Gender Differences in Adjustment To Bereavement: An Empirical and Theoretical Review

Review of General Psychology Copyright 2001 by the Educational Publishing Foundation
2001, Vol. 5, No. 1, 62-83 1089-2680/01/$5.00 DOI: 10.1037//1089-2680.5.1.62
Gender Differences in Adjustment to Bereavement:

An Empirical and Theoretical Review
Margaret Stroebe, Wolfgang Stroebe, and Henk Schut
Utrecht University
The loss of a marital partner results in substantial increases in morbidity and mortality
among both men and women, but the effects are relatively greater for widowers than
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
for widows in the acute grieving period. Evidence is reviewed, and explanations of the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
pattern are examined. An interpretation in terms of gender differences in social support

(cf. M. Stroebe & W. Stroebe, 1983), although plausible, has not yet been empirically
confirmed. Likewise, with respect to gender differences in coping styles, women are
more confrontive and expressive of their emotions than men, but there has been little
validation of the generally accepted grief work hypothesis (working through grief by
women brings about their better recovery). Cognitive processes underlying effective
coping with bereavement are analyzed, and a stressor-specific framework, the dual-
process model of coping with loss, is suggested to help explain gender differences in
health outcomes.
One of the most intriguing phenomena to ences in the levels of social support received by
emerge in the short history of the scientific widowers and widows. Thus, in this article, we
study of bereavement is the existence of pat- develop a new explanation based on gender
terns of gender differences in reactions. An differences in coping with grief.
early review of the literature (M. Stroebe & Our focus is on partner loss, because this is
Stroebe, 1983) suggested that men suffer rela- the type of bereavement for which there is the
tively greater health consequences than women most empirical research, and on early bereave-
and that this difference is due to higher levels of ment, because, over time, levels of distress and
social support received by widows than widow- debility generally decline to become equivalent
ers. There are good reasons to reassess these to those for nonbereaved samples (M. Stroebe,
claims. A substantial body of new evidence has Stroebe, & Hansson, 1993). As in our earlier
become available with regard to both the pattern review (M. Stroebe & Stroebe, 1983), we have
of gender differences and the psychological conceived of gender differences in terms of
mechanism assumed to account for these differ- relative risk. The relative risk of a disease is
ences. We argue here that the new empirical denned in epidemiology as the ratio of the
evidence is generally consistent with our claim chance of disease in individuals exposed to a
that men suffer relatively greater health conse- risk factor to the risk of disease in individuals
quences from bereavement than women but not without exposure (Jeffery, 1989). Thus, by
with the explanation in terms of gender differ- comparing the disease risk of bereaved men or
women with that of a matched nonbereaved
control group, we can assess the relative risk
attributable to bereavement for a given health
Margaret Stroebe, Wolfgang Stroebe, and Henk Schut,
Research Institute for Psychology & Health, Utrecht Uni- risk among widowers and widows. Our analysis
versity, Utrecht, the Netherlands. of gender differences is based on comparison of
We wish to thank Susan Folkman and Vicki Helgeson for the relative magnitude of widowed-married ra-
their valuable comments on an earlier version of this article tios (i.e., relative risks for widowers and wid-
and Maaike Terheggen for assistance in the preparation of
the article. ows) across different health consequences.
Correspondence concerning this article should be ad- These patterns provide important clues for the
dressed to Margaret Stroebe, Research Institute for Psy- analysis of theoretical mechanisms.
chology & Health, Department of Psychology, Utrecht Uni-
versity, P.O. Box 80.140, 3508 TC Utrecht, the Nether- It is worth noting that risk can also be as-
lands. Electronic mail may be sent to M.Stroebe® sessed from a population perspective. Popula-
fss.uu.nl. tion-attributable risk is the number of excess
62
GENDER DIFFERENCES IN BEREAVEMENT 63
cases of a disease in a population that can be Distress and Depressive Symptomatology

attributed to a particular risk factor. The number
of excess cases in a given population attribut- Most bereaved people feel distressed and de-
able to a given risk factor is only partly depen- pressed after the loss of a loved person (Parkes,
dent on relative risk. Another important deter- 1996; W. Stroebe & Stroebe, 1987). Although
minant is the frequency with which a risk factor these reactions may, in some cases, be severe
or combination of risk factors occurs in a given enough to require psychiatric or medical treat-
population. The widow to widower ratio fre- ment, they are generally considered to be nor-
quently reaches 4:1 (U.S. Department of Health mal reactions to the death of a loved person,
& Human Services, 1985), because women live ones that will abate over the course of time,
usually without intervention (cf. Clayton,

longer than men, tend to marry men older than
1990). Thus, it is important to consider studies

