Posttraumatic Stress Disorder in a National Sample of Female and Male Vietnam Veterans: Risk Factors, War-Zone Stressors, and Resilience-Recovery Variables

Abstract

Relationships among pretrauma risk factors (e.g., family instability, childhood antisocial behavior), war-zone stressors (e.g., combat, perceived threat), posttrauma resilience-recovery variables (e.g., hardiness, social support), and posttraumatic stress disorder (PTSD) symptom severity were examined. Data from a national sample of 432 female and 1,200 male veterans were analyzed using structural equation modeling. For both genders, direct links to PTSD from pretrauma, war-zone, and posttrauma variable categories were found; several direct associations between pretrauma and posttrauma variables were documented. Although war-zone stressors appeared preeminent for PTSD in men, posttrauma resilience-recovery variables were more salient for women. Researchers, policymakers, and clinicians are urged to take a broad view on trauma and its sequelae, especially regarding possible multiple exposures over time and the depletion and availability of important resources.

Full text

Journal
of
Abnormal Psychology
1999, Vol. 108,
No. 1,
164-170
In
the
public domain
SHORT
REPORTS
Posttraumatic Stress Disorder
in a
National Sample
of
Female
and
Male
Vietnam Veterans: Risk Factors, War-Zone Stressors,
and
Resilience-Recovery Variables
Daniel
W.
King
and
Lynda
A.
King
Boston
Veterans
Affairs
Medical Center
and
Boston University School
of
Medicine
David
W. Foy
Pepperdine University
Terence
M.
Keane
Boston Veterans
Affairs
Medical Center
and
Boston University School
of
Medicine
John
A.
Fairbank
Research Triangle Institute
and
Duke University Medical Center
Relationships among
pretrauma
risk
factors
(e.g.,
family instability, childhood antisocial behavior),
war-zone
stressors
(e.g.,
combat,
perceived threat),
posttrauma
resilience-recovery
variables
(e.g.,
har-
diness,
social
support),
and
posttraumatic
stress
disorder
(PTSD)
symptom severity were examined.
Data
from
a
national sample
of 432
female
and
1,200
male
veterans were analyzed using structural equation
modeling.
For
both genders,
direct
links
to
PTSD
from pretrauma, war-zone,
and
posttrauma variable
categories
were found;
several
direct
associations
between pretrauma
and
posttrauma
variables
were
documented. Although war-zone
stressors
appeared
preeminent
for
PTSD
in
men, posttrauma resilience-
recovery variables
were
more
salient
for
women.
Researchers,
policymakers,
and
clinicians
are
urged
to
take
a
broad view
on
trauma
and its
sequelae,
especially
regarding possible multiple exposures over
time
and
the
depletion
and
availability
of
important
resources.
Posttraumatic
stress disorder (PTSD)
is an
anxiety disorder
observed
in
persons
who
have been exposed
to an
extreme
stressor
that
evokes feelings
of
"intense fear, helplessness,
or
horror"
(American
Psychiatric Association, 1994,
p.
428). Symptoms
in-
clude
reexperiencing
the
event through frightening dreams
and
intrusive
recollections, avoidance
of
circumstances that might trig-
ger a
reexperiencing episode, emotional numbing
and
retreat
from
intimate
relationships,
and
increased arousal.
The
earliest
efforts
to
elucidate
the
etiology
of
PTSD evoked
a
debate
as to
whether symptoms were attributable primarily
to the
trauma
(e.g., Figley, 1978)
or to a
condition that predated exposure
(Worthington,
1978).
The
cumulative research findings over
2
decades
emphasize
the
preeminence
of the
stressor (e.g., Green,
Daniel
W.
King, Lynda
A.
King,
and
Terence
M.
Keane, National
Center
for
Posttraumatic
Stress
Disorder,
Boston Veterans Affairs Medical
Center
and
Boston University
School
of
Medicine; David
W.
Foy, Grad-
uate School
of
Education
and
Psychology,
Pepperdine
University; John
A.
Fairbank, Department
of
Psychiatry
and
Behavioral
Sciences,
Research
Triangle
Institute, Research Triangle Park, North Carolina
and
Duke Uni-
versity Medical Center.
This
research
was
supported
by
National Institute
of
Mental Health
Grant
MH49168
(Violence
and
Traumatic Stress Program).
We
express
our
appreciation
to
Casey Taft
for his
assistance.
Correspondence
concerning
this article should
be
addressed
to
Daniel
W.
King, Behavioral
Science
Division, National
Center
for
Posttraumatic
Stress
Disorder, Boston
Veterans
Affairs Medical Center, Boston, Massa-
chusetts
02130.
Electronic mail
may be
sent
to
king.daniel@boston.va.gov.
1994) with
an
approximate dose-response relationship: Higher
levels
of
exposure
are
associated with greater symptomatology.
Yet
the
relationship
is not
perfect,
and
there
has
been increasing
attention
given
to
risk
and
resilience factors that appear
to
exac-
erbate
or
mitigate
the
stress response. Such factors have included
pretrauma
demographic
and
psychosocial characteristics (e.g.,
Me-
Cranie, Hyer, Boudewyns,
&
Woods, 1992)
and
posttrauma
re-
sources
and
features
of the
recovery environment (e.g., Solomon,
Waysman,
&
Mikulincer, 1990). Thus,
a
better explanation
for the
occurrence
or
persistence
of
PTSD
in
trauma victims
may
very
well
lie
with
the
testing
of
more sophisticated
multivariate
models.
To
this end,
we
embarked
on a
series
of
studies intended
to
systematically
examine
a
network
of
potentially explanatory vari-
ables associated with PTSD, applying structural equation modeling
methodologies
to
data
from
the
National Vietnam Veterans Read-
justment
Study
(NVVRS;
Kulka
et
al.,
1990).
