The Impact of Psychological Aggression on Women's Mental Health and Behavior: The Status of the Field

Abstract

Potential effects of psychological aggression have been investigated in terms of mental health disorders/symptoms, physical manifestations, personality traits, increases in physical and psychological aggression, and relationship changes and cessation. Probably because of methodological difficulties inherent in assessing this complex phenomenon, even the most commonly assessed outcome variables have produced contradictory results, with the exception of leaving or desiring to leave the psychologically abusive relationship. Some of the significant findings require more corroboration due to the small number of studies addressing them. The relative role of physical abuse in relation to psychological abuse has yet to be disentangled. Mediator/moderator variables are likely extremely important for explaining why particular consequences do not consistently result from psychological aggression. Although the hypothesized consequences/outcomes seem appropriate for further study, the most significant contribution may lie in intensive investigation of long-term, enduring psychological abuse that produces major personality and affective shifts in the recipient.

Full text

THE IMPACT OF PSYCHOLOGICAL AGGRESSION
ON WOMEN’S MENTAL HEALTH AND BEHAVIOR
The Status of the Field
DIANE R. FOLLINGSTAD
University of Kentucky
Potential effects of psychological aggression have been investigated in terms of
mental health disorders/symptoms, physical manifestations, personality traits,
increases in physical and psychological aggression, and relationship changes and
cessation. Probably because of methodological difficulties inherent in assessing
this complex phenomenon, even the most commonly assessed outcome variables
have produced contradictory results, with the exception of leaving or desiring to
leave the psychologically abusive relationship. Some of the significant findings
require more corroboration due to the small number of studies addressing them.
The relative role of physical abuse in relation to psychological abuse has yet to be
disentangled. Mediator/moderator variables are likely extremely important for
explaining why particular consequences do not consistently result from psycho-
logical aggression. Although the hypothesized consequences/outcomes seem
appropriate for further study, the most significant contribution may lie in inten-
sive investigation of long-term, enduring psychological abuse that produces
major personality and affective shifts in the recipient.
Key words: psychological aggression; mental health effects; psychological and physical abuse
271
AUTHOR’S NOTE: Please address correspondence to Diane R. Follingstad, PhD, Women’s Circle Endowed Chair in the Center for
Research on Violence Against Women, Department of Psychiatry, University of Kentucky College of Medicine, 3470 Blazer Parkway,
Lexington, KY 40509; e-mail: Follingstad@uky.edu.
TRAUMA, VIOLENCE, & ABUSE, Vol. 10, No. 3, July 2009 271-289
DOI: 10.1177/1524838009334453
© 2009 The Author(s)
THE STUDY OF PSYCHOLOGICAL ABUSE,
following the groundswell of research investi-
gating the parameters and impact of physical
abuse, was a natural segue in the mental health
field. A number of factors lent credence to the
focus on these problematic nonphysical inter-
personal tactics as an independent research
arena. In particular, these included the high
prevalence of psychological maltreatment that
appeared to accompany domestic violence, the
literature which suggested that psychological
maltreatment could predict the occurrence of
physical abuse (e.g., Murphy & O’Leary, 1989),
and the reports that many battered women
perceived psychological abuse to be worse
than physical abuse (e.g., Follingstad, Rutledge,
Berg, Hause, & Polek, 1990).
The major focus for the mental health field
has been to learn the ways in which psycho-
logical abuse impacts recipients, which in turn

272 TRAUMA, VIOLENCE, & ABUSE / July 2009
may suggest types of intervention or treatment
or prevention efforts. Certainly, the term abuse
suggests that egregious actions have occurred
interpersonally which would inevitably produce
harmful outcomes. In fact, some researchers/
theorists have included negative psychological
effects on the recipient as part and parcel of the
actual definition of psychological abuse. For
example, Loring (1994) emphasized the con-
sequences of “diminishment” and “destruc-
tion of the inner self” in her definition.
Hoffman (1984) characterized abuse as pro-
ducing consequences on the woman’s capac-
ity to work or interact with family/society, or
as affecting her physical or mental health.
Murphy and Hoover (1999) incorporated such
effects as fear, dependency, and damage to
the self-concept into their definition of psy-
chological abuse.
Although most persons would assume nega-
tive outcomes from psychologically abusive
actions by an intimate partner, there are many
problems and issues that must be addressed
before we clearly know whether a direct link
exists between this type of intimate partner
violence (IPV) and significant negative out-
comes. It is important to understand the state of
the research methodology to use a realistic and
necessary lens through which the results must
be considered. The major problems for this field
of study fall into the categories of definitional/
conceptual issues and methodological issues.
(A brief discussion of these is found below,
but for a thorough analysis of significant prob-
lems regarding measurement of psychological
aggression, see Follingstad, 2007.)
DEFINITIONAL AND CONCEPTUAL
PROBLEMS
Definitional and conceptual problems have
resulted from inconsistencies in defining what
behaviors/actions are included, assumptions
that current measures truly assess a person’s
experience of psychological abuse, awkward
application of physical abuse models to this
more complex phenomenon, and varied lan-
guage used by researchers, theorists, and practi-
tioners. Problems defining “psychological
abuse” and definitional confusion have been
acknowledged by most researchers in this area
(e.g., Maiuro, 2001; Schumacher, Slep, &
Heyman, 2001), but inconsistencies remain.1
Discrepant views and definitional approaches
still prevent development of a coherent concep-
tualization, much less operationalization of this
phenomenon. For example, scales purporting
to measure psychological abuse range widely
in terms of the categories represented in the
literature. Follingstad, Coyne, and Gambone
(2005) delineated psychological abuse items
into 17 categories which appeared to be rela-
tively independent from each other, for example,
treatment as an inferior, monitoring behavior,
public embarrassment, verbal abuse. Even the
term psychological abuse presents its own prob-
lems as our lack of sophisticated measurement
cannot tell us exactly, except for extreme cases,
when behavior has moved beyond boorish,
inept, unskilled, or even aversive actions into a
realm unambiguously considered “abusive.”
Because of this, a number of researchers have
recently begun to promote the use of psychologi-
cal aggression as a more useful term which rep-
resents the full range of potentially negative
intimate interpersonal behaviors without imply-
ing, at this point, that all psychological aggres-
sion is abusive. (For the purpose of this article,
however, the term psychological abuse will be
used to indicate the more severe end of the
range of psychological aggression—i.e., sug-
gesting a greater level of seriousness, negative
intent, and emotional harm—whereas the terms
psychological aggression and maltreatment will be
used interchangeably to cover the full range of
psychological actions that have been proposed
to be problematic.)
The major difficulties for definition and con-
ceptualization of psychological aggression
include the following: (a) knowing exactly
which major categorizations of negative inti-
mate relationship behavior (i.e., verbal abuse,
emotional abuse, intimidation/dominance, con-
trol) constitute or should be included in the defi-
nition; (b) determining whether there are
unique subcategories of psychological abuse
which function as independent factors; (c) know-
ing whether specification of more distinct sub-
types of psychological abuse is important or
useful for prediction of impact; (d) knowing

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 273
whether persons who report the presence of
particular behaviors actually experienced them
as significantly harmful or problematic (which
is assumed by researchers); (e) understanding
whether our current operationalization of psy-
chological abuse through listing aversive inter-
personal behaviors validly captures the
phenomenon; and (f) determining the implica-
tions of frequency and/or severity of the psycho-
logically aggressive behaviors.
METHODOLOGICAL PROBLEMS
Methodological problems not only result
directly from problems in conceptualization but
also arise from many other sources. Certainly,
one can imagine how a study’s results might
differ when one researcher views psychologi-
cal aggression solely as verbal criticism and
ridicule whereas another researcher uses a very
broad conceptualization in which verbal
aggression is only one of many types of aver-
sive actions. However, a number of conceptual-
ization inconsistencies might be handled
relatively easily were it not for other method-
ological problems that significantly hinder our
ability to validly indicate if, and when, and
how often, psychological abuse has occurred in
a particular relationship or within a particular
group. Basically, researchers have approached
the measurement of this complex interactional
process occurring within a system by collecting
self-report information from one member of a
couple as to whether certain behaviors (desig-
nated by the individual researcher as psycho-
logical abuse) have occurred. This unsophisticated
approach has to be acknowledged as highly ques-
tionable for establishing validity. Furthermore,
current measures are not designed to provide
the necessary information for interpretation,
for research purposes, as to whether an action
is psychological aggression or abuse. Factors
which are highly likely to affect interpretation
include the following: (a) the context in which
a behavior occurred; (b) when a behavior
occurred in a sequence; (c) how the occurrence
was interpreted by the recipient; (d) whether
the recipient experienced the behavior as
“abusive”; (e) whether a simple description of
a behavior on a checklist captures the desired
elements for accurate labeling; (f) whether par-
ticular personality constellations, cognitive
sets, interpersonal styles, and the like influ-
ence the interpretation of events in inaccurate
ways; (g) whether the language used on check-
lists of behaviors is subjectively interpreted
similarly by all respondents; (h) which psycho-
logically aversive behaviors might be viewed
as representative of intimate conflict with mini-
mal harm rather than abuse; (i) our current
inability to handle inconsistencies in interpre-
tations among the recipient, initiator, and
observer as to whether “abuse” actually
occurred; (j) lack of knowledge as to whether
particular frequencies and/or more severe
types of psychological aggression should be
weighted in scoring formats; (k) potential prob-
lem with quantitative scores when the same
score could mean the person experienced high
frequencies of mild psychological aggressive
behaviors or that they experienced low fre-
quencies of severe behaviors; (l) whether con-
sistency (i.e., reliability) in responding to
psychological abuse checklists is equivalent to
validity; (m) whether those respondents with
minimal endorsement of psychological aggres-
sion items are more similar to individuals
reporting no psychological aggression or
whether they can/should be included in the
“psychologically abused” group for analy-
ses; (n) whether psychological abuse occurs
reciprocally more often than unilaterally, and,
if so, whether these patterns suggest signifi-
cantly different outcomes; (o) whether psycho-
logical behaviors should only be classified as
“abuse” if a negative impact on the recipient
is evident; and (p) whether scores on known
measures can validly establish a person as
“psychologically abused” or as a “psycho-
logical abuser.”
Researchers in this area typically state that
they are aware that their cross-sectional, corre-
lational research results do not establish cau-
sality. However, almost every study which has
investigated the impact of psychological
aggression on outcome variables used this
research methodology and almost every con-
clusion is based on correlational data. The lack
of longitudinal studies is a critical problem in
this field.

