinically,
the central role of present-day romantic attachment style is encouraging and offers an
optimistic outlook for change in spite of traumatic early experiences.
Alexander, P. (1992). Application of attachment theory to the study of sexual abuse. Journal of
Consulting and Clinical Psychology, 60, 185–195.
Alexander, P. C. (1993). The differential effects of abuse characteristics and attachment in the predic-
tion of long-term effects of sexual abuse. Journal of Interpersonal Violence, 9, 346–362.
Alexander, P. C., Anderson, C. L., Brand, B., Schaeffer, C. M., Grelling, B., & Kretz, L. (1998). Adult
attachment and longterm effects in survivors of incest. Child Abuse & Neglect, 22, 45–61.
Allen, J. (2001). Traumatic relationships and serious mental disorders. Chichester, England: Wiley &
Allen, J., Coyne, L., & Huntoon, J. (1998). Complex posttraumatic stress disorder in women from a
psychometric perspective. Journal of Personality Assessment, 70, 277–298.
596 Riggs et al.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders
(4th ed., text revision). Washington, DC: Author.
Anderson, C. L., & Alexander, P. C. (1996). The relationship between attachment and dissociation
in adult survivors of incest. Psychiatry, 59, 240–254.
Asarnow, J. R., Tompson, M., Hamilton, E. G., Goldstein, M. J., & Guthrie, D. (1994). Family-
expressed emotion, childhood-onset depression, and childhood-onset schizophrenia spectrum
disorders: Is expressed emotion a nonspecific correlate of child psychopathology or a specific
risk factor for depression? Journal of Abnormal Child Psychology, 22, 129–147.
Baker, J. D., Capron, E. W., & Azorlosa, J. (1996). Family environment characteristics of persons
with histrionic and dependent personality disorders. Journal of Personality Disorders, 10,
Baron, R. M., & Kenny, D. A. (1986). The mediator-moderator variable distinction in social psycho-
logical research: Conceptual, strategic and statistical considerations. Journal of Personality and
Social Psychology, 51, 1173–1182.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a
four-category model. Journal of Personality and Social Psychology, 61, 226–244.
Beck, A. T., & Steer, R. A. (1990). Manual for the Beck Anxiety Inventory. San Antonio, TX:
Beitchman, J. H., Zucker, K. J., Hood, J. E., DaCosta, G. A., Akman, D., & Cassavia, E.
(1992). A review of the long-term effects of child sexual abuse. Child Abuse & Neglect,
Benedict, L. L. W., & Zautra, A. A. J. (1993). Family environmental characteristics as risk factors of
childhood sexual abuse. Journal of Clinical Child Psychology, 22, 363–374.
Berenbaum, H., Valera, E. M., & Kerns, J. G. (2003). Psychological trauma and schizotypal symp-
toms. Schizophrenia Bulletin, 29, 143–152.
Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability, and validity of a dissociation
scale. Journal of Nervous and Mental Disease, 174, 727–735.
Blatt, S. J., & Levy, K. N. (2003). Attachment theory, psychoanalysis, personality development and
psychopathology. Psychoanalytic Inquiry, 23, 102–150.
Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss. New York: Basic Books.
Bowlby, J. (1988). Developmental psychiatry comes of age. American Journal of Psychiatry, 145,
Boys, A., Marsden, J., & Strang, J. (2001). Understanding reasons for drug use amongst young
people: A functional perspective. Health Education Research, 16, 457–469.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment:
An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close
relationships (pp. 46–76). New York: Guilford.
Brennan, K. A., & Shaver, P. R. (1998). Attachment styles and personality disorders: Their connec-
tions to each other and to parental divorce, parental death, and perceptions of parental caregiv-
ing. Journal of Personality, 66, 835–878.
Brennan, K. A., Shaver, P. R., & Tobey, A. E. (1991). Attachment styles, gender, and parental prob-
lem drinking. Journal of Social and Personality Relationships, 8, 451–466.
Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of post-traumatic
stress disorder. Psychological Review, 103, 670–686.
Briere, J., & Runtz, M. (1990). Differential adult symptomatology associated with three types of
child abuse histories. Child Abuse & Neglect, 14, 357–364.
Carlson, E. A. (1998). A prospective longitudinal study of attachment disorganization/disorientation.
Child Development, 69, 1107–1128.
Carlson, E. B., & Putnam, F. W. (1993). An update on the Dissociative Experiences Scale. Disso-
ciation, 6, 16–27.
Carlson, E. B., Putnam, F. W., Ross, C. A., Torem, M., Coons, P., Dill, D., et al. (1993). Validity of
the Dissociative Experiences Scale in screening for multiple personality disorder: A multicenter
study. American Journal of Psychiatry, 150, 1030–1036.
Family, Attachment, and Psychopathology 597
Chassin, L., Pitts, S. E., DeLucia, C., & Todd, M. (1999). A longitudinal study of children of alco-
holics: Predicting young adult substance use disorders, anxiety, and depression. Journal of
Abnormal Psychology, 108, 106–119.
Choca, J. P., & Van Denburg, E. (1997). Interpretive guide to the Millon Clinical Multiaxial Inventory
(MCMI) (2nd ed.). Washington, DC: American Psychological Association.
Cloitre, M., Scarvalone, P., & Difede, J. (1997). Posttraumatic stress disorder, self, and interpersonal
dysfunction among sexually retraumatized women. Journal of Traumatic Stress, 10, 437–482.
Craig, R. J. (Ed.). (1993). The Millon Clinical Multiaxial Inventory: A clinical research information
synthesis. Hillsdale, NJ: Erlbaum.
Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual differences in
adolescent and adult attachment. In J. Cassidy & P. Shaver (Eds.), Handbook of attachment:
Theory, research, and clinical applications (pp. 434–465). New York: Guilford.
Cummings, E. M., & Davies, P. T. (1994). Maternal depression and child development. Journal of
Child Psychology and Psychiatry, 35, 73–112.
Diehl, M., Elnick, A. B., Bourbeau, L., & Labouvie-Vief, G. (1998). Adult attachment styles: Their
relations to family context and personality. Journal of Personality and Social Psychology, 74,
Ellis, D. A., & Zucker, R. A. (1997). The role of family influences in development and risk. Alcohol
Health and Research World, 21, 218–227.
Engels, M., Moisan, D., & Harris, R. (1994). MMPI indices of childhood trauma among 110 female
outpatients. Journal of Personality Assessment, 63, 135–147.
Fossati, A., Raine, A., Carretta, I., Leonardi, B., & Maffei, C. (2003). The three-factor model of
schizotypal personality: Invariance across age and gender. Personality and Individual Difference,
Gauthier, L., Stollak, G., Messe, L., & Aronoff, J. (1996). Recall of childhood neglect and physical
abuse as differential predictors of current psychological functioning. Child Abuse & Neglect,
Gold, S. N., Hyman, S. M., & Andres-Hyman, R. C. (2004). Family of origin environments in two
clinical samples of survivors of intra-familial, extra-familial, and both types of sexual abuse.
Child Abuse & Neglect, 28, 1199–1212.
Harter, S. L., & Vanecek, R. J. (2000). Cognitive assumptions and long-term distress in survivors
of childhood abuse, parental alcoholism, and dysfunctional family environments. Cognitive
Therapy and Research, 24, 445–472.
Head, S. B., Baker, J. D., & Williamson, D. A. (1991). Family environment characteristics and de-
pendent personality disorder. Journal of Personality Disorders, 5, 256–263.
Heim, C., Ehlert, U., Hanker, J. P., & Hellhammer, D. H. (1998). Abuse-related post-traumatic stress
disorder and alterations of the hypothalamic-pituitary-adrenal axis in women with chronic pel-
vic pain. Psychosomatic Medicine, 60, 309–318.