A first
endeavor
(L. A.
King
&
King, 1994) examined
the
latent structure
of
PTSD
as
represented
by the
NVVRS's
primary survey instrument,
the
Mississippi Scale
for
Combat-Related PTSD (Keane, Caddell,
&
Taylor, 1988). Confirmatory factor analysis
of
Mississippi Scale
scores
yielded
a
higher-order global PTSD factor
that
subsumed
the
symptom categories
of
reexperiencing
and
situational avoid-
ance, withdrawal
and
numbing, arousal
and
lack
of
control,
and
guilt
and
suicidality. Items within these categories then served
as
manifest
indicators
of the
PTSD construct
in the
later studies.
We
next investigated
how
several war-zone stressor dimensions
(exposure
to
traditional combat events, exposure
to
atrocities
or
164

SHORT
REPORTS
165
episodes
of
extraordinarily abusive violence, subjective
or
per-
ceived threat,
and the
general milieu
of a
harsh
or
malevolent
environment)
were differentially associated with
PTSD
(D. W.
King, King,
Gudanowski,
&
Vreven, 1995).
An
interesting
finding
was
the
potency
of the
malevolent
environment latent
variable,
a
relatively
low-magnitude
but
continuous
or
unremitting stressor
encompassing irritations, deprivations,
and
daily
pressures
related
to
life
in a war
zone;
it had
both direct
and
indirect
effects
on
PTSD. Traditional combat made
its
impact indirectly through
the
perceived threat stressor,
and
atrocities-abusive
violence demon-
strated
a
direct association with
PTSD.
The
third
and
fourth
studies
in the
sequence incorporated this
multidimensional representation
of
war-zone stressors along with
prewar
risk factors
(D. W.
King, King, Foy,
&
Gudanowski, 1996)
and
postwar resilience
and
recovery variables
(L. A.
King, King,
Fairbank,
Keane,
&
Adams, 1998), respectively.
In
both studies,
as
expected,
the
war-zone
stressors
demonstrated
a
strong
association
with
PTSD,
and the
model
for
female
veterans differed
from
the
model
for
male veterans. With regard
to
salient prewar factors
for
women, instability within
the
family
of
origin
had the
largest total
effect
on
PTSD,
and
early trauma history (accidents, assaults,
and
natural
disasters)
had a
direct link
to
PTSD.
For
men, large total
effects
were
found
for
family
instability, childhood antisocial
behavior,
and age at
entry
to
Vietnam; younger
age at
entry
and
early
trauma history were directly associated with PTSD. More-
over, early trauma history interacted with war-zone stressor level
to
exacerbate
PTSD
symptoms
for
high
combat-exposed
male
veterans.
Thus, prewar risk factors played
a
role
in
explaining
PTSD, albeit more
so for men
than
for
women.
In
the
study involving postwar variables,
the
final
models
for
both genders identified robust mediation
effects
for
social support
(structural
and
functional),
the
intrapersonal resource
of
hardiness,
and
additional negative
life
events following
the
Vietnam experi-
ence. Though the patterns of relationships among war-zone stres-
sors,
postwar variables,
and
PTSD differed
for
women
and
men,
the
most important mediator
for
both
was
functional
social
sup-
port,
followed
by
hardiness. Overall,
the
findings suggested that
an
appreciation
of the
mechanisms
by
which war-zone stressors
are
related
to
PTSD should take into consideration important
resilience-recovery influences.
The
present study formulated
and
tested
an
integrated model
using
all
categories
of
variables: prewar
risk
factors,
war-zone
stressors, postwar resilience-recovery variables,
and
PTSD.
Al-
though
each
of the
prior studies addressed important substantive
concerns regarding PTSD,
our
intent here
was to
simultaneously
consider
the
system
of
associations among variables
within
and
across
the
several categories.
We
postulated that direct relation-
ships
to
PTSD documented
in the
earlier studies would
be
sup-
ported
in the
integrated model.
In
addition,
the
testing
of the
integrated
model permitted
for the
first
time
the
evaluation
of
relationships between prewar
risk
factors
and
postwar resilience-
recovery variables.
In
this regard,
we had
several specific hypoth-
eses.
First, drawing
from
contemporary perspectives
on
early
disorganized
attachment
and
psychopathology
(Jones, 1996; Main,
1996),
the
prewar factors
of
relationship with father
and
family
instability
were expected
to
have direct associations with postwar
social support
and
hardiness. Second, childhood antisocial behav-
ior was
proposed
to
directly relate
to
social support
and
negative
life
events
in the
postwar period. This proposition derived
from
previous work suggesting that
the
aggressive, abrasive posture
characterizing antisocial behavior leads
to
relationship
difficulties
and
social isolation (Loeber
&
Dishion, 1983)
as
well
as
suscep-
tibility
to a
plethora
of
additional adverse circumstances
and
misfortunes (Caspi,
Elder,
&
Bern,
1987).
Third, early trauma
history
was
hypothesized
to
have
a
direct link
to
postwar negative
life
events, consistent
with
Bremner,
Southwick,
and
Charney's
(1995)
observation that "individuals with
a
history
of
exposure
to
extreme psychological stress appear
to
have
an
increased vulner-
ability
to
exposure
to
subsequent
stress"
(p.
152). Then, following
the
findings
of
Green, Grace, Lindy, Gleser,
and
Leonard (1990),
we
proposed that war-zone stressors would emerge
as the
most
powerful
contributor
to
PTSD, followed
by
postwar
resilience-
recovery variables,
and
then prewar
risk
factors.
