Download full-text PDF

The Myths of Coping with Loss

Article · Literature Review · July 1989with1,613 Reads
DOI: 10.1037/0022-006X.57.3.349 · Source: PubMed
Abstract
Drawing from theory and clinical lore, we consider how individuals are assumed to cope following irrevocable loss. Several assumptions are reviewed reflecting beliefs concerning the grieving process. Specifically, we examine the expectation that depression is inevitable following loss; that distress is necessary, and failure to experience it is indicative of pathology; that it is necessary to "work through" or process a loss; and that recovery and resolution are to be expected following loss. Although limited research has examined these assumptions systematically, available empirical work fails to support and in some cases contradicts them. Implications of our analysis for theoretical development and research are explored. Finally, we maintain that mistaken assumptions held about the process of coping with loss fail to acknowledge the variability that exists in response to loss, and may lead others to respond to those who have endured loss in ways that are unhelpful.
Journal
of
Consulting
and
Clinical Psychology
1989,
Vol.
57, No. 3,
349-357
Copyright
1989
by the
American Psychological Association, Inc.
0022-006X/89/S00.75
The
Myths
of
Coping
With Loss
Camille
B.
Wortman
The
University
of
Michigan
Roxane Cohen Silver
University
of
Waterloo
Drawing
from
theory
and
clinical lore,
we
consider
how
individuals
are
assumed
to
cope
following
irrevocable
loss.
Several assumptions
are
reviewed
reflecting
beliefs
concerning
the
grieving process.
Specifically,
we
examine
the
expectation that depression
is
inevitable
following
loss;
that
distress
is
necessary,
and
failure
to
experience
it is
indicative
of
pathology; that
it is
necessary
to
"work
through"
or
process
a
loss;
and
that recovery
and
resolution
are to be
expected
following
loss.
Al-
though
limited research
has
examined these assumptions systematically, available empirical
work
fails
to
support
and in
some cases contradicts them. Implications
of our
analysis
for
theoretical
development
and
research
are
explored. Finally,
w e
maintain that mistaken assumptions held about
the
process
of
coping with
loss
fail
to
acknowledge
the
variability
that
exists
in
response
to
loss,
and
may
lead others
to
respond
to
those
who
have
endured loss
in
ways
that
are
unhelpful.
In
this article,
we
focus
on how
people cope with loss events
that
involve permanent change
and
cannot
be
altered
or un-
done.
It is our
belief that such experiences provide
an
excellent
arena
in
which
to
study basic
processes
of
stress
and
coping.
In
the
health
and
medical areas, many
specific
losses might
be
considered irrevocable:
the
permanent
loss
of
bodily
function,
the
loss
of
particular body
parts,
the
loss
of
cognitive capacity,
the
death
of a
loved one,
or
one's
own
terminal illness.
In an
attempt
to
advance theoretical development
in
this rich
and
complex
area, this article updates
an
earlier
review
we
com-
pleted
on
reactions
to
undesirable
life
events
(Silver
&
Wort-
man,
1980). Because
the
most rigorous empirical studies
have
been
in the
areas
of
physical disability
and
bereavement,
we
shall focus
on
these
two
areas
in
this article.
When
a
person experiences
an
irrevocable loss, such
as the
death
of a
loved
one or
permanent paralysis,
how
will
he or she
react?
We
maintain that people hold strong assumptions about
how
others should respond
to
such losses.
As we
have discussed
in
more detail elsewhere (Silver
&
Wortman,
1980;
Wortman
&
Silver,
1987), such assumptions
are
derived
in
part
from
the
theories
of
loss
offered
by
prominent writers
in the
area,
and in
part
from
clinical lore about coping with loss
and our
cultural
understanding
of the
experience.
As
detailed below, individuals
who
encounter
a
loss
are
expected
to go
through
a
period
of
intense distress;
failure
to
experience such distress
is
thought
to
The
order
of
authorship
was
arbitrary.
Research
and
preparation
of
this article
were
supported
by
U.S.
Pub-
lic
Health Service Grant
MCJ-260470
and by
National Institute
on
Aging
Program Project Grant A605561
to
Camille
B.
Wortman
and
Roxane
Cohen
Silver.
