University of Southern California Dissertations and Theses

Longitudinal relationships between bereavement and physical health‐related quality of life in middle‐ to older‐aged women

(USC Thesis Other)

Longitudinal relationships between bereavement and physical health‐related quality of life in middle‐ to older‐aged women

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content AGE, BEREAVEMENT, AND HEALTH IN WOMEN 1

Longitudinal relationships between bereavement and
physical health-related quality of life in middle- to older-aged women

Doctoral Dissertation
Uta Maeda, MA
Department of Psychology, Clinical Sciences
University of Southern California

August 2014

Dissertation Committee:
Margaret Gatz, PhD (chair)
Biing-Jiun Shen, PhD
Wendy Mack, PhD
Beth Meyerowitz, PhD
Stephen Read, PhD

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 2

Table of Contents
Page
Abstract 5

Introduction 6
Bereavement and Physical Health 6
Bereavement in women 9
Age-Related Differences in Bereavement 11
Context of Bereavement 13
Other factors impacting the relationship between stressful life events and physical health 14
Study Aims 15

Methods 16
Participants 16
Procedure 18
Measures 19
Age
Time elapsed since bereavement
Physical HRQOL
Life events
Context of bereavement
Moderators 21
1. Depression
2. Social Support
3. Perceived stress
Covariates 23
1. Ethnicity
2. Marital/partner status
3. Socioeconomic status
4. Education
5. ELITE randomization status
6. Total number of negative stressful life events
7. Total perceived undesirability of negative stressful life events
8. Perceived undesirability of the bereavement event
Analyses 25
Preliminary and Descriptive Analyses
Modeling trajectories of physical HRQOL
Results
Descriptive Analyses 28
Sample characteristics 28
Attrition analyses 29
Preliminary Analyses
Identification of covariates 30
Stressful Life Events 31
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 3

Modeling Trajectory of Physical HRQOL
Main analyses 33
Short-term effects of bereavement 34
Age moderation 35
Other psychosocial moderators 35
Bereavement x depression
Bereavement x social support
Bereavement x perceived stress
Types of Bereavement 38
Context of Bereavement 39
Secondary Psychosocial Outcomes 40
Depression
Perceived stress
Social support
Posthoc Analyses 43
Secondary health outcome
Length of randomized treatment
Stressful Life Events

Discussion 46
Reexamining study design 50
Measurement considerations
Sample considerations
Predictors of physical HRQOL 54
Conclusion & Future Directions 56

References 57

Tables
1. Demographic, clinical, and psychosocial characteristics of the sample 66
2. Means of physical HRQOL and other key psychosocial variables 68
3. Correlations between physical HRQOL and key study variables 70
4. Results of longitudinal growth modeling 71
5. Top ten most commonly reported stressful life events 72

Figures
1. Hypothesized physical HRQOL trajectories 73
2. Hypothesized interaction between age and bereavement 74
3. Flowchart of participant exclusion and attrition 75
4. Path model 76
5. Mean physical HRQOL for bereaved vs. non-bereavedgroups 77
6. Cross-sectional relationship between physical HRQOL and time since bereavement 78
7. Interaction between age and bereavement 79
8. Interaction between depression and bereavement 80
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 4

9. Interaction between social support and bereavement 81
10. Interaction between perceived stress and bereavement 82
11. Mean CES-D scores for bereaved vs. non-bereaved groups 83
12. Mean social support scores for bereaved vs. non-bereaved groups 84
13. Mean perceived stress scores for bereaved vs. non-bereaved groups 85

Appendices
A. The Women’s Health Questionnaire 86
B. PERI Life Events Scale 88
C. Context of Bereavement 94
D. CES-D 96
E. Medical Outcomes Study – Social Support Scale 97
F. Perceived Stress Scale 98