they are themselves, and remarry after bereave-
of distress and depression separately from the
ment less frequently than men; thus, there are general category of mental illness. Most be-
likely to be more excess cases among widows reaved people who are depressed do not seek
than among widowers. As a result, the popula- treatment and would not be classified as men-
tion-attributable risk is higher for widows than tally ill.
for widowers. Thus, the fact that clinicians are
Measures used to assess such effects vary
likely to encounter a larger proportion of de- from established, validated depression scales,
pressed widows than widowers because of the such as the Beck Depression Inventory (Beck,
higher proportion of widows in the population 1967) or the Center for Epidemiologic Studies
(population-attributable risk) is not at all incon- Depression Scale (CES-D; Radloff, 1977), to
sistent with the fact that widowers have a higher measures of distress constructed specifically for
relative risk of depression after bereavement a particular study. Likewise, studies vary from
than widows. small cohort investigations conducted longitu-
dinally to cross-sectional surveys of depression
rates. Because the aim here is to identify general
Gender Differences in Health patterns of gender differences by widowed ver-
Consequences of Bereavement: sus married mental health status, this broad
range of study types is covered in this section.
The Empirical Evidence
A further observation is necessary before we
There is a fairly sound body of evidence on review the studies. Reactions of distress and
the detrimental health consequences of bereave- depression are not only grief specific, showing
ment for widowed persons in general. When endorsement of items on grief symptom scales
comparisons are made with still-married indi- such as those focusing on upset at the loss of
viduals, as suggested earlier, detrimental effects one's spouse, distress about being left alone, or
to physical as well as mental health for both loneliness in the absence of one's partner (e.g.,
widowed men and women are evident. Rates of the Grief Experience Inventory; see Sanders,
Mauger, & Strong, 1991), but also generalized,
distress and depression, use of medication,
showing elevated rates among bereaved indi-
physical illness health measures (e.g., number
viduals on depression scales (e.g., the Beck
of days sick and visits to doctors), and even Depression Inventory; see Beck, 1967) that do
mortality are generally reported to be higher for not make any specific reference to grief.
those who are widowed than for comparable Clearly, no comparison of bereaved and married
still-married individuals (Carter & Glick, 1976; individuals' responses is possible when grief-
Gove, 1972a, 1972b, 1973; Osterweis, Solomon, specific scales are used, whereas such compar-
& Green, 1984; Parkes, 1996; W. Stroebe & isons are possible on general scales of depres-
Stroebe, 1987; W. Stroebe, Stroebe, & Schut, sion. Can we draw conclusions about relative
1993). The gender difference emerges when one upset among widows and widowers from re-
examines the relative magnitude of the wid- sponses on specific scales? This is indeed prob-
owed to married ratios, and this too pertains lematic, because women in general and wid-
across a broad spectrum of mental and physical owed women in particular not only have higher
health consequences. depression levels but are also more acknowl-
64 STROEBE, STROEBE, AND SCHUT
edging and expressive of their emotions than tact with friends and church membership) into
men (Derlega, Metts, Petronio, & Margulis, the regression equation.
1993; Notarious & Johnson, 1982; Pennebaker Similar gender differences were also found in
& Roberts, 1992; Shields, 1991). With these a large-scale, methodologically sophisticated
assessment problems in mind, we turn to the study conducted by Siegel and Kuykendall
empirical studies. (1990), even though this study focused on the
Support for the hypothesis that widowers are impact of the recent death of a close (non-
relatively more distressed than widows by the spouse) family member on depressed mood.
loss of their spouse was provided in two early Respondents were more than 800 elderly men
studies, each of which has become a classic and women of whom 14% had recently suffered
contribution although involving very different the loss of a close family member. When loss
types of design. In a community survey of de- status was entered into a regression analysis,
pression involving nearly 2,500 White persons, higher levels of depressive symptomatology
Radloff (1975) found a Gender X Marital Status were found among men but not among women.
interaction on level of depression. Although Thus, losing a close family member within the
previous 6 months resulted in a greater increase
married women were more depressed than mar-
in depressed mood among men than women.
ried men, widowed men were more depressed
Important here is a further pattern that emerged
than widowed women. Thus, despite the higher
when the researchers assessed whether marital
depression rates of women, men became rela-
status would moderate the impact of the recent
tively more depressed on bereavement. In the death of a (nonspouse) family member. This
second study, a small, in-depth investigation of analysis resulted in a significant Marital Sta-
a sample of younger bereaved individuals (49 tus X Loss interaction for men but not for
widows and 19 widowers), Glide, Weiss, and women; it was the widowed men who, in com-
Parkes (1974) found, 2 to 4 years after bereave- parison with their married counterparts, reacted
ment, that widowers had taken longer to recover with elevated depression levels to the loss of a
than widows. Widows had higher depression close family member. Siegel and Kuykendall
scores than married women 1 year after be- (1990) thus identified elderly widowers as a
reavement; at the later follow-up, however, they particularly vulnerable group, one that is at
were no more depressed than the married greater relative risk of depression (and, they
women, whereas widowers remained signifi- argued, consequent physical health detriments)
cantly more depressed than married men. than their female counterparts.
More recent studies have confirmed this pat- The greater relative risk of depression for
tern, although it must be noted that surprisingly widowers has also been demonstrated outside
few studies have compared depression levels by the United States in studies conducted in Great
gender for widowed and married individuals. A Britain (Cramer, 1993) and the Netherlands
stringent test of the gender difference was pro- (Nieboer, Lindenberg, & Ormel, 1998; Van
vided by Umberson, Wortman, and Kessler Grootheest, Beekman, Broesse van Groenou, &
(1992), who used the results of a large national Deeg, 1999). Using data from a nationally rep-
survey to examine long-term differences in vul- resentative cross-sectional survey of 9,003 Brit-
nerability to depression among those who were ish adults and controlling for age, education,
conjugally bereaved. These authors found that and income, Cramer (1993) found that widow-
in comparisons with same-gender married indi- ers had significantly higher rates of psycholog-
viduals, widowers were relatively more affected ical distress (as measured by the General Health
by loss than widows. The same pattern was Questionnaire; Goldberg & Hillier, 1979) rela-
reported in a telephone survey of 746 elderly tive to married men than did widows relative to
men (222 married and 38 widowed) and women married women. In fact, the rates of widows did
(253 married and 233 widowed) conducted in not show a significant elevation. Similarly, Van
two U.S. states. Widowhood had "a substan- Grootheest et al. (1999), in a large-scale com-
tially stronger effect on depression for men than munity study of 2,626 widowed and married
women" (Lee, Willetts, & Seccombe, 1998, p. men and women 55 to 85 years of age con-
622), and the interaction was not reduced by the ducted in the Netherlands, found widowhood to
introduction of measures of social support (con- be associated with higher levels of depressive
symptomatology (as measured by the CES-D). was based on a direct comparison of the scores
This association was stronger for men than for on an eight-item measure of "adjustment-
women. Interestingly, the effect was mediated depression" of 78 widows and 41 widowers.
by different types of environmental stressors for This finding, therefore, might merely be a re-
widows (e.g., network size) and widowers (e.g., flection of the higher depression rate among
emotional support received). Furthermore, wid- women in the population. As such, the interpre-
ows appeared to adapt to widowhood more suc- tation may in fact be correct but should not be
cessfully over time. Finally, Nieboer et al. interpreted in terms of more extreme reactions
(1998), who analyzed the consequences of wid- of women than men to the loss of a partner.
ow(er)hood in regard to depressive symptom- Likewise, Gilbar and Dagan (1995) concluded,
atology in a sample of 1,252 widowed respon- from their study of 43 widows and 24 widow-
dents in the Netherlands, also reported rela- ers, that widows suffered more than widowers.
tively higher depression scores for widowers They found no statistically significant differ-
than for widows. The data were cross-sectional ences between widowers' and widows' scores
and were analyzed according to a cutoff point on a depression measure, but they found that
indicating recency of bereavement (less than vs. widowers reported fewer difficulties in coping
more than 2 years). The gender difference in with their loss than did widows, as measured
bereavement outcome was significant for those with the Texas Revised Inventory of Grief
recently bereaved but not for those who had lost (Faschingbauer, Zisook, & De Vaul, 1987).
their partners more than 2 years earlier. This is There were no nonbereaved controls. The con-
in contrast to the results of Van Grootheest et al. clusion drawn by Gilbar and Dagan (1995) that
(1999), Glick et al. (1974), and recent compar- men cope better may be more in line with the
isons made by Bierhals et al. (1996; see also observed general tendency of men to acknowl-
Chen et al., 1999) focusing on the temporal edge or report less depressive symptomatology
course of grief symptomatology among widow- (cf. Meshot & Leitner, 1993) than an indication
ers and widows (with no control groups). The that they are having a less difficult time with
reason for this discrepancy is unclear but is not their bereavement. Control group comparisons
of central concern, because our focus is on the would have clarified this issue.
acute grieving period.
Other studies have claimed similar levels of
A few authors have argued that widows suf- distress for bereaved men and women (Born-
fer greater depression on bereavement than wid- stein, Clayton, Halikas, Maurice, & Robbins,
owers, concluding that female elevations on de- 1973; Clayton, Halikas, & Maurice, 1972; Gal-
pression during bereavement exceed the corre- lagher, Breckenridge, Thompson, & Peterson,
sponding male elevations (Carey, 1977; Gilbar 1983; Lund, Caserta, & Dimond, 1986; Weis-
& Dagan, 1995; Jacobs, Kasl, Ostfeld, Berk- man & Klerman, 1977). Lund et al. (1986)
man, & Charpentier, 1986). On examination, it made gender comparisons among widowed in-
becomes evident that these studies failed to dividuals but failed to compare rates with those
include nonbereaved control groups. Thus, they for nonbereaved men and women. These au-
probably confounded the general gender differ- thors concluded that there were no substantial
ence in depression (i.e., absolute risk)—namely, differences with respect to depression between
that women have higher depression rates than widowed men and women in their study. Al-
men—with the bereavement-specific effect (i.e., though literally this is a correct inference from
relative risk). As argued earlier, the general the data, the suggestion it raises is that there are
gender difference in the base rate for depression no differences in reactions to loss. Therefore, a
must be controlled to reveal the gender differ- criticism similar to that raised earlier applies:
ence that can be attributed to the impact of Because women in general display more dis-
bereavement. Failure to do so results in a con- tress and symptoms of depression, a finding of
founding of the absolute risk in depression with no difference could indeed indicate relative in-
the bereavement-specific or relative risk. For creases for widowers. The early Clayton studies
example, Carey (1977, 1979) has frequently suffered from the same methodological short-
been cited in the literature as providing evi- coming (see M. Stroebe & Stroebe, 1983).
dence that widows are at higher risk of depres- A number of interview studies that included
sion than widowers. However, this conclusion married control groups but still failed to reveal
a Gender X Marital Status interaction may have depressed widows may have been selected out
been subject to selection bias. For example, of the study. Mortality patterns, discussed later,
there is some indication that the failure of our support this hypothesis (see Gallagher-Thomp-
own study to show a Gender X Marital status son et al., 1993).
interaction may have been due to selection bias Hays, Kasl, and Jacobs (1994) reported on
among interview participants (W. Stroebe & the course of distress among spouses (N = 440)
Stroebe, 1993; W. Stroebe, Stroebe, & Domitt- of patients hospitalized for serious illness or
ner, 1988). Because our study had a relatively surgery, some of whom died during the 2-year
low response rate, we asked individuals who follow-up period. Few gender differences were
refused to be interviewed whether they were observed, but, at 25 months, the widows who
willing to complete a mailed questionnaire, and had remained in the study were more depressed
a reasonable proportion of respondents agreed. than widowers. This study involved extensive
