J Fam Viol (2014) 29:287297
DOI 10.1007/s10896-014-9589-2
RESEARCH ON PERPETRATORS OF INTIMATE PARTNER VIOLENCE
Risk Factors for Mens Intimate Physical Aggression in Spain
Anita Jose & K. Daniel OLeary &
Jose Luis Graa Gomez & Heather M. Foran
Published online: 5 April 2014
# Springer Science+Business Media New York 2014
Abstract This study tests a model of physical aggression in a
sample of 194 men being treated for physical or psychological
aggression in the greater Madrid area of Spain. The prevalence
of aggression in this sample was lower than in a US batterer
sample. In the path model highlighted here, borderline per-
sonality traits and alcohol problems were identified as risk
factors for physical aggression perpetration and the role of
borderline traits was mediated by psychological aggression. A
reciprocal relationship was found between psychological per-
petration and victimization but not physical perpetration and
victimization. These findings are discussed within a cross-
cultural context; conceptual and treatment implications are
also addressed.
Keywords Alcohol . Borderline personality traits . Intimate
partner violence . Physical aggression . Psychological
aggression
In the broadest sense, relationship aggression includes phys-
ical acts (e.g., pushing, slapping, kicking), psychological acts
(e.g., name-calling, threatening), sexual acts (i.e., using force
or coercion to engage in sexual activity), and related injuries
(e.g., cuts, bruises) (Straus et al. 1996). Many representative
studies conducted in the US suggest the 12-month prevalence
A. Jose : K. D. OLeary
Stony Brook University, Stony Brook, NY, USA
J. L. Graa Gomez
Complutense University of Madrid, Madrid, Spain
H. M. Foran
University of Braunschweig, Braunschweig, Germany
A. Jose (*)
University Behavioral Associates/Montefiore Medical Center, 334 E.
148th St. (2nd Floor), Bronx, NY 10451, USA
e-mail: ajose@montefiore.org
of physical aggression is around 10 % (Jose and OLeary
2009). Physical aggression is even more prevalent among
men arrested for domestic violence, as these batterer sam-
ples are, by definition, at high risk for relationship violence.
Although the criteria for mandated intervention may differ by
region, one New York State study (N=82) found that the
majority of men mandated to treatment reported psychological
(89.0 %) and physical (70.7 %) perpetration in the year prior
to starting treatment. Additionally, 24.4 % reported sexual
aggression, and 48.8 % reported injuring the partner within
that time frame (Vega and OLeary 2007).1
Despite the amount of research on intimate partner aggres-
sion in the US, there is much less international research on the
topic; and much of the extant data focuses on lifetime aggres-
sion rates which may be especially susceptible to memory
decay. However, according to a World Health Organization
study with data from ten countries, 13 % (Japan) to 62 % (Peru
province) of women reported physical victimization in their
lifetime (Garcia-Moreno et al. 2006). A separate study (De-
mographic and Health Surveys) analyzed female reports of
male-perpetrated violence in nine countries where lifetime
prevalence ranged from 17.5 % in Cambodia to 44.1 % in
Colombia (Kishor and Johnson 2004). Finally, the Worldsafe
study found that womens self-reported lifetime rates of phys-
ical victimization ranged from 11 % (Egypt) to 43 % (Trivan-
drum, India) in six communities across four countries (Hassan
et al. 2004). As the above findings indicate, rates of partner
aggression around the world are often as high as, or higher
than, rates of 2025 % found in the US (e.g., Nisonoff and
Bittman 1979; Tjaden and Thoennes 2000).
Given the high international rates of partner violence, the
study of physical aggression in countries other than the US
1
The rate of male victimization was similar to perpetration in the Vega
and O'Leary (2007) study for psychological (90.2 %), physical (74.4 %),
and sexual (24.4 %) aggression as well as injury (45.1 %).
288
seems especially timely. The present study identifies risk
factors for physical perpetration in a sample of male batterers
from Spain. Research on aggression in Spain is particularly
relevant as significant changes in Spanish law over the last
two decades have affirmed the importance of psychological
treatment for perpetrators (e.g., Organic Act 1/2004, 2004;
Integrated Protection Measures against Gender Violence).
Spain and the United States are similar on a number of
demographic, educational, occupational, and health indica-
tors. For example, both countries have high rates of adult
literacy (99 % in the US and 98 % in Spain; United Nations
Development Programme 2009), and high life expectancies
(77 years in the US and 91 years in Spain; UNICEF 2009a, b).
Further, the majority of people living in both nations are
urbanized (81 % in the US and 77 % in Spain; UNICEF
2009a, b); and each nations health system attainment is
ranked in the top 20 internationally (World Health
Organization 2000).
The similarity between development indicators in Spain
and the US is mirrored by similarities in the prevalence of
intimate partner aggression in these two countries. Using a
large community sample (N=4,550) in the Madrid area, Graa
(2009) found the rate of both male-to-female and female-to-
male physical aggression was approximately 12 %almost
identical to prevalence rates found in one landmark represen-
tative US study (Straus and Gelles 1990). Another study
explored victimization rates in a sample of 2,015 women
living in urban areas throughout Spain (Medina-Ariza and
Barberet 2004). The prevalence of male-to-female physical
aggression was lower in this sample (8 %) than in the Straus
and Gelles (1990) data, perhaps because the Spanish study
measured point prevalence while the US study measured
annual prevalence. Medina-Ariza and Barberet (2004) also
recorded rates for sexual victimization (11.5 %) that overlap-
ped with US findings, whereas the rates of injury (5.8 %) were
almost double rates in the Straus and Gelles (1990) or Graa
(2009) samples.
Despite the similarities of demographic, educational and
health indicators across Spain and the United States, and
despite the initial community studies which suggest the prev-
alence of intimate partner aggression appears similar in Spain
and the US; these countries differ in at least two important
respects related to partner violence, namely, the attention that
has been directed to the problem of violence, and the extent of
alcohol use in each country.