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 5

Abstract
Although the detrimental effects of bereavement on structural aspects of physical health
(e.g., mortality, disease) have been well documented in the literature, its impact on functional
health patterns has been less studied. This study examined the longitudinal relationship between
non-spousal bereavement and physical health-related quality of life (HRQOL) in a large sample
of community-dwelling women over a period of two years. Each participant had a pre-
bereavement baseline (if applicable) and was followed every 6 months thereafter. Potential
moderators of the relationship between bereavement and physical HRQOL, including age,
depression, perceived stress, social support, and context of bereavement, were explored.
Contrary to expectations, the occurrence of a non-spousal bereavement event did not predict
change over time in physical HRQOL and there was no difference in the pattern of physical
HRQOL over time between bereaved and non-bereaved women. Within the bereaved group,
those who had lost a family member had slower recovery of physical HRQOL over 24 months
compared to those who had lost a friend. Descriptive data suggested that participants had better
physical HRQOL and emotional functioning 6 months after bereavement if the deceased
individual was of older age or had ongoing serious health concerns prior to death. The age of the
participant did not moderate the relationship between bereavement and physical HRQOL.
Depression, perceived stress, and social support were each significant predictors of physical
HRQOL trajectories but themselves were unaffected by bereavement, and they did not moderate
the effects of bereavement on physical HRQOL. Overall, these findings were inconsistent with
the existing bereavement literature, which largely focuses on spousal bereavement. Potential
explanations for this discrepancy, including frequency or length of measurement, choice of
physical HRQOL measure, and biases of the obtained sample were systematically explored.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 6

Introduction

Bereavement, defined as the situation of having lost a significant person through death,
has been classified as one of the most significant and most traumatic events that one can
experience during his or her lifetime (Holmes & Rahe, 1967). According to the Social
Readjustment Rating Scale (SRRS), a standard measure of stressful life events, death of a spouse
ranks at the top as the event requiring the most intense psychological and lifestyle readjustment.
Accordingly, the existing research on detrimental consequences of bereavement has focused
primarily on spousal bereavement (Stroebe et al., 2007; Van den Berg et al., 2011; Carnelley et
al., 2006; Chen et al., 1999; Thompson et al, 1984; Lee, Willets, & Seccombe, 1998; Matthews,
1991; Moon et al., 2011; Lichtenstein, Gatz, & Berg, 1998; Mendes de Leon, Kasl, & Jacobs,
1993). Although some scholars have called attention to the special case of loss of a child
(Littlefield & Rushton, 1986; Miles, 1985; Li et al., 2003) or to other types of losses such as the
bereavement of a sibling or parent (D’Epinay et al., 2010; Middleton et al., 1998), non-spousal
bereavement generally remains a relatively understudied area of the literature. The purpose of the
current study is to examine the detrimental effects of non-spousal bereavement, primarily on
physical health-related outcomes and secondarily on emotional and social outcomes.
Bereavement and Physical Health
It has been consistently reported that stressful life events play a critical role in the
etiology of various physical disorders (Tosevski & Milovancevic, 2006; Glaser, 2005; Carroll et
al., 2005; Philips et al., 2005; Rozanski et al., 2005; Baum & Grunberg, 1991). As the top-rated
stressful life event, spousal bereavement in particular has gained significant attention in the
literature as a potential medical problem. Compared to their non-bereaved counterparts, bereaved
spouses report a greater number of physical symptoms, such as headaches, indigestion, dizziness,
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 7