The
first
published works
from
the
NVVRS database, emanating
from
the
primary Research Triangle Institute researchers, were
largely descriptive
in
nature:
Schlenger
et
al.
(1992)
and
Weiss
et
al.
(1992)
reported
prevalence rates
for
current,
lifetime,
and
partial PTSD
for
various cohorts
and
subgroups,
and
Jordan
et al.
(1991)
provided current
and
lifetime rates
for
other psychiatric
disorders
and
related them
to
war-zone exposure. Prior multivar-
iate analyses
of the
NVVRS database have also been conducted
by
Fontana
and
Rosenheck
(1994;
a
study
of
PTSD
in the
male
veteran subsample)
and
Fontana, Schwartz,
and
Rosenheck
(1997;
a
study
of
PTSD
in the
female veteran subsample). They likewise
documented
the
importance
of
combat exposure,
family
back-
ground,
and
homecoming
and
other
social
supports.
Yet,
the
spe-
cific
structural equation modeling strategies that
they
used could
have better optimized
the
potential
of the
methodology. Many
of
their endogenous,
"down-stream,"
or
dependent variables were
simple dichotomies,
to
which were applied linear regression-based
procedures that presuppose continuous dependent variables.
A
more appropriate approach would involve
the use of
logistic
regression-based procedures designed
for
dichotomous outcomes
(e.g.,
Agresti
&
Finlay, 1986; Cohen
&
Cohen, 1983). Further-
more,
in
both
the
model
for
women
and the
model
for
men, very
few
constructs were formulated
as
latent variables
with
multiple
manifest
indicators.
The
consequence
was a
limited ability
to
capitalize
on a
major
benefit
of
structural equation modeling,
namely,
the
generation
of
unbiased parameter estimates
or
path
coefficients
in the
structural model that derive
from
the
specifica-
tion
of
perfectly reliable latent variables (Bollen, 1989; James,
Mulaik,
&
Brett, 1982;
Joreskog
&
Sorbom,
1993).
Our use in the
present study
of
continuous endogenous variables
and
models that
fully
use
latent variables
was
intended
to
enhance
the
statistical
conclusion
validity (Cook
&
Campbell, 1979)
of
findings
from
the
NVVRS
data.
Method
Data
Source
Data
were
taken
from the
responses
of
1,632 (432
women
and
1,200
men)
veterans
who
participated
in the
NVVRS
(Kulka
et
al.,
1990).
All had
served
in the
Vietnam
theater
of
operations
sometime
between
August
1964
and
March
1975.
In an
extended
interview
and
self-report
session,
each
veteran
supplied
information
on a
broad
array
of
topics
related
to
prewar
background
and
functioning,
military
and
war-zone
experiences,
and
postwar
circumstances,
life
events,
and
mental
health
status.
Further
information
on the
sampling
methodology
and
sample
characteristics
is

166
SHORT REPORTS
available
in a
number
of
other sources (e.g., Jordan
et
al.,
1991;
Kulka
et
al.,
1990;
Schlenger
et
al.,
1992; Weiss
et
al.,
1992).
Variables
and
Their Operational Definitions
Table
1
identifies
and
briefly
defines
the
collection
of 15
variables used
in
the
current study. Details
on
item composition, scoring,
and
psychomet-
ric
properties
may be
found
in the
references provided
in the
last column
or
obtained directly
from
the
first
author (Daniel
W.
King).
Overview
of
Analyses
In
prior work with
the
NVVRS data
(D. W.
King
et
al.,
1995;
D. W.
King
et
al.,
1996;
L. A.
King
et
al.,
1998),
we
conducted confirmatory
factor
analyses
and
successfully demonstrated
and
cross-validated
the
viability
of
measurement models
for all 15
latent variables used
in the
present study. Accordingly,
the
analyses
for
this final study
in the
series
concentrated
on the
specification
and
evaluation
of the
structural model
representing
a
full
network
of
relationships among these latent variables.
Moreover,
building
on our
previous
findings
that demonstrated gender-
based
differences
in the
measurement models,
we
necessarily specified
and
evaluated separate structural models
for
female
and
male veterans (Marsh,
1994).
In
all
analyses, matrices
of
covariances among
observed
or
manifest
indicators
were submitted
to the
LISREL
8
program
(Joreskog
&
Sorbom,
1993). Generalized least squares estimation
was
used,
and
residual covari-
ances
for all
manifest
indicators were fixed
at 0.
Results
and
Discussion
Copies
of the
matrices
that
were analyzed
and
full
descriptions
of
the
decision rules
for
model
trimming
may be
obtained
from
Daniel
W.
King.
The
final
accepted model
for
women
yielded
^(915,
N =
423)
=
1343.84,
p <
.001;
for the
final
model
for
men,
^(909,
N =
1,150)
=
2214.97,
p <
.001.
All
other
fit
indices
endorsed
the
accepted models.
For
women,
the
LISREL
goodness-
of-fit
index
(GFI;
Joreskog
&
Sorbom,
1993)
was
.86;
the
com-
parative
fit
index
(CFI;
Rentier,
1990)
was
.99;
the
parsimony
normed
fit
index
(PNFI;
James
et
al.,
1982)
was
.89;
and the
root
mean
square
error
of
approximation
(RMSEA;
Steiger,
1990)
was
.033. This
last
value
is an
estimate
of the
discrepancy
between
the
population
covariance
matrix
and the
fitted
covariance
matrix
for
the
accepted model;
it is
well
below
the .05
standard
for
close
fit
(Browne
&
Cudeck,
1989),
with
an
associated
probability
ap-
proaching
1.00
that
its
true
value
falls
below
this
standard.