First, researchers and lawmakers have focused on the prob-
lem of intimate partner aggression in the United States since
the 1970s, reflecting the view that aggression is a broad
public health problem (e.g., the 1994 Violence against Women
Act and its successors). However, cultural awareness of rela-
tionship aggression only began to gain momentum in Spain in
the mid-to-late 1990s when the murder of a woman garnered
national and international media attention (Death of a Wife
J Fam Viol (2014) 29:287297
1997). The tragic case of Ana Orantes, who was murdered by
her husband after she made multiple unsuccessful attempts to
obtain orders of protection, was extensively covered by the
media at the time and is frequently identified as a catalyst for
change in Spanish laws and health initiatives such as Organic
Act 1/2004 (2004), mentioned previously.
Second, a meta-analysis shows that problematic alcohol
use is a predictor of aggression perpetration (e.g., Foran and
OLeary 2008). Traditionally wet and dry countries can be
differentiated by average per capita alcohol use. Wet countries
(including Spain) typically have greater access to alcohol and
have a population that drinks more frequently during daily life
events such as meals when compared to dry countries (includ-
ing the US). Wet countries also have less abstinence and,
importantly, less intoxication compared to dry countries
(Bloomfield et al. 2003). Indeed, some research suggests the
average per capita alcohol use in the US is only about 70 % of
what is found in Spain (Organisation of Economic
Cooperation and Development 2009). Despite differing per
capita rates of consumption, alcohol use is prevalent in both
the US and Spain. One large-scale US study found that 65.6 to
68.1 % of participants reported drinking in the year prior
(Maxwell 2008), whereas a Spanish study found that alcohol
use in the year prior ranged from 75 to 79 % (Gual 2006). In
short, although alcohol use rates are higher in Spain, it may be
less related to partner aggression than in the US as a result of
the normative role of alcohol and lower intoxication rate in
wet countries. The present study identifies risk factors for
physical perpetration in males engaged in a Spanish batterer
treatment program with a particular focus on the bi-
directionality of aggression, and the roles of personality pa-
thology and alcohol problems in relationship violence.
The Dyadic Nature of Aggression
Aggression is often viewed as the result of complex interac-
tions between partners; maladaptive conflict tactics (e.g., re-
ciprocating a partners negative behavior) are associated with
an increased likelihood of relationship violence (Cordova
et al. 1993). Psychological aggression has been consistently
identified as one of the strongest risks for physical aggression
across settings, samples, and methods, with one meta-analysis
finding an effect size of r=0.49 for psychological aggression
across 15 studies (Stith et al. 2004). The dyadic associations
underlying partner aggression can be modeled using the
Actor-Partner Interdependence Model (APIM; Kenny 1996).
APIM is typically used to model longitudinal interactions and
seems appropriate for the current study as psychological ag-
gression is widely viewed as a temporal precedent to physical
aggression (e.g., Murphy and OLeary 1989; OLeary 1993).
When applied to the current research, APIM would indicate
that psychological perpetration predicts physical perpetration
J Fam Viol (2014) 29:287297
and victimization, and that psychological victimization pre-
dicts physical perpetration and victimization (Fig. 1). As
aggression is often bilateral, reciprocal paths between psycho-
logical perpetration and victimization, as well as between
physical perpetration and victimization, are modeled.
The Role of Mental Health and Substance Use
in Aggression
Personality Pathology Various models have suggested an
association between personality and aggression (e.g., OLeary
et al. 2007; Stuart and Holtzworth-Munroe 2005). One well
known tripartite typology suggests the severity, frequency,
generality of violence, and presence of Axis II pathology,
can predict mens membership into one of three groups of
perpetrators: generally violent/antisocial, dysphoric/
borderline, and family-only (Holtzworth-Munroe and Stuart
1994). Research has corroborated the role of Axis II pathology
in relationship violence, finding that both antisocial and bor-
derline traits are risk factors for relationship violence (e.g.,
Holtzworth-Munroe and Stuart 1994; Mauricio et al. 2007;
Murphy et al. 1993, 2001; Porcerelli et al. 2004; South et al.
2008). The model tested in the current study conceptualizes
borderline and antisocial traits as independent constructs,
consistent with Holtzworth-Munroe and Stuarts (1994) latent
class analysis.
Antisocial Personality Disorder in men has been associated
with male-to-female physical and psychological aggression,
both concurrently and prospectively (Kim and Capaldi 2004).
In another study, antisocial characteristics were correlated
significantly with psychological (r=0.47; p<0.01) and phys-
ical (r = 0.36; p < 0.01) aggression in a forensic sample
(Mauricio et al. 2007). Finally, antisocial traits were found to
predict trait anger and alcohol problems, each of which inde-
pendently predicted psychological aggression (which in turn
predicted physical aggression) in men arrested for partner
aggression (Stuart et al. 2006). Psychopathy, a construct
which overlaps with antisociality (Lilienfeld 2010), is de-
scribed in the literature as reflecting impulsivity, manipula-
tion, and lack of empathy for others, without assuming the
presence of significant criminal behavior. Psychopathic be-
havior, a hallmark of which is the pervasive disregard for
others, can also set the stage for violent outcomes.
Traits associated with Borderline Personality Disorder have
also been associated with relationship aggression. Recent
findings indicate a correlation between borderline traits and
psychological (r=0.57; p<0.01) as well as physical (r=0.36;
p<0.01) aggression (Mauricio et al. 2007). Dutton (1994)
found that Borderline Personality Organization (BPO), a con-
struct indicative of lower-level pathology on the spectrum of
borderline traits, was correlated with mens physical and ver-
bal aggression. Further, Tweed and Dutton (1998) found that
289
men categorized as impulsively aggressive (as opposed to
instrumentally aggressive) had higher BPO scores. Theoreti-
cally, borderline traits such as emotional lability, dependency,
and relationship instability may be related to a greater likeli-
hood of relationship conflict.
Alcohol Problems Not only has alcohol been associated with
an increase in general aggression (Murdoch et al. 1990), it has
also been associated with higher rates of physical aggression
in the context of romantic relationships. Two meta-analyses
suggest a modest effect (0.23<r<0.24) for the association
between alcohol and male perpetration (Foran and OLeary
2008; Stith et al. 2004). When considering the association
between personality pathology and alcohol use, there is em-
pirical evidence that borderline traits predict alcohol use prob-
lems longitudinally (Stepp et al. 2005). Antisocial traits and
substance use have also been linked in a number of studies
(e.g., Stuart et al. 2006, 2008). As alcohol problems tend to
diminish impulse control, and as impulsivity is a feature of
both psychopathy and borderline traits, it is hypothesized that
the combination of alcohol problems and personality pathol-
ogy intensifies aggressive behavior.