Higgins, D. J., & McCabe, M. P. (2001). Multiple forms of child abuse and neglect: Adult retrospec-
tive reports. Aggression and Violent Behavior, 6, 547–578.
Higgins, D. J., & McCabe, M. P. (2003). Maltreatment and family dysfunction in childhood and the
subsequent adjustment of children and adults. Journal of Family Violence, 18, 107–120.
Hill, J., Fonagy, P., Safier, E., & Sargent, J. (2003). The ecology of attachment in the family. Family
Process, 42, 205–221.
Hughes, T. L., Johnson, T., & Wilsnack, S. C. (2001). Sexual assault and alcohol abuse: A compari-
son of lesbians and heterosexual women. Journal of Substance Abuse, 13, 515–532.
Jacob, T., & Leonard, K. (1994). Family and peer influences in the development of adolescent alco-
hol abuse. In R. A. Zucker, G. Boyd, & J. Howard (Eds.), The development of alcohol problems:
Exploring the biopsychosocial matrix of risk (pp. 123–156). National Institute on Alcohol
Abuse and Alcoholism Research Monograph No. 26. NIH Pub. No. 94–3495. Bethesda, MD:
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma. New York:
598 Riggs et al.
Johnson, V., & Pandina, R. (1991). Effects of the family environment on adolescent substance use,
delinquency and coping styles. American Journal of Drug and Alcohol Abuse, 17, 71–88.
Johnson, J. J., Smailes, E. M., Cohen, P., Brown, J., & Bernstein, D. P. (2000). Associations 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, 14, 171–187.
Jumper, S. A. (1995). A meta-analysis of the relationship of child sexual abuse to adult psychological
adjustment. Child Abuse & Neglect, 19, 715–728.
Justice, B., & Calvert, A. (1990). Family environment factors associated with child abuse. Psy-
chological Reports, 66, 458.
Kamsner, S., & McCabe, M. P. (2000). The relationship between adult psychological adjustment
and childhood sexual abuse, childhood physical abuse, and family-of-origin characteristics.
Journal of Interpersonal Violence, 15, 1243–1261.
Krupnick, J. L., Green, B. L., Stockton, P., Goodman, L., Corcoran, C., & Petty, R. (2004). Mental
health effects of adolescent trauma exposure in a female college sample: Exploring differen-
tial outcomes based on experiences of unique trauma types and dimensions. Psychiatry, 67,
Lillehoj, C. J., Trudeau, L., Spoth, R., & Wickrama, K. A. S. (2004). Internalizing, social compe-
tence, and substance initiation: Influence of gender moderation and a preventive intervention.
Substance Use & Misuse, 39, 963–991.
Ludolph, P. S., Westen, D., Misle, B., Jackson, A., Wixom, J., & Wiss, F. C. (1990). The borderline di-
agnosis in adolescents: Symptoms and developmental history. American Journal of Psy chiatry,
Main, M., & Hesse, E. (1990). Parents’ unresolved traumatic experiences are related to infant
disorganized attachment status: Is frightened and/or frightening parental behavior the linking
mechanism? In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the
preschool years (pp. 161–182). Chicago: University of Chicago Press.
Marvin, R. S. (2003). Implications of attachment research for the field of family therapy. In
P. Erdman & T. Caffery (Eds.), Attachment and family systems: Conceptual, empirical, and
therapeutic relatedness (pp. 3–27). New York: Brunner-Routledge.
Marvin, R. S., & Stewart, R. B. (1990). A family systems framework for the study of attachment. In
M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years:
Theory, research, and intervention (pp. 51–86). Chicago: University of Chicago Press.
Mickelson, K., Kessler, R., & Shaver, P. (1997). Adult attachment in a nationally representative
sample. Journal of Personality and Social Psychology, 73, 1092–1106.