This mailed questionnaire contained a measure interviewing and may thus be subject to the
of depression. Findings indicated that the wid- same Gender X Attrition interaction effect
owers who refused interviews (but filled in the noted earlier (selection out of more distressed
postal questionnaire) were significantly more men). There were, in fact, high nonresponse and
depressed than those who accepted, whereas the dropout rates, with an estimated 20% of those
opposite pattern prevailed for widows (M. initially approached remaining in the sample at
Stroebe & Stroebe, 1989). We accounted for the 2-year follow-up (of whom an even smaller
this finding in terms of social norms that make percentage were widowers). Furthermore, al-
it embarrassing for depressed men but not though rates were compared with rates for non-
women to show strong emotions (e.g., cry) in an bereaved groups and with distress rates while
interview setting. This "dropout" effect, in the spouses were still alive, it must be noted
which less depressed widowers and more de- that, at all times, all of the study participants
pressed widows were selected into the inter- were facing a life-threatening situation; thus, it
views, would account for the fact that inter- would be expected that the distress "baseline"
viewed widowers did not show the excesses in would be high. However, it is indeed possible
comparison with widows observed in the stud- that the widow excess that emerged only after 2
ies reviewed earlier. years indicates a longer term reversal; that is,
One other study, an investigation of elderly widows in this sample were relatively more
widows and widowers conducted by Gallagher poorly adjusted than widowers in the longer
et al. (1983), compared depression levels of 212 term (which, as noted earlier, is not the central
recently widowed individuals (99 men and 113 interest of this article).
women) with those of married, same-gender Finally, on the basis of data derived from a
controls (84 men and 78 women) and revealed large-scale, nationally representative longitudi-
no Gender X Marital Status interaction. This nal study of 13,008 noninstitutionalized adults
study, then, also observed the methodological in the United States, Marks and Lambert (1998)
requirements with respect to controls and still found that becoming widowed between the two
failed to support the pattern of relatively higher points of measurement was associated with
levels of depression among widowers reported marginally greater increases in symptoms of
in the other methodologically sound studies. Is depression among women than men (p < .10).
this strong evidence against the hypothesis of The problem with this finding, which runs
relatively greater excesses for widowers? Al- against the results of all of the other method-
though this could be the case, there are also ologically strong studies described earlier, is
reasons to argue otherwise. The study had a that it was based on only 10 widowers (as
high rejection rate (only 30% responded to the compared with 82 widows).
first mailing; for more information, see Gal- In conclusion, there is no large and unequiv-
lagher-Thompson, Futterman, Farberow, Thomp- ocal body of research confirming the pattern of
son, & Peterson, 1993), so selection biases must relative widower excesses in levels and rates of
be considered. As was the M. Stroebe and depression observed nearly two decades ago
Stroebe (1989) investigation, it was an inter- (M. Stroebe & Stroebe, 1983). Nevertheless, the
view study, and the results would suggest that large-scale, methodologically more sophisti-
here, too, more depressed widowers and less cated studies of Umberson et al. (1992), Siegel
and Kuykendall (1990), and Cramer (1993) pro- found across all marital status categories if data
vide substantial evidence in favor of widower are analyzed in regard to untreated mental ill-
excesses and support the conclusions from ear- nesses. Feinson (1986) concluded, on the basis
lier studies conducted by Glick et al. (1974) and of a study involving a cross-sectional design,
Radloff (1975). Studies making claims to the that there were no gender differences in the
contrary have typically failed to take into ac- prevalence of "mental disorders" (assessed with
count such factors as baseline rates (relatively the Symptom Checklist 90; Derogatis, 1977)
more widows and higher depression rates for among a small sample (40 widowers and 119
women) and the potential selection factor. Fur- widows) of bereaved elderly people. However,
ther research is needed not only to confirm these it was unclear whether Gender X Marital Status
conclusions but to establish subgroup patterns. ratios were computed and, given the small sam-
In particular, patterns for different age groups ple, unlikely that they would be found. In con-
are still unclear, for example, whether older trast, Gove's findings have been confirmed in
widowers are at especially high risk, as could be more recent, methodologically sounder surveys
inferred from the Gallagher et al. (1983) results. (e.g., Bebbington, 1987; Gove, Hughes, &
Also needed is further exploration of patterns Style, 1983). Bebbington (1987) presented data
across the duration of bereavement. Do the det- from English national statistics for first admis-
rimental consequences for widowers really per- sions with affective disorders for the years
sist longer than the consequences for widows, 1982-1985. Although widowed men had lower
as suggested by some but not all studies? absolute admission rates than widows, they
were at higher relative risk. Bebbington (1987)
Mental Illness concluded that, in terms of risk of affective
disorder, being widowed makes more difference
It is evident that "mental illness," as a cate- to a man than it does to a woman.
gory, combines many diverse and unrelated dis- Analyses have recently been extended to
orders, and to treat all phenomena within a comparisons between Black and White wid-
common framework has long been recognized owed men and women, revealing intriguing
as problematic (cf. Gove, 1980). However, be- variations (study of the health consequences of
cause the interest here is in differences between bereavement among Black persons having been
widowed and married individuals in the rates of generally neglected in the past). It seems that,
psychiatric problems severe enough to reach although the pattern of relative male excess in
criteria for professional treatment, it seems jus- mental illness during bereavement is still a ro-
tifiable to group together diverse psychiatric bust phenomenon for Whites, a different pattern
illnesses, disregarding for the moment differ- may pertain for Black persons. Williams,
ences in etiology or symptomatology. The em- Takeuchi, and Adair (1992) conducted a cross-
phasis here, unlike the previous section, is on sectional survey of psychiatric disorders among
diagnosed categories of mental disorders. both Whites and Blacks by marital status
There is long-standing evidence that mental (N = 18,571). They found that widowed Black
illness incidence rates are consistent with find- men and women, like their White counterparts,
ings on various measures of distress and depres- had higher rates of disorder than married indi-
sion. Rates of mental illness are typically higher viduals. In this survey, excesses for White wid-
for widowed than married individuals and owed women (vs. married women) did not
higher for women than men (Bebbington, 1987; reach significance when adjustments were made
Bloom, Asher, & White, 1978; Fox, 1984; for socioeconomic status and household size,
Parkes, 1964; Regier et al., 1988). In addition, whereas excesses did reach significance for
the difference in mental health status between widowed Black women. Gender ratios were cal-
married and widowed individuals is greater for culated, that is, the rate of widowed to married
men than for women. A classic review of this men divided by the rate of widowed to married
literature was conducted by Gove (1972b), who women. Opposite patterns of relative risk were
identified the pattern just described very consis- found for Blacks and Whites with respect to
tently across a variety of studies. On the basis of these comparisons. Patterns for White widowers
three early data sets used by Gove (1972b), Fox consistently indicated that widowhood was
(1980) argued that female excesses are to be worse for them than for widows; among Blacks,
widowhood was worse for widows than for In conclusion, information on mental illness
widowers (apart from anxiety disorders). derives from cross-sectional surveys and is thus
How can this discrepancy in the gender dif- subject to methodological shortcomings associ-
ference pattern for Blacks versus Whites be ated with this technique. In particular, more
explained? Williams et al. (1992) suggested an information is needed on patterns among the
explanation in terms of differences in the social recently bereaved; in this case, large-scale sur-
and psychological contexts of marital dissolu- veys would be necessary, because the preva-
tion for Whites and Blacks. Whereas widowed lence of mental disorders is low (e.g., as com-
White men experience great social isolation and pared with undiagnosed depression rates). Re-
loneliness, widowed Black men are likely to cent evidence confirms the conclusion of early
receive high levels of support from family and research that men may suffer relatively more
friendship networks. Thus, although this study from mental illness during widowhood than
women. The findings may not extend to other
provides confirmation of the relative excess that
cultural groups with different marital and social
we identified previously (M. Stroebe & Stroebe,
support patterns. Although some diagnosis-spe-
1983), it indicates that the pattern may indeed
cific information is available, more studies to
be culturally specific. Intriguing—and impor-
establish specific diagnoses in recently be-
tant—though this is potentially, it is advisable reaved widows and widowers would be useful.
to wait for further empirical evidence before
drawing general conclusions.
Some recent studies, including that by Wil- Physical Symptoms and Illnesses
liams et al. (1992) just described, have also
Physical health detriments are frequent
provided diagnosis-specific information in the
among recently bereaved persons, affecting
mental illness area by gender and marital status.
both men and women (Parkes, 1996; W. Stroebe
There are indications from a number of studies
& Stroebe, 1987). They not only suffer from a
that male excesses during bereavement may fall
variety of physical symptoms and illnesses but
particularly excessively within the diagnostic have higher rates of disability than married in-
category of alcoholic disorders, whereas be- dividuals and exhibit increased use of medical
reaved women's mental illness patterns may services, such as consultations with doctors,
rather be classified within depressive disorders consumption of prescribed medicines, and hos-
(see W. Stroebe & Stroebe, 1987). This would pitalizations (Joung, van der Meer, & Macken-
seem to be in accordance with gender differ- bach, 1995; M. Morgan, 1980; Verbrugge,
ences in ways of coping (cf. de Ridder, 2000). 1979). In addition to this pattern, a gender dif-
Women express their emotionality, vent their ference has been reported: Women suffer phys-
distress, confide in others, and use formal re- ical health problems and illnesses and use health
sources (including psychotherapy), whereas services more than men (Verbrugge, 1989; Ver-
men remain silent and keep feelings of distress brugge & Wingard, 1987). A number of pro-
and anxiety to themselves (though, neverthe- cesses have been suggested as explanations of
less, as shown earlier, widowers do have rela- this gender difference in morbidity, including
tively higher self-reported depression rates than the possibility that men are less willing to ac-
widows). We discuss these issues in greater knowledge or report symptoms than are women
detail subsequently. It has in fact been shown (see, e.g., Verbrugge, 1989; Wingard, 1984).
that widowers have a higher alcohol intake than This underlines the need, as described earlier,
married men, but this is not the case for widows for within-gender comparisons of bereaved and
versus married women (Cramer, 1993; Wil- nonbereaved individuals.
liams et al., 1992). In this context, it is also Glick et al. (1974), in their classic study of
interesting to note that emotional distress and Boston widows and widowers, found higher
alcohol use have recently been viewed as gen- scores on a specially constructed physical health
der-linked responses to the same stressful expe- questionnaire for widows than for widowers 14
rience, with men increasing alcohol intake but months after bereavement, but only for the latter
reporting less emotional distress during severe was there a significant difference in comparison
stress and women experiencing the opposite with married controls. In other words, although
effects (Horwitz & Davies, 1994). absolute levels of symptomatology were higher
among widows, relative excesses were greater In conclusion, the available evidence on
for widowers. This pattern was similar to that of physical health status of widows and widowers
an early (less well-designed) study by Gerber tends to support the patterns found for mental
and colleagues (Gerber, Rusalem, Hannon, Bat- health reviewed earlier. The evidence so far
tin, & Arkin, 1975; Gerber, Wiener, Battin, & shows widowers to be relatively more vulnera-
Arkin, 1975). However, Thompson, Brecken- ble than widows (i.e., as compared with their
ridge, Gallagher, and Peterson (1984), in the married counterparts) on physical health in-
same study that failed to reveal differences be- dexes ranging from self-reported health to
tween the genders on depression (discussed ear- chronic conditions, consultation rates, and dis-
lier), also found no Marital Status X Gender ability. Surprisingly few studies are available.
interactions on indexes of physical health Methodologically sound investigations of the