The Present Study
The present study uses path analysis to test a model of key
variables theorized to play a role in understanding relationship
aggression. In particular, the tested model includes two major
risk factors for relationship aggressionpersonality patholo-
gy and alcohol problems; and takes into account the interre-
lations between victimization and perpetration. As reviewed
above, relationship aggression involves both partners by def-
inition; recognizing the interrelation between victimization
and perpetration is an essential, but often neglected compo-
nent in models of relationship aggression. The novelty of the
current research lies in the fact that although a number of
individual risk factors (especially substance use and Axis II
pathology) have been connected to aggression perpetration,
there has been less research focused on creating a theoretical
model combining these variables.
Method
Participants
Men from the greater Madrid area of Spain participated in the
current study between January 2008 and December 2009 (N=
194). Participants had been referred to a court-mandated treat-
ment program for physical (e.g., punching, hair pulling) or
psychological (e.g., threats to kill) aggression as defined by
290
Fig. 1 Hypothesized model
Psychopathic Traits
Borderline Traits
the Madrid legal system and had typically been on probation
for 6 to 10 months prior to participating in the study.
Exclusion criteria for involvement in the treatment program
at which these data were collected included reporting current
drug abuse, current heavy drinking (six or more drinks per
day) or exhibiting acute psychotic symptoms. Men who did
not meet any treatment program exclusion criteria, and who
could read, write, and speak in Spanish were asked to partic-
ipate in this study. Of those who attended at least one assess-
ment session, 21 % did not complete the full assessment (e.g.,
because of scheduling or language barriers which would ne-
cessitate a referral to other programs, leaving the country, or
refusing to complete the treatment program); only men who
completed all of the assessments were included in these
analyses.
Measures
English-language versions of the questionnaires were trans-
lated into Spanish by a bilingual psychologist at the
Complutense University of Madrid, with extensive profes-
sional experience translating psychological research papers
from English into Spanish. After initial translation, mea-
sures were back-translated by a Professor from the De-
partment of English. A subset of the measures was then
applied to a pilot sample of 15 men to determine whether
potential participants had any difficulty with item compre-
hension. Pilot participants reported understanding the
meaning of all items.
Alcohol Use Disorders Identification Test (AUDIT; Saunders
et al. 1993) This 10-item measure of alcohol consumption
J Fam Viol (2014) 29:287297
Psychological Physical
Victimization Victimization
Psychological Physical
Perpetration Perpetration
Alcohol Use
patterns (e.g., frequency) and related consequences (e.g., so-
cial, occupational) is widely used for both research and clin-
ical purposes. Response values for each question range from 0
to 4, with a maximum score of 40. It has good internal
consistency and excellent sensitivity and specificity as an
alcohol screen (Allen et al. 1997). In the present study, internal
consistency was adequate (=0.80). Close to one-third of this
sample (29.6 %) obtained an AUDIT score of eight or higher,
consistent with problem drinking.
Levenson Primary and Secondary Psychopathy Scale (LPSP;
Levenson et al. 1995) This 26 item self-report scale measures
the domains of manipulation and callousness (primary psy-
chopathy) and impulsivity (secondary psychopathy). The par-
ticipant chooses options ranging from strong agreement to
strong disagreement (rated on a 5-point scale) in response to
statements measuring each domain. It has adequate psycho-
metric qualities in independent research (e.g., Lynam et al.
1999). This measure is viewed as an acceptable and appropri-
ate self-report measure of psychopathic traits (Lilienfeld 2010;
Lilienfeld and Fowler 2006). In the present study, alpha for
the full scale was 0.53; alpha for the primary subscale was
0.30 and alpha for the secondary subscale was 0.42.
McLean Screening Instrument for Borderline Personality Dis-
order (MSI-BPD; Zanarini et al. 2003) This 10-item self-
report scale measures Borderline Personality Disorder (BPD)
based on its DSM-IV (American Psychiatric Association
2000) diagnostic criteria (e.g., impulsivity, self-harm, feelings
of emptiness). Each response receives one point if scored in
the keyed direction. Both internal consistency (=0.74) and
test-retest reliability (=0.79) were high (Zanarini et al. 2003).
J Fam Viol (2014) 29:287297
In this study, Cronbachs alpha was adequate at 0.78. The
authors suggest a score of 7 or greater reflects a BPD diagno-
sis. In this sample, 12 people (6 %) met the cutoff.
Revised Conflict Tactics Scale (CTS2; Straus et al. 1996) The
CTS2 is a 78-item self-report questionnaire assessing behav-
iors during relationship conflict. Thirty nine items ask about
perpetration and 39 items ask about victimization within the
past year, across five subscales (negotiation, psychological
aggression, physical assault, injury, sexual coercion). Re-
sponses are scored from 0 (no incidents in the past year) to 6
(over 20 incidents in the past year) for each item. According to
Straus and colleagues (1996), coefficient alpha ranged from
0.79 to 0.95 for subscales of the CTS2. In the present study,
alpha for the full scale was 0.94 with subscales ranging from
0.68 (injury) to 0.91 (physical assault, negotiation). For psy-
chological aggression, =0.90 and for sexual coercion, =
0.75.
Procedure
Study Procedure Participants meeting all inclusion criteria
attended a series of sessions with two Masters-degree level
therapists. During the first session, the lead therapist explained
the study in detail, and obtained the participants informed
consent to participate in the present study which involved
confidential self-report assessments. Those who decided not
to participate (21 %; n=52) were able to complete the man-
dated intervention with no penalty.
Including the first session, participants met individually
with the therapists for eight weekly sessions, each lasting
approximately 60 minutes. Beginning in the second session,
time was set aside for participants to complete each of the
questionnaires for this study and additional assessment mea-
sures, with a therapist available to answer questions. To min-
imize socially desirable responding, participants were aware
that therapists leading the groups would be blind to their
responses. All questionnaires were self-report, and all ques-
tions about the relationship referred to the relationship leading
to domestic violence charges (i.e., not necessarily the relation-
ship at the time of questionnaire completion). The rest of the
time during these sessions was devoted to individual therapy
focused on building a therapeutic alliance between the partic-
ipant and therapist and preparing participants for the group
component of treatment, not evaluated here.