For
men,
these
fit
indices
were
also
quite
satisfactory;
GFI of
.91,
CFI
of
.99, PNFI
of
.90,
and
RMSEA
of
.035,
again
with
an
associated
probability
approaching
1.00.
In
line
with
the
present
study's
first
proposition, direct
relation-
ships
to
PTSD
in
these
integrated
models
for
women
and men
generally
mirror
direct
effects
found
in the
prior
component
stud-
ies of
prewar, war-zone,
and
postwar
variables.
As
Figure
1
shows,
six
variables
were
directly
linked
to
PTSD
for
women:
the
prewar
Table
1
Variables
in the
Model
Variable
No. of
manifest
indicators
Description
Source
of
more detailed
information
Prewar risk factors
Socioeconomic status
Relationship
with father
Family instability
Early trauma
history"
Childhood antisocial behavior
Age at
entry
to
Vietnam
War-zone stressors
Traditional
combat"
Atrocities-abusive
violence
3
Perceived threat
Malevolent
environment
Postwar resilience-recovery variables
Additional
stressful
life
events
a
Hardiness
Structural
social
support"
Functional
social support
PTSD
3
Father's education, mother's education; veteran's perception
of
family financial well-being
2
Closeness
to
primary father figure
3
Disruptive
family
environment; presence
of
interparental
and
parent-to-child abuse
1
Cumulative index
of
traumagenic
or
life-threatening
experiences
5
Disciplinary problems prior
to age
15
1
Veteran's
age in
years upon arrival
in the war
zone
1
Stereotypical war-zone events (firing
a
weapon, receiving
fire)
1
War-zone events considered deviant
(e.g.,
mutilation, killing
civilians)
3
Subjective judgments
of
fear
6
Day-to-day discomforts (e.g.,
the
heat,
poor
living facilities)
1
Highly stressful experiences
in the
postwar period
4
Sense
of
control, commitment
to
self, viewing change
as
challenge
1
Size
and
complexity
of
social network
7
Perceived emotional sustenance
and
instrumental assistance
from
others
6
Categories
of
symptoms formed from
Keane
et
al.'s
(1988)
Mississippi
Scale
items; symptom count
from
the
NVVRS's
adaptation
of the
Diagnostic Interview
Schedule (Robins, Helzer,
Croughan,
&
Ratcliff,
1981);
Kulka
et
al.'s
(1990)
predicted probability
of
PTSD
D. W.
King
et al.
(1996)
D. W.
King
et al.
(1996)
D.
W.
King
et al.
(1996)
D. W.
King
et al.
(1996)
D. W.
King
et al.
(1996)
D. W.
King
et al.
(1996)
D. W.
King
et al.
(1995)
D. W.
King
et al.
(1995)
D. W.
King
et al.
(1995)
D. W.
Kinget
al.
(1995)
L. A.
King
et al.
(1998)
L. A.
King
et al.
(1998)
L. A.
King
et al.
(1998)
L. A.
King
et al.
(1998)
L. A.
King
&
King (1994);
D. W.
King
et al.
(1995)
Note.
NVVRS
=
National Vietnam Veterans Readjustment Study;
PTSD
=
posttraumatic
stress disorder.
a
These variables
had
causal indicators;
all
others
had
effect
indicators.

SHORT REPORTS
167
Figure
1.
Simplified
representation
of the
structural models
of the
associations among prewar
risk
factors,
war-zone
stressors,
postwar resilience-recovery variables,
and
posttraumatic
stress disorder (PTSD).
All
between-category
coefficients
for
women
and men are
standardized
within
genders.
SES =
socioeconomic
status;
FI =
family
instability;
ETH
=
early trauma history;
CAB =
childhood antisocial behavior;
AGE = age
at
entry
to
Vietnam;
TC =
traditional combat;
AAV =
atrocities-abusive violence;
PT
=
perceived threat;
ME
=
malevolent environment; ASLE
=
additional
stressful
life
events;
HAR
=
hardiness;
SSS =
structural
social support;
FSS =
functional
social support. Relationships with asterisks were
not
predicted
in
this study
or
observed
in the
prior component studies.
risk factor
of
early trauma history,
the
war-zone stressors
of
atrocities-abusive
violence
and
perceived threat,
and the
postwar
resilience-recovery variables
of
additional
stressful
life
events,
hardiness,
and
functional
social support. They accounted
for 11%
of
the
variance
in
PTSD.
For
men, nine variables showed direct
relationships
to
PTSD:
the
same
six as for
women, plus
the
prewar
risk
factor
of age at
entry
to
Vietnam,
the
malevolent environment
war-zone
stressor,
and the
postwar resilience-recovery variable
of
structural
social support. These nine variables accounted
for
70%
of
the
variance.
The
only inconsistency
in
direct relationships
to
PTSD over
the
several studies
in the
sequence involved
the
asso-
ciation between malevolent environment
and
PTSD.
In the
earlier
study
of
prewar
and
war-zone variables, this relationship
was
obtained
for
both genders.
In the
prior study
of
postwar
and
war-zone
variables,
the
relationship
was not
observed
for
either
gender, likely
as a
consequence
of
multicollinearity
between
ma-
levolent environment
and one or
more postwar variables.
In the
current study,
the
association
was not
present
in the
final
model
for
women,
and a
rather modest association
was
retained
in the
final
model
for
men, with
the
large
majority
of
malevolent environ-
ment's
total
effect
(74%) attributable
to its
indirect relationship
to
PTSD
through other war-zone stressors
and
postwar variables.
Support
for the
specific
hypotheses relating prewar
risk
factors
to
postwar resilience-recovery variables
may be
found
in the
associations that
label
the
horizontal line
in the
uppermost portion
of
Figure
1.