among their sample of older widowed persons. course of health among widowers and widows
Possibly the same reasons noted before to ex- across the duration of their bereavement are
plain the negative results with respect to depres- urgently needed. Next, we turn to the body of
sion apply with respect to physical health as evidence on the most extreme outcome of be-
well. Also, Gallagher-Thompson et al. (1993), reavement, namely mortality. With respect to
discussing this study, suggested the explana- this outcome, a more substantial literature has
tion that men are likely to underestimate their accumulated.
health problems during the early months of
bereavement, because, indeed, they found gen-
der differences in mortality (see subsequent
Mortality
discussion). Mortality patterns found in cross-sectional
Joung et al. (1997; see also Joung, 1996) studies are quite consistent (Hu & Goldman,
examined marital status differences in self-re- 1990; M. Stroebe, Stroebe, Gergen, & Gergen,
ported health in a prospective investigation 1981), and, with few exceptions, large-scale
of 3,510 men and women in the Netherlands. longitudinal studies have confirmed the pattern
The usual pattern emerged, with widowed per- of relative widower excesses (Bowling & Wind-
sons generally reporting poorer perceived sor, 1995; Helsing & Szklo, 1981; Lillard &
health than married persons, although chronic Waite, 1995; Mellstrom, Nilsson, Oden, Rund-
conditions were not reported to be excessive. gren, & Svanborg, 1982; for a more detailed
With respect to relative excesses, differences review, see M. Stroebe & Stroebe, 1993). The
were apparent in perceived general health (wid- overall picture is that mortality rates for wid-
ower-married ratio: 2.11; widow-married ratio: owers (vs. married men) are relatively higher
1.05) and subjective health complaints (widower- than those for widows (vs. married women). As
married ratio: 1.47; widow-married ratio: described subsequently (see also M. Stroebe et
0.83). This pattern suggests a relatively higher al., 1981; M. Stroebe & Stroebe, 1993), there
excess on self-reported health indexes for wid- are discrepancies in a few studies for some
owers than for widows. Similar results were subgroups (e.g., certain age groups or duration
reported in the study by Cramer (1993) re- periods), and occasionally widows' rates have
viewed earlier. Widowed and married persons' not been found significantly excessive in com-
consultation rates for physical symptoms were parison with those of married women. How-
assessed. There was a relatively greater excess ever, in general, the mortality patterns provide
for widowers than for widows. the strongest evidence of all health indexes that
That this pattern extends to indicators of dis- widowers are indeed at relatively higher risk
ability (including limitations and assistance re- than widows, and, given that death is the most
quired for daily living and for work) was dem- extreme consequence of bereavement, much
onstrated in a recent study by Goldman, Koren- weight may be attached to this finding.
man, and Weinstein (1995) using data from a Recent studies have become highly sophisti-
longitudinal U.S. survey of aging and health. It cated in controlling for confounding variables
was found that widowed men were at much and artifacts, and the relatively higher excess of
higher risk of being disabled than married men, widowers in comparison with widows has re-
whereas the corresponding differential for mained robust. Goldman et al. (1995) included
women was considerably smaller. extensive baseline controls (total baseline sam-
pie: 7,500) and still found widowers to have decreasing somewhat after 2 years, still re-
higher odds of dying (by about 25%) than mar- mained elevated over the subsequent years fol-
ried men. They concluded that the same may be lowing bereavement. Similar results were re-
true for widows, but the coefficients for women ported in the Goldman et al. (1995) study. The
were smaller, and the estimated effects were increased mortality risks of widowhood among
insignificant. Similarly, in a recent survey (total men were not concentrated in the early dura-
sample: 36,142) incorporating control for in- tions of widowhood, suggesting to the research-
come (which varied between the gender sub- ers that widowhood is a state of chronic rather
groups), Rogers (1995) found that being conju- than acute stress for widowers. This is contrary
gally bereaved was much more detrimental for to the main body of research, including studies
men than for women. Using data from their reviewed here, which has established most ex-
large prospective cohort study in the Nether- treme effects early on.
lands, Joung, van de Mheen, Stronks, van Pop- Information is accumulating on causes of
pel, and Mackenbach (1998; see also Joung, death that may account for the relatively higher
1996) successively adjusted for sociodemo- mortality excess of widowers (see Mergenha-
graphic confounders and selection factors, gen, Lee, & Gove, 1985; Rogers, 1995; M.
health behaviors, and material circumstances in Stroebe & Stroebe, 1993). In an early cross-
calculating relative mortality risks by marital sectional survey, Gove (1972a) found greater
status. Examination of this data set (N = suicide (as well as attempted suicide) excesses
27,000) showed consistently higher relative risk for widowers relative to married men than for
ratios for widowers than for widows (vs. their widows relative to married women. Longitudi-
married counterparts). Finally, in a prospective nal studies of the distribution of suicides over
study of mortality in Finland among all 35-84- the duration of bereavement indicate that wid-
year-old married persons, Martikainen and owers are more at risk during the first few
Valkonen (1996) found higher excess mortality months, whereas the risk for widows spreads
for widowers (17%) than for widows (6%) even more evenly over a longer period (Bojanowsky
after controlling for confounding factors (e.g., & Bojanowsky, 1976; MacMahon & Pugh,
homogamy, common accidents, and common 1965).
unfavorable environments). In a recent report, Li (1995) provided a de-
In line with these general patterns, dropout tailed examination of the relative risk of suicide
through mortality was significantly more exces- among widowed elderly people. Li included a
sive for widowers than for widows (same-gen- large cohort of White married (n = 6,266) and
der nonbereaved controls were included) over widowed (n = 3,486) persons and conducted a
the course of the longitudinal study conducted 12-year follow-up survey. Adjustments were
by Gallagher-Thompson and her colleagues made in this study for housing, education level,
(Gallagher-Thompson et al., 1993). This result age, church attendance, and smoking. The risk
is remarkable, given the small sample size for of suicide for widowers was more than five
the identification of mortality differences. These times that for married men, whereas the relative
investigators were able to identify social isola- risk of suicide for widows was near unity. In-
tion and interpersonal difficulties as character- formation has also recently been collected on
izing the decedents rather than the survivors in gender differences in bereavement across the
the study. broad spectrum of causes of death (e.g., Joung,
The duration of excessive mortality risk may Glerum, van Poppel, Kardaun, & Mackenbach,
also be longer for widowers than for widows. In 1996). Among widowers, 32.1% of excess mor-
a well-controlled, large-scale cohort study (N = tality was due to diseases of the circulatory
12,522 spouse pairs), Schaefer, Quesenberry, system, whereas, among widows, the percent-
and Soora (1995) found excessive relative risks age was only 21.7%. Malignant tumors ac-
for both bereaved men and women in the first counted for 22.3% of widows' excess mortality,
year of loss (particularly the second half of this whereas this cause was listed for only 13.5% of
year). However, whereas the effects among the excess of widowers. Other causes were
women appeared to be limited to this period roughly equivalent across the genders. In the
(after adjustment for other predictors of mortal- survey by Rogers (1995) mentioned earlier, the
ity) the risk of mortality among men, though exceptionally high mortality of widowers was
found to be attributable to all causes, but par- bereavement, it was accounted for by only 8
ticularly to those causes associated with social individuals. Thus, this result may not be reli-
pathology, such as accidents, suicide, and cir- able. Furthermore, it was found that old-old
rhosis of the liver. widows had a lower than expected mortality
Ethnic group comparisons in mortality risk (only one death in the first 6 months of
among widowed individuals have also recently bereavement). Finally, examination of the data
been made. We noted earlier a reversal in rela- indicates that, overall, widowers' rates were in
tive excess among Black men and women, the fact relatively more excessive than those of
latter having relatively higher rates of psychiat- widows. Thus, this study does not provide
ric disorders than the former (Williams et al., strong contradictory evidence.
1992). It is interesting to note that, in a recent In summary, the mortality patterns provide
longitudinal study of mortality risk (N = strong support for the claim that men suffer
11,112), Lillard and Waite (1995) found that relatively more (in comparison with their same-
Black men and women faced risks of dying gender controls) from the consequences of wid-
approximately equal to those faced by their owhood than women. There have been more
White counterparts when controls were made studies conducted, and the patterns are, for the
for potentially confounding variables such as most part, clear-cut.
income differences. Both Black and White wid-
owers had relatively higher risks of dying than
Black or White widows, as compared with their Conclusions
married counterparts. Why there should be a
relatively higher excess for Black widowed Conjugal bereavement results in increased
women than for Black widowed men with re- morbidity and mortality for men and women,
spect to psychiatric disorders but not mortality with men being relatively more vulnerable to
remains to be clarified. the health risks than women, particularly—the
data have shown—during the period of acute
Negative results (of which there are few) grief. Although these conclusions have not been
have typically been reported for studies with unanimously shared, discrepant conclusions are
very small samples (e.g., Rees & Lutkins, 1967; due, as we have argued, to methodological over-
Ward, 1976). Relatively large samples are nec- sights such as the failure to base conclusions on
essary for significant differences in mortality to within-gender comparisons or to consider selec-
emerge. This can be illustrated with the study of tion effects. More longitudinal research would
Young, Benjamin, and Wallis (1963), in which be useful, for example, to establish relative risks
214 of the 4,486 widowers died during the first within and across the specific areas (e.g., de-
half year of bereavement, 66 more than would pressive symptomatology and mental and phys-
have been expected on the basis of a compara- ical illness) among recently bereaved widows
ble married sample. This reflects a mortality and widowers (vs. their respective nonbereaved
rate of 4.8% for the widowed as compared counterparts) and high-risk subgroups of both
with 3.2% for the married. If one were to study men and women. Nevertheless, the convergence
a hundred such widowers, one could not expect of evidence across the various manifestations
to find significant differences. For example, does allow confidence in our conclusion that
Jagger and Sutton (1991) reported a signifi- there are, in fact, relative male versus female
cantly elevated risk of mortality among a group excesses among those suffering from acute
of 161 elderly widows in the first 7 years of grief.
bereavement, but among the 41 widowers that
could be followed, the increased relative risk
did not reach an acceptable level of signifi- Theoretical Interpretation of Gender
cance. Another recent study revealed a rela- Differences in Health Consequences
tively higher risk for 76 young-old widows of Bereavement
between 65 and 74 years of age than for 45
young-old widowers and 56 old-old widowers How can this pattern of relative male ex-
(Mendes de Leon, Kasl, & Jacobs, 1993). Al- cesses in health detriments during bereavement
though the authors described this as a "threefold be explained? In an earlier theoretical discus-
increase" in mortality risk during the 1st year of sion (M. Stroebe & Stroebe, 1983), it was ar-
gued that gender differences in social support consequences of stressors and to help maintain
after bereavement offered the most plausible one's emotional equilibrium" (Billings & Moos,
explanation. In the meantime, increasing evi- 1981, p. 141). Problem-focused coping "in-
dence has emerged to indicate the necessity for cludes attempts to modify or eliminate the
an extension of this perspective to include anal- sources of stress through one's own behavior"
ysis of gender differences in methods of coping (Billings & Moos, 1981, p. 141). It is important
with bereavement. In this section, we reevaluate to note that emotion-focused and problem-
the different determinants using cognitive stress focused coping are not regarded as mutually
theory as a heuristic framework. exclusive. The second dimension of coping has
The basic assumption of stress theory is that been identified through a variety of different
stressful life events play an important role in the formulations, including vigilance-cognitive
etiology of various somatic and psychiatric dis- avoidance (Krohne, 1993); monitoring-blunt-
orders (see, e.g., Lazarus & Folkman, 1984). ing (Miller, Combs, & Kruus, 1993) and
More specifically, it is assumed that a stressful avoidant-non-avoidant coping (Suls & Fletcher,
life event may precipitate the onset of a physical 1985).
or mental disorder, particularly if a predisposi- In the context of marital bereavement, gender
tion toward that disorder already exists. Further- differences could be predicted with respect to
more, research has identified neurophysiologi- different aspects of the stress-coping equation
cal mechanisms linking stress with various outlined earlier. First, the demands of situations
detrimental consequences for the immune, gas- could be less stressful for women than for men
trointestinal, and cardiovascular systems (see (Dressier, 1985). Second, there could be gender
Baum & Grunberg, 1991). differences in the way the situation is appraised.
According to cognitive stress theory, critical For example, women might appraise the conse-
life events such as bereavement are stressful quences of partner loss as less stressful than
because they require major readjustment (e.g., men. Third, there could be gender differences in
Lazarus & Folkman, 1984; W. Stroebe & coping resources, particularly in social support,
Stroebe, 1987). The intensity of stress created because men are likely to rely exclusively on
by a life event depends on the extent to which their wives as confidants, whereas women fre-
trie perceived demands of the situation are ap- quently have confidants outside their marriage
praised by individuals as taxing or exceeding (cf. M. Stroebe & Stroebe, 1983; Thoits, 1986).