Results
Descriptive Statistics
SPSS 17 (SPSS Inc 2008) was used to report descriptive
statistics and conduct basic analyses (e.g., correlation). On
291
average, men in this sample were 38.7 years (SD =
10.93 years). At evaluation, 26.8 % of participants were
married, 32.3 % were separated or divorced, and the remain-
der was single or reported their relationship status as other.
In Madrid, individuals can be referred to batterer programs for
physical, sexual, or psychological perpetration. In this sample
91.1 % were referred for physical aggression and 8.9 % were
referred for psychological aggression. No participants were
referred for sexual aggression.
The vast majority of participants reported being Spanish
(n=102). Of the remaining participants, 56 were from nations
in Central or South America, 11 were from Africa, seven were
Caribbean, six were from other European countries, two were
from Asia, and six reported multiple nationalities. Demo-
graphically, Spanish men were significantly younger than
non-Spanish men, t (192)=3.23; p<.01. Nevertheless, no
significant difference emerged between Spanish and immi-
grant men in the prevalence of any aggression perpetration
variable (psychological aggression, physical assault, sexual
coercion, or injury).
Prevalence of Aggression
Table 1 illustrates the annual prevalence rates for psycholog-
ical and physical aggression, as well as mens self-reports of
annual aggression frequency. The majority of men reported
the occurrence of psychological perpetration or victimization
in the year prior, with over half reporting physical perpetra-
tion, victimization, or both. Close to half of men in this sample
reported perpetrating injury against their partners, and roughly
a quarter reported being injured by their partners. Almost one
fifth of the sample also reported perpetrating sexual aggres-
sion in the past year; a smaller number reported being sexually
victimized by their partner. Men reported that psychological
and physical victimization occurred more frequently than
perpetration, and that injury perpetration occurred more fre-
quently than victimization.2
Bivariate Correlations
Table 2 includes the mean and standard deviation of each
variable on the diagonal. Above the diagonal are the bivar-
iate correlations between specified variables. Patterns of
correlation generally reflect paths hypothesized in the tested
models. Specifically, MSI-BPD (Zanarini et al. 2003) scores
exhibited significant bivariate associations with alcohol
problems, psychological aggression, and physical aggres-
sion. The AUDIT (Saunders et al. 1993) was significantly
2
As this study is about aggression, with a specific focus on physical and
psychological violence, data on the CTS2 Negotiation subscale are not
presented here.
292 J Fam Viol (2014) 29:287297

Table 1 12-month prevalence

and frequency of aggression Prevalence of aggression
(N=194) Male perpetration (%) Male victimization (%)
Psychological 80.8 81.2
Physical 58.9 54.2
Sexual 19.6 12.4
Injury 45.6 23.8
Frequency of aggression
Male perpetration mean (SD) Male victimization mean (SD) P-value
Psychological 8.62 (8.31) 10.49 (9.81) .00
Physical 3.67 (6.19) 5.36 (8.89) .00
Sexual 0.77 (2.14) 0.51 (2.69) .22
Injury 1.21 (1.81) 0.84 (1.94) .00

correlated with psychological and physical perpetration. Path Analytic Results

Finally, physical and psychological aggression variables
were highly correlated with each other, as expected.
Path Analytic Strategy and Model Fit
Path analyses were conducted to test the hypothesized model
illustrated in Fig. 1, using Mplus 6.0 (Muthn and Muthn
2010). The robust maximum likelihood (MLR) estimate,
which does not assume multivariate normality, was used
for these analyses. Following Hu and Bentlers (1998)
guidelines, fit was evaluated using multiple indices: the
Comparative Fit Index (CFI; Bentler 1990), Standardized
Root Mean Square Residual (SRMR; Bentler 1995), Root
Mean Square Error of Approximation (RMSEA; Steiger
and Lind 1980), and chi-square tests. For these analyses,
CFI > 0.95, SRMR < 0.05, RMSEA < 0.05, and a non-
significant chi-square each signified good fit.
The original model is illustrated in Fig. 1. The data fit the
model well: n=180, CFI=1.00, SRMR=0.01, RMSEA=0.00
(90 % CI=0.000.05) and 2 (4)=1.05, p=.90. Unexpectedly,
psychopathy was not related to any hypothesized paths (e.g.,
alcohol problems, psychological perpetration, or physical per-
petration). This finding remained when overall LSRP scores,
primary subscale scores, or secondary subscale score were
used to measure the construct. In addition, neither of the paths
linking one partners psychological aggression to the other
partners physical aggression was significant. The paths
linking physical perpetration and victimization were also
non-significant, though the path from perpetration to victim-
ization approached significance. Finally, alcohol problems
were not associated with psychological perpetration. Trim-
ming the model to remove non-significant paths did not
improve fit so all paths were retained (see Fig. 2 for
the model including all standardized path coefficients).

Table 2 Means, standard deviations, and correlations among all variables (N=194)

LPSP MSI- AUDIT Psychological Psychological Physical Physical

BPD perpetration victimization perpetration victimization

LPSP 51.20 (8.60) 0.45* 0.30* 0.24** 0.07 0.30** 0.12

MSI-BPD 1.95 (2.20) 0.52** 0.50** 0.26** 0.55** 0.28**
AUDIT 5.97 (5.70) 0.36** 0.19** 0.48** 0.23**
Psychological perpetration 8.44 (8.16) 0.69** 0.69** 0.54**
Psychological victimization 10.43 (9.86) 0.46** 0.72**
Physical perpetration 3.61 (6.09) 0.64**
Physical victimization 5.33 (8.92)

Means (standard deviations) appear across the diagonal. For correlations: *p<.05; **p<.01
J Fam Viol (2014) 29:287297
Fig. 2 Final model (n=180)
Psychopathic Traits
.14
Borderline Traits
.45**
As can be seen in Fig. 2, the model supported the bi-
directionality of psychological aggression, as well as the
links between psychological and physical perpetration,
and between psychological and physical victimization.
Notably, borderline traits were significantly related to all
of the hypothesized variables.