In the
model
for
women, there were three such
associations,
two of
which were hypothesized
and one of
which
was
unanticipated;
in the
model
for
men, three
of the
four
associ-
ations were hypothesized
and one was
unanticipated.
For
both
genders, early trauma history
was
directly linked
to
additional

168
SHORT REPORTS
stressful
life
events
in the
postwar period, thus
affirming
Bremner
et
al.'s
(1995) commentary that prior
life
stress predicts later
life
stress. Also, consistent with Loeber
and
Dishion
(1983),
the
pro-
posed direct
and
negative association between childhood antisocial
behavior
and
later
functional
social
support
was
observed
for
women.
And the
hypothesized direct
and
negative relationships
between
family
instability
and
both forms
of
social support were
found
for
men, suggesting that
a
chaotic
family
environment
characterized
by
parental dysfunction
and
disorganized attachment
may
very well compromise
one's
ability
to
build, foster,
and
benefit
from
a
support network
in
later years (e.g.,
Jones,
1996;
Main,
1996).
In the
end,
5 of 20
specific hypotheses regarding
prewar-postwar direct relationships were supported.
These
results
are
not
inconsequential, especially when
one
considers that
the
prewar predictors were competing with
a
rather powerful
and
salient
set of
other predictors,
the
war-zone stressors,
and
that
the
prewar risk
factors
themselves were associated with
the
war-zone
stressors. Thus, given
the
complex interplay
of
relationships
hy-
pothesized
in the
model,
one
would anticipate that some would
necessarily diminish
in the
presence
of
others.
A
final
expectation
in
this study
was a
hierarchy among
the
categories
of
variables contributing
to
PTSD,
with war-zone stres-
sors proposed
to be
preeminent, followed
by
postwar resilience-
recovery variables,
and
then prewar
risk
factors.
To
address this
research issue,
we
reasoned
that
the
most appropriate index
of a
single
variable's
contribution would
be its
total
effect,
a
composite
of
its
direct
effect
on
PTSD
and all
indirect
effects
on
PTSD
through
other variables. Accordingly,
we
summed
the
absolute
values
of
total
effects
of
variables within each
of the
categories
to
ascertain relative
influence.
For
men,
the
observed
order
was
consistent
with
our
proposition.
The sum of the
total
effects
for
war-zone stressors
was
1.36;
for
postwar resilience-recovery vari-
ables,
the
value
was
.96;
and for
prewar
risk
factors,
it was
.80.
For
women, however,
the
order
did not
conform
to our
expectation.
The
sums were 1.10
for the
war-zone stressors, 1.33
for the
postwar resilience-recovery variables,
and .53 for the
prewar
risk
factors.
It
appears, therefore,
that
postwar variables
are
somewhat
more important
in
accounting
for
PTSD symptomatology
in fe-
male veterans. These women
may
have been more adept
at
mar-
shaling
intrapersonal
and
interpersonal resources
in
times
of
need
than
their male veteran counterparts.
On the
other hand,
one
must
keep
in
mind that
the
war-zone experience
for
many male veterans
was
likely more directly
life
threatening than
the
experience
of
women
who
served
in
Vietnam. Consequently, war-zone incidents
and
conditions
may be
more salient
for
men.
As
with much
of the
research
on
stress
and
trauma,
the
cross-
sectional design
of the
NVVRS
and the
retrospective self-report
nature
of its
data necessarily mandate careful interpretation
of our
findings.
The
principal concern
is
ambiguity about
the
direction
of
relationships among variables, which
may
derive from
a
number
of
sources: poor recall, particularly
for
some
experiences
in an
emo-
tionally charged
or
highly stressful environment (Burke, Heuer,
&
Reisberg, 1992;
Christiansen,
1992);
an
associated tendency
to
reconstruct events
and
assign meaning based
on
one's
current
psychological
state
(Metis,
Sprecher,
&
Cupach, 1991; Nisbett
&
Wilson, 1977);
and
possible recognition-seeking, compensation-
seeking,
or the
need
to
present oneself
in a
socially desirable
manner
(D. W.
King
&
King,
1991).
Many theoreticians
and
researchers could
rightfully
argue that paths
from
PTSD
to
other
constructs
in the
system might
be
operating: Level
of
current
PTSD
symptomatology
may
very
well
influence replies
to
inqui-
ries
about hardiness
and
perceived support
from
others. Indeed,
a
veteran's
PTSD
symptomatology
may
work
to
drive away rela-
tives
and
friends.
In
addition,
both
McFarlane
(1992)
and
South-
wick,
Morgan, Nicolaou,
and
Charney
(1997) provided empirical
support
for the
proposition that PTSD
and
associated mood
and
arousal
states
may at
least partially determine
how
strongly
one
endorses memories
of
prior traumatic experiences. Furthermore,
we
duly
recognize that structural equation modeling does
not
confirm
a
model. Rather,
as
emphasized
by
Breckler (1990),
among
others,
it
simply asserts that there
are no
available data
to
disconfirm
the
model.
The
best strategy
is for
putative relation-
ships
in any
model
to be
informed
by
theory
and
substantive
issues,
an
approach that guided decisions
in the
present
study.
We
also acknowledge that
not all
factors potentially associated with
PTSD
were incorporated into this study, including
the
comorbid
condition (and possible coping mechanism)
of
substance abuse.
With
these cautions
in
mind, what conclusions
may be
drawn
from
this examination
of an
integrated network
of
relationships
among variables explaining PTSD symptomatology? Without
doubt,
the
findings endorse
a
multivariate
perspective
on
PTSD.