274 TRAUMA, VIOLENCE, & ABUSE / July 2009
Even with this abbreviated review of the
many methodological problems, caution is rec-
ommended when interpreting results from the
existing literature. The literature reviewed
below has been critiqued for its methodology,
but attention is still paid to significant or con-
sistent findings.
ORGANIZATION OF THE
LITERATURE REVIEW
To arrive at meaningful conclusions, the lit-
erature is organized with three major dimen-
sions in mind. First of all, investigations of the
impact of aversive intimate psychological
actions on women’s mental health vary widely
as to which outcome variables are used.
Considering what would constitute appropri-
ate mental health outcomes for this area of
research would lead most clinicians to think in
terms of diagnosable mental disorders. However,
the “outcome” or “response” to psychological
aggression has also been variously measured
in terms of: relationship changes; relationship
cessation; physical/somatic manifestations;
trait development; impact on the victim’s
behavior in terms of her own physical and/or
psychological aggression; other emotional
reactions often modeled after characteristics
associated (sometimes erroneously) with being
a recipient of long-term physical battering such
as guilt, fear, decrease in self-esteem, and
Posttraumatic Stress Disorder (PTSD) symp-
toms. Thus, articles will be reviewed in rela-
tion to others investigating similar dependent
variables.
Second, a large proportion of the studies
include assessment of not only psychological
aggression but also physical (and sometimes
sexual) abuse in the respondents’ relationships.
Although some studies use regression tech-
niques to partial out the contribution of psycho-
logical aggression on mental health variables,
fewer studies have investigated only psycho-
logical aggression. Because physically battered
women are likely to also report psychological
aggression in their relationships, they may
more likely be a convenience sample for collect-
ing this data. Or, it may be difficult to find a
sample of women who have been the recipient
of only psychological and not physical aggres-
sion. Marshall’s (1994) study indicated that out
of 640 women solicited for research purposes,
only 14% had absolutely no history of physical
violence, such that women with very mild his-
tories of physical force were assigned to the
“psychological abuse only” group. Tolman
(1992) has suggested that the impact of psy-
chological abuse within a physically abusive
relationship is likely to be qualitatively differ-
ent from its occurrence in a violence-free rela-
tionship because of the implicit threat of
violence behind the psychological aggression.
Thus, studies including both forms of aggres-
sion (as well as sexual abuse) will include that
information.
Third, certain demographics of the samples
are likely to influence responses on the out-
come variables. For example, women who are
dating (especially college samples) appear to
experience qualitatively and quantitatively less
psychological aggression than women in long-
term married or cohabiting relationships.
Women solicited for research participation
while residing in a battered women’s shelter
may be a unique group or have skewed demo-
graphics (e.g., fewer financial alternatives, less
of a support system) than abused women liv-
ing in the community, thus potentially con-
founding research findings if discrepancies are
not considered as covariates or generalizations
are made to the larger population of battered
women or women in general. In fact, Sackett
and Saunders (2001) specifically contrasted
battered women in shelters versus battered
women in the community, all of whom were
seeking help from a domestic violence agency
and found significant differences in education
and income. Thus, where demographics might
influence findings, these will be indicated.
Because numerous studies assessed more
than one impact of psychological aggression,
the first mention of a study will report on its
methodological strengths/weaknesses. After
that, brief conclusions about that study’s find-
ings will be reported taking into account the
methodological issues. Finally, only published
articles were reviewed, eliminating results from
conference presentations and dissertations due
to the lack of peer review.

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 275
IMPACTS OF PSYCHOLOGICAL
AGGRESSION
Mental Disorders/Symptoms—Depression
Possibly due to the prior research literature
linking physical abuse with mental disorders
(e.g., Gleason, 1993), depression was the psy-
chological symptom most frequently hypothe-
sized to be a consequence of psychological
aggression. Although the reviewed studies
investigated whether a relationship exists
between the two variables, few studies reported
whether the depression reached clinical or
diagnosable elevations. Although rationales for
why depression would occur due to psycho-
logical aggression (e.g., diminishment of self-
esteem; feelings of hopelessness regarding
change) seem plausible, certain types of psy-
chological aggression could potentially result
in alternative emotional responses (e.g., the
man’s attempts to control the woman resulting
in her feeling anger; intimidation by the man
producing fear in the woman).
Orava, McLeod, and Sharpe’s (1996) com-
parison of physically abused versus nonabused
women concluded that one of the best predic-
tor variables to distinguish the groups was
depression level. When they statistically con-
trolled for the amount of verbal abuse received
by the women, the relationship between depres-
sion and physical abuse was no longer signifi-
cant, suggesting that verbal abuse explained
the level of depression rather than physical
abuse. However, severity of physical abuse was
positively correlated with depression as well.
Lower educational levels for the physically
abused group and other coexisting variables
(e.g., less personal power, less internally ori-
ented) that could also explain levels of depres-
sion bolster the need for further corroboration
of this finding. Furthermore, the authors relied
on the limited items from the Conflict Tactics
Scale (CTS) tapping psychological “conflict tac-
tics” rather than more typical psychological
abuse items, thereby potentially missing exist-
ing relationships. They also had a relatively
small number of physically abused women
who had all been shelter residents, potentially
resulting in findings that are not generalizable.
Dutton, Goodman, and Bennett (2001) included
all types of intimate abuse in regression analy-
ses to predict court-involved African American
battered women’s current level of depression.
Psychological abuse, in the form of Emotional/
Verbal and Dominance/Isolation subscales of
the Psychological Maltreatment of Women
Inventory (PMWI; Tolman, 1989) significantly
explained 27% of the variance, but physical
aggression, sexual abuse, and injury level did
not remain significantly related to depression
when all were considered simultaneously.
Though this study used typical, standardized
measures, generalizability of the results is a
concern due to the unique sample and the large
unexplained variance. Coker, Davis, et al.
(2002), using data from a national survey, con-
cluded that abuse of power and control were
more strongly predictive of depression than
was verbal abuse, although verbal abuse con-
tributed significantly. In addition, Ali, Oatley,
and Toner (1999) concluded that “emotional
abuse can function as a precipitating factor for
depression in women” after determining that
the greater portion of women with major
depressive episode (40% of the sample) were in
the category of those with the most emotional
abuse, and Beck Depression Inventory (BDI)
scores were significantly higher for participants
with major emotional abuse. A strength of this
study was their multimodal assessment, which
included a semistructured interview, a self-
administered measure, and independent rat-
ings of the women’s written descriptions of
stressful events. However, their method for
determining emotional abuse raises concerns
about validity, numerous women identified the
domain of their emotional abuse as the work-
place, and all women were coming in for thera-
peutic services. Migeot and Lester’s (1996) brief
research report concerning psychological aggres-
sion in dating relationships of college men and
women also concluded that psychological abuse
was associated with depression scores on the
BDI and that dating abuse was predicted by
depression scores. More specifically, another
study of dating college women concluded
that both Emotional/Verbal and Dominance/
Isolation types of psychological abuse were sig-
nificantly related to depressive symptoms,
although when the researchers controlled for