For a more nuanced understanding of this association,
endorsement of individual MSI-BPD items was reviewed.
The most frequently endorsed item was Item 1 (46.6 % en-
dorsed), which asks whether close relationships have been
troubled by a lot of arguments or repeated breakups. Item
6 (endorsed by 33.3 %) was the next most frequent, and asks
about distrust towards others. The least frequently endorsed
items were Items 2 (6.3 %) and 9 (7.3 %) which ask about self-
harm and identity instability, respectively.
Mediation
Post-hoc mediation analyses were conducted on the associ-
ation between borderline traits and physical aggression
using bias-corrected bootstrapping with 500 re-samples in
Mplus 6.0 (Muthn and Muthn 2010). Bootstrapping is
currently the preferred method to determine mediation.
The full information maximum likelihood (FIML) estima-
tor, which accounts for missingness, was used for these
analyses. Tests of mediation suggested that psychological
perpetration is a significant partial mediator of the associa-
tion between borderline traits and physical perpetration
(0.20; 95%CI=0.11 to 0.30). However, alcohol problems
were not a significant partial mediator of the association
between borderline traits and physical aggression (0.07;
95%CI=0.01 to 0.14). The sum of indirect effects was
0.27 (95%CI=0.14 to 0.39).
293
Psychological .65** Physical
Victimization Victimization
.-.22
.00 .49** .33** .35 .28
-.02
-.14
Psychological Physical
Perpetration .55** Perpetration
.37**
.16*
.11
.14*
Alcohol Use
Discussion
Prevalence of Aggression
Rates of psychological and physical aggression in the present
study were lower than rates found in a US batterer sample (e.g.,
Vega and OLeary 2007), and men in the present study report-
ed more victimization than perpetration in the year prior.
Although rates for sexual aggression in this sample were also
lower than rates found by Vega and OLeary (2007), men in
this sample did not report any difference between perpetration
and victimization rates for this variable. Injury perpetration
was quite high in this sample with close to half the sample
(45.6 %) reporting perpetration, remarkably similar to rates in
Vega and OLearys (2007) batterer sample (48.8 %). These
high self-reported injury rates are consistent with the idea that
those who are mandated for batterer programs engage in severe
forms of violence that lead to negative physical repercussions.
It is somewhat surprising that greater physical and psycho-
logical victimization than perpetration is reported in this sam-
ple of men court-referred for perpetrating aggression against
their romantic partners. This finding may reflect mens
underreporting of their own aggression (e.g., Szinovacz
1983). It may also be because men were asked to report on
violence in the relationship they were in when they became
involved with the legal system for domestic violence, and
their memory for the details of that relationship may have
diminished with time. Another possibility is that aggression
may have occurred bi-directionally in these relationships.
Participant path model data indicates reciprocal paths for
psychological (though not physical) aggression. Nevertheless,
it is clear that men in this sample reported perpetrating injury
more frequently than they were victimized, suggesting clear
negative physical consequences for these mens partners. This
294
result is consistent with Archers (2000) meta-analysis, which
found that the negative physical impact of aggression is great-
er on female victims than on male victims.
Path Analyses
As hypothesized, borderline traits put participants at higher
risk for psychological and physical aggression, and at risk for
alcohol use. Further, psychological aggression partially medi-
ated the link between borderline traits and physical aggres-
sion. Analysis of the specific MSI-BPD (Zanarini et al. 2003)
items endorsed by participants suggest that the interpersonal
aspects of BPD related to relationship instability such as
arguments and distrust may be driving the association of
BPD with aggression and alcohol problems in this sample.
Conceptually, the distress caused by perceived and actual
relationship instability may lead to impulsive, reactive, and
violent responses in an individual with borderline traits. Fur-
thermore, personality pathology and other mental health prob-
lems may lead to a need for self-medication with alcohol or
illicit substances which are known to decrease inhibitions and
increase impulsivity, in turn leading to conflict.
Also as hypothesized, and consistent with a plethora of
research on the topic, psychological aggression was a risk
factor for physical aggression, when either perpetration or
victimization was measured. A reciprocal association was
found between psychological perpetration and victimization;
this result seems reasonable as bilateral psychological aggres-
sion occurs within the context of couple arguments or conflict
(e.g., trading insults) according to scores of studies (OLeary
2001). The path from physical perpetration to physical vic-
timization trended towards significance. Although it should be
interpreted with much caution, this may reflect females
responding to their partners aggression in self-defense.
Another finding that trended towards significance suggests
that greater levels of male psychological victimization are
associated with lower levels of male physical perpetration.
However, this finding is not significant at the 0.05-level and
should be interpreted cautiously. Results linking alcohol prob-
lems to aggression were mixed in this sample. Alcohol was
not associated with psychological aggression, but was a risk
factor for physical aggression. Despite a ceiling because of
study exclusion for excessive alcohol use, alcohol problems
were a risk factor for physical aggression.
Whereas the AUDIT includes some general questions that
are open to interpretation based on cultural factors, this mea-
sure also invites respondents to answer specific questions
about alcohol, thereby requesting behavioral reports rather
than subjective comparison to what an individual believes to
be normative behavior. In a society where drinking is norma-
tive, more accurate representations of problem drinking may
be obtained using behaviorally specific questions. In addition,
the AUDIT was developed for cross-cultural use (Saunders
J Fam Viol (2014) 29:287297
et al. 1993) and thereby seems an appropriate tool to measure
alcohol use in the Spanish context. Alcohol use measured in
this manner is indeed associated with physical aggression.
Finally, and contrary to predictions, psychopathy was not
associated with any other variable in the model, despite sig-
nificant first-order correlations with alcohol use and violence
perpetration. This was the case when the full scale measure
was used, and when each subscale was used individually. This
finding indicates psychopathy does not predict aggression in
this sample after taking into account other relevant predictors.
Alternately, the measure used (or perhaps self-report in gen-
eral) might not be ideal to assess psychopathy in this popula-
tion; the non-significant result may also be as a result of
cultural differences in the conceptualization of psychopathy
or the presence of a third variable.
