402 Chapter 11show equal health profiles, can we conclude
that widowhood has the same effects on the
health of women and men? No, because
women and men who are not widowed differ
in health. For example, married women are
more depressed than married men. A study
that shows no sex differences in depression
among widowed women and men could im-
ply that widowhood increased men’s distress
levels to those of women or lowered wom-
en’s distress levels to those of men. In other
words, widowhood could have very different
effects on women’s and men’s distress. Let’s
take another example. In general, men have
higher suicide rates compared to women. A
study that shows no sex differences in sui-
cide rates among the widowed could imply
that widowhood increased women’s suicide
rates to those of men or decreased men’s sui-
cide rates to those of women. The most ap-
propriate comparison group to use in a study
of widowhood is married women and men
because both widowed and married people
share the experience of having entered into
marriage. It would not be appropriate to
compare widowed persons to never-married
persons because we know there are differ-
ences between the kinds of people who do
and do not get married.
The second way to examine the effects of
widowhood on health is to conduct a longi-
tudinal study. Longitudinal studies typically
examine people shortly after widowhood and
then follow them over time to assess changes
in their health. The disadvantage of this
methodology is that we do not know peo-
ple’s level of health before widowhood. The
advantage, however, is that we know peo-
ple’s initial health status immediately after
widowhood so we can truly examine changes
in health over time.
The ideal study of widowhood would use
aprospective designin which people’s healthEffect of Bereavement on Health
If marriage is good for health, presumably
losing a spouse has negative effects on health.
These negative effects could stem from the
loss of resources that the deceased spouse
provided as well as the general experience of
bereavement. Determining the effects of be-
reavement on health is not easy. Two kinds of
studies have been conducted to address this
issue: cross-sectional and longitudinal. Cross-
sectional studies evaluate people who are wid-
owed at a single point in time. The advantage
of this methodology is that large representa-
tive samples can be studied. There are three
disadvantages. First, people are widowed for
varying lengths of time, and the length of
time since widowhood is bound to influence
health. Second, the healthiest people are more
likely to remarry after widowhood. Thus the
people who remain widowed are not rep-
resentative of all widowed people and may
be more unhealthy than the widowed who
have remarried. Third, causality cannot be
inferred. In other words, we will not know if
widowhood caused the decline in health or if
unhealthy people were more likely to be wid-
owed. At first glance, this latter possibility
may seem unlikely. However, recall that peo-
ple are attracted to similar others and marry
people who are similar to themselves. One
characteristic on which matching could occur
is health. It is possible, then, that less healthy
people are more likely to lose a spouse.
An important methodological issue to
keep in mind when evaluating cross-sectional
studies of the effect of widowhood on health
is whether an appropriate comparison group
of nonwidowed persons was used. This is es-
pecially important when evaluating sex differ-
ences in the effects of widowhood on health.
Why? If widowed women and widowed menM11_HELG0185_04_SE_C11.indd 402 6/21/11 12:43 PM