Mikulincer, M., & Florian, V. (1999). The association between parental reports of attachment style and
family dynamics and offspring’s reports of adult attachment style. Family Process, 38, 243–257.
Mikulincer, M., Florian, V., & Weller, A. (1993). Attachment styles, coping strategies, and posttrau-
matic psychological distress: The impact of the Gulf War in Israel. Journal of Personality and
Social Psychology, 58, 273–280.
Mikulincer, M., & Shaver, P. R. (2003). The attachment behavioral system in adulthood: Activation,
psychodynamics, and interpersonal processes. In M. P. Zanna (Ed.), Advances in experimental
social psychology, Vol. 25 (pp. 56–152). San Diego, CA: Academic Press.
Miller, L. S., & Burns, S. A. (1995). Gender differences in schizotypic features in a large sample of
young adults. Journal of Nervous and Mental Disease, 183, 657–661.
Millon, T. (1990). Toward a new personology. New York: Wiley.
Millon, T. (1994). Millon Clinical Multiaxial Inventory-III manual. Minneapolis, MN: National Com-
Millon, T., Davis, R. D., & Millon, C. (1997). MCMI-III manual (2nd ed.). Minneapolis, MN: Na-
tional Computer Systems.
Millon, T., Davis, R. D., Millon, C., Escovar, L., & Meagher, S. (2000). Personality disorders in
modern life. New York: Wiley.
Family, Attachment, and Psychopathology 599
Molnar, B. E., Buka, S. L., & Kessler, R. C. (2001). Child sexual abuse and subsequent psychopa-
thology: Results from the National Comorbidity Survey. American Journal of Public Health,
Moos, R. H., & Moos, B. S. (2002). Family Environment Scale Manual: Development, applications,
research (3rd ed.). Palo Alto, CA: Center for Health Care Evaluation, Department of Veterans
Affairs and Stanford University Medical Centers.
Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., & Herbison, G. P. (1996). The long-term
impact of the physical, emotional, and sexual abuse of children: A community study. Child
Abuse & Neglect, 20, 7–21.
Muller, R., Lemieux, K., & Sicoli, L. (2001). Attachment and psychopathology among formerly
maltreated adults. Journal of Family Violence, 16, 151–169.
Myerson, L. A., Long, P. J., Miranda, R., & Marx, B. P. (2002). The influence of childhood sexual
abuse, physical abuse, family environment, and gender on the psychological adjustment of
adolescents. Child Abuse & Neglect, 26, 387–405.
Nash, M. R., Hulsey, T. L., Sexton, M. C., Harralson, T. O., & Lambert, W. (1993). Long-term
sequelae of childhood sexual abuse: Perceived family environment, psychopathology, and dis-
sociation. Journal of Consulting and Clinical Psychology, 61, 276–283.
Nijenhuis, E. R. S., Vanderlinden, J., Spinhoven, P. (1998). Animal defensive reactions as a model for
trauma-induced dissociative reactions. Journal of Traumatic Stress, 11, 243–260.
Ogawa, J. R., Sroufe, L. A., Weinfield, N. S., Carlson, E. A., & Egeland, B. (1997). Development
and the fragmented self: Longitudinal study of dissociative symptomatology in a nonclinical
sample. Development and Psychopathology, 9, 855–879.
Onishi, M., Gjerde, P. F., & Block, J. (2001). Personality implications of romantic attachment
patterns in young adults: A multi-method, multi-informant study. Personality and Social Psy-
chology Bulletin, 27, 1097–1110.
Pfaller, J., Kiselica, M., & Gerstein, L. (1998). Attachment style and family dynamics in young
adults. Journal of Counseling Psychology, 45, 353–357.
Polusny, M. A., & Follette, V. M. (1995). Long-term correlates of child sexual abuse: Theory and
review of the empirical literature. Applied and Preventive Psychology, 4, 143–166.