For
a
more detailed discussion
of
these issues,
the
reader
is
referred
to
Wortman
and
Silver
(1987).
The
authors
wish
to
thank
two
anonymous reviewers
for
helpful
com-
ments
on an
earlier version
of
this article.
Correspondence concerning this article should
be
addressed
to Ca-
mille
B.
Wortman, Social Environment
and
Health Program, Institute
for
Social Research,
The
University
o f
Michigan,
Ann
Arbor, Michigan
48106-1248.
be
indicative
of a
problem. Moreover,
i t is
assumed that success-
ful
adjustment
to
loss requires that individuals "work
through"
or
deal with their
feelings
of
grief rather than
"denying"
or
"re-
pressing"
them. Within
a
relatively brief
period
of
time, how-
ever,
people
are
expected
to
resolve their loss
and
recover their
earlier
level
of
functioning.
Because
it is
generally assumed that
the
coping process
un-
folds
in a
particular
way,
others
may
evaluate
or
judge those
who
do not
conform
to
these expectations
as
reacting abnormally
or
inappropriately.
For
example, because they believe that people
should
recover relatively soon
after
the
loss,
outsiders might
re-
act
judgmentally
to
continuing signs
of
distress (cf.
Silver
&
Wortman,
1980;
Tail
&
Silver,
1989).
In
fact,
if
laypersons hold
unrealistically
narrow
views
of
what constitutes
a
normal grief
response, they
may
have
difficulty
offering
the
appropriate
forms
of
assistance
to
friends
and
family
members
who are
try-
ing
to
cope with loss. Moreover, because they
too may
hold
as-
sumptions
about
how one
should react when
a
loss
is
experi-
enced,
individuals
who
have
encountered loss
may
harshly eval-
uate
their
own
responses
and may
believe them
to
indicate
underlying
problems
or
pathology (Silver
&
Wortman,
1980).
Because assumptions about
the
grieving process
are
likely
to
have
a
pervasive impact
on how
reactions
to
loss
are
evaluated,
we
feel
it is
important
to
identify
those assumptions that
are
most prevalent
in our
culture
and to
consider systematically
the
available
research data
in
support
of
each one.
We
have
identi-
fied five
assumptions that
we
believe
to be
very
prevalent
in the
grief
literature.
In the
following
sections,
the
validity
of
each
assumption
is
evaluated against
the
available research
data.
While
much
of the
early work
in
this
area
suffered
from
serious
methodological shortcomings (e.g., reliance
on
subjective
im-
pressions
of
unstructured interview
data,
unstandardized mea-
surements,
and
biased samples, etc.;
see
Silver
&
Wortman,
1980,
for a
review),
recent research
in the
bereavement
and
physical
disability areas
has
improved
o n the
deficiencies
of
pre-
vious
literature.
In the
following
discussion,
we
review only
what
we
believe
to be the
best empirical work available
to
test
the
assumptions
we
have
identified.
Except where indicated,
all
of
this
work
has
used standardized outcome measures
and
349
350
CAMILLE
B.
WORTMAN
AND
ROXANE COHEN
SILVER
structured interviews
o f
relatively
large, unbiased samples
and
followed
them over time.
In the
concluding sections
of
this arti-
cle,
we
explore
the
implications
of the
available
data
for
theory,
research,
and
intervention
following
loss,
and
consider
why
such
myths about coping with loss
may
have been perpetuated
despite
the
absence
of
validating
data.
Distress
or
Depression
Is
Inevitable
It
is
widely
assumed
in our
culture that when
a
major loss
is
experienced,
the
normal
way
t o
react
is
with intense
distress
or
depression.
The
most prevalent theories
in the
area
of
grief
and
loss, such
as the
classic psychodynamic models (e.g., Freud,
1917/1957)
and
Bowlby's
(1980)
attachment model,
are
based
on
the
assumption that
at
some point, individuals will confront
the
reality
of
their loss
and go
through
a
period
of
intense dis-
tress
or
depression.
In the
recent authoritative report
on be-
reavement
published
by the
Institute
of
Medicine,
it was
stated
that there
is a
"near-universal occurrence
of
intense emotional
distress
following
bereavement, with
features
similar
in
nature
and
intensity
to
those
of
clinical
depression"
(Osterweis,
Solo-
mon,
&
Green,
1984,
p.