276 TRAUMA, VIOLENCE, & ABUSE / July 2009
the women’s level of perceived interpersonal
control, neither type remained significant (Katz
& Arias, 1999). Complicating any conclusions
for this study was the fact that the average
level of depressive symptoms was low, inter-
personal control was generally perceived as
high, and only low levels of psychological
aggression were reported.
A study by Taft et al. (2006) is deserving of
individual attention in that it was method-
ologically stronger than most of the other studies
assessing the relationship between psychologi-
cal aggression and depression. This group of
researchers used a community sample in
Massachusetts that was expected to have low
rates of serious physical aggression that would
allow the researchers to examine the correlates
of psychological aggression more explicitly.
While controlling for any physical aggression,
ruling out substance abuse problems, and
insuring that the couples were not planning to
divorce, male-to-female psychological aggres-
sion was significant for predicting women’s
report of depression on the Brief Symptom
Inventory (BSI), although physical aggression
was still implicated for the presence of depres-
sion. Unfortunately, the authors combined
items from the CTS and the PMWI (creating
problems for establishing validity), most par-
ticipants were White (impacting generalizabil-
ity), and the study, relying on a measure which
was not designed specifically for the purpose of
measuring depression, did not establish whether
the depression reached clinical/diagnosable
levels.
Marshall’s 2001 study investigated rela-
tionships between different types of psycho-
logical abuse and physical/psychological
health outcomes. Regarding depression, she
reported that severe depression was partially
explained by “undermining,” “isolation,” and
“discounting” types of psychological aggres-
sion. However, these same types (plus others)
also predicted numerous other health and
emotional health indicators, suggesting that
these types do not necessarily uniquely pro-
duce depression. She concluded that subtle
psychological abuse factors in this study were
more often responsible for the variance in
women’s mental and physical health. (Later
research by Jones, Davidson, Bogat, Levendosky,
& von Eye, 2005, questioned Marshall’s distinc-
tion of subtle vs. overt forms of psychological
abuse when they analyzed her scale and found
it to be basically unidimensional.) Marshall’s
extensive 2001 study recruited volunteers
which resulted in excellent racial diversity
(although this was not used as a variable) but a
group that was characterized by low income.
This factor is likely to limit generalization of
the findings to the general population. Also, the
author did not offer a rationale as to why some
items were eliminated based on high or low
frequency of endorsement, making it difficult
to know whether the findings accurately reflect
the collected data (e.g., low frequency for items
may actually be indicative of the most severe
forms of psychological aggression). Because
this was one of few studies which had excel-
lent representation for Whites and African
Americans, the lack of analysis by race pre-
vented the researcher from knowing whether
that variable needed to be used as a covariate,
and findings indicating racial differences might
have been informative and aided interpreta-
tion of results.
In contrast to studies reporting an associa-
tion between psychological aggression and
depression, Campbell, Kub, Belknap, and
Templin’s (1997) modified the Index of Spouse
Abuse (ISA; Hudson & McIntosh, 1981) and
found that psychological aggression was not a
significant predictor of depression in battered
women. Rather, multivariate analysis indicated
that physical abuse was an important predictor
of depression along with daily hassles and
childhood physical abuse. The woman’s ability
to take care of herself was deemed to be a sig-
nificant protective factor against depression.
Tuel and Russell’s (1998) findings also cast
doubt as to the direct, necessary, and certain
outcome of depression from psychological
aggression. Their study, also using a revision of
the ISA measure, reported that depression
scores of battered women on the BDI were sig-
nificantly predicted by physical abuse but not
psychological abuse. Two cautions for this
study and others using the BDI are (a) the mea-
sure’s face validity makes it eminently able to
be manipulated; and (b) even though specific

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 277
scores have been labeled as representative of
increasing levels of depression, the BDI is typi-
cally used in correlational analyses rather than
dichotomizing the variable to indicate whether
the woman’s score reached clinical levels of
depression. Similar to Campbell et al.’s (1997)
study, Pape and Arias’ (1995) analysis of depres-
sion (and other mental health symptoms using
the Symptom Checklist-90-R [SCL-90-R]) as
related to both physical and psychological
aggression did not find distress levels to be
related to the level of abuse they experienced.
But several factors may account for the lack of
significance. The participants were dating col-
lege students (suggesting a lower intensity or
commitment level of the relationship), the
amount of physical and psychological aggres-
sion was quite low and typically not severe, the
limited items from the CTS were used, and the
sample was extremely homogeneous.
Coming from a different angle, Rollstin and
Kern (1998) hypothesized that psychopathol-
ogy (i.e., MMPI elevations) resulted from the
experience of abuse, and, therefore, women
who are no longer in these relationships would
have lower scores than women currently being
abused. In this study, the mean score on the
depression scale was not clinically elevated,
although several more “characterological” ones
were elevated and significantly associated with
physical and psychological aggression. The
psychological abuse scale was an amalgam of
the CTS subscale plus items added by the
authors, creating concerns about validity. Also,
the authors eliminated one fifth of the sample
based on an exaggerating or malingering
response set, which makes sense for conduct-
ing research on valid profiles, but which was
not discussed for the potential implications of
eliminating a large subset of their participants.
Sackett and Saunders (2001), accepting the
premise that psychological abuse has a signifi-
cant impact on a recipient, but interested in
determining the impact of different types of psy-
chological abuse, compared battered women
using a shelter versus a community sample of
women with violence in their relationships. For
both groups, the best predictor of depression
(measured by the BDI) was the amount of physi-
cal violence received followed by the severity of
psychological abuse. Specifically, “ignoring”
and “ridicule” were the types of psychological
aggression most significantly related to depres-
sion, but the unique variance by these types
was only 5% to 13%. Although a strength of
this study was the factor analysis conducted
using the psychological aggression items, the
items were originally drawn from prior descrip-
tions in the literature and the authors’ own
clinical work, such that the validity of the mea-
sure has not been established.
If suicidal behavior is included as a subcate-
gory of depression, less is known about the
relationship of psychological aggression for
prediction of suicide attempts. The brief report
of Migeot and Lester (1996) suggested no rela-
tionship of psychological aggression in dating
relationships to current or prior suicide ide-
ation or to suicide threats or attempts, but the
study by Kaslow et al. (1998) suggests a more
complicated picture. Among African American
women, both physical and psychological
aggression rates were higher for suicidal than
nonsuicidal women, but even this relationship
was mediated by levels of psychological dis-
tress, hopelessness, and drug use, while being
moderated by social support. Coker, Smith, et al.
(2002) found suicide ideation/actions to be pre-
dicted by an inclusive variable of all types of
IPV but not for psychological abuse alone.
In summary, although the overall findings
present contradictory results, some of the stron-
ger studies suggest a relationship exists between
psychological aggression and depression.
Unfortunately, these have not been longitudi-
nal studies but rather cross-sectional correla-
tional analyses which cannot establish causality.
Studies on dating populations, with often short-
term relationships, are probably less important to
consider than studies using couples in long-term
cohabiting relationships because of the low lev-
els of either psychological aggression or depres-
sion. Taft et al.’s (2006) study using long-term
couples with low levels of physical aggression
appears to be the most promising information to
date as to the impact of psychological aggres-
sion, even with its few methodological limita-
tions. It appears more difficult to disentangle
the source of depression when both physical
and psychological aggression are present. We