Implications
Results suggest evidence for a multivariate, theoretically-
based conceptualization of partner aggression. Prevalence
rates for psychological and physical aggression in this Spanish
batterer sample are similar to (though slightly lower than) US
batterer samples. Further, despite the two major cultural dif-
ferences highlighted at the start of this paper related to the rise
of relationship aggression as a public health problem and the
role of alcohol use, borderline traits, alcohol problems, and
psychological aggression have been associated with aggres-
sion this Spanish sample, as they have been in other research
with US samples.
The consistent association between alcohol problems and
physical aggression across countries, and despite the ceiling
for alcohol misuse imposed in this study, indicates that alcohol
use is a real problem in regards to aggression perpetration.
Consistent evidence for such a link has also been found in
other international research from the United States (e.g., Stuart
et al. 2006) and the Ukraine (OLeary et al. 2008).
A major treatment implication of this study is that assess-
ment and intervention around individual risk factors for ag-
gression (particularly borderline traits and alcohol problems)
may play a role in decreasing the prevalence of psychological
and physical aggression in this at-risk group. This is especially
promising as Dialectical Behavior Therapy (Linehan 1993)
has great empirical support in the treatment of borderline traits
and motivational enhancement (Miller and Rollnick 1991) has
been successfully used to address drinking problems. As men
in batterer programs may not be able to address their aggres-
sion in a dyadic context (nor is this recommended when safety
is a concern), individual treatments targeting dyadic and indi-
vidual issues play a special role in the treatment of individuals
arrested for domestic violence.
A second significant treatment implication of this research
is related to perpetration rates. Although a number of studies
focus on recidivism in a batterer population (e.g., Gondolf
J Fam Viol (2014) 29:287297
1999), there is extremely limited prevalence data for aggres-
sion prior to treatment in a batterer population. Although this
may be because physical assault perpetration is assumed to be
100 % in a batterer sample, this data suggests that assumption
may be inaccurate since a proportion of men were referred by
the court for psychological aggression. Reporting the relative
prevalence of physical, psychological, and sexual aggression
in batterer samples is recommended. Furthermore, since the
topography of an individuals aggressive behaviors (e.g., type,
frequency, and severity) may be related to treatment outcome,
recidivism, and other factors; this area is well worth further
study.
In a related vein, although caution must be taken when
comparing US and Spanish batterer samples because of dif-
ferent legal requirements for court mandated treatment, the
lower prevalence of physical aggression in the current sample
compared to US samples may be explained by retrospective
reporting bias, or by participants faking good. It is possible
the discrepant findings may also reflect a real difference in the
prevalence of aggression across these two countries. This last
perspective is supported by the high perpetration reports for
some types of violence, despite the fact that men in this study
are in the Spanish legal system and are therefore unlikely to
benefit from over-reporting aggression (faking bad). For
example, although none of the participants in this study were
referred for sexual aggression, 19.6 % reported perpetration.
Furthermore, participants reported that mens injury perpetra-
tion was about twice as prevalent, and significantly more
frequent, than mens victimization.
Limitations and Future Directions
All data were collected from the male partner only, who was
asked about the relationship he was in when he became
mandated for batterer treatment. As these data are retrospec-
tive, cross sectional, and utilize self-report measures, future
directions may include the use of prospective and/or longitu-
dinal designs, and the use of multiple reporters (e.g., partner
report, police reports). In addition, because of the exclusion
criteria for the treatment program in which participants of this
study were enrolled, men with significant alcohol use or any
substance use were excluded from these analyses.
In addition to addressing the above limitations, future
research should clarify the role of other variables of interest
in aggression perpetration. For instance, psychopathy was
hypothesized to predict aggression, but did not do so. Further
research on the specific aspects of psychopathy that may be
associated with relationship violence, as well as the best
method for measuring this variable, can add to the literature
associating personality pathology and relationship conflict.
It would also be beneficial to further incorporate the socio-
cultural context of the country that is being studied (e.g.,
cultural attitudes toward masculinity, toward women, or
295
toward intimate partner aggression). Future research identify-
ing the prevalence and risk factors of aggression across sam-
ples (community, treatment samples) across cultural contexts;
and using male, female, and dyadic reports can reveal univer-
sal and culturally specific risk factors for relationship vio-
lence. For practical reasons, as well as reasons of statistical
power, this study assessed specific, limited types of psycho-
pathology. As a range of mental illness has been linked to
physical perpetration, one suggestion for future research is to
identify common variables (e.g., impulsivity, the experience
of trauma) or explore a broader range of psychopathology to
determine theoretical and empirical interrelationships between
variables.
Acknowledgments The authors would like to thank Gregory L. Stuart
for his valuable suggestions; E. David Klonsky, Bonita London-
Thompson, and Evelyn Bromet for their excellent feedback; and Thomas
Olino for his statistical expertise. We would also especially like to thank
Natalia Redondo for organizing the data and Virginia Navascues for her
translations. Finally, we would like to thank our anonymous reviewers for
their helpful suggestions.
References
Allen, J. P., Litten, R. Z., Fertig, J. B., & Barbor, T. (1997). A review of
research on the Alcohol Use Disorders Identification Test (AUDIT).
Alcoholism: Clinical and Experimental Research, 21, 613619.
American Psychiatric Association. (2000). Diagnostic and statistical
manual of mental disorders (4th ed., text rev.). Washington, DC:
Author.
Archer, J. (2000). Sex differences in aggression between heterosexual
partners: a meta- analytic review. Psychological Bulletin, 126, 651
680. doi:10.1037/0033-2909.126.5.651.
Bentler, P. M. (1990). Comparative fit indexes in structural models.
Psychological Bulletin, 107, 238246. doi:10.1037/0033-2909.
107.2.238.
Bentler, P. M. (1995). EQS structural equations program manual.
Encino: Multivariate Software.
Bloomfield, K., Stockwell, T., Gmel, G., & Rehn, N. (2003). International
comparisons of alcohol consumption. Alcohol Research & Health,
27, 95109. Retrieved from: http://www.niaaa.nih.gov/Publications/
AlcoholResearch/.
Cordova, J. V., Jacobson, N. S., Gottman, J. M., Rushe, R., & Cox, G.
(1993). Negative reciprocity and communication in couples with a
violent husband. Journal of Abnormal Psychology, 102, 559564.
doi:10.1037/0021-843X.102.4.559.