18).
Similarly, depression
has
been
the
foremost
reaction reported
and
discussed
in the
literature
on
spinal cord
injury
(Bracken
&
Shepard, 1980;
Deegan,
1977;
Gunther,
1971;Knorr&Bull,
1970).
As
empirical evidence
has
begun
to
accumulate, however,
i t
is
clear that
the
assumption
of
intense universal distress
follow-
ing
a
major
loss
such
as
bereavement
or
spinal
cord
injury
may
be
unwarranted.
It is
true that
in the
bereavement literature,
some studies have reported
that
feelings
o f
sadness
or
depressed
mood
are
fairly
common.
For
example,
Glick,
Weiss,
and
Parkes
(1974)
have noted that
88% of the
widows they studied
experienced
depressed mood (see also Clayton,
Halikas,
&
Maurice,
1971).
However,
i n
those
investigations
that
have
in-
cluded
a
more systematic
and
rigorous assessment
of
depression
or
distress,
it is
clear that such
a
reaction
is by no
means univer-
sal.
In one
study,
Clayton, Halikas,
and
Maurice
(1972)
inter-
viewed
widows within
30
days
of
losing their spouse. Using
strict diagnostic criteria
to
assess depression, Clayton
et
al.
found
that only
a
minority
o f
respondents
(35%)
could
be
classi-
fied a s
definitely
or
probably depressed. Similarly, Vachon, Rog-
ers,
et al.
(1982)
found
that
1
month
after
the
loss,
30% of the
widows
they studied scored below
5 on the
General Health
Questionnaire
(GHQ)—a
score considered
insufficient
to
war-
rant
further
psychiatric
assessment.
In
their sample
of
primar-
ily
Mormon elderly bereaved individuals, Lund,
Caserta,
and
Dimond(
1986)
reported
that
only 14.6%
o f
the
men
and
19.2%
of
the
women they studied
at 3
weeks postloss evidenced
"at
least
mild"
depression
on the
Zung Depression Scale.
In
fact,
only
12.5%
to 20% o f
this sample reported scores exceeding
the
cutoff
score delineated
as
indicating depression
at any of six
different
assessment points
from
3-4
weeks
t o 2
years postloss.
Examination
of
empirical
data
in the
spinal
cord injury
liter-
ature reveals
a
similar pattern.
For
example, Howell,
Fullerton,
Harvey,
and
Klein
(1981)
conducted
a
careful
assessment
of
22
patients
who had
been injured approximately
1
month,
and
followed
them
for an
average
o f 9
weeks. Each patient
was
inter-
viewed
utilizing
the
Schedule
of
Affective
Disorders
and
Schizo-
phrenia
and
completed
the
Beck Depression Inventory weekly.
Only
a
minority
of
patients
(22.7%)
experienced
a
depressive
disorder
following
injury
that
met
Research Diagnostic Criteria
(see
also
Fullerton,
Harvey,
Klein,
&
Howell,
1981).
Similarly,
Lawson
(1976)
studied
10
spinal-cord-injured patients
5
days
a
week
for the
entire length
of
their rehabilitation stay. Despite
a
multimethod
assessment
of
depression (self-report, profes-
sional ratings,
and
psychoendocrine
and
behavioral measures),
there
was no
clear period
of at
least
a
week
in
which
measures
were consistently
in the
depressive range
for any
patient
(see
also
Malec
&
Neimeyer,
1983).
Thus,
the
few
systematic investi-
gations that
are
available
have
failed
to
demonstrate
the
inevita-
bility
of
depression
following
loss.
Distress
Is
Necessary,
and
Failure
to
Experience Distress
Is
Indicative
of
Pathology
The
clinical literature
is
clear
in
suggesting that those
who
fail
to
respond
to
loss with intense distress
are
reacting abnor-
mally
(e.g.,
Deutsch, 1937; Marris, 1958). Bowlby (1980)
has
identified
"prolonged absence
of
conscious grieving"
(p.
138)
as one of two
types
of
disordered mourning.
In the
previously
mentioned
Institute
of
Medicine
report,
"absent
grief
was
classified
as one of two
forms
of
"pathologic"
mourning (Oster-
weis
et
al.,
1984,
p.