Taken together,
all
categories
of
variables (pretrauma, trauma,
and
posttrauma)
appear important
to our
understanding
of
individual
differences
in the
display
of
symptoms. With
due
regard
to
pos-
sible common method
effects,
it
appears that
a
large amount
of
variance
(72%
for
women;
70% for
men)
was
accounted
for by the
collection
of
variables that were directly linked
to
PTSD
in the
models. Also, each
of
these variables, drawn
from
all
three cate-
gories, yielded
a
unique contribution
to
reported symptom sever-
ity. Thus, events
and
circumstances
that
preceded
the
focal
trauma
as
well
as
events
and
circumstances that
characterize,
the
post-
trauma environment must
be
recognized,
a
conclusion consistent
with
the
work
of
Fontana
and
Rosenheck (1994; Fontana
et
al.,
1997).
The
pretrauma
risk
factors
in
this
study—especially
(low)
so-
cioeconomic status, (poor) relationship with father,
family
insta-
bility,
and
early trauma
history—portray
a
stress-filled
and
possi-
bly
disadvantaged existence
for
some veterans even prior
to the
Vietnam
experience.
As
Green (1994) suggested, current symp-
tomatology
may be
tied
to
multiple exposures
to
stressful
life
events
across
time. Evidence
for
this position lies
in the
path
coefficient
for the
relationship between early trauma history
and
PTSD
that
was
retained
in the
models
for
both genders;
not
only
are the
focal war-related
experiences
associated with PTSD symp-
tom
severity,
but so are
traumas that preceded such experiences.
Also,
the
underlying mechanism
by
which
the
pretrauma factors
relate
to
current
PTSD
symptomatology
may be a
depletion
of
available resources
to
deal with subsequent stressors,
not
only
those encountered
in the war
zone
but
also those characterizing
the
recovery environment. This possibility
is
consistent
with
the
tenets
of
Hobfoll's (1989) conservation
of
resources conceptualization
of
stress. Indeed,
our
examination
of the
associations between
risk
factors
and
resilience-recovery variables supports this assertion,
with
several documented path coefficients between features
of
prewar life circumstances (e.g., socioeconomic status
and
family
instability)
and
postwar intrapersonal
and
social
resources
(e.g.,
hardiness
and
social support).
It
should
be
noted that these
prewar-

SHORT
REPORTS
169
postwar
associations
are
statistically independent
of the
influence
of
war-zone stressors.
In
addition,
for
both women
and
men,
the
resilience-recovery
variables were quite potent. Particularly relevant were
the
large
associations
of
hardiness
and
functional social support with
PTSD
for
both genders. Each
of
these resource variables
may
serve
to
uniquely
offset
the
deleterious consequences
of
stressors
on
PTSD.
Again,
we see the
salience
of
multiple stressful life events across
the
life
span. Continuing
the
argument
for
depletion
of
resources,
the
additional stressful
life
events variable
was
directly related
to
PTSD
but was
also negatively associated with both hardiness
and
functional
social support.
And of
course,
one or
more
of the
war-zone
stressors likewise
had
negative direct relationships with
hardiness
and
social support
for
both women
and
men. Last,
the
findings
endorsed
a
multifactorial
representation
of the
traumatic
experience
itself,
in
this
case,
several
war-zone
stressor
dimen-
sions
being differentially associated with both pretrauma
risk
factors
and
posttrauma
resilience-recovery variables.
To
reiterate,
the
study
reported
here
is the
culmination
of a
series
of
inquiries into possible explanatory variables accounting
for
chronic PTSD symptomatology following exposure
to an
iden-
tified,
potentially traumatizing experience. Perhaps
the
major les-
son
to be
learned
from
this examination
of a
national sample
of
Vietnam veterans
is
that responses
to
highly stressful negative life
events
may
derive
from
a
complex interplay
of
factors that stretch
backward
in
time
from
the
focal experience
and
forward
in
time
to
the
present. Therefore, those interested
in the
study
of
trauma
and
its
aftermath
and
those charged with planning, implementing,
and
evaluating
service delivery
to
trauma victims need
to
adopt
a
broad
outlook.
We
would recommend that this
perspective
be
motivated
by
two
overriding themes. First, exposure
to
multiple stressful
events
over
an
extended period
of
time, perhaps many years,
may
drive
current symptomatology
and
mandates more than
a
narrow
investigation
of a
single traumatic event. Second,
access
to and
activation
of
intrapersonal,
interpersonal,
and
environmental
re-
sources
may
serve
to
assuage
the
occurrence
or
persistence
of
reactions
to
severe stressors
and
thereby work
to
neutralize their
negative impact
on
psychological well-being.
Of
course,
our
data
suggest
that these
two
features
of the
potential
etiological
net-
work—revictimization
or
cumulative trauma
and
resource
avail-
ability—are
associated with
one
another:
A
consequence
of any
given
incident
of
exposure
may be the
dissipation
of
resources,
leaving
the
victim more vulnerable
to
stress
reactions
when faced
with future
traumatic
events.
References
Agresti,
A.,
&
Finlay,
B.
(1986). Statistical methods
for the
social sciences
(2nd
ed.).
New
York: MacMillan.
American
Psychiatric Association.
(1994).
Diagnostic
and
statistical
man-
ual
of
mental disorders
(4th
ed.).
Washington,
DC:
Author.
Bentler,
P.
M.
(1990).
Comparative
fit
indexes
in
structural models.
Psy-
chological Bulletin,
107,
238-246.
Bollen,
K. A.
(1989). Structural equations with latent variables.
New
York:
Wiley.
Breckler,
S. J.
(1990). Applications
of
covariance
structure modeling
in
psychology: Cause
for
concern? Psychological Bulletin,
107, 260-273.
Bremner,
J. D.,
Southwick,
S.