Reichertz, D., & Frankel, H. (1990). Family environments and problematic adolescents: Toward an
empirically based typology. Community Alternatives: International Journal of Family Care, 2,
Roche, D., Runtz, M., & Hunter, M. (1999). Adult attachment: A mediator between child sexual
abuse and later psychological adjustment. Journal of Interpersonal Violence, 14, 184–207.
Ross, C. A., Duffy, M. M., & Ellason, J. W. (2002). Prevalence, reliability, and validity of dissocia-
tive disorders in an inpatient setting. Journal of Trauma & Dissociation, 3, 7–17.
Ross, C. A., Heber, S., Norton, G. R., Anderson, D., Anderson, G., & Barchet, P. (1989). The
Dissociative Disorders Interview Schedule: A structured interview. Dissociation, 2, 169–189.
Rutter, M., & Quinton, D. (1984). Parental psychiatric disorder: Effects on children. Psychological
Medicine, 14, 853–880.
Shapiro, D. L., & Levendosky, A. A. (1999). Adolescent survivors of childhood sexual abuse: The
mediating role of attachment style and coping in psychological and interpersonal functioning.
Child Abuse & Neglect, 23, 1175–1191.
Simpson, J. A., & Rholes, W. S. (2002). Fearful-avoidance, disorganization, and multiple working
models: Some directions for future theory and research. Attachment and Human Development,
Stein, J. A., Burden, L. M., & Nyamathi, A. (2002). Relative contributions of parent substance use
and childhood maltreatment to chronic homelessness, depression, and substance abuse prob-
lems among homeless women: Mediating roles of self-esteem and abuse in adulthood. Child
Abuse & Neglect, 26, 1011–1027.
Styron, T., & Janoff-Bulman, R. (1997). Childhood attachment and abuse: Long-term effects on
adult attachment, depression, and conflict resolution. Child Abuse & Neglect, 10, 1015–1023.
600 Riggs et al.
Thakkar, R. R., & McCanne, T. R. (2000). The effects of daily stressors on physical health in women
with and without a history of childhood sexual abuse. Child Abuse and Neglect, 24, 209–221.
Trull, T. (2001). Relationships of borderline features to parental mental illness, childhood abuse,
Axis I disorder, and current functioning. Journal of Personality Disorders, 15, 19–32.
Tyler, K. A. (2002). Social and emotional outcomes of childhood sexual abuse: A review of recent
research. Aggression and Violent Behavior, 7, 567–589.
van IJzendoorn, M., & Schuengel, C. (1996). The measurement of dissociation in normal and
clinical populations: Meta-analytic validation of the Dissociative Experiences Scale. Clinical
Psychology Review, 16, 365–383.
Wekerle, C., & Wolfe, D. A. (1998). The role of child maltreatment and attachment style in adoles-
cent relationship violence. Development and Psychopathology, 10, 571–586.
Yama, M., Tovey, S. L., & Fogas, B. S. (1993). Childhood family environment and sexual abuse as
predictors of anxiety and depression in adult women. American Journal of Orthopsychiatry,
Yama, M., Tovey, S. L., Fogas, B. S., & Morris, J. (1995). The relationship among childhood sexual
abuse, parental alcoholism, family environment, and suicidal behavior in female college stu-
dents. Journal of Child Sexual Abuse, 4, 19–93.
Zerbe, K. J. (1999). Women’s mental health in primary care. Philadelphia: Saunders.
Acknowledgments. This project was partially funded by the Department of Psychology and Phi-
losophy at Texas Woman’s University and the Ross Institute. We are grateful to the staff and patients
of the Timberlawn Psychiatric Hospital Trauma Program for their time.
Correspondence regarding this article should be directed to Shelley A. Riggs, PhD, University of North
Texas, Department of Psychology, P.O. Box 311280, Denton, TX 76203–1280. E-mail: firstname.lastname@example.org