their coping resources, given that failure to cope Finally, there could be differences in coping
leads to important negative consequences strategies between men and women (Barnett,
(Lazarus & Folkman, 1984). Stress theory pro- Biener, & Baruch, 1987; Belle, 1987; Hobfoll,
vides the theoretical underpinning for the so- Dunahoo, Ben-Porath, & Monnier, 1994; Miller
called "buffering model," which suggests that & Kirsch, 1987).
high levels of social support protect individuals
against the deleterious impact of stress on
health. Situational Demands and Stress Appraisal
When a situation has been appraised as
stressful, individuals must do something to mas- Gender differences in situational demands
ter the situation or to control their emotional have mainly been analyzed in the context of role
reactions to the situation. Coping is a complex theory (e.g., Bernard, 1972; Gove, 1972b,
and multidimensional phenomenon encompass- 1973). Role theorists have argued that, as a
ing a variety of strategies, skills, and behaviors result of the greater advantages men have in
(cf. Carpenter, 1992; de Ridder, 2000; Krohne, marriage, the loss of a partner is more stressful
1993). Two dimensions of coping have been for men than for women. It has also been sug-
identified as central within the general domain gested that there may be different stressors in-
of stress-coping theory: emotion-focused ver- volved for men than for women. Umberson et
sus problem-focused coping and confronta- al. (1992), in their study of bereaved spouses,
tional versus avoidant strategies of coping (de found that, among widows, the primary mech-
Ridder, 1997). Emotion-focused coping "in- anism causing depression appeared to be finan-
cludes behavioral or cognitive responses whose cial strain, whereas for widowers it seemed to
primary function is to manage the emotional be strains associated with household management.
It is important to note, however, that the reported evidence of buffering (Krause, 1986;
extent to which a situation is experienced as Norris & Murrell, 1990; Schwarzer, 1992), al-
stressful is determined not by the objective beit using measures of social integration or re-
characteristics of the situation but by processes ceived social support, others have not (Greene
of cognitive appraisal. Bereaved men and & Feld, 1989; Murphy, 1988; W. Stroebe,
women may differ in their (primary) appraisal Stroebe, Abakoumkin, & Schut, 1996). The role
of a loss as harmful to their well-being. They of social support in accounting for gender dif-
may also differ in their (secondary) appraisal of ferences in bereavement outcomes has received
their ability to cope with their loss. Although minimal research attention. There are two parts
there have been few studies on gender differ- to this hypothesis, namely (a) that there are
ences in appraisal of stressful situations, the gender differences in the levels of social support
evidence that is available suggests that women, perceived or received by widowed individuals
relative to men, exhibit higher ratings of how and (b) that these differential levels of social
"upsetting" an event is or how much adjustment support are responsible for gender differences in
would be required (Bradley, 1980; de Ridder, health outcomes. There is some evidence—
2000; Sigmon, Stanton, & Snyder, 1995). It is mainly from studies of the elderly—that wid-
less clear, however, whether their higher reac- ows receive more social support than widowers
tivity to stressful situations results in less ade- (e.g., Berardo, 1970; Bock & Webber, 1972;
quate coping. In her extensive review of gender Perlman, Gerson, & Spinner, 1978); however,
differences in stress and coping, de Ridder before an analysis conducted as part of the
(2000) suggested that the higher reactivity of Tubingen Longitudinal Study of Bereavement,
women may by a disadvantage mainly when it had never been tested empirically whether
transient stressors are involved, but may be use-
differential social support is, in fact, responsible
ful in coping with more chronic stressors. Given
for gender differences in bereavement outcomes.
that bereavement has a long-term impact, this
In the course of the latter study (e.g., W.
would suggest an advantage for women relative
Stroebe, Stroebe, & Abakoumkin, 1999; W.
to men.
Stroebe et al., 1988, 1996), data were collected
that allowed a test of these hypotheses. In this
Resources: Social Support and Buffering study, 30 widows and 30 widowers were as-
sessed after their loss experiences, and their
Wives have been identified as their husbands' levels of perceived social support as well as
main and sometimes only confidants, whereas depressive symptomatology were repeatedly
women are more likely to have someone in measured over a 2-year period. These measure-
addition to their husband in whom to confide ments were then compared with the same set of
and turn to: notably, another woman (cf. Belle, measurements collected with a matched sample
1987; Fischer & Phillips, 1982; Lin & Westcott, of married controls. Although there was some
1991; Umberson et al., 1992). Thus, women are evidence that widows received more social sup-
more likely to have close relationships with port than widowers, there was no indication that
persons other than their spouses and to be the
this gender difference was responsible for gen-
ones to nurture and sustain the couple's social
der differences in bereavement outcomes (W.
relationships with others (Belle, 1987; Thoits,
Stroebe et al., 1999). This failure to find con-
1986, 1991). These differential patterns are
firmation for the stress-theoretical hypothesis
likely to be operant in adjustment to bereave-
ment: It has been suggested that widows gener- that social support is responsible for gender
ally receive more social support than widowers differences in bereavement outcomes is consis-
and that their higher level of social support tent with attachment theory, which rejects the
protects or buffers them against the deleterious notion that supportive friends can compensate
effect of partner loss (L. A. Morgan, 1984). for the loss of an attachment figure (Bowlby,
Research on the role of social support in 1969; Weiss, 1975; see also W. Stroebe et al.,
adjustment to loss has focused exclusively on 1996). Thus, even though it would be premature
testing the buffering model against the main on the basis of one study to rule out social
effect model. The results of these studies have support as a factor contributing to gender dif-
been inconsistent. Although some have indeed ferences in bereavement outcome, it has to be
acknowledged that there is little empirical sup- 1994); also, as noted earlier, women are more
port for this interpretation. likely to confide their troubles to others and to
be more expressive of their emotions (e.g., Der-
Coping Styles lega et al., 1993).
It has further been argued that the effective-
In the last decade, differences in coping ness of confrontive versus avoidant strategies
styles have become a major focus of research. varies according to the nature of the stressor,
Applied to gender differences in the health con- its controllability, and the duration of coping
sequences of bereavement, an interpretation in efforts required in dealing with it (Suls
terms of coping would be based on two assump- & Fletcher, 1985; Weidner & Collins, 1993).
tions: (a) that there are gender differences in When the particular stressor of bereavement is
coping with bereavement and (b) that the strat- considered, it becomes evident that the distinc-
egies used by women are more effective than tions made between problem- and emotion-fo-
those used by men. cused coping, avoidant and confrontive strate-
gies, and the controllable versus uncontrollable
General Patterns of Coping nature of a stressor need more detailed, precise
specification.
A substantial number of studies have re-
vealed that men tend, on the whole, to use more Implications for Coping With
problem-focused coping than women, whereas Bereavement
women are more likely to use emotion-focused
coping strategies (e.g., Billings & Moos, 1981; Although many of the stressful consequences
de Ridder, 2000; Endler & Parker, 1990; Folk- of partner loss (e.g., financial strain or problems
man & Lazarus, 1980; Pearlin & Schooler, in household management) that aggravate the
1978; Ptacek, Smith, & Zanas, 1992; Stone & distress resulting from bereavement are amena-
Neale, 1984; Vingerhoets & van Heck, 1990). ble to problem-focused coping, the most funda-
Conclusions are not completely unanimous, mental source of stress in bereavement—
however: In one study involving a student sam- namely, grief over the loss of a loved person—
ple, differences between the genders were not cannot be changed in the sense that separation
found on emotion- and problem-focused coping from the deceased cannot be reversed. As noted
(Thoits, 1991). earlier, unchangeable, uncontrollable situations
It has been claimed that, in general, problem- are said to be better dealt with through emotion-
focused strategies are more effective than emo- focused coping. However, to cope in an emo-
tion-focused ones (Billings & Moos, 1984; tion-focused way entails anything from wishful
Hobfoll et al., 1994; Hovanitz & Kozora, 1990), thinking to rumination, seeking information,
which—if we accept the gender difference pat- venting emotions, suppressing or denying neg-
tern found in the majority of studies—would ative or positive emotions, engaging in pleasant
imply better functioning (and consequences) activities, and engaging in dangerous activities
among men than among women. Other studies (e.g., drinking alcohol). It is evident that not all
have indicated, however, that the effectiveness of these emotion-focused strategies would be
of problem- versus emotion-focused strategies predictors of good adjustment to bereavement,
is dependent on the particular stressor or stress- and (as described shortly), for some, the oppo-
ful situation; for example, problem-focused site prediction has been substantiated (e.g., for
coping is less useful if the situation is uncon- rumination).
trollable (see de Ridder, 2000; Folkman, 1992). A more limited construct than emotion-
With respect to the confrontation-avoidance focused coping that has become familiar within
dimension, there is some agreement that men the field of bereavement research is the notion
are more avoidant in coping with stressors than of "working through" grief. Although subject to
are women (de Ridder, 2000; Krohne, 1993). criticism, this has been a central concept in
For example, men are more likely to engage in traditional grief theories (for reviews, see M.
distracting behavior (avoidance coping) when Stroebe, 1992; M. Stroebe & Stroebe, 1991;
depressed, whereas women are more likely to Wortman & Silver, 1987, 1989). According to
ruminate (Nolen-Hoeksema, Parker, & Larson, the so-called grief work hypothesis, which de-
rives from psychoanalytic theory and attach- (M. Stroebe & Stroebe, 1989). It seemed that
ment theory (Bowlby, 1980; Freud, 1917/1957; the widowers avoided the opportunity to share
Lindemann, 1944), one needs to work though their deep emotions, whereas, for widows, the
grief and confront the reality of loss, and not interviews provided an opportunity to talk about
to do so is unhealthy. This, then, incorporates their feelings with a sympathetic other. This
a confrontation-avoidance coping dimension. interpretation was supported by additional find-
This notion is also consistent with a body of ings on "talking to others" (widows also talked
research conducted by Pennebaker and his col- more about their loss; M. Stroebe et al., 1993).
leagues (e.g., Pennebaker, 1990; Pennebaker & Differential effectiveness of coping styles.
O'Heeron, 1984; Traue & Pennebaker, 1992) There was also evidence that individuals who
that has consistently shown the benefits of dis- talked about their loss had fewer depressive and
closing emotions, including those relating to somatic symptoms and were in better health
bereavement (note, however, recent reviews by than those who did not talk about their loss.
Kelly & McKillop, 1996, and Zech, 2000, ar- However, this relationship is correlational and
guing that disclosure may not always be bene- thus not necessarily indicative of a causal role
ficial, and positive effects may be a function of of talking about grief in reducing depressive
the type or manner of disclosure and the recep- symptomatology. It could simply be that those
tivity of the social environment). with higher symptomatology need to communi-
cate more (i.e., the opposite direction of causal-
Do Gender Differences in Coping With ity). Further analyses showed no evidence that
Bereavement Mediate the Differential confiding in others actually reduced symptom-
Health Pattern? atology over time. Similarly, in an investigation
conducted in the Netherlands, there was no in-
The prediction from the preceding discussion dication that the social sharing of emotions,
would be that confronting one's grief, as fol- whether defined as confiding or more widely in
lowed by women, would be more effective than the sense'of allowing others to see one's emo-
avoiding it, as preferred by men. This would be tions, had any beneficial impact on symptom
in favor of women's way of coping with stress levels (Schut, 1992; Schut, Stroebe, Stroebe,
in general and could, if valid, explain gender van den Bout, & de Keijser, 1994; M. Stroebe et
differences in bereavement outcomes. Women al., 1993).
cope better and have fewer health consequences One shortcoming of the preceding analysis is
because they confront and express their grief that it assesses only overt strategies of coping. It
more than men. is possible to confront emotions and "do one's
Gender differences in coping with bereave- grief work" in other ways, many of them intra-
ment. There are very few studies that have personal. Further analyses of the impact across
directly addressed the topic of how widows and time of adopting confrontational versus noncon-
widowers go about their grieving and what im- frontational coping strategies showed that wid-
pact this has on outcomes. Thus, we cannot say owers who avoided grief work were more de-
to what extent the patterns of gender differences pressed later in the study than widowers who
in coping with stressful events in general (re- confronted their grief. For widows, the style
viewed earlier) apply specifically to bereave- adopted seemed to make little difference (M.
ment. However, some indications can be de- Stroebe & Stroebe, 1991). Thus, there was only
rived from research on related issues. Results limited support for the notion that working
from the Tubingen Longitudinal Study of Be- through grief fosters adjustment. A potential
reavement showed gender differences in will- explanation is that men who distract do so more
ingness to communicate that were related to completely and to the detriment of their
depressive symptomatology (M. Stroebe & health—they, in fact, do need to confront their
Stroebe, 1989, 1991). There was a disinclina- emotions somewhat more—whereas women, as
tion among widowers who were highly dis- a result of the context in which they grieve and
tressed to participate in interviews, whereas the their more open grieving style, typically have
opposite was the case for widows, for whom more occasion to work through grief. On the
interview participation was associated with other hand, women will be prevented, by house-