chest pain, and weight loss (Shahar et al., 2001; Stroebe et al., 2007). In particular, Bradbeer et
al., (2003) reported that widowed individuals were three times more likely than non-bereaved
individuals to endorse current severe activity-limiting pain. Spousal bereavement has also been
associated with onset or aggravation of a variety of illnesses and disability, including heightened
risk for cancer and cardiovascular diseases (Buckley et al., 2010; Chen et al., 1999). Two months
after the loss of a spouse, bereaved older adults were 1.4 times more likely than their non-
bereaved counterparts to experience a new or worsened illness, and reported 1.73 times higher
medication use (Thompson et al., 1984). Self-rated health was significantly lower among
bereaved individuals, especially in those who considered the life event to be a negative turning
point (Sutin et al., 2010; Thompson et al., 1984).
The literature has consistently reported that spousal bereavement is associated with
increased rates of mortality due to a variety of causes (Schaefer et al, 1995; Hart et al., 2007;
Espinosa & Evans, 2008; Stroebe et al., 2007; Ott & Lueger, 2002; van den Berg et al., 2007).
This impact cannot be attributed simply to selection effects or the concordance of health,
attitudes, and socioeconomic standing between partners (Boyle et al., 2011). In particular, it has
been estimated that the death of a spouse leads to approximately 12% loss in residual life
expectancy and a 10-40% increase in the mortality risk in the surviving spouse (van den Berg et
al., 2011; Elwert & Christakis, 2008; Elwert & Christakis, 2006; Manzoli et al., 2007). Often
referred to as “death from a broken heart,” this increase in mortality is largely attributed to
psychological distress and secondary health consequences due to changes in living arrangements,
financial status, eating habits, and social networks (Stroebe et al., 2007).
The effects of spousal bereavement on physical health have generally been found to
follow a curvilinear trend, with detrimental impact increasing immediately after bereavement and
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 8

gradually declining to pre-loss levels over the course of one to two years. Heightened risks for
new illness and disability were found to peak in the first few weeks after the death event and
gradually taper off beyond that point (Buckley et al., 2010; Thompson et al., 1984). Similarly,
increase in mortality risk has been reported to be the greatest in the weeks following the death of
a spouse, and remains significantly elevated for approximately six months before declining
(Christakis & Iwashyna, 2003; Cottington et al., 1980; Hart et al., 2007; Lichtenstein et al., 1998;
Martikainen & Valkonen, 1996; Mendes de Leon et al., 1993). In terms of functional aspects of
health, bereavement had a strong negative effect on ability to perform activities of daily living
such as toileting, eating and preparing food, dressing, bathing, and mobility in very old (age 80-
84) adults at 12-18 months but dissipated by 24-30 months (d’Epinay, Cavalli, & Guillet, 2010).
A prospective study of widowed women found that 70% of participants endorsed clinically
significant distress symptoms on the General Health Questionnaire (GHQ; Goldberg, 1978) at
one month after bereavement, which tapered off over the course of 24 months for 41% while
persisting at high levels for the remaining 29% (Vachon et al., 1980).
In contrast to the abundance of literature regarding widowhood, there has been relatively
little research regarding non-spousal bereavement in mid- and late life, although these losses can
also constitute stressful life events and by extension would be expected to have a significant
negative impact on health. In recent years, the literature has called for the need to understand the
specific challenges faced by survivors of different types of loss (Lieberman, 1989). Comparisons
of spousal and non-spousal bereavement have found that the deaths of closer kin may have a
stronger emotional impact on survivors than the deaths of more distant individuals. The loss of
an immediate family member has been associated with significantly more intense distress
symptoms than the loss of a friend or extended family member (Holland & Neimeyer, 2011).
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 9

Similarly, a study of bereaved middle-aged women reported that levels of depression were
significantly higher for those who had lost a child, followed by those who had lost a spouse, then
by those who had lost a parent (Arbuckle & de Vries, 1995; Leahy, 1992; Sanders, 1979;
Middleton et al., 1998). A study of bereaved middle-aged individuals by Murrell, Himmelfarb, &
Phifer (1989) found that “attachment” losses (death of a child, spouse, or parent) were associated
with higher levels of depression than “nonattachment” losses (death of a sibling, grandchild, or
close friend). However, another study suggested that the death of a sibling is just as detrimental
as the death of a spouse: as both individuals are typically of the same generation as the bereaved
individual, their deaths often reflect the loss of one’s closest peer ties and provide the most direct
representation of a person’s own position in the “natural order” of mortality (Perkins & Harris,
1990). While relationship to the deceased has been consistently linked to differential effects on
emotional health, whether and how these findings extend to physical health remains a topic that
needs more investigation.
Bereavement in women
Studies have repeatedly demonstrated that the loss of a marital partner results in a
relatively greater increase in morbidity and mortality for men than for women, particularly
during the initial period of acute grief (Lee, Willets, & Seccombe, 1998; Matthews, 1991;
Stroebe & Stroebe, 1993; Stroebe et al., 2001). Findings regarding the impact of bereavement on
health have not been consistent for women; in fact, some studies have found no negative effect
(van den Berg et al., 2007; Moon et al., 2011; Helsing & Szklo, 1981), and still others have
shown that spousal bereavement actually confers a protective effect in the long term
(Lichtenstein, Gatz, & Berg, 1998; Mendes de Leon, Kasl, & Jacobs, 1993). Many of the
suggested explanations for the lack of detrimental impact of bereavement in women have been
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 10