Death of a wife in Spain brings outcry on domestic violence. (1997,
December 26). The NewYork Times, p. A5.
Dutton, D. G. (1994). Behavioral and affective correlates of borderline
personality organization in wife assaulters. International Journal of
Law and Psychiatry, 17, 265277. doi:10.1016/0160-2527(94)
90030-2.
Foran, H. M., & OLeary, K. D. (2008). Alcohol and intimate partner
violence: a meta-analytic review. Clinical Psychology Review, 28,
12221234. doi:10.1016/j.cpr.2008.05.001.
Garcia-Moreno, C., Jansen, H. A. F. M., Ellsberg, M., Heise, L., & Watts.
(2006). Prevalence of intimate partner violence: findings from the
WHO multi-country study on womens health and domestic vio-
lence. The Lancet, 368, 12601269. doi:10.1016/S0140-6736(06)
69523-8.
296
Gondolf, E. W. (1999). A comparison of four batterer intervention sys-
tems: do court referral, program length, and services matter?
Journal of Interpersonal Violence, 14, 4161. doi:10.1177/
088626099014001003.
Graa, J. L. (2009, November). Prevalence of partner aggression in a
Spanish sample belonging to the Comunidad of Madrid. In K. D.
OLeary (Chair), Universal or country-specific vulnerability factors
for partner violence: intervention implications. Symposium con-
ducted at the 43rd annual meeting of the Association for
Behavioral and Cognitive Therapies, New York, NY.
Gual, A. (2006). Alcohol in Spain: is it different? Addiction, 101, 1073
1077. doi:10.1111/j.1360-0443.2006.01518.x.
Hassan, F., Sadowski, L. S., Bangdiwala, S. I., Vizcarra, B., Ramiro, L.,
De Paula, C. S., et al. (2004). Physical intimate partner violence in
Chile, Egypt, India, and the Philippines. Injury Control and Safety
Promotion, 11, 111116.
Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male
batterers: three subtypes and the differences among them.
Psychological Bulletin, 116(3), 476497. doi:10.1037/0033-2909.
116.3.476.
Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure
modeling: sensitivity to underparametrized model misspecification.
Psychological Methods, 3, 424453. doi:10.1037/1082-989X.3.4.
424.
Jose, A., & OLeary, K. D. (2009). Prevalence of partner aggression in
representative and clinic samples. In K. D. OLeary & E. Woodin
(Eds.), Psychological and physical aggression in couples: Causes
and Interventions (pp. 1535). Washington, DC: American
Psychological Association.
Kenny, D. A. (1996). Models of non-independence in dyadic research.
Journal of Social and Personal Relationships, 13, 279294. doi:10.
1177/0265407596132007.
Kim, H. K., & Capaldi, D. M. (2004). The association of antisocial
behavior and depressive symptoms between partners and risk for
aggression in romantic relationships. Journal of Family Psychology,
18, 8296. doi:10.1037/0893-3200.18.1.82.
Kishor, S., & Johnson, K. (2004). Profiling domestic violencea multi-
country study. Retrieved January 2, 2013, from: http://www.
measuredhs.com/pubs/pdf/OD31/OD31.pdf.
Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing
psychopathic attributes in a noninstitutionalized population. Journal
of Personality and Social Psychology, 68(1), 151158. doi:10.1037/
0022-3514.68.1.151.
Lilienfeld, S. O. (2010, February). Diagnosing psychopathic personality.
Grand Rounds presented at Montefiore Medical Center, Albert
Einstein College of Medicine, Bronx, NY.
Lilienfeld, S. O., & Fowler, K. A. (2006). The self-report assessment of
psychopathy: Problems, pitfalls, and promises. In C. J. Patrick (Ed.),
Handbook of psychopathy (pp. 107132). New York: Guilford
Press.
Linehan, M. M. (1993). Cognitive-behavioral treatment of Borderline
Personality Disorder. New York: Guilford Press.
Lynam, D. R., Whiteside, S., & Jones, S. (1999). Self-reported psychop-
athy: A validation study. Journal of Personality Assessment, 73(1),
110132. doi:10.1207/S15327752JPA730108.
Mauricio, A. M., Tein, J., & Lopez, F. G. (2007). Borderline and antiso-
cial personality scores as mediators between attachment and inti-
mate partner violence. Violence and Victims, 22, 139157. doi:10.
1891/088667007780477339.
Maxwell, J. C. (2008). Are we becoming more alike? Comparison of
substance use in Australia and the United States as seen in the 1995,
1998, 2001 and 2004 national household surveys. Drug and Alcohol
Review, 27, 473481. doi:10.1080/0959230802090055.
Medina-Ariza, J., & Barberet, R. (2004). Intimate partner violence in
Spain: findings from a national survey. Violence Against Women, 9,
302322. doi:10.1177/1077801202250073.
J Fam Viol (2014) 29:287297
Miller, W. R., & Rollnick, S. (1991). Motivational interviewing:
Preparing people to change addictive behavior. New York:
Guilford Press.
Murdoch, D., Pihl, R. O., & Ross, D. (1990). Alcohol and crimes of
violence: present issues. International Journal of the Addictions, 25,
10651081.
Murphy, C. M., & OLeary, K. D. (1989). Psychological aggression
predicts physical aggression in early marriage. Journal of
Consulting and Clinical Psychology, 57, 579582. doi:10.1037/
0022-006X.57.5.579.
Murphy, C. M., Meyer, S., & OLeary, K. D. (1993). Family of origin
violence and MCMI- II psychopathology among partner assaultive
men. Violence and Victims, 8(2), 165176.
Murphy, C. M., OFarrell, T. J., Fals-Stewart, W., & Feehan, M. (2001).
Correlates of intimate partner violence among male alcoholic pa-
tients. Journal of Consulting Psychology, 69, 528540. doi:10.1037/
0022-006X.69.3.528.
Muthn, B. O., & Muthn, L. K. (2010). Mplus (version 6.0) [Computer
software]. Los Angeles: Muthn & Muthn.
Nisonoff, L., & Bittman, I. (1979). Spouse abuse: incidence and relation-
ship to selected demographic variables. Victimology, 4, 131140.
OLeary, K. D. (1993). Through a psychological lens: Personality traits,
personality disorders, and levels of violence. In R. J. Gelles & D. R.