greater distress than was refusal to participate hold and family tasks, from exclusively engag-
ing in their preferred style of emotion-focused (namely, the changes in life that occur because
coping. the deceased is no longer present, such as role
and identity changes). Thus, they proposed an
Coping With Bereavement: A orthogonal dimension of coping with loss. The
Reevaluation dimension, designated loss-restoration orienta-
tion, ranges from the orientation toward the loss
The findings just described raise doubts about of an attachment figure to an orientation toward
the basic assumption underlying the grief work the secondary stresses due to the loss. It is
hypothesis that confronting one's grief is more evident, following the discussion in the pre-
effective than avoiding it (M. Stroebe, 1992). vious section, that loss orientation is not
On the other hand, too complete an avoidance equivalent to emotion-focused coping and that
of grief also may not be conducive to recovery. restoration orientation is not equivalent to prob-
What is suggested by these results is the need lem-focused coping. To illustrate, restoration-
for a nonlinear model of recovery: Too much or oriented coping differs from problem-focused
too little grieving is maladaptive. Recent studies coping in that it also subsumes emotion-focused
have begun to confirm this suggestion. Nolen- coping associated with secondary stress man-
Hoeksema et al. (1994) reported disadvantages agement (one has fears about mastering the
of ruminating about bereavement. Rumination skills lost with the deceased, which could be
was defined in terms of concentration on de- tackled either by trying out the task or by work-
pressive thoughts and their meaning and passive ing on one's fears).
worrying. People with a ruminative style early Individuals who confront their loss would be
in bereavement were found to have higher de- engaging in grief work, or ruminating, or in-
pression levels at 6 months (depression level at dulging in wishful thinking, whereas those who
Time 1 was controlled). Those with a more avoid it would not. Individuals could also differ
distractive style became less depressed over in the extent to which they confront or avoid the
time. Bereaved women were found to ruminate secondary stresses associated with bereavement
more than bereaved men. Compatible with this (e.g., the skills to be mastered may or may not
general pattern, in a study also focusing on be attempted). According to this perspective,
bereavement (though not on gender differences), there are both benefits and costs involved with
Bonanno, Keltner, Holen, and Horowitz (1995) confronting and avoiding grief, just as there are
provided evidence that avoidant strategies may be benefits and costs involved with tackling versus
more functional than had previously been as- ignoring all of the additional tasks that arise as
sumed in the bereavement literature. the result of loss.
To understand gender differences in bereave- A central component of the model that dis-
ment, reexamination not only of the nature of tinguishes it from classic stress-coping theory
the stressor, but also of processes of coping and is a dynamic process fundamental to successful
of the efficacy of different strategies of coping, coping, namely, "oscillation" (for a discussion
would seem necessary. M. Stroebe and Schut of the difference between this concept and that
(1995, 1999) suggested a framework for con- of "intrusion-avoidance" in the trauma litera-
ceptualizing these components in their dual- ture, see M. Stroebe, Schut, & Stroebe, 1998).
process model of coping with loss, which is an This refers to the alternation between loss-ori-
adaptation of cognitive stress theory to the be- ented and restoration-oriented coping. At times,
reavement situation. bereaved individuals will be confronted by their
The model postulates that dealing with both loss; at other times, they will avoid memories,
the direct emotional consequences of loss and be distracted, or seek relief by concentrating on
concurrently occurring life changes is essential other things, or there may simply be no alter-
for adjustment to loss and that preoccupation native but to attend to the additional stressors
with one of these aspects, to the neglect of the (e.g., managing household chores or earning a
other, slows down this process. They argued living). The model proposes that oscillation is
that the loss of a partner results in two sources necessary for optimal adjustment over time (see
of stress: (a) stress directly associated with the the work of Helgeson [e.g., 1994; Helgeson &
loss of the loved person and (b) stress that Fritz, 1998] identifying the poor health conse-
comes about as a secondary consequence of loss quences of unmitigated communion, a trait as-
sociated with femininity, or unmitigated agency, tress. It seems plausible that what was achieved
associated with masculinity). Direct empirical in- in intervention was assistance with confronting
vestigation of this proposition is still needed. emotions for men and dealing with the second-
The major prediction suggested by the dual- ary stressors of bereavement (or perhaps over-
process model is that both forms of coping are focus on emotions) for women. In each case,
essential for effective coping. Thus, confronta- there were health benefits.
tion of a loss (i.e., including grief work) is The argument can be summarized as follows:
essential for a healthy recovery, and this comes Grief over the loss of a loved person is the
easily to women. However, if it is also relentless fundamental source of stress in bereavement,
(no oscillation), no progress toward recovery is and if this loss is not processed, health deficits
made; it is necessary to attend to other things as occur. Widows confront the emotional impact
well. As noted earlier, this, however, is de- of their bereavement, which is advantageous for
manded of women who have the caring role and them; in doing so, however, they may tend to
must attend to the household and to other tasks. neglect confrontation with the variety of sec-
It is our contention that men can more easily ondary stressors that are integral to bereave-
and completely block their emotions (being also ment. Men are more avoidant of the emotional
generally more avoidant in their coping style) impact of loss, focusing instead on dealing with
and not attend to the tasks defined in grief work secondary problems that arise through loss. For
by adopting a fairly exclusive restoration-ori- a healthy outcome, attention-avoidance of both
ented approach. We suggest that it is this dif- stressor types is essential. Following this line of
ference in role constraints of men and women argument, gender differences in health out-
that is a major cause of gender differences in comes of bereavement would be due to the fact
health outcomes of bereavement. In line with that role constraints are more likely to prevent
this argument, research has shown that men and women than men from engaging exclusively in
women have different problems to cope with their preferred way of coping.
(cf. de Ridder, 2000; Folkman & Lazarus, 1980; According to this model, the gender differ-
Porter & Stone, 1995), men reporting more ence to be expected in the acute grieving phase
work-related problems and more miscellaneous would diminish, or even reverse, across the
problems and women reporting more problems longer duration of bereavement. In the course of
focused on the self, parenting problems, and time, restoration orientation becomes more cen-
problems with other people (Porter & Stone, tral, and this is easier for men. The disadvan-
1995). In other words, differences between the tages for widowers that have been identified
genders may be due to the nature of the stressor here may become more advantageous as time
in addition to differences in appraisal and goes on. There is some support for the assump-
coping. tion that, after intense reactions to loss have
Indirect support for the preceding interpreta- abated, widowers' lifestyles and situations,
tion of gender differences in bereavement out- roles, and ways of coping may become as adap-
come comes from the study of Schut (1992; tive as, or even more adaptive than, widows'
Schut, Stroebe, van den Bout, & de Keijser, (cf. Hansson & Carpenter, 1994; Hays et al.,
1997). This study offered grief counseling to 1994; Nieboer et al., 1998; Stevens, 1995).
widows and widowers for mildly disturbed However, we have also drawn attention to stud-
grief. The participants had been bereaved for ies that appear to be less supportive of this
some time (ranging from 11 to 18 months after pattern (e.g., Glick et al., 1974; Van Grootheest
their loss). The program guided these men and et al., 1999). Just as the central parameters of
women in their use of confrontive versus the model need further testing, so too do pre-
avoidant strategies in coping not only with the dictions such as these, about gender differences
emotional aspects of bereavement but also with in the short and long term.
the secondary stressors with which they had
difficulty. "Teaching" bereaved men and women Conclusion
to cope in the way that the opposite gender
usually copes (teaching men to be more emotion As indicated by the review of empirical stud-
oriented and women to be more problem ori- ies on the morbidity and mortality of conjugal
ented) was associated with a lowering of dis- bereavement presented in the first part of this
article, there is now a reasonably sound body of then, correct. The confusion enters when abso-
evidence to support the conclusion that men lute risks are interpreted as relative risks. If one
suffer relatively higher consequences of partner is interested in the impact of bereavement on
loss than do women. It is much less clear, how- levels of distress among men and women, as we
ever, how to interpret this pattern. The assump- were in this analysis, then only one conclusion
tion made in an earlier review (M. Stroebe & is correct: Men suffer more.
Stroebe, 1983) that gender differences in the
social support received by widows and widow-
ers were responsible for differential bereave- References
ment outcomes has not, in the meantime, re-
ceived strong empirical support. Although the Barnett, R. C , Biener, L., & Baruch, G. K. (Eds.).
evidence as to the existence of such gender (1987). Gender and stress. New York: Free Press.
differences in social support after bereavement Baum, A., & Grunberg, N. E. (1991). Gender, stress
is quite persuasive, it has not been demonstrated and health. Health Psychology, 10, 80-85.
that these differences mediate gender differ- Bebbington, P. (1987). Marital status and depression:
A study of English national admission statistics.
ences in health outcomes.
Ada Psychiatrica Scandinavica, 75, 640-650.
Although there is some support for the as- Beck, A. T. (1967). Depression: Clinical, experimen-
sumption that women are more likely to con- tal and theoretical aspects. New York: Hoeber.
front their emotions than men, there is little Belle, D. (1987). Gender differences in the social
evidence that this type of confrontive coping, in moderators of stress. In R. C. Barnett, L. Biener, &
itself, confers a health advantage. Therefore, we G. K. Baruch (Eds.), Gender and stress (pp. 257-
argued for a revision of the so-called grief work 277). New York: Free Press.
hypothesis and suggested the following as- Berardo, F. M. (1970). Survivorship and social iso-
sumptions: (a) that both coping orientations, lation: The case of the aged widower. Family Co-
those directed toward the tasks of loss and of ordinator, 19, 11-25.
restoration, are essential in coming to terms Bernard, J. (1972). The future of marriage. Harmond-
sworth, Middlesex, England: Penguin Books.
with the loss of a loved one; (b) that attention to
Bierhals, A., Frank, E., Prigerson, H., Miller, M.,
and avoidance of both the direct emotional im- Fasiczka, A., & Reynolds, C. (1996). Gender dif-
pact of loss itself and its secondary conse- ferences in complicated grief among the elderly.
quences are necessary (the oscillation process); Omega: Journal of Death and Dying, 32, 303-317.
and (c) that gender differences in bereavement Billings, A. G., & Moos, R. H. (1981). The role of
outcomes are due to the fact that external con- coping responses and social resources in attenuat-
straints prevent women but not men from ex- ing the stress of life events. Journal of Behavioral
clusively engaging in their preferred style of Medicine, 4, 139-157.
coping. Billings, A. G., & Moos, R. H. (1984). Coping,
stress, and social resources among adults with
The "who suffers more" issue is still a com-
unipolar depression. Journal of Personality and
plex one. Interpreted in terms of relative ex- Social Psychology, 46, 877-891.
tremity of reactions to the loss or "relative risk" Bloom, B. L., Asher, S. J., & White, S. W. (1978).
(in which case nonbereaved control groups or Marital disruption as a stressor: A review and
prebereavement data must be included in the analysis. Psychological Bulletin, 85, 867-894.
analysis), research has clearly demonstrated that Bock, E. W., & Webber, I. L. (1972). Suicide among
widowers suffer more than widows. However, if the elderly: Isolating widowhood and mitigating
one is only interested in levels of distress among alternatives. Journal of Marriage and the Fam-
the bereaved (i.e., absolute risk), comparing ily, 34, 24-31.
rates of distress of widows and widowers di- Bojanowsky, J., & Bojanowsky, A. (1976). Zur
rectly without reference to nonbereaved con- Risikozeit des Selbstmordes bei Geschiedenen und
trols, then the conclusion would be that men and Verwitweten [Period of risk for suicides among the
divorced and widowed]. Nervenarzt, 47, 307-309.
women suffer to the same extent (or that suf-
Bonanno, G. A., Keltner, D., Holen, A., & Horowitz,
fering would be more pronounced for women). M. J. (1995). When avoiding unpleasant emotions
Finally, from a population perspective, the fact may not be such a bad thing: Verbal-autonomic
that there are many more widows than widow- response dissociation and midlife conjugal be-
ers means that there are more depressed widows reavement. Journal of Personality and Social Psy-
than widowers. All of these conclusions are, chology, 69, 975-989.
Bornstein, P. E., Clayton, P. J., Halikas, J. A., Mau- Dressier, W. W. (1985). The social and cultural con-
rice, W. L., & Robbins, E. (1973). The depression texts of coping: Action, gender and symptoms in a
of widowhood after thirteen months. British Jour- southern Black community. Social Science and
nal of Psychiatry, 122, 561-566. Medicine, 21, 499-506.
Bowlby, J. (1969). Attachment and loss: Vol. 1. At- Endler, N. S.,.& Parker, J. D. A. (1990). Multidimen-
tachment. New York: Basic Books. sional assessment of coping: A critical evaluation.
Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Journal of Personality and Social Psychology, 58,
Sadness and depression. New York: Basic Books. 844-854.
Bowling, A., & Windsor, J. (1995). Death after wid- Faschingbauer, T. R., Zisook, S., & DeVaul, R.
owhood: An analysis of mortality rates up to 13 (1987). The Texas Revised Inventory of Grief. In
years after bereavement. Omega: Journal of Death S. Zisook ((Ed.), Biopsychsocial aspects of grief
and Dying, 31, 35-39. and bereavement {pp. 109-124). Washington, DC:
Bradley, C. (1980). Gender differences in reporting American Psychiatric Press.