M.,
&
Charney,
D. S.
(1995). Etiological
factors
in the
development
of
posttraumatic stress disorder.
In C. M.
Mazure
(Ed.),
Does stress cause psychiatric illness? Progress
in
psy-
chiatry,
No. 46
(pp. 149-185).
Washington,
DC:
American Psychiatric
Press.
Browne,
M.
W.,
&
Cudeck,
R.
(1989).
Single sample cross-validation
indices
for
covariance
structures. Multivariate Behavioral Research,
24,
445-455.
Burke,
A.,
Heuer,
F., &
Reisberg,
D.
(1992). Remembering emotional
events. Memory
&
Cognition,
20,
277-290.
Caspi,
A.,
Elder,
G. H., &
Bern,
D. J.
(1987). Moving against
the
world:
Life-course patterns
of
explosive children. Developmental Psychol-
ogy,
23,
308-313.
Christianson,
S. A.
(1992). Emotional stress
and
eyewitness memory:
A
critical review. Psychological Bulletin,
112, 284-309.
Cohen,
J.,
&
Cohen,
P.
(1983). Applied multiple
regression/correlation
analysis
for the
behavioral sciences
(2nd
ed.).
Hillsdale,
NJ:
Erlbaum.
Cook,
T.
D.,
&
Campbell,
D. T.
(1979).
Quasi-experimentation:
Design
and
analysis issues
for
field settings. Chicago: Rand McNally.
Figley,
C. R.
(1978). Symptoms
of
delayed combat
stress
among
a
college
sample
of
Vietnam veterans. Military Medicine,
143,
107-110.
Fontana,
A.
F.,
&
Rosenheck,
R.
(1994). Posttraumatic stress disorder
among Vietnam
theater
veterans:
A
causal model
of
etiology
in a
community
sample. Journal
of
Nervous
and
Mental Disease,
182,
677-
684.
Fontana,
A.,
Schwartz,
L.
S.,
&
Rosenheck,
R.
(1997) Posttraumatic stress
disorder among female Vietnam veterans:
A
causal model
of
etiology.
American Journal
of
Public Health,
87,
169-175.
Green,
B. L.
(1994).
Psychosocial research
in
traumatic stress:
An
update.
Journal
of
Traumatic Stress,
7,
341-362.
Green,
B.
L.,
Grace,
M.
C.,
Lindy,
J.
D..
Gleser,
G.
C.,
&
Leonard,
A.
(1990).
Risk factors
for
PTSD
and
other diagnoses
in a
general sample
of
Vietnam veterans. American Journal
of
Psychiatry,
147, 729-733.
Hobfoll,
S. E.
(1989).
Conservation
of
resources:
A new
attempt
at
con-
ceptualizing stress. American Psychologist,
44,
513-524.
James,
L.
R.,
Mulaik,
S.
A.,
&
Brett,
J. M.
(1982). Causal analysis:
Assumptions, models,
and
data. Beverly Hills,
CA:
Sage.
Jones,
E. E.
(1996). Introduction
to the
special section
on
attachment
and
psychopathology:
Part
I.
Journal
of
Consulting
and
Clinical Psychol-
ogy,
64,
5-7.
Jordan,
B.
K.,
Schlenger,
W.
E.,
Hough,
R.
L.,
Kulka,
R.
A.,
Weiss,
D.
S.,
Fairbank,
J. A. &
Marmar,
C. R.
(1991). Lifetime
and
current prevalence
of
specific psychiatric disorders among Vietnam veterans
and
controls.
Archives
of
General Psychiatry,
48,
207-215.
Joreskog,
K.
G.,
&
Sorbom,
D.
(1993).
L1SREL
8
user's
reference
guide.
Chicago: Scientific Software.
Keane,
T.
M.,
Caddell,
J.
M.,
&
Taylor,
K. L.
(1988). Mississippi Scale
for
Combat-Related Posttraumatic Stress
Disorder:
Three studies
in
reliabil-
ity
and
validity. Journal
of
Consulting
and
Clinical Psychology,
56,
85-90.
King,
D.
W.,
&
King,
L. A.
(1991).
Validity issues
in
research
on
Vietnam
veteran adjustment. Psychological Bulletin,
109, 107-124.
King,
D.
W.,
King,
L.
A.,
Foy,
D.
W.,
&
Gudanowski,
D. M.
(1996).
Prewar factors
in
combat-related posttraumatic stress disorder: Struc-
tural
equation modeling with
a
national sample
of
female
and
male
Vietnam
veterans. Journal
of
Consulting
and
Clinical Psychology,
64,
520-531.
King,
D.
W.,
King,
L.
A.,
Gudanowski,
D.
M.,
&
Vreven,
D. L.
(1995).
Alternative representations
of
warzone stressors: Relationships
to
post-
traumatic
stress
disorder
in
male
and
female Vietnam veterans. Journal
of
Abnormal Psychology,
104, 184-196.
King,
L.
A.,
&
King,
D. W.
(1994). Latent
structure
of the
Mississippi
Scale
for
Combat-Related Post-Traumatic Stress Disorder: Exploratory
and
higher-order confirmatory factor analyses. Assessment,
1,
275-291.
King,
L.
A.,
King,
D.
W.,
Fairbank,
J.
A.,
Keane,
T.
M.,
&
Adams,
G.
(1998). Resilience-recovery factors
in
posttraumatic stress disorder
among
female
and
male Vietnam veterans: Hardiness, postwar social

170
SHORT REPORTS
support,
and
additional
stressful
life
events. Journal
of
Personality
and
Social
Psychology,
74,
420-434.
Kulka,
R.
A.,
Schlenger,
W. E.,
Fairbank,
J.
A.,
Hough,
R. L.,
Jordan,
B. K.,
Marmar,
C.