278 TRAUMA, VIOLENCE, & ABUSE / July 2009
cannot state that psychological aggression
always directly produces depression; however,
with more sophisticated measurement of this
phenomenon, we may be able to better under-
stand what mediates or moderates its effects.
Mental Disorders/
Symptoms—Anxiety/PTSD
Several of the investigations already reviewed
used anxiety as an outcome variable as well.
Although Rollstin and Kern’s (1998) study
using MMPI scores found that levels of vio-
lence and psychological aggression were not
associated with a clinical elevation of Scale 7
(i.e., characterological anxiety), and Pape and
Arias (1995) found lower levels of anxiety asso-
ciated with participants’ appraisals of control
over violence rather than a direct relationship
between anxiety and psychological aggression,
the most significant study again was the inves-
tigation by Taft et al. (2006). Again, controlling
for potentially confounding variables, male-to-
female psychological aggression was predic-
tive of anxiety as measured by the (BSI) subscale
of anxiety. The subscales of the BSI are poten-
tially problematic for making solid research
statements, however, as they were derived
from the original SCL-90-R items, which inves-
tigators have reported are not well-supported
as independent factors but rather that the over-
all measure is better used as a univariate global
distress scale (e.g., Rauter, Leonard, & Swett,
1996; Schmitz et al., 2000).
Researchers who have assessed for PTSD
symptoms in relation to psychological aggres-
sion typically do so when the sample consists of
physically battered women. Because PTSD is an
anxiety disorder assessed when events more
horrendous than life’s usual major stressors
occur (i.e., when one’s life or bodily integrity is
threatened), assessing for PTSD when physical
assault occurs fits the definition. However,
except for particular categories of psychological
aggression, such as threats of physical harm/
death or certain forms of intimidation, almost all
forms of psychological aggression do not meet
the first criterion for diagnosing PTSD. Dutton
et al.’s (2001) final regression model included all
types of abuse experienced by court-involved
battered women to predict PTSD symptoms,
and the analysis determined that both the level
of injury and the Emotional/Verbal scale from
the PMWI fairly equally explained 23% of the
variance. The researchers concluded that “in the
context of physical assault . . . it is the psycho-
logical abuse experience that is more determina-
tive of stress disorder symptoms.” Although
Arias and Pape (1999) reported that psychologi-
cal aggression was a significant predictor of
PTSD symptoms, which in turn were related to
a high level of distress, they warned that it may
be premature to conclude that physical victim-
ization will not account for PTSD symptoms.
Street and Arias (2001) concluded that the
Emotional/Verbal factor of the PMWI was a
significant and independent predictor of PTSD
symptoms whereas physical abuse was not.
However, Coker, Smith, et al. (2002) found that
their sample’s PTSD symptom scores were not
associated with psychological aggression but
rather with sexual and physical abuse.
In summary, it is surprising that more
researchers have not considered anxiety in gen-
eral as an outcome variable when anxiety, differ-
ent from fear, can be conceptualized as a threat
to one’s self-view. Both a partner’s denigration
of her as well as a recipient’s belief that she
should be able to change/handle/control the
partner’s emotional tactics toward her would
theoretically raise her anxiety level. Therefore,
more investigation is required to be able to make
a definitive statement regarding whether psycho-
logical abuse directly or indirectly causes clini-
cally significant anxiety. Even the finding by Taft
et al. (2006), although suggestive of a relation-
ship, has to be viewed cautiously due to measure-
ment issues. PTSD symptoms should probably be
measured only in relation to physical/sexual
abuse or limited more specifically to its associa-
tion with the types of psychological aggression
fitting PTSD criteria, even though the concept
of anxiety would still be highly relevant as an
outcome variable for the occurrence of psycho-
logical aggression.
Mental Disorders/Symptoms—
Miscellaneous
Only one article investigated the role of guilt
and shame in relation to psychological aggression
and PTSD in battered women (Street & Arias,

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 279
2001). Although Emotional/Verbal abuse was
related to shame, the Dominance/Isolation fac-
tor was not. Guilt was not related to either form
of psychological aggression.
Straight, Harper, and Arias (2003) conducted
a study with single college women which
found a relationship between psychological
aggression from a partner and illegal drug use,
although no relationship was found for alcohol
problems or binge drinking. Interestingly
enough, other health-risk or promotion
behaviors were not related to experiencing psy-
chological aggression except for self-reported
limitations on, and perceptions of, health.
Unfortunately, whether drug use preceded the
experience of psychological abuse is unknown.
Only two other studies reported on a relation-
ship between substance problems and psycho-
logical aggression which produced contradictory
results. The first was the public health study
using national survey statistics by Coker, Davis,
et al. (2002) who reported that abuse of power
and control, as a subtype of psychological
aggression, was more strongly associated with
heavy alcohol use, therapeutic drug use, as
well as recreational drug use than was verbal
abuse. However, Coker, Smith, et al. (2002),
with their sample of 1,152 women recruited
from family practice clinics, found that drug
and alcohol abuse were not associated with
psychological aggression although they were
associated with sexual or physical abuse.
Several authors reported on an association
between fear (of physical harm) and psycho-
logical aggression, although all three studies
involved physically battered women, compli-
cating any understanding of the mechanism as
to how psychological aggression, in and of
itself, would result in fear (except for the cate-
gory of direct threats of harm). Marshall (2001)
found a fear of injury or death to be explained
by the man’s violence as well as “monitoring,”
“undermining,” “isolation,” and “discounting”
forms of psychological aggression. Henning
and Klesges (2003) determined that more psy-
chological aggression was related to women
feeling seriously threatened, but their list of
items included those specific items of threaten-
ing behavior intended to produce fear (i.e.,
“threaten to kill you or children if you end the
relationship” and “threaten to kill self if you
end the relationship”). Tolman and Bhosley
(1991), as cited by Sackett and Saunders (2001),
also reported that psychological aggression was
a stronger predictor of fear than physical abuse
in a sample of physically abused women.
Some researchers reported on “stress/distress,”
which presents definitional problems, espe-
cially in light of the earlier review of PTSD as a
specific form of stress with designated criteria.
The vagueness of the term raises concerns that
the outcome being measured may be too dif-
fuse for understanding exactly what has been
measured or what mental health intervention is
required. Marshall (2001) used a measure as to
how well women handle stress which was
explained by “dominating,” “indifference,” “dis-
crediting,” “monitoring,” and “undermining”
forms of psychological aggression. A combina-
tion of psychological aggression and level of
injury predicted 37% of the acute stress in
Dutton et al.’s (2001) sample of court-involved,
mostly African American battered women. In
addition, although Vitanza, Vogel, and Marshall
(1995) reported that the “women in this sample
were seriously distressed,” they concluded
that the physical violence experienced by
women in their sample appeared to affect the
association between psychological aggression
and any specific effects.
One study suggested “cognitive impairment”
as a potential result of psychological aggres-
sion, although the authors did not define it
technically as a disability or thought disorder
but used author-constructed items such as con-
fusion about the date or getting lost when going
to familiar places, and participants reported the
occurrence of these in the preceding 4 weeks
(Straight et al., 2003). There was no validity
information presented for these items. More
specifically, Vitanza et al. (1995) administered
the “strongest items” of the Cognitive Failures
Questionnaire, which purports to measure fail-
ures in perception, memory, and motor func-
tion. However, psychological aggression was
not implicated in cognitive failure, but rather
sexual aggression was a better predictor.
With regard to the impact of psycholo gical
aggression on self-esteem, Aguilar and
Nightingale’s (1994) study was the only one
specifically focused on that variable. Altho ugh
battered women had lower self-esteem than

280 TRAUMA, VIOLENCE, & ABUSE / July 2009
nonbattered women, only emotional/control-
ling (as compared with physical, sexual/
emotional, and miscellaneous) abuse was sig-
nificantly related to lower self-esteem scores.
However, the author-devised emotional abuse
items with yes/no response options failed to
consider potentially confounding demo-
graphic differences between the battered
women seeking assistance versus the control
group of women coming from the general uni-
versity population and community. Thus, cor-
roboration of these findings is necessary. Tolman
and Bhosley (1991) found only a small effect on
self-esteem in their sample, in that the highest
correlation with self-esteem was the amount of
physical violence incurred, although the ignor-
ing and ridiculing types of psychological
aggression were related but accounted for little
variance. Orava et al. (1996) suggested that ver-
bal abuse “contributed” to the difference in
scores between abused and nonabused women
in self-esteem, because when the influence of
verbal abuse was removed, most of the between-
group differences were lessened. Lending some
support to this relationship, Baldry (2003; as
cited in Katz, Arias, & Beach, 2000) cited psy-
chological aggression as predictive of low self-
esteem. Two studies focused on whether
particular forms of psychological abuse were
more likely to be related to lower self-esteem.
Dutton and Painter (1993) reported that this
form of self-evaluation was related to the
Dominance/Isolation factor of the PWMI, while
Sackett and Saunders (2001) again found that
the amount of physical violence incurred was
an important factor for predicting low self-
esteem, but also that “ignoring” and “ridicul-
ing” were explanatory for small amounts of the
variance. Stets (1991), assessing only psycho-
logical aggression in a college dating sample,
reported finding no differences in frequency
between men and women for inflicting psycho-
logical aggression but that women reported
sustaining more psychological aggression than
men (mostly in terms of upset or hurt feelings).
When Stets statistically controlled for one’s
perceived level of controlling one’s dating part-
ner, the results indicated that women with low
self-esteem both inflicted and sustained psy-
chological aggression more. Important to note
for these findings is that Stets’ devised ques-
tionnaire consisted of verbal aggression items
of which the milder forms were more fre-
quently reported.
Physical Manifestations
of Psychological Distress
A number of authors have assessed the
physical health status of individuals on the
assumption that health variables can be affected,
either directly or indirectly, from the psycho-
logical distress resulting from the experience of
psychological aggression. Unfortunately,
“physical health” is defined in many ways, at
times by specification of physical disorders and
often by the measure used. For example,
Straight et al. (2003) did not find psychological
aggression to be associated with specific health-
risk behaviors (e.g., sleep, exercise, smoking,
alcohol problems), but they did find a relation-
ship with self-reported limitations on activities
and roles due to their physical health, as well as
on the college women’s overall perceptions of
their health. More important, their assessment
of approach versus avoidance coping styles
moderated the effects of psychological aggres-
sion on most of the specific health behaviors as
well as the participants’ perceptions of their
health. Taft et al.’s (2006) well-controlled study
of a representative community group sug-
gested that male-to-female psychological
aggression was uniquely associated with an
increase in physical health symptoms, although
these were measured by yes/no answers to
whether they experienced 12 health symptoms
“fairly often” in the prior 6 months (a subscale
of the Health and Daily Living Form). However,
specifics were not given, so what type of health
symptoms were addressed or the seriousness
of them is unknown. In contrast, Coker, Smith,
Bethea, King, and McKeown (2000) investi-
gated many specific physical health problems.
They reported that psychological aggression
was associated with chronic neck or back pain,
arthritis, headaches, chronic pelvic pain, stom-
ach ulcers, sexually transmitted infection, spastic
colon and other gastrointestinal upsets, and “dis-
abilities preventing work.” While concluding
that psychological aggression was as strongly