and rating of life events: A comparison of diabetic Feinson, M. (1986). Aging widows and widowers:
and healthy subjects. Journal of Psychosomatic Are there mental health differences? International
Research, 24, 35-37. Journal of Aging and Human Development, 28,
Carey, R. G. (1977). The widowed: A year later. 251-260.
Journal of Counseling Psychology, 24, 125-131. Fischer, C. S., & Phillips, S. L. (1982). Who is alone:
Carey, R. G. (1979). Weathering widowhood: Prob- Social characteristics of people with small net-
lems and adjustment of the widowed during the works. In L. A. Peplau & D. Perlman (Eds.),
first year. Omega: Journal of Death and Dying, 10, Loneliness: A sourcebook of current theory, re-
163-174. search and therapy (pp. 21-33). New York:
Carpenter, B. N. (Ed.). (1992). Personal coping: The- Wiley.
ory, research, and application. Westport, CT: Folkman, S. (1992). Making the case for coping. In
Praeger. B. N. Carpenter (Ed.), Personal coping: Theory,
research, and application (pp. 31-46). Westport,
Carter, H., & Glick, P. C. (1976). Marriage and
CT: Praeger.
divorce: A social and economic study. Cambridge,
MA: Harvard University Press. Folkman, S., & Lazarus, R. (1980). An analysis of
coping in a middle-aged community sample. Jour-
Chen, J. H., Bierhals, A. J., Prigerson, H. G., Kasl,
nal of Health and Social Behavior, 2, 219-239.
S. V., Mazure, C. M., & Jacobs, S. (1999). Gender
Fox, J. W. (1980). Gove's specific gender-role theory
differences in the effects of bereavement-related
of mental illness: A research note. Journal of
psychological distress in health outcomes. Psycho-
Health and Social Behavior, 21, 260-267.
logical Medicine, 29, 367-380.
Fox, J. W. (1984). Gender, marital status, and age as
Clayton, P. J. (1990). Bereavement and depression. social selection factors in recent psychiatric treat-
Journal of Clinical Psychiatry, 51, 34-40. ment. Journal of Health and Social Behavior, 25,
Clayton, P. J., Halikas, J. A., & Maurice, W. L. 394-405.
(1972). The depression of widowhood. British Freud, S. (1957). Mourning and melancholia. In J.
Journal of Psychiatry, 120, 71-76. Strachey (Ed. and Trans.), Standard edition of the
Cramer, D. (1993). Living alone, marital status, gen- complete works of Sigmund Freud (pp. 243-258).
der and health. Journal of Applied and Community London: Hogarth Press, (original work published
Social Psychology, 3, 1-15. 1917).
de Ridder, D. (2000). Gender, stress and coping: Do Gallagher, D. E., Breckenridge, J. N., Thompson,
women handle stressful situations differently from L. W., & Peterson, J. A. (1983). Effects of be-
men? In L. Sherr & J. S. St. Lawrence (Eds.), reavement on indicators of mental health in elderly
Women, health and the mind (pp. 115-135). widows and widowers. Journal of Gerontol-
Chichester, England: Wiley. ogy, 38, 565-571.
de Ridder, D. (1997). What is wrong with coping Gallagher-Thompson, D., Futterman, A., Farberow,
assessment? A review of conceptual and method- N., Thompson, L. W., & Peterson, J. (1993). The
ological issues. Psychology and Health, 12, 417- impact of spousal bereavement on older widows
431. and widowers. In M. Stroebe, W. Stroebe, & R. O.
Derlega, V. J., Metts, S., Petronio, S., & Margulis, Hansson (Eds.), Handbook of bereavement: The-
S. T. (1993). Self-disclosure. Newbury Park, CA: ory, research, and intervention (pp. 227-239).
Sage. New York: Cambridge University Press.
Derogatis, L. R. (1977). SCL-90: Administration, Gerber, I., Rusalem, R., Hannon, N., Battin, D., &
scoring and procedures manual for the revised Arkin, A. (1975). Anticipatory grief and aged wid-
version. Baltimore: John Hopkins University ows and widowers. Journal of Gerontology, 30,
Press. 225-229.
Gerber, I., Wiener, A., Battin, D., & Arkin, A. Horwitz, A. D., & Davies, L. (1994). Are emotional
(1975). Brief therapy to the aged bereaved. In B. distress and alcohol problems differential out-
Schoenberg, A. C. Carter, A. H. Kutscher, D. comes to stress? An exploratory test. Social Sci-
Peretz, & I. Goldberg (Eds.), Bereavement: Its ence Quarterly, 75, 607-621.
psychosocial aspects (pp. 310-333). New York: Hovanitz, C. A., & Kozora, E. (1990). Life stress and
Columbia University Press. clinically elevated MMPI scales: Gender differ-
Gilbar, O., & Dagan, A. (1995). Coping with loss: ences in the moderating influence of coping. Jour-
Differences between widows and widowers of de- nal of Clinical Psychology, 45, 166-111.
ceased cancer patients. Omega: Journal of Death Hu, Y., & Goldman, N. (1990). Mortality differen-
and Dying, 31, 207-220. tials by marital status: An international compari-
Glick, I., Weiss, R. S., & Parkes, C. M. (1974). The son. Demography, 27, 233-250.
first year of bereavement. New York: Wiley. Jacobs, S., Kasl, S. V., Ostfeld, A. M., Berkman, L.,
Goldberg, D. P., & Hillier, V. F. (1979). A scale & Charpentier, P. (1986). The measurement of
version of the General Health Questionnaire. Psy- grief: Age and gender variation. British Journal of
chological Medicine, 19, 139-145. Medical Psychology, 59, 305-310.
Goldman, N., Korenman, S., & Weinstein, R. (1995). Jagger, C , & Sutton, C. J. (1991). Death after marital
Marital status and health among the elderly. Social bereavement: Is the risk increased? Statistics in
Science and Medicine, 40, 1717-1730. Medicine, 46, 395-404.
Gove, W. R. (1972a). Gender, marital status, and Jeffery, R. W. (1989). Risk behaviors and health:
suicide. Journal of Health and Social Behav- Contrasting individual and population perspec-
ior, 13, 204-213. tives. American Psychologist, 44, 1194-1202.
Gove, W. R. (1972b). The relationship between gen- Joung, I. (1996). Marital status and health: Descrip-
der roles, marital roles, and mental illness. Social tive and explanatory studies. The Hague: Konin-
Forces, 51, 34-44. klijke Bibliotheek.
Gove, W. R. (1973). Gender, marital status and mor- Joung, I., Glerum, J. J., van Poppel, F. W., Kardaun,
tality. American Journal of Sociology, 79, 45-67. J. W., & Mackenbach, J. P. (1996). The contribu-
Gove, W. R. (1980). Mental illness and psychiatric tion of specific causes of death to mortality differ-
treatment among women. Psychology of Women ences by marital status in the Netherlands. Euro-
Quarterly, 4, 345-362. pean Journal of Public Health, 6, 142-149.
Gove, W. R., Hughes, M., & Style, C. (1983). Does Joung, I., Stronks, K., van de Mheen, H., van Poppel,
marriage have positive effects on the psychologi- F. W., van der Meer, J. B., & Mackenbach, J. P.
cal well-being of the individual? Journal of Health (1997). The contribution of intermediary factors to
and Social Behavior, 24, 122-131. marital status differences in self-reported health.
Greene, R. W., & Feld, S. (1989). Social support Journal of Marriage and Family, 59, 476-490.
coverage and the well-being of elderly widows and Joung, I., van de Mheen, H., Stronks, K., van Poppel,
married women. Journal of Family Issues, 10, F. W., & Mackenbach, J. P. (1998). A longitudinal
33-51. study of health selection in marital transitions.
Hansson, R. O., & Carpenter, B. N. (1994). Relation- Social Science and Medicine, 46, 425-435.
ships in old age: Coping with the challenge of Joung, I., van der Meer, J. B., & Mackenbach, J. P.
transition. New York: Guilford Press. (1995). Marital status and health care utilization.
Hays, J. C , Kasl, S. V., & Jacobs, S. C. (1994). The International Journal of Epidemiology, 24, 569-
course of psychological distress following threat- 575.
ened and actual conjugal bereavement. Psycholog- Kelly, A. E., & McKillop, K. J. (1996). Conse-
ical Medicine, 24, 917-927. quences of revealing personal secrets. Psycholog-
Helgeson, V. S. (1994). Relation of agency and com- ical Bulletin, 120, 450-465.
munion to well-being: Evidence and potential ex- Krause, N. (1986). Social support, stress, and well-
planations. Psychological Bulletin, 116, 412-428. being among older adults. Journal of Gerontol-
Helgeson, V. S., & Fritz, H. L. (1998). A theory of ogy, 41, 512-519.
unmitigated communion. Personality and Social Krohne, H. W. (Ed.). (1993). Attention and avoid-
Psychology Review, 2, 173-183. ance. Seattle, WA: Hogrefe & Huber.
Helsing, K. J., & Szklo, M. (1981). Mortality after Lazarus, R. S., & Folkman, S. (1984). Stress, ap-
bereavement. American Journal of Epidemiology, praisal, and coping. New York: Springer.
114, 41-52. Lee, G. R., Willetts, M. C , & Seccombe, K. (1998).
Hobfoll, S. E., Dunahoo, C. L., Ben-Porath, Y., & Widowhood and depression: Gender differences.
Monnier, J. (1994). Gender and coping: The dual- Research on Aging, 20, 611-630.
axis model of coping. American Journal of Com- Li, G. (1995). The interaction effect of bereavement
munity Psychology, 22, 49-82. and gender on the risk of suicide in the elderly: An
historical cohort study. Social Science and Medi- Murphy, S. A. (1988). Mental distress and recovery
cine, 40, 825-828. in a high-risk bereavement sample three years after
Lillard, L. A., & Waite, L. J. (1995). 'Til death do us untimely death. Nursing Research, 37, 30-35.
part: Marital disruption and mortality. American Nieboer, A. P., Lindenberg, S. M., & Ormel, J.
Journal of Sociology, 100, 131-156. (1998). Conjugal bereavement and well-being of
Lin, N., & Westcott, J. (1991). Marital engagement/ elderly men and women: A preliminary study.
disengagement, social networks, and mental Omega: Journal of Death and Dying, 38, 113-142.
health. In J. Eckenrode (Ed.), The social context of Nolen-Hoeksema, S., Parker, L. E., & Larson, J.
coping (pp. 213-237). New York: Plenum. (1994). Ruminative coping with depressed mood
Lindemann, E. (1944). Symptomatology and man- following loss. Journal of Personality and Social
agement of acute grief. American Journal of Psy- Psychology, 67, 92-104.
chiatry, 101, 141-148. Norris, F. H., & Murrell, S. A. (1990). Social support,
Lund, D. A., Caserta, M. S., & Dimond, M. F. life events, and stress as modifiers of adjustment to
(1986). Gender differences through two years of bereavement by older adults. Psychology and Ag-
bereavement among the elderly. Gerontologist, 26, ing, 5, 429-436.
314-320. Notarious, C , & Johnson, J. (1982). Emotional ex-
MacMahon, B., & Pugh, T. F. (1965). Suicide in the pression in husbands and wives. Journal of Mar-
widowed. American Journal of Epidemiology, 81, riage and the Family, 44, 483-489.
23-32. Osterweis, M., Solomon, F., & Green, M. (1984).
Marks, N. F., & Lambert, J. D. (1998). Marital status Bereavement: Reactions, consequences, and care.
continuity and change among young and midlife Washington, DC: National Academy Press.
adults: Longitudinal effects on well-being. Journal Parkes, C. M. (1964). Recent bereavement as a cause
of Family Issues, 19, 652-686. of mental illness. British Journal of Psychiatry,
Martikainen, P., & Valkonen, T. (1996). Mortality 110, 198-204.
after death of spouse in relation to duration of Parkes, C. M. (1996). Bereavement: Studies of grief
bereavement in Finland. Journal of Epidemiology in adult life (3rd ed.). New York: International
and Community Health, 50, 264-268. Universities Press.
Mellstrom, D., Nilsson, A., Oden, A., Rundgren, A., Pearlin, L., & Schooler, C. (1978). The structure of
& Svanborg, A. (1982). Mortality among the wid- coping. Journal of Health and Social Behavior, 19,
owed in Sweden. Scandinavian Journal of Social 2-21.
Medicine, 10, 33-41. Pennebaker, J. W. (1990). Opening up: The healing
Mendes de Leon, C , Kasl, S. V., & Jacobs, S. (1993). power of confiding in others. New York: Morrow.
Widowhood and mortality risk in a community Pennebaker, J. W., & O'Heeron, R. C. (1984). Con-
sample of the elderly: A prospective study. Journal fiding in others and illness rates among spouses of
of Clinical Epidemiology, 46, 519-527. suicide and accidental death victims. Journal of
Mergenhagen, P. J., Lee, B. A., & Gove, W. R. Abnormal Psychology, 93, 473-476.
(1985). Till death us do part: Recent changes in the Pennebaker, J. W., & Roberts, T. (1992). Toward a
relationship between marital status and mortality. his and hers theory of emotion: Gender differences
Sociology and Social Research, 70, 53-56. in visceral perception. Journal of Social and Clin-
Meshot, C. M., & Leitner, L. M. (1993). Adolescent ical Psychology, 11, 199-212.
mourning and parental death. Omega: Journal of Perlman, D., Gerson, A. C , & Spinner, B. (1978).
Death and Dying, 26, 287-299. Loneliness among senior citizens: An empirical
Miller, S. M., Combs, C , & Kruus, L. (1993). Tuning report. Essence, 2, 239-248.
in and tuning out: Confronting the effects of con- Porter, L. S., & Stone, A. S. (1995). Are there really
frontation. In H. W. Krohne (Ed.), Attention and differences in coping? A reconsideration of previ-
avoidance (pp. 51-69). Seattle, WA: Hogrefe & ous data and results from a daily study. Journal of
Huber. Social and Clinical Psychology, 14, 184-202.
Miller, S. M., & Kirsch, N. (1987). Gender differ- Ptacek, J. T., Smith, R. E., & Zanas, J. (1992).
ences in cognitive coping with stress. In R. C. Gender, appraisal, and coping: A longitudinal
Barnett, L. Biener, & G. K. Baruch (Eds.), Gender analysis. Journal of Personality, 60, 747-770.
and stress (pp. 278-307). New York: Free Press. Radloff, L. (1975). Gender differences in depression:
Morgan, L. A. (1984). Changes in family interaction The effects of occupation and marital status. Gen-
following widowhood. Journal of Marriage and der Roles, 1, 249-265.
the Family, 46, 323-331. Radloff, L. S. (1977). The CES-D Scale: A self-
Morgan, M. (1980). Marital status, health, illness and report depression scale for research in the general
service use. Social Science and Medicine, 14A, population. Applied Psychological Measure-
633-643. ment, 1, 385-401.
Rees, W., & Lutkins, S. (1967). Mortality of bereave- Stroebe, M. (1992). Coping with bereavement: A
ment. British Medical Journal, 4, 13-16. review of the grief work hypothesis. Omega: Jour-
Regier, D. A., Boyd, J. H., Burke, J. D., Rae, D. S., nal of Death and Dying, 26, 19-42.
Myers, J. K., Kramer, M., Robins, L. N., George, Stroebe, M., & Schut, H. A. W. (1995, April). Dif-
L. K., Karno, M., & Locke, B. Z. (1988). One- ferential patterns of coping with bereavement be-
month prevalence of mental disorders in the tween widows and widowers. Paper presented at
United States. Archives of General Psychiatry, 45, the annual conference of the British Psychological
977-986. Society, Warwick, England.
Rogers, R. G. (1995). Marriage, gender, and mortal- Stroebe, M., & Schut, H. A. W. (1999). The dual
ity. Journal of Marriage and the Family, 57, 515— process model of coping with bereavement: Ratio-
526. nale and description. Death Studies, 23, 197-224.
Sanders, C. M., Mauger, P. A., & Strong, P. A. Stroebe, M., Schut, H., & Stroebe, W. (1998).
(1991). A manual for the Grief Experience Inven- Trauma and grief: A comparative analysis. In J. H.
tory. Palo Alto, CA: Consulting Psychologists Harvey (Ed.), Perspectives on loss: A sourcebook
Press. (pp. 81-96). Philadelphia: Brunner/Mazel.
Schaefer, C , Quesenberry, C. P., & Soora, W. Stroebe, M., & Stroebe, W. (1983). Who suffers
(1995). Mortality following conjugal bereavement more? Sex differences in health risks of the wid-
and the effects of a shared environment. American owed. Psychological Bulletin, 93, 297-301.
Journal of Epidemiology, 141, 1142-1152. Stroebe, M., & Stroebe, W. (1989). Who participates
Schut, H. A. W. (1992). Omgaan met de dood van de in bereavement research? A review and empirical
partner: Effecten of gezondheid en effecten van study. Omega: Journal of Death and Dying, 20,
rouwbegeleiding. [Coping with conjugal bereave- 1-29.
ment: Effects on psychological functioning and Stroebe, M., & Stroebe, W. (1991). Does "grief
effects of grief counseling]. Amsterdam: Thesis work" work? Journal of Consulting and Clinical
Publishers. Psychology, 59, 479-482.
Schut, H. A. W., Stroebe, M., Stroebe, W., van den Stroebe, M., & Stroebe, W. (1993). The mortality of
Bout, J., & de Keijser, M. A. (1994, June). Are bereavement: A review. In M. Stroebe, W. Stro-
accepted ways of coping with loss really effica- ebe, & R. O. Hansson (Eds.), Handbook of be-
cious? Paper presented at the Fourth International reavement: Theory, research, and intervention
Conference on Grief and Bereavement in Contem- (pp. 175-195). New York: Cambridge University
porary Society, Stockholm, Sweden. Press.
Schut, H. A. W., Stroebe, M., van den Bout, J., & de Stroebe, M., Stroebe, W., Gergen, K., & Gergen, M.
Keijser, J. (1997). Gender differences in the effi- (1981). The broken heart: Reality or myth? Ome-
cacy of grief counseling. British Journal of Clini- ga: Journal of Death and Dying, 12, 87-105.
cal Psychology, 36, 63-72. Stroebe, M., Stroebe, W., & Hansson, R. O. (Eds.).
Schwarzer, C. (1992). Bereavement, received social (1993). Handbook of bereavement: Theory, re-
support, and anxiety in the elderly: A longitudinal search, and intervention. New York: Cambridge
analysis. Anxiety Research, 4, 287-298. University Press.
Shields, S. A. (1991). Gender in the psychology of Stroebe, W., & Stroebe, M. (1987). Bereavement and
emotion: A selective research review. In K. T. health. New York: Cambridge University Press.
Strongman (Ed.), International review of studies Stroebe, W., & Stroebe, M. (1993). Determinants of
on emotion (Vol. 1, pp. 227-245). Chichester, En- adjustment to bereavement in younger widows and
gland: Wiley. widowers. In M. Stroebe, W. Stroebe, & R. O.
Siegel, J. M., & Kuykendall, D. H. (1990). Loss, Hansson (Eds.), Handbook of bereavement: The-
widowhood, and psychological distress among the ory, research, and intervention (pp. 208-226).
elderly. Journal of Consulting and Clinical Psy- New York: Cambridge University Press.
chology, 58, 519-524. Stroebe, W., Stroebe, M., & Abakoumkin, G. (1999).
Sigmon, S. T., Stanton, A. L., & Snyder, C. R. Does differential social support cause gender dif-
(1995). Gender differences in coping: A further ferences in bereavement outcome? Journal of Ap-
test of socialization and role constraint theories. plied Social and Community Psychology, 9, 1-12.
Gender Roles, 33, 25-35. Stroebe, W., Stroebe, M., Abakoumkin, G., & Schut,
Stevens, N. (1995). Gender and adaptation in later H. A. W. (1996). The role of loneliness and social
life. Ageing and Society, 15, 37. support in adjustment to loss: A test of attachment
Stone, A. A., & Neale, J. M. (1984). New measures versus stress theory. Journal of Personality and
of daily coping: Development and preliminary re- Social Psychology, 70, 1241-1249.
sults. Journal of Personality and Social Psychol- Stroebe, W., Stroebe, M., & Domittner, G. (1988).
ogy, 46, 892-906. Individual and situational differences in recovery
from bereavement: A risk group identified. Jour- Verbrugge, L. M., & Wingard, D. L. (1987). Gender
nal of Social Issues, 44, 143-158. differentials in health and mortality. Women and
Stroebe, W., Stroebe, M , & Schut, H. A. W. (1993, Health, 12, 103-145.
October). Working through loss: Does it help? Vingerhoets, A. J. J. M., & van Heck, G. L. (1990).
Paper presented at the annual conference of the Gender, coping and psychosomatic symptoms.
Society for Experimental Social Psychology, Santa Psychological Medicine, 20, 125-135.
Barbara, CA. Ward, A. W. (1976). Mortality of bereavement. Brit-
Suls, J., & Fletcher, B. (1985). The relative efficacy ish Medical Journal, 1, 700-702.
of avoidant and nonavoidant coping strategies: A Weidner, G., & Collins, L. (1993). Gender, coping
meta-analysis. Health Psychology, 4, 249-288. and health. In H. W. Krohne (Ed.), Attention and
Thoits, P. A. (1986). Multiple identities: Examining avoidance (pp. 241-265). Seattle, WA: Hogrefe &
gender and marital status differences in distress. Huber.