R.,
&
Weiss,
D. S.
(1990). Trauma
and the
Vietnam
War
generation: Report
on the findings
from
the
National Vietnam
Veterans
Readjustment
Study.
New
York:
Brunner/Mazel.
Loeber,
R.,
&
Dishion,
T.
(1983). Early predictors
of
male delinquency:
A
review.
Psychological Bulletin,
94, 68
-99.
Main,
M.
(1996). Introduction
to the
special section
on
attachment
and
psychopathology:
2.
Overview
of the field of
attachment. Journal
of
Consulting
and
Clinical Psychology,
64,
237-243.
Marsh,
H. W.
(1994). Confirmatory factor analysis models
of
factorial
invariance:
A
multifaceted approach. Structural Equation Modeling:
A
Multidisciplinary
Journal,
1,
5—34.
McCranie,
E.
W.,
Hyer,
L.
A.,
Boudewyns,
P.
A.,
&
Woods,
M. G.
(1992).
Negative
parenting behavior, combat exposure,
and
PTSD
symptom
severity:
Test
of a
person-event
interaction model. Journal
of
Nervous
and
Mental Disease, 180,
431-438.
McFarlane,
A. C.
(1992). Avoidance
and
intrusion
in
posttraumatic
stress
disorder. Journal
of
Nervous
and
Mental Disease, 180,
439-445.
Metis,
S.,
Sprecher,
S.,
&
Cupach,
W. R.
(1991). Retrospective self reports.
In B. M.
Montgomery
& S.
Duck
(Eds.),
Studying
interpersonal inter-
action.
New
York: Guilford Press.
Nisbett,
R.
E.,
&
Wilson,
T. D.
(1977). Telling more than
we can
know:
Verbal
reports
on
mental processes. Psychological Review,
84,
231-
259.
Robins,
L. N.,
Helzer,
J.
E.,
Croughan,
J.,
&
Ratcliff,
K. S.
(1981). National
Institute
of
Mental Health Diagnostic Interview Schedule:
Its
history,
characteristics,
and
validity. Archives
of
General Psychiatry,
38,
381—
389.
Schlenger,
W.
E.,
Kulka,
R.
A.,
Fairbank,
J.
A.,
Hough,
R.
L.,
Jordan,
B.
K.,
Marmar,
C.
R.,
&
Weiss,
D. S.
(1992).
The
prevalence
of
posttraumatic
stress disorder
in the
Vietnam generation:
A
multimethod,
multisource assessment
of
psychiatric disorder. Journal
of
Traumatic
Stress,
5,
333-363.
Solomon,
Z.,
Waysman,
M.,
&
Mikulincer,
M.
(1990). Family
functioning,
perceived societal support,
and
combat-related psychopathology:
The
moderating
role
of
loneliness. Journal
of
Social
and
Clinical Psychol-
ogy,
2,
241-256.
Southwick,
S.
M.,
Morgan,
C.
A.,
Nicolaou,
A.
L.,
&
Charney,
D. S.
(1997). Consistency
of
memory
for
combat-related traumatic events
in
veterans
of
Operation Desert Storm. American Journal
of
Psychiatry,
154,
173-177.
Steiger,
J. H.
(1990). Structural model evaluation
and
modification:
An
interval
estimation approach. Multivariate Behavioral Research,
25,
173-180.
Weiss,
D.
S.,
Mannar,
C.
R.,
Schlenger,
W.
E.,
Fairbank,
J.
A.,
Jordan,
B.
K.,
&
Kulka,
R. A.
(1992).
The
prevalence
of
lifetime
and
partial
post-traumatic
stress disorder
in
Vietnam theater veterans. Journal
of
Traumatic
Stress,
5,
365-376.
Worthington,
E. R.
(1978). Demographic
and
pre-service variables
as
predictors
of
postmilitary
service adjustment.
In C. R.
Figley
(Ed.),
Stress disorders among Vietnam veterans (pp.
173-187).
New
York:
Brunner/Mazel.
Received September
5,
1997
Revision received February
20,
1998
Accepted April
8,
1998
New
Editors Appointed, 2000-2005
The
Publications
and
Communications
Board
of the
American
Psychological
Associa-
tion
announces
the
appointment
of
three
new
editors
for
6-year
terms
beginning
in
2000.
As
of
January
1,1999,
manuscripts
should
be
directed
as
follows:
•
For
Experimental
and
Clinical
Psychopharmacology,
submit manuscripts
to
Warren
K.
Bickel,
PhD,
Department
of
Psychiatry, University
of
Vermont,
38
Fletcher
Place,
Burlington,
VT
05401-1419.
•
For
the
Journal
of
Counseling
Psychology,
submit
manuscripts
to
Jo-Ida
C.
Hansen,
PhD,
Department
of
Psychology,
University
of
Minnesota,
75
East
River
Road,
Minneapolis,
MN
55455-0344.
•
For
the
Journal
of
Experimental
Psychology:
Human
Perception
and
Perfor-
mance,
submit
manuscripts
to
David
A.
Rosenbaum,
PhD,
Department
of
Psy-
chology,
Pennsylvania
State
University,
642
Moore
Building, University
Park,.
PA
16802-3104.
Manuscript
submission
patterns make
the
precise
date
of
completion
of the
1999 volumes
uncertain.
Current
editors,
Charles
R.
Schuster,
PhD; Clara
E.
Hill,
PhD;
and
Thomas
H.
Carr,
PhD,
respectively,
will
receive
and
consider
manuscripts through
December
31,
1998.
Should
1999
volumes
be
completed
before that
date,
manuscripts
will
be
redirected
to the new
editors
for
consideration
in
2000
volumes.