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 281
associated with the majority of negative health
outcomes as physical abuse, their measure of
psychological aggression raises questions as to
validity. If women received a particular score on
the Women’s Experience with Battering (WEB)
Scale but did not report concurrent physical or
sexual IPV, they were given a default designa-
tion as “psychologically battered.” However,
items on this measure are the women’s percep-
tions of how she has been treated or how she
feels, such that even some type of objective
verification of psychological events occurring
is not required.
Two studies looked at psychological
aggression’s impact on highly specific health
conditions—irritable bowel syndrome and
depression surrounding pregnancy. Ali et al.
(2000) reported that, after they controlled for
physical and sexual abuse (which were both
significant) in their sample, a significant differ-
ence still remained for emotional abuse, self-
blame, and self-silencing with fairly large mean
differences between their patients with IBS and a
comparison group of women with organic gastro-
intestinal disorders. Trotter, Bogat, and Levendosky
(2004) investigated risk and protective factors for
their sample of pregnant women who were expe-
riencing physical and psychological aggression.
Above and beyond the impact of physical vio-
lence, the authors reported that the depressive
symptoms present in the third trimester of preg-
nancy mediated the relationship between how
severe the psychological aggression was at that
time and depressive symptoms 2 to 3 months
postpartum. Encouragingly, the amount of emo-
tional support the women received prior to the
baby’s birth moderated the relationship of severe
psychological abuse at that time and the concur-
rent level of depression. Not surprisingly, the
depression level at the prenatal period was the
best predictor of postnatal depression.
In an effort to determine which types of psy-
chological aggression might produce specific
outcomes, Marshall (2001) reported that health
quality was explained by “undermining,” “dis-
counting,” and “monitoring.” However, she used
one item asking for the person’s perception of
the overall quality of her health, raising con-
cerns as to the validity of this finding due to the
highly subjective nature of the item, the lack of
specificity in asking about one’s “health,” and
even the lack of temporal considerations in the
item (e.g., Today’s perception of one’s health?
One’s health over the past year? One’s health
during the psychologically aggressive relation-
ship?). Coker, Davis, et al. (2002) also separated
out different forms of psychological aggression
and concluded that abuse of power and control
was more strongly associated with current
poor health, substance use, and developing a
chronic disease than verbal abuse in their anal-
ysis of national statistics.
In summary, although impact on physical
functioning appears to be a viable area for
investigation, the literature providing the link-
age between decrements in physical health and
psychological aggression is problematic from
several angles: (a) The measurement has typi-
cally been overly broad, vague, or highly sub-
jective, thus impairing interpretation; (b) There
has not been enough focus on the potential
theoretical mechanisms by which psychologi-
cal aversive events impact physical function-
ing; and (c) Self-report of ailments needs to be
corroborated by documentation for strengthen-
ing validity.
Personality Traits/Styles
Little research has been conducted on per-
sonality traits as impacted by psychological
aggression possibly because traits are assumed
to precede particular experiences, possibly
because traits are not believed to be shaped
much by adult experiences, and likely because
some mental states such as hopelessness and
low self-esteem are considered to be mental
health symptoms rather than traits. The results
from a few studies are of some interest. Two
studies were conducted to determine whether
psychological aggression resulted in elevations
on MMPI scales. Using the CTS subscale on
nonphysical conflict tactics and adding a few
items of their own, Rollstin and Kern (1998)
found that psychological aggression was sig-
nificantly associated with MMPI scores,
although only three scales were in the clinical
range for the battered women. As stated earlier,
the elevated scales were not the ones measuring
affective states of depression and anxiety.

282 TRAUMA, VIOLENCE, & ABUSE / July 2009
Rather, the elevations were on Scales 4 (repre-
sented by antisocial attitudes and a sense of
alienation from others), 6 (represented by per-
secutory feelings and interpersonal sensitivity),
and 8 (suggesting alienation, difficulty control-
ling thought processes and oddities of percep-
tion). In addition, scores on scales measuring
passivity and social introversion (i.e., 5 and 0)
were within normal ranges, suggesting that the
women experiencing abuse were not, as a group,
dependent, inactive, passive, and identified with
a highly traditional female role. Rollstin and
Kern also determined that being out of the abu-
sive relationship for a longer period was not
related to scores diminishing over time. Kahn,
Welch, and Zillmer (1993) also conducted a
study using the MMPI and concluded that the
African American battered women’s experience
of receiving psychological aggression was
related to their MMPI scores. In line with Rollstin
and Kern (1998), this study reported clinically
significant elevations on Scales 4, 6, and 8. It is
important to note that these scales are more in
line with personality styles rather than affective
reactions, but it is impossible to know whether
these women developed persecutory feelings,
interpersonal sensitivity, a sense of alienation,
and so on before or as a result of being in an
abusive relationship.
Impact on One’s Own
Behavior Regarding Aggression
Murphy and O’Leary (1989) investigated
patterns of psychological and physical aggres-
sion with couples who were newly married.
Because couples were only included in the
study if they had not engaged in any physical
aggression as of a month before the marriage,
the authors could follow more carefully when
physical incidents first took place. Contributing
to the idea that these two forms of interper-
sonal behavior are related (for both men and
women), both the person’s own and their
partner’s psychological aggression predicted
the person’s initial incident of physical aggres-
sion, whereas marital dissatisfaction was not a
good longitudinal predictor. Thus, higher lev-
els of inflicting and sustaining psychological
aggression in a marriage appear to predict
movement into physical incidents over time.
Stets (1991) stated that inflicting psychological
aggression was correlated with sustaining
psychological aggression in her sample of dat-
ing college men and women, such that she
concluded that this phenomenon is often
reciprocal.
Relationship Changes
Fortunately, a number of researchers seem
aware that the impact of aversive psychological
actions could be an impact on the relationship
itself, rather than necessarily producing nega-
tive mental health consequences. For example,
a recipient of psychological aggression could
begin to devalue the partner, to view the part-
ner as problematic, to detach from the relation-
ship, to change his or her investment in the
relationship, or to revise his or her expectations
or structure of the relationship. Kasian and
Painter (1992) found that both men and women
experiencing high psychological aggression
reported lower satisfaction with the dating rela-
tionship. In another study, psychological aggres-
sion was negatively correlated with relationship
stability and intimacy of dating couples (Katz,
Arias, & Beach, 2000), and this association was
not a function of the women’s own levels of
self-esteem. Raghavan, Swan, Snow, and Mazure
(2005) employed a psychological aggression
measure consisting of some CTS-2 items and
some PMWI items. Their contribution was the
finding that greater levels of psychological
aggression were related to decreased relation-
ship efficacy, defined by them as “the extent to
which a woman believes she can manage the
relationship successfully.” However, a reduced
sense of efficacy was not related to the woman
leaving the relationship. Once again, Marshall
(2001) attempted to explain an outcome of psy-
chological aggression (i.e., marital well-being)
by her hypothesized overt and subtle subtypes.
Although she reported that relationship quality
was partially explained by “discounting,” “iso-
lation,” “undermining,” and “indifference,” the
physical violence the women experienced was
also predictive.
In summary, while this area appears to be a
highly fruitful one for understanding the impact