65). This report emphasized that
it is
com-
monly
assumed, particularly
by
clinicians,
"that
the
absence
of
grieving
phenomena
following
bereavement represents some
form
of
personality pathology"
(p.
18). Although
the
authors
noted
that
there
is
little
empirical
evidence
in
support
of
this
assumption, they concluded nonetheless that "professional help
may
be
warranted
for
persons
who
show
no
evidence
of
having
begun
grieving"
(p.
65).
The
assumption that
distress
or
depres-
sion
is a
necessary part
of
the
grieving
process
i s
also quite prev-
alent
in the
literature
on
spinal cord
injury
(e.g.,
Karney,
1976;
Kerr
&
Thompson, 1972;
Nemiah,
1957;
see
Trieschmann,
1978,
1980,
for
reviews).
I n
fact,
authors have maintained that
depression
is
therapeutic because
it
signals that
the
person
is
beginning
to
confront
the
realities
of his or her
situation (e.g.,
Cook,
1976;Dinardo,
1971;
Nemiah, 1957).
The
belief that distress should occur
is s o
powerful
that
it
also
leads
to
negative attributions toward those
who do not
show
evidence
of it. One
such attribution
is
that
the
person
is
denying
the
loss.
As
Siller
(1969)
has
maintained regarding
the
disabled,
occasionally
a
newly disabled person
does
not
seem
to be
particu-
larly
depressed,
and
this should
be a
matter
of
concern.
. ..
A
person should
be
depressed because something significant
has
hap-
pened,
and not to
respond
as
such
is
denial. Such obvious denial
is
rare except
in the
case
of a
retarded person
or in the
very
young,
(p.
292)
A
second attribution
is
that
the
person
is
emotionally
too
weak
to
initiate
the
grieving process. Drawing
from
clinical experi-
ence
with
patients
undergoing psychiatric treatment, Deutsch
(1937)
maintained that grief-related
affect
was
sometimes omit-
ted
among individuals
who
were
not
emotionally strong enough
to
begin grieving.
A
third attribution
is
that individuals
who
fail
to
grieve
are
simply unable
to
become attached
to
others.
For
example,
Raphael
(1983)
suggested that among those
who do
not
show signs
of
grief,
the
preexisting relationship
may
have
been "purely narcissistic with little recognition
of the
real per-
son
who was
lost"
(pp.
205-206).
THE
MYTHS
OF
COPING
351
If,
in
fact,
depression
is
necessary following
loss,
those
people
who
experience
a
period
of
depression should adapt more suc-
cessfully
than those
who do not
become depressed.
However,
this
view
has not
been substantiated empirically.
In
contrast,
several
studies
have
found
that those
who are
most distressed
shortly
following
loss
are
among those
likely
to be
most dis-
tressed
1 to 2
years later.
For
example, Vachon, Rogers,
et
al.
(1982)
found
that among
162
widows,
an
elevated score
1
month
postloss
on the
GHQ,
a
measure
of
distress
and
social
functioning,
was the
most
powerful
predictor
of
high distress
24
months later. Similarly, Lund
et al.
(1985-1986)
found
that
the
best
predictor
of
long-term coping
difficulties
among elderly
widows
and
widowers
was the
presence
of
strong negative emo-
tional
responses
to the
loss (such
as
expressing
a
desire
to die
and
crying) during
the
early bereavement period (see also
Bornstein,
Clayton, Halikas, Maurice,
&
Robins,
1973;
Parkes
&
Weiss, 1983,
for
similar
findings).1
Comparable results have
been obtained
by
investigators studying spinal cord
injury.
In a
cross-sectional study
of 53
male spinal-cord-injured patients,
Dinardo
(1971)
assessed
depressed mood
by
self-report
and
professional
assessments. Results indicated that
the
absence
of
depression
was
associated with higher self-concepts
and
with
staff
ratings
of
successful adjustment
to the
disability, leading
the
author
to
conclude that
"those
individuals
who
react
to
spi-
nal
cord
injury
with depression
are
less
well
adjusted
at any
given
point
in
their rehabilitation than
the
individuals
who do
not
react with
depression"
(p. 52)
(see
Lawson,
1976,
and
Malec
&
Neimeyer, 1983,
for
comparable
findings).