American Sociological Review, 51, 259-272. Weisman, M. M., & Klerman, G. L. (1977). Gender
Thoits, P. A. (1991). Gender differences in coping differences and the epidemiology of depression.
with emotional distress. In J. Eckenrode (Ed.), The Archives of General Psychiatry, 34, 98—111.
social context of coping (pp. 107-138). New York: Weiss, R. (1975). Loneliness: The experience of emo-
Plenum. tional and social isolation. Cambridge, MA: MIT
Thompson, L. W., Breckenridge, J. N., Gallagher, D., Press.
& Peterson, J. A. (1984). Effects of bereavement Williams, D. R., Takeuchi, D. T., & Adair, R. K.
on self-perceptions of physical health in elderly (1992). Marital status and psychiatric disorders
widows and widowers. Journal of Gerontol- among Blacks and Whites. Journal of Health and
ogy, 39, 309-314. Social Behavior, 33, 140-157.
Traue, H. C , & Pennebaker, J. W. (Eds.). (1992). Wingard, D. L. (1984). The gender differential in
Emotion inhibition and health. Seattle, WA: morbidity, mortality, and lifestyle. Annual review
Hogrefe & Huber. of public health (Vol. 5, pp. 433-458). Palo Alto,
Umberson, D., Wortman, C. B., & Kessler, R. C. CA: Annual Reviews.
(1992). Widowhood and depression: Explaining Wortman, C , & Silver, R. (1987). Coping with irre-
long-term gender differences in vulnerability. vocable loss. In A. Baum, C. Frederick, I. Frieze et
Journal of Health and Social Behavior, 33, 10-24. al. (Eds.), Cataclysms, crises, and catastrophes:
U.S. Department of Health and Human Services, Psychology in action (pp. 189-235). Washington,
Public Health Service. (1985). Vital statistics of DC: American Psychological Association.
the United States, 1985. Washington, DC: U.S. Wortman, C , & Silver, R. (1989). The myths of
Government Printing Office. coping with loss. Journal of Consulting and Clin-
Van Grootheest, D. S., Beekman, A. T. F., Broesse ical Psychology, 57, 349-357.
van Groenou, M. I., & Deeg, D. J. H. (1999). Young, M., Benjamin, B., & Wallis, C. (1963). Mor-
Gender differences in depression after widow- tality of widowers. Lancet, 2, 454-456.
hood: Do men suffer more? Social Psychiatry and Zech, E. (2000). The effects of the communication of
Psychiatric Epidemiology, 34, 391-398. emotional experiences. Unpublished doctoral dis-
Verbrugge, L. M. (1979). Marital status and health. sertation, University of Louvain-la-Neuve, Lou-
Journal of Marriage and the Family, 41, 267-285. vain, Belgium.
Verbrugge, L. M. (1989). The twain meet: Empirical
explanations of gender differences in health and Received November 22, 1999
mortality. Journal of Health and Social Behav- Revision received May 16, 2000
ior, 30, 282-304. Accepted May 16, 2000 •

Gender Differences in Adjustment To Bereavement: An Empirical and Theoretical Review

Uploaded by

Copyright:

Available Formats

You might also like

Gender Differences in Adjustment To Bereavement: An Empirical and Theoretical Review

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gender Differences in Adjustment To Bereavement: An Empirical and Theoretical Review

Uploaded by

Copyright:

Available Formats

Review of General Psychology Copyright 2001 by the Educational Publishing Foundation

2001, Vol. 5, No. 1, 62-83 1089-2680/01/$5.00 DOI: 10.1037//1089-2680.5.1.62

Gender Differences in Adjustment to Bereavement:

pattern are examined. An interpretation in terms of gender differences in social support

cases of a disease in a population that can be Distress and Depressive Symptomatology

usually without intervention (cf. Clayton,

1990). Thus, it is important to consider studies

interactions on indexes of physical health Methodologically sound investigations of the

Bradley, C. (1980). Gender differences in reporting American Psychiatric Press.

gender and marital status differences in distress. Huber.

You might also like