Loseke (Eds.), Current controversies on family violence (pp. 730).
Newbury Park: Sage.
OLeary, K. D. (2001). Psychological abuse: A variable deserving
critical attention in domestic violence. In K. D. OLeary & R.
D. Maiuro (Eds.), Psychological abuse in violent domestic
relations (pp. 328).
OLeary, K. D., Slep, A. M. S., & OLeary, S. G. (2007). Multivariate
models of mens and womens partner aggression. Journal of
Consulting and Clinical Psychology, 75, 752764. doi:10.1037/
0022-006X.75.5.752.
OLeary, K. D., Tintle, N., Bromet, E. J., & Gluzman, S. F. (2008).
Descriptive epidemiology of intimate partner aggression in
Ukraine. Social Psychiatry and Psychiatric Epidemiology, 43,
619626. doi:10.1007/s00127-008-0339-8.
Organic Act 1/2004. (2004). Integrated protection measures against gen-
der violence. Boletin Oficial del Estado [Official State Gazette], no.
313 (December 29, 2004). English version retrieved January 12,
2013 from: http://www.coe.int/t/dghl/standardsetting/violence/
Organic%20Act%20on%20Violence.pdf.
Organisation of Economic Cooperation and Development. (2009).
OECD health data 2009. Retrieved May 3, 2010 from: http://
www.oecd.org/document/16/0,3343,en_2649_34631_2085200_1_
1_1_1,00.html.
Porcerelli, J. H., Cogan, R., & Hibbard, S. (2004). Personality character-
istics of partner violent men: a Q-sort approach. Journal of
Personality Disorders, 18, 151162. doi:10.1521/pedi.18.2.151.
32776.
Saunders, J. B., Aasland, O. G., Babor, T. F., de la Puente, J. R., & Grant,
B. F. (1993). Development of the Alcohol Use Disorders
Identification Test (AUDIT): WHO collaborative project on early
detection of persons with harmful alcohol consumption. Addiction,
88(6), 791804. doi:10.1111/j.1360-0443.1993.tb02093.x.
South, S. C., Turkheimer, E., & Oltmanns, T. F. (2008). Personality
disorder symptoms and marital functioning. Journal of Clinical
and Consulting Psychology, 76, 769780. doi:10.1037/a0013346.
SPSS Inc. (2008). SPSS statistics for windows (version 17.0) [computer
software]. Chicago: Author.
Steiger, J. H., & Lind, J. C. (1980, May). Statistically based tests for the
number of common factors. Paper presented at the annual meeting of
the Psychometric Society, Iowa City, IA.
Stepp, S. D., Trull, T. J., & Sher, K. J. (2005). Borderline personality
features predict alcohol use problems. Journal of Personality
Disorders, 19, 711722. doi:10.1521/pedi.2005.19.6.711.
J Fam Viol (2014) 29:287297
Stith, S. M., Smith, D. B., Penn, C. E., Ward, D. B., & Tritt, D. (2004).
Intimate partner physical abuse perpetration and victimization risk
factors: a meta-analytic review. Aggression and Violent Behavior, 10,
6598. doi:10.1016/j.avb.2003.09.001.
Straus, M. A., & Gelles, R. J. (Eds.). (1990). Physical violence in
American families: Risk factors and adaptations to violence in 8,
145 families. New Brunswick: Transaction Publishers.
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996).
The revised conflict tactics scales (cts2): development and preliminary
psychometric data. Journal of Family Issues, 17(3), 283316.
Stuart, G. L., & Holtzworth-Munroe, A. (2005). Testing a theoretical
model of the relationship between impulsivity, mediating variables,
and husband violence. Journal of Family Violence, 20, 291303.
doi:10.1007/s10896-005-6605-6.
Stuart, G. L., Meehan, J. C., Moore, T. M., Morean, M., Hellmuth, J., &
Follansbee, K. (2006). Examining a conceptual framework of inti-
mate partner violence in men and women arrested for domestic
violence. Journal of Studies on Alcohol, 67, 102112. Retrieved
at: http://www.jsad.com/.
Stuart, G. L., Temple, J. R., Follansbee, K. W., Buccosi, M. M., Hellmuth,
J. C., & Moore, T. M. (2008). The role of drug use in a conceptual
model of intimate partner violence in men and women arrested for
domestic violence. Psychology of Addictive Behaviors, 22, 1224.
doi:10.1037/0893-164X.22.1.12.
Szinovacz, M. E. (1983). Using couple data as a methodological tool: the case
of marital violence. Journal of Marriage and the Family, 45, 633644.
297
Tjaden, P., & Thoennes, N. (2000). Prevalence and consequences of
male-to-female intimate partner violence as measured by the
National Violence Against Women Survey. Violence Against
Women, 6, 142161. doi:10.1177/10778010022181769.
Tweed, R. G., & Dutton, D. G. (1998). A comparison of impulsive and
instrumental subgroups of batterers. Violence and Victims, 13(3),
217230.
UNICEF. (2009a). At a glance: Spain. Retrieved December 23, 2009,
from http://www.unicef.org/infobycountry/spain_statistics.html.
UNICEF. (2009b). At a glance: United States of America. Retrieved
December 23, 2009, from http://www.unicef.org/infobycountry/
usa.html.
United Nations Development Programme. (2009). Human development
report 2009: Overcoming barriers: Human mobility and
development. Retrieved December 23, 2009, from http://hdr.undp.
org/en/media/HDR_2009_EN_Complete.pdf.
Vega, E. M., & OLeary, K. D. (2007). Test-retest reliability of the revised
Conflict Tactics Scales (CTS2). Journal of Family Violence, 22,
703708. doi:10.1007/s10896-007-9118-7.
World Health Organization. (2000). The world health report 2000.
Retrieved from http://www.who.int/whr/2000/en/annex09_en.pdf.
Zanarini, M. C., Vujanovic, A. A., Parachini, E. A., Boulanger, J. L.,
Frankenburg, F. R., & Hennen, J. (2003). A screening measure for
BPD: the Mclean Screening Instrument for Borderline Personality
Disorder (MSI-BPD). Journal of Personality Disorders, 17(6), 568
573. doi:10.1521/pedi.17.6.568.25355.