An
important component
of the
view
that depression
is
nec-
essary
is
that
if
individuals
fail
to
experience
distress
shortly
after
the
loss, symptoms
of
distress
will
erupt
at a
later point.
Marris
(1958)
has
commented that "much later,
in
response
to
a
less important
or
trivial
loss,
the
death
of a
more distant rela-
tive,
a
pet—the
bereaved person
is
overwhelmed
by
intense
grief
(p. 27)
(see also
Bowlby,
1980; Rando,
1984).
It is
also
widely
believed that
the
failure
to
grieve
will
result
in
subse-
quent health problems.
The
Institute
of
Medicine report (Oster-
weis
et
al.,
1984) reviewed
the
work
of
several clinicians
who
suggested
that those
who
fail
to
grieve outwardly
may
manifest
their depression through
a
variety
of
physical symptoms
or
so-
matic complaints.
Despite
its
prevalence, available evidence provides little sup-
port
for the
assumption that those
who
fail
to
experience dis-
tress shortly
after
loss
will
have
difficulties
later.
In the
pre-
viously
mentioned study
of
bereavement
by
Clayton
et al.
(see
Bornstein
et
al.,
1973),
interviews were conducted with
109
widows
and
widowers
at 1
month,
4
months,
and
13
months
postloss.
As
noted earlier, only
35% of
these respondents were
classified
as
either
definitely
or
probably depressed
at the 1-
month interview.
However,
only
3 of the
remaining
71
respon-
dents
had
become depressed
by the
4-month interview. More-
over,
only
1
subject evidenced depression
for the first
time
at
13
months
postloss,
leading
the
investigators
to
conclude that
"delayed"
grief
is
relatively
rare.
Similar
findings
were obtained
in a
longitudinal study
of 99
widows conducted
by
Vachon,
Sheldon,
et al.
(1982).
Thirty-two
of
these women were classi-
fied as
"low
distress"
by
virtue
of
their scores
on the GHQ 1
month postloss,
and 94%
continued
to
evidence "low
distress"
when
interviewed
2
years later.
In
fact,
only
2
women
in the
study
who had low
distress
scores
at 1
month
had
high
distress
scores
at the
2-year interview.
A
very
similar pattern
of findings was
obtained
in our
recent
longitudinal
study
of
124
parents
who had
lost
an
infant
to
Sud-
den
Infant
Death Syndrome
(SIDS;
Silver
&
Wortman,
1988;
Wortman
&
Silver,
1987,
in
press). Parents were classified
as
exhibiting
low or
high distress
at the
initial interview
3
weeks
postloss
on the
basis
of
their scores
on the
depression
subscale
of
the
Symptom Check
List-90
(SCL-90;
Derogatis,
1977).
As
in
the
previously described research, most respondents
in
this
study
who
showed
low
distress
at 3
weeks also showed
low
dis-
tress
18
months later. Only
a
small percentage
of the
sample
moved
from
low to
high distress
over
the
period
of
study. Those
parents
who
were evidencing
low
distress shortly
after
their
ba-
by's death were
no
more
likely
than parents
who
reported high
distress
to
indicate that
the
pregnancy
had
been unplanned
or
difficult
to
accept;
nor did
they
differ
in
their evaluation
of
their
babies
as
having
been
beautiful,
intelligent,
and
happy while
alive.
Thus, these data
fail
to
support
the
notion that absence
of
distress
may be due to
insufficient
attachment
to the
lost
loved
one.2
In
summary,
the
bulk
of
research
to
date provides little sup-
port
for the
widely
held
view
that those
who
fail
to
exhibit early
distress
will
show
subsequent
difficulties.
The
data clearly sug-
gest
that
"absent
grief"
is not
necessarily problematic, and,
at
least
as it is
assessed
in the
studies conducted
to
date, "delayed
grief"
is far
less common than clinical lore would suggest.