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 283
of psychological aggression, viewing the rela-
tionship as a system which likely changes as a
result of aversive interpersonal actions has not
received much attention.
Relationship Cessation
Whether a woman on the receiving end of
psychological aggression actually leaves that
partner does appear of interest to researchers.
A number of researchers have assessed the rela-
tive contributions of the two forms of maltreat-
ment to determine whether one or the other is
more likely to trigger the relationship break-up.
Jacobson, Gottman, Gortner, Berns, and Shortt
(1996) conducted intensive investigations with
maritally violent couples, and from their
research they reported that one of the six vari-
ables predicting separation or divorce 2 years
after involvement in the research was the hus-
band’s “global negative affect.” However,
Gortner, Jacobson, Berns, and Gottman (1997)
made an intensive effort to locate couples who
had dropped off their rolls to make a more
definitive statement about reasons for marital
disruption with aggressive couples. Their more-
inclusive findings stated that the husband’s
emotional abuse was a stronger predictor of the
marriage ending than the physical violence the
woman incurred—in fact, emotional abuse was
singled out as the most important factor in
leaving, particularly when the types of emo-
tional abuse the husband engaged in were
“degrading” and “attempting to isolate the
woman from others.” Two other important
findings were that the women who left actually
were more likely to respond assertively to the
husband’s emotional abuse, and they were
more likely to be emotionally abusive them-
selves. Strongly supporting this research was
Raghavan et al.’s (2005) study claiming that a
woman appears to leave her abusive partner
directly related to levels of psychological
aggression and not due to physical abuse or
feelings of efficacy within the relationship. The
authors cautioned, however, that the fact that a
woman did not leave strictly due to higher
severity of physical abuse might be because the
woman was afraid to leave in such cases. With
a dating population, Kasian and Painter (1992)
discovered that relationship termination was
due to not only more negative actions present
but also fewer positive behaviors occurring in
the relationship. Intentions to leave a relation-
ship also seemed linked to increases in psycho-
logical aggression in two studies (Arias & Pape,
1999; Henning & Klesges, 2003), although Arias
and Pape found that neither physical nor psy-
chological aggression alone resulted directly in
the intention to leave the relationship, but
rather the woman’s psychological well-being
mediated this relationship. Only Dutton et al.
(2001) did not find psychological aggression to
be predictive of a woman leaving her relation-
ship, but their sample of court-involved bat-
tered women may represent a group with
severe violence and many more complications
and dilemmas for decision making.
DISCUSSION AND IMPLICATIONS
FOR FUTURE RESEARCH
Although a number of published articles
report deleterious effects as a function of expe-
riencing psychological aggression, the rela-
tively small number of studies, the problematic
methodology, and the typical co-occurrence of
physical violence prevent strong statements of
causal effect or clear, uncontested results.
Probably due to methodological issues, contra-
dictory results are frequently found. The most
compelling problem, however, is how and
whether to interpret results from studies for
which the primary measurement technique
appears flawed. Therefore, this discussion will
focus on the identification of probable relation-
ships of mental health outcomes when psycho-
logical aggression is present as well as to
provide suggestions for improved research in
this area (Table 1).
Research on depression—the outcome vari-
able most often studied for its relationship with
psychological aggression—appeared to dem-
onstrate relatively good support for depressive
symptoms with increasing psychological aggres-
sion, but the studies measuring clinically rele-
vant levels of depression did not show a
significant relationship. Also, results were mixed
as to whether physical or psychological aggres-
sion was the more predictive variable or the

284 TRAUMA, VIOLENCE, & ABUSE / July 2009
only contributing source for the emergence of
depression. Results from studies attempting to
use “types” of psychological aggression for
prediction of depression did not clearly and
uniformly identify categories which could then
be used to create interventions. Unfortunately,
the research to date has not clarified whether a
third variable causes both psychological aggres-
sion and depression (e.g., relationship dissatis-
faction), what dose produces diagnosable
depression, and whether frequency, severity, or
actual “type” of psychological aggression pro-
duces the most harm. These limitations are
present for almost all of the studies reviewed in
this article and apply to almost all of the out-
come variables.
Contradictory results are also present when
anxiety is the measured outcome variable,
although certain forms of psychological aggres-
sion would seem highly likely to impact anxi-
ety levels. However, anxiety, as a clinical concept
reflecting anticipated threats to one’s self-view,
needs to be delineated from and measured in
different formats from “fear” regarding one’s
physical safety and also from PTSD, as the
group of symptoms likely to result from a
trauma. Because the definition of PTSD requires
that the diagnosed individual must have “expe-
rienced, witnessed, or was confronted with an
event . . . that involved actual or threatened
death or serious injury, or a threat to the physi-
cal integrity of self or others,” there is likely to
be controversy as to whether assessing PTSD as
an outcome variable for psychological aggres-
sion is a viable option. Except for the category
wherein one’s partner “threatens to kill or
harm” them or others significant to them
(which could include the partner having
destructive rages in the woman’s vicinity as
well as direct verbal threats), other forms of
psychological aggression do not seem to fit the
definition for a “traumatic event.” Interestingly,
some researchers found “threats to kill/harm”
to load significantly on a physical violence fac-
tor rather than the psychological aggression
factor. However, not to ignore the very high
likelihood of anxiety resulting from many of
the other forms, assessment of anxiety needs to
be thoroughly investigated as an outcome
variable, possibly using measures such as the
State-Trait Anxiety Inventory for Adults
(Spielberger, 1983).
A number of outcome variables might show
greater consistency and clarity if future research
more specifically defined them when measur-
ing their relationship with psychological
aggression. For example, mixed results were
found regarding the association of self-esteem
and experiencing psychological aggression.
Because individuals often exhibit varying self-
views across roles, activities, and relationships,
future research might profit from specifically
and functionally defining this broad concept to
render it a more useful variable for research
and for clinical applications. In addition, medi-
ating variables seemed important for explain-
ing the relationship of one’s self-view with
psychological maltreatment rather than there
being a direct link between the two.
The relationship between psychological
aggression and “health” also remains to be dis-
entangled, potentially through increased speci-
ficity of health variables. Broad-based questions
as to a person’s perception of their physical
health or limitations result in highly subjective
responses which are difficult to interpret, espe-
cially when those ratings conflict with the per-
son’s reports of specific health-risk behaviors.
Actual documentation of reported health prob-
lems or treatment (e.g., medical records, doctor
appointments) would also strengthen these
studies beyond self-report. Current health vari-
ables are often correlated with self-report of
psychological aggression without assessment
as to whether the health problems existed prior
to the onset of the psychological aggression.
Coker and her colleagues provided a good
model by their elimination of individuals whose
health problems originated prior to the first
IPV. Medical/psychological theoretical models
need to be developed through future research
to explain why specific health problems appear
to be associated with physical versus psycho-
logical aggression (e.g., why arthritis is associ-
ated with psychological aggression rather than
physical abuse).
Probably the most consistently found impact of
psychological aggression was not a psychological
impact per se but rather a behavior—leaving or
desiring to leave the relationship. Only one

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 285
study assessing this connection did not find
psychological abuse to be an important factor,
although several suggested certain variables
significantly mediated this association: (a) vari-
ables descriptive of the women, such as their
psychological well-being, responding assert-
ively to the husband’s emotional abuse, or
engaging in emotional abuse themselves; and
(b) variables descriptive of the relationship,
such as fewer positive behaviors along with the
psychological aggression, the partner engaging
in degradation, and the partner attempting to
isolate the woman.
Although the variables listed above had a
reasonable body of studies to comment on,
there were quite a few variables without enough
corroboration for strong associations to be
claimed, especially in light of methodological
limitations. One study investigated the role of
guilt and shame and did not find strong rela-
tionships of those variables with psychological
aggression. The choice of these variables was
intriguing because, although some psychologi-
cally abusive strategies intend to diminish a per-
son’s self-view and to undermine self-confidence,
guilt and shame would not necessarily occur
from those machinations. However, guilt and
shame might be used as mediating variables in
future research. Suicidal behavior, also assessed
in only a few studies, did not seem strongly
associated with psychological abuse alone.
This outcome variable might be better studied
by using broader samples of suicidal patients
or even community samples to try to establish
whether psychological maltreatment is suffi-
cient for inducing suicidal ideation or behavior.
Substance “use” and abuse, although associ-
ated with psychological aggression in two
studies, is another outcome variable requiring
careful delineation of the use of legal and ille-
gal substances before definitive statements can
be made.
Two outcome variables posed in the litera-
ture require definition and operationalization
before they are likely to be useful parameters to
assess. “Stress” as an outcome variable is dif-
fuse and lacks specificity, and, for it to be a use-
ful concept, it will need well-defined parameters
for documenting its presence and a method for
establishing a clinically defined criterion level.
The same critique applies to any designation in
this literature to date of the outcome variable of
“cognitive impairment.” This variable needs to
be measured in terms of specific mental func-
tions of interest, such as memory, concentra-
tion, or attention, and researchers need to use
standardized measures that designate levels of
impairment significant for consideration.
Finally, personality traits, if they are to be used
as outcome variables in future studies, will
require specification and operationalization as
a more effective means to assess them in con-
trast with interpreting their presence from scale
elevations measuring psychopathology. While
there has been little attention directed toward
personality factors in the literature, anecdotal
and research literature at least obliquely men-
tion behavioral manifestations of them. For
example, sociability, optimism, trust, depen-
dency, and others are certainly viable aspects of
psychological functioning that could be nega-
tively affected from long-term exposure to fre-
quent, enduring, and severe psychological
aggression (Table 2).
Several other conclusions are briefly offered
here. First, due to the relatively infrequent and
mild (possibly normative) forms occurring in
college dating populations, psychological
aggression in dating relationships needs to be
assessed separately from its occurrence in mari-
tal or long-term cohabiting relationships. Second,
researchers investigating psychological aggres-
sion should probably always ask their samples
as to the presence and frequency/severity of
physical/sexual abuse. Because populations
with both types endorse more serious forms of
psychological aggression, the threat of poten-
tial physical violence likely influences the
impact of any psychological abuse, and our con-
clusions would likely be incomplete without
that information. Third, we are at the stage
where more investigation into mediating/
moderating factors seems paramount. Because
of large percentages of unexplained variance
present even when there is a significantly
established association between psychological
aggression and hypothesized consequences, it
is highly likely that mediating and moderating
factors are influencing the outcomes. Identifying
what prevents some women from developing