related to gender differences in traditional social roles in marriage. For instance, women may be
more able to handle daily household chores and thus have less difficulty readjusting and
reordering their daily lives after bereavement (Pearlin, Menaghan, Lieberman, & Mullan, 1981).
Furthermore, since caregiving responsibilities are more likely to be put on females compared to
males, when the care recipient dies, bereavement serves to reduce burden and may actually
improve health (Schulz et al., 1995).
It is unclear, however, whether women experience similar advantages in following the
loss of a non-spouse. There is some suggestion in the literature that, due to their longer life
expectancy, women are anticipated to experience a greater number of bereavement events over
the course of their lives and thus may become better equipped to cope with such losses (Lee at el.,
1998). Furthermore, due to their tendency to have richer social networks compared to men,
women may generally be less isolated after any type of bereavement. At the same time, there are
more potentially more losses, especially as the members of their social network become older.
However, protective factors and risk factors for the physical health consequences of non-spousal
bereavement have not been studied specifically in women and still need much clarification.
The current sample focuses on middle- to older-aged women in the postmenopausal stage,
which is a period of life characterized by increased psychological and physiological symptoms.
Traditionally, the severity of such symptoms during climacterium has been attributed to the
hormonal changes of menopause. However, evidence suggests that stressful life events may have
a stronger influence on the severity of “menopausal” symptoms than the actual event of
menopause itself (Green & Cooke, 1980; Cooke & Green, 1981; Dennerstein et al, 1999). This
midlife transition is associated with greater occurrence of stressful negative life events compared
to other stages of a woman’s life, and this rise is almost exclusively attributed to interpersonal
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 11

losses and exit events, particularly the death of significant persons from one’s social network
(Cooke & Greene, 1981; Green & Cooke, 1980; Ballinger, 1985). Different types of stressful life
events were shown to have differential associations with psychological and somatic symptoms;
for example, while psychological symptoms were directly related to miscellaneous non-exit
stressors, somatic symptoms were predicted only by concomitant effects of exit stress and
miscellaneous stress (Cooke & Green, 1981). At the same time that stress has been shown to
exacerbate the mood and somatic symptoms of peri- and post-menopause (Alexander et al.,
2007), the changing hormonal environment during this life stage may in turn increase
vulnerability to stress-induced negative symptoms (Dennerstein et al., 1999; Dennerstein et al.,
2007). For example, compared to their premenopausal counterparts, postmenopausal women
experience a more intense physiological response (e.g., increased heart rate, blood pressure) to
acute stress (Seeman et al., 2001, Owens et al., 1993). Stressful life events have dynamic and
reciprocal interactions with mood and somatic symptoms during post-menopause, and this study
aims to shed light on these relationships specifically related to bereavement, the most prominent
type of exit event.
Age-Related Differences in Bereavement
Whether and how age impacts the relationship between bereavement and physical health
have also been a matter of debate in the literature. With age, individuals experience increased
numbers of bereavements, from spouses to parents to other kin and friends similar to their own
age. Older adults are likely to experience these bereavements in conjunction with other stressful
life events that occur more frequently with age (e.g. financial difficulty, relocation, cognitive
decline, etc.), resulting in the accumulation of negative effects that may interfere with their
ability to cope with the death of loved ones (Kraajj, Arensman, & Spinhoven, 2002; Moss, Moss,
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 12