The
Importance
of
"Working
Through"
the
Loss
It
is
widely
assumed that
a
period
of
depression
will
occur
once
the
person
confronts
the
reality
of his or her
loss. Then,
it
is
commonly believed,
the
person must "work through"
or
pro-
cess what
has
happened
in
order
to
recover successfully (see,
e.g., Brown
&
Stoudemire, 1983; Doyle, 1980). Implicit
in
this
assumption
is the
notion that individuals need
to
focus
on and
"process"
what
has
happened
and
that
attempts
to
deny
the
implications
of the
loss,
or
block
feelings
or
thoughts about
it,
will
ultimately
be
unproductive (cf. Bowlby, 1980; Parkes
&
Weiss,
1983).
Marris
(1958)
maintained that
"if the
bereaved
cannot
work through this process
of
grieving they
may
suffer
' Of
course,
this
does
not
necessarily
imply
that
it is
detrimental
t o
go
through
a
period
of
depression
following
loss. Undoubtedly, some
o f
the
people
who are
most
depressed
after
the
loss
may be
those
who
have
enduring
psychological
difficulties
that
were
present prior
to the
loss.
In
the
research that
has
been
conducted
t o
date,
it is not
possible
to
discriminate
between
those
respondents
who
became depressed
primar-
ily
as
a
result
of
their
loss
and
those
who had a
lifelong
history
of
psycho-
logical
disturbance.
Only
prospective
designs,
which
assess
such
con-
founding
variables
prior
to the
loss,
can
help
to
shed
light
on
this issue.
Because
most
loss
events
are
relatively
infrequent,
prospective
studies
are
prohibitively
costly
and
thus
have
rarely
been
attempted.
2
The
only
evidence
for
delayed
grief
that
we
have
been
able
to un-
cover
comes
from
a
study
by
Parkes
and
Weiss
(1983).
These
investiga-
tors
found
that
widows
who
reported
having
had
marriages character-
ized
by
high
conflict
displayed
little
or no
emotional distress
in the
weeks
following
their
loss.
However,
at the
13-month
and 2- to
4-year
follow-up
interviews,
these
widows
were
evidencing
greater
difficulties
than
those
women
whose
marriages
were
characterized
b y low
conflict.
352
CAMILLE
B.
WOKTMAN
AND
ROXANE COHEN SILVER
lasting
emotional damage"
(p.
29).
Rando
(1984) concurred
with
this assessment, stating that "for
the
griever
who has not
attended
to his
grief,
the
pain
is a s
acute
and
fresh
ten
years later
as it was the day
after"
(p.
114).
On
the
basis
of
statements such
as
these,
it
might
be
expected
that those
who
show evidence
of
"working through" their loss
in
the
weeks
o r
months
following
it
will
be
more
successful
in
resolving
the
loss than those
who do
not.
However,
the
limited
evidence
that
is
available suggests
that
this
may not be the
case.
Although
they
did not
assess "working
through"
the
loss spe-
cifically,
Parkes
and
Weiss
(1983) conducted
a
comprehensive
study
of
bereavement
that
provided
data
of
possible
relevance
to
this concept. Respondents were rated
by
coders
on the
degree
to
which
they evidenced yearning
or
pining
for the
deceased
3
weeks
after
their loss. Subjects were divided into
a
"High
Yearn-
ing" group, composed
of
those respondents
who
appeared
to
yearn
or
pine constantly,
frequently,
or
whenever inactive,
and
a
"Low Yearning" group,
who
yearned
never,
seldom,
or
only
when
reminded
of the
loss.
In
fact,
high initial yearning
was
found
to be
predictive
of
poor mental
and
physical health out-
comes
at 13
months postloss.
As
Parkes
and
Weiss
(1983)
ex-
pressed
it, "We
might suppose that people
who
avoid
or
repress
grief
are the
most
likely
to
become disturbed
a
year
later,
yet
this
is not the
case"
(p.
47). Interestingly, high initial yearning
was
associated with poor outcome
even
at the final
interview
conducted
2 to 4
years
after
the
loss.
In
our
previously described study
of
parents
who
suffered
a
SIDS
loss
(Silver
&
Wortman,
1988;
Wortman
&
Silver,
1987),
we
also examined
the
impact
of
early evidence
of
"working
through"
the
loss
on
subsequent adjustment. "Working
through"
was
operationalized
as
active attempts
by the
parent
to
make sense
of and
process
the
death, including searching
for