286 TRAUMA, VIOLENCE, & ABUSE / July 2009
particular negative consequences would be
very useful information. And, last, if research-
ers are interested in emotional states arising
from psychological aggression, future research
might target a broader range of feeling states,
including anger, sadness, “hurt,” emotional
withdrawal, and the like that are probably dis-
tinguished by internalizing versus externaliz-
ing orientations of psychological aggression
recipients.
To have increased confidence regarding find-
ings in future research studies, the conceptual-
ization of psychological aggression may benefit
from using theory about aggression in general
from the social psychological literature rather
than operationalizing psychological aggression
by establishing “types” of actions or treating it
as a unique phenomenon. Above and beyond,
in theory, our measurement of psychological
aggression must be improved so that we are
not uncertain what our data mean when an
individual reports being on the receiving end
of a specific behavior, currently the standard
approach of scales in use, and assessment must
become much more explicit to accomplish the
actual goals of measurement. For example, if
someone checks the item “partner criticized
you,” the recipient could consider this con-
structive feedback, or he or she might not like
it but consider it valid, or he or she was both-
ered by it but did not consider it “abusive” or
“harmful,” or he or she experienced it as a
reciprocal criticism following their own nega-
tive communication, or he or she could con-
sider the criticism as unwarranted and hurtful,
and so on. Producing measurement which is
more representative of the range of psycho-
logically aggressive behaviors and more inter-
pretable as to the meaning and impact of those
behaviors (such as Follingstad et al., 2005) is an
important initial step. Measurement which
includes the likely impact along with a descrip-
tion of a psychologically aggressive behavior
would also allow for better interpretation of
the results. Respondents could also be required
to consider whether any consequences occurred
to them as a result of the particular behaviors.
Although the findings above have been cri-
tiqued as to methodological shortcomings, many
of the outcome variables which researchers
have studied have reasonable theoretical ratio-
nales for predicting their occurrence, especially
as many have been previously linked to other
forms of IPV and negative treatment between
partners. Even with strong statements as to the
significant problem of defining and measuring
psychological abuse and these outcome vari-
ables, the feeling states chosen by researchers as
likely to result from psychological aggression
have some initial support. For better establish-
ment of the validity of these findings, however,
future research may need to change the criteria
from measuring the existence of “any” emo-
tional change in a person on the receiving end
of psychologically unpleasant or aggressive
behavior to measuring “clinical” levels of
impact, that is, moving from statistical signifi-
cance to clinical significance. As part of the
human experience, we know (without research
studies) that persons will be upset, emotionally
hurt, temporarily depressed, saddened, and so
on, when treated badly in an interchange with
an intimate partner. Almost all humans have to
navigate negative interpersonal experiences to
recover from the incidents, determine responses
to them, make decisions for their lives, and inte-
grate them into belief systems about them-
selves, others, and the world. These are the
experiences for which counseling and support
groups, self-help books, and good friends seem
geared to assist the usual person negotiating
life’s interpersonal difficulties. However, our
research, rarely requiring stringent criteria, clin-
ical elevations, or outside documentation, is
often willing to label any measure of “distress”
which is correlated with behaviors (that range
from unpleasant through highly aggressive)
that researchers identify as important for inves-
tigation. What may be much more illuminating
and significant for the advancement of this field,
would be to more thoroughly investigate identi-
fiably egregious psychological abuse cases in
which more enduring and life-changing impacts
have resulted. Research questions of interest
might include the following: (a) Do structural,
internal changes occur to a person’s personality
due to being on the receiving end of such psy-
chological abuse which, in turn, impacts their
decisions about the relationship itself? (b) Are
certain persons, by virtue of their beliefs, life

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 287
schemas, or expectations, more vulnerable to
this form of exploitation? and (c) Do women
experiencing psychological abuse have particu-
lar schemas or relationship expectations or
societal pressures which render them more
able to be manipulated and coerced in this way
than men?
Although all of our efforts to understand the
consequences of psychological aggression are
useful for better understanding the human
condition, what will likely make our investiga-
tions most noteworthy are those which find
markers for identification of and intervention
with individuals at high risk for developing
significant psychological impairments or who
are already experiencing severe victimization
through psychological abuse. Specifically, when
we are able to appropriately define it, measure
it in sophisticated ways, and study its long-
term and enduring impact on the psychological
structures of women, we will be more qualified
to design treatment and prevention efforts to
treat the effects of, and to inoculate women
from, psychological abuse.
IMPLICATIONS FOR PRACTICE, POLICY, AND RESEARCH
• Psychological aggression in dating relationships needs
to be evaluated separately from its occurrence in mari-
tal or long-term cohabiting relationships.
• Researchers should likely always include assessment
of physical and sexual abuse when studying psycho-
logical aggression.
• The inclusion of mediating and moderating variables
in future research will likely allow for better interpreta-
tion as to why the hypothesized consequences of psy-
chological aggression are not uniformly seen in
recipients.
• Methodological shortcomings are significant in the
assessment of psychological aggression/abuse and
more sophisticated measurement is necessary before
we will be able to make statements about its effects.
• A more significant impact for research in this field is
likely to come from intensive study of long-term,
enduring patterns of serious psychological abuse for
clues as to effective treatment and prevention efforts.
NOTE
1. There has never been a well-validated model or measure
which designates when particular types of actions, or a particular
frequency of a behavior, or the severity of a psychological behavior
can definitively be classified as psychological abuse from a simple
description. There certainly are a small number of behaviors, that
because of their egregiousness, are consistently viewed as abu-
sive, but many do not receive consistent ratings along this dimen-
sion (see Follingstad & DeHart, 2000; Follingstad, Helff, Binford,
Runge, & White, 2004).
REFERENCES
Aguilar, R. J., & Nightingale, N. N. (1994). The impact of
specific battering experiences on the self-esteem of
abused women. Journal of Family Violence, 9, 35-46.
Ali, A., Oatley, K., & Toner, B. B. (1999). Emotional abuse
as a precipitating factor for depression in women.
Journal of Emotional Abuse, 1(4), 1-13.
Ali, A., Toner, B. B., Stuckless, N., Gallop, R., Diamant, N. E.,
Gould, M. I., et al. (2000). Emotional abuse, self-blame,
and self-silencing in women with irritable bowel syn-
drome. Psychosomatic Medicine, 62, 76-82.
Arias, I., & Pape, K. T. (1999). Psychological abuse:
Implications for adjustment and commitment to leave
violent partners. Violence and Victims, 14(1), 55-67.
Baldry, A. C. (2003). “Sticks and stones hurt my bones
but his glance and words hurt more:” The impact of
psychological abuse and physical violence by former
and current partners on battered women in Italy.
International Journal of Forensic Mental Health, 2, 47-57.
Campbell, J. C., Kub, J., Belknap, R. A., & Templin, T. N.
(1997). Predictors of depression in battered women.
Violence Against Women, 3, 271-293.
Coker, A. L., Davis, K. E., Arias, I., Desai, S., Sanderson, M.,
Brandt, H. M., et al. (2002). Physical and mental health
effects of intimate partner violence for men and women.
American Journal of Preventive Medicine, 23, 260-268.
Coker, A. L., Smith, P. H., Bethea, L., King, M. J., &
McKeown, R. E. (2000). Physical health consequences
of physical and psychological intimate partner vio-
lence. Archives of Family Medicine, 9, 451-457.
Coker, A. L., Smith, P. H., Thompson, M. P., McKeown, R. E.,
Bethea, L., & Davis, K. E. (2002). Social support pro-
tects against the negative effects of partner violence on
mental health. Journal of Women’s Health and Gender-
Based Medicine, 11, 465-476.
Dutton, D. G., & Painter, S. (1993). Emotional attachment in
abusive relationships. Violence and Victims, 8, 105-120.
Dutton, M. A., Goodman, L. A., & Bennett, L. (2001). Court-
involved battered women’s responses to violence: The
role of psychological, physical, and sexual abuse.
Violence and Victims, 14(1), 89-105.
Follingstad, D. R. (2007). Rethinking current approaches
to psychological abuse: Conceptual and methodologi-
cal issues. Aggression and Violent Behavior, 12, 439-458.

288 TRAUMA, VIOLENCE, & ABUSE / July 2009
Follingstad, D. R., Coyne, S., & Gambone, L. (2005). A
representative measure of psychological aggression
and its severity. Violence and Victims, 20(1), 25-38.
Follingstad, D. R., & DeHart, D. D. (2000). Defining psy-
chological abuse of husbands toward wives: Contexts,
behaviors, and typologies. Journal of Interpersonal
Violence, 15, 891-920.
Follingstad, D. R., Helff, C. M., Binford, R. V., Runge, M. M.,
& White, J. D. (2004). Lay persons’ versus psycholo-
gists’ judgments of psychologically aggressive actions.
Journal of Interpersonal Violence, 19, 916-942.
Follingstad, D. R., Rutledge, L. L., Berg, B. J., Hause, E. S.,
& Polek, D. S. (1990). The role of emotional abuse in
physically abusive relationships. Journal of Family
Violence, 5, 107-120.
Gleason, W. J. (1993). Mental disorders in battered women:
An empirical study. Violence and Victims, 8, 53-68.
Gortner, E., Jacobson, N. S., Berns, S. B., & Gottman, J. M.
(1997). When women leave violent relationships:
Dispelling clinical myths. Psychotherapy, 34, 343-352.
Henning, K., & Klesges, L. M. (2003). Prevalence and
characteristics of psychological abuse reported by
court-involved battered women. Journal of Interpersonal
Violence, 18, 857-871.
Hoffman, P. (1984). Psychological abuse of women by
spouses and live-in lovers. Women and Therapy, 3(1),
37-47.
Hudson, W. W., & McIntosh, S. R. (1981). The assessment
of spouse abuse: Two quantifiable dimensions. Journal
of Marriage & the Family, 43, 873-885.
Jacobson, N. S., Gottman, J. M., Gortner, E., Berns, S., &
Shortt, J. W. (1996). Psychological factors in the longi-
tudinal course of battering: When do the couples split
up? When does the abuse decrease? Violence & Victims,
11, 371-392.
Jones, S., Davidson, W. S., II, Bogat, G. A., Levendosky, A.,
& von Eye, A. (2005). Validation of the Subtle and
Overt Psychological Abuse Scale: An examination of
construct validity. Violence and Victims, 20, 407-416.
Kahn, F. I., Welch, T. L., & Zillmer, E. A. (1993). MMPI-2
profiles of battered women in transition. Journal of
Personality Assessment, 60, 100-111.
Kasian, M., & Painter, S. L. (1992). Frequency and severity
of psychological abuse in a dating population. Journal
of Interpersonal Violence, 7, 350-364.
Kaslow, N. J., Thompson, M. P., Meadows, L. A., Jacobs, D.,
Chance, S., Gibb, B., et al. (1998). Factors that mediate
and moderate the link between partner abuse and sui-
cidal behavior in African American women. Journal of
Consulting and Clinical Psychology, 66, 533-540.
Katz, J., & Arias, I. (1999). Psychological abuse and
depressive symptoms in dating women: Do different
types of abuse have differential effects? Journal of
Family Violence, 14, 281-295.
Katz, J., Arias, I., & Beach, S. R. H. (2000). Psychological
abuse, self-esteem, and women’s dating relationship
outcomes: A comparison of the self-verification and
self-enhancement perspectives. Psychology of Women
Quarterly, 24, 349-357.
Loring, M. T. (1994). Emotional abuse. New York: Lexington
Books.
Maiuro, R. D. (2001). Sticks and stones may break my
bones, but names will also hurt me: Psychological abuse
in domestically violent relationships. In K. D. O’Leary
& R. D. Maiuro (Eds.), Psychological abuse in violent
domestic relations. New York: Springer.
Marshall, L. L. (1994). Physical and psychological abuse.
In W. R. Cupach & B. H. Spitzberg (Eds.), The dark side
of interpersonal communication (pp. 281-311). Hillsdale,
NJ: Lawrence Erlbaum.
Marshall, L. L. (2001). Effects of men’s subtle and overt psy-
chological abuse on low-income women. In K. D. O’Leary
& R. D. Maiuro (Eds.), Psychological abuse in violent domes-
tic relations. New York: Springer.
Migeot, M., & Lester, D. (1996). Psychological abuse in
dating, locus of control, depression, and suicidal pre-
occupation. Psychological Reports, 79, 682.
Murphy, C. M., & Hoover, S. A. (1999). Measuring emo-
tional abuse in dating relationships as a multifactorial
construct. Violence and Victims, 14(1), 39-53.
Murphy, C. M., & O’Leary, K. D. (1989). Psychological
aggression predicts physical aggression in early mar-
riage. Journal of Consulting and Clinical Psychology, 57,
579-582.
Orava, T. A., McLeod, P. J., & Sharpe, D. (1996). Perception
of control, depressive symptomatology, and self-
esteem in women in transition from abusive relation-
ships. Journal of Family Violence, 11, 167-186.
Pape, K. T., & Arias, I. (1995). Control, coping, and victim-
ization in dating relationships. Violence and Victims,
10(1), 43-54.
Raghavan, C., Swan, S. C., Snow, D. L., & Mazure, C. M.
(2005). The mediational role of relationship efficacy
and resource utilization in the link between physical
and psychological abuse and relationship termination.
Violence Against Women, 11(1), 65-88.
Rauter, U. K., Leonard, C. E., & Swett, C. P. (1996). SCL-
90-R factor structure in an acute, involuntary, adult
psychiatric inpatient sample. Journal of Clinical
Psychology, 52, 625-629.
Rollstin, A. O., & Kern, J. M. (1998). Correlates of battered
women’s psychological distress: Severity of abuse and
duration of the postabuse period. Psychological Reports,
82, 387-394.
Sackett, L. A., & Saunders, D. G. (2001). The impact of
different forms of psychological abuse on battered
women. In K. D. O’Leary & R. D. Maiuro (Eds.),
Psychological Abuse in Violent Domestic Relations
(pp. 197-210). New York: Springer.
Schmitz, N., Hartkamp, N., Kiuse, J., Franke, G. H., Reister,
G., & Tress, W. (2000). The Symptom Check-List-90-R
(SCL-90-R): A German validation study. Quality of Life
Research, 9, 185-193.
Schumacher, J. A., Slep, A. M. S., & Heyman, R. E. (2001).
Risk factors for male-to-female partner psychological
abuse. Aggression and Violent Behavior, 6, 255-268.
Spielberger, C. D. (1983). Manual for the State-Trait
Anxiety Inventory (STAI). Palo Alto, CA: Consulting
Psychologists Press.
Stets, J. E. (1991). Psychological aggression in dating rela-
tionships: The role of interpersonal control. Journal of
Family Violence, 6(1), 97-114.

Follingstad / PSYCHOLOGICAL AGGRESSION EFFECTS 289
Straight, E. S., Harper, F. W., & Arias, I. (2003). The impact
of partner psychological abuse on health behaviors
and health status in college women. Journal of
Interpersonal Violence, 18, 1035-1054.
Street, A. E., & Arias, I. (2001). Psychological abuse and
PTSD in battered women: Examining the roles of
shame and guilt. Violence and Victims, 16(1), 65-78.
Taft, C. T., O’Farrell, T. J., Torres, S. E., Panuzio, J., Monson,
C. M., Murphy, M., et al. (2006). Examining the corre-
lates of psychological aggression among a community
sample of couples. Journal of Family Psychology, 20,
581-588.
Tolman, R. M. (1989). The development of a measure of
psychological maltreatment of women by their male
partners. Violence and Victims, 4, 159-177.
Tolman, R. M. (1992). Psychological abuse of women. In
R. T. Ammerman & M. Hersen (Eds.), Assessment of
family violence: A clinical and legal sourcebook. New York:
Wiley.
Tolman, R. M., & Bhosley, G. (1991). The outcome of par-
ticipation in a shelter-sponsored program for men who
batter. In D. Knudsen & J. Miller (Eds.), Abused and bat-
tered: Social and legal responses. New York: Aldine de
Gruyter.
Trotter, J. L., Bogat, G. A., & Levendosky, A. A. (2004). Risk
and protective factors for pregnant women experienc-
ing psychological abuse. Journal of Emotional Abuse,
4(2), 53-70.
Tuel, B. D., & Russell, R. K. (1998). Self-esteem and
depression in battered women: A comparison of
lesbian and heterosexual survivors. Violence Against
Women, 4, 344-362.
Vitanza, S., Vogel, L. C. M., & Marshall, L. L. (1995).
Distress and symptoms of PTSD in abused women.
Violence and Victims, 10, 23-34.
Diane R. Follingstad, PhD, cur-
rently serves as the Women’s Circle
Endowed Chair in the Center for
Research on Violence Against Women
and as a professor in the Department of
Psychiatry, College of Medicine at the
University of Kentucky. This appointment follows her
career as a distinguished professor emerita in clinical and
forensic psychology at the University of South Carolina.
Her research in the area of intimate partner violence has
covered issues related to battered women, physical dating
violence, factors impacting jury verdicts in cases where
battered women killed a partner, and most recently, her
research efforts have led to more sophisticated measure-
ment of psychological aggression. She is board certified as
a forensic psychologist by American Board of Professional
Psychology and has held such positions as president of
that national certifying board (American Board of
Forensic Psychology) and secretary of APA’s Division of
Psychology and Law, which awarded her with honorary
fellow status.
For reprints and permissions queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.