& Hansson, 2001). There is some evidence that age-related declines in physical health and
functioning may increase vulnerability to stressful life events such as bereavement (Hansson &
Stroebe, 2007). It is possible that an accumulation of loss and the subsequent erosion of social
support, combined with disadvantage in baseline health, may lead to a depletion of adaptive
resources and result in “bereavement overload” (Moss, Moss, & Hansson, 2001; Parkes, 2001;
Stroebe & Schut, 2001).
However, most researchers have argued for the opposite pattern, suggesting that older age
is associated with a less intense deleterious effect of bereavement on physical health
(Martikainen & Valkonen, 1996; Schaefer et al., 1995; Lichtenstein et al. 1998). The death of a
parent, spouse, or similar-aged peer is usually an “on-time” event for older adults and is more
likely to occur peacefully and/or after a long and fulfilling life. It has been suggested that older
adults are thus more likely to have adapted to such events in their social networks and are more
prepared to deal with future deaths (Sherbourne et al., 1992; Hansson, Remondet, & Galusha,
1993; Wortman & Silver, 1992). In comparison, death at younger ages is more likely to occur
suddenly and under tragic circumstances (e.g. accident, violence), and thus may confer more
severe health consequences to the bereaved (Parkes, 2001; Stroebe & Schut, 2001). For example,
compared to older adults, younger individuals report greater need for medications or medical
care following the death of a loved one (Parkes, 2001).
The issue of age-related differences in the consequences of bereavement is complex, and
several reasons for these discrepancies in the literature have been suggested. It may be possible
that different age groups demonstrate different patterns of symptoms, such as being more
psychological than physical in nature or vice versa, peaking at different points in time, or
persisting for different durations. For example, while Sherbourne et al. (1992) reported age
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 13

differences in the impact of bereavement on emotional health, they found neither a positive or
negative impact on physical health. The study of age differences is further complicated by the
fact that such comparisons often assume within-group homogeneity that indeed does not exist.
Greater variability often exists within an age group rather than between, and heterogeneity only
increases with age. Such diversity in the bereavement experience is well illustrated in studies of
widowed older adults; while some individuals are highly resilient and adjust well to their new
circumstances, others enter a downward spiral into chronic emotional and physical distress
(Lund et al., 1993; Bonanno et al., 2004; Fry, 1998).
Context of Bereavement
Simply assessing whether an individual has been bereaved or not is insufficient, as this
method would treat all deaths as equal in weight, regardless of differences in the surrounding
circumstances such as the timing and predictability of the death, the quality of the relationship
with the deceased, and functional dependency on the deceased. For example, the death of a loved
one after a long term battle with a terminal illness may draw a drastically different response than
the sudden death of a healthy individual due to a car accident or murder. While many studies
have noted the importance of context for future research directions, few studies have actually
taken this into account (Christakis & Iwashyna, 2003; Lee & Carr, 2007). Of these limited
findings, deaths occurring under high-stress circumstances were reported to have more adverse
consequences for physical health compared to deaths occurring under relatively low-stress
circumstances. For example, widows and widowers whose spouses had been in hospice care
prior to death had a reduced risk of mortality after bereavement compared to those whose
spouses were not in hospice care, perhaps due to relatively low caregiving demands (Christakis
& Iwashyna, 2003) or that hospice care adds support prior to the death. It is possible that these
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 14

contextual differences account for the evidence that older adults are less intensely impacted by
bereavement, as losses experienced during later life stages may be associated with greater
warning and preparation time as well as higher likelihood of long-term in-home caregiving or
hospice care. On the other hand, some studies have found that context of death does not
significantly influence physical health-related outcomes (Carr et al. 2001, Carr 2003).
Specifically, there was no evidence that whether a death was anticipated or not made a difference
in in cardiovascular disease risk or mortality (Christakis & Iwashyna 2003; Buckley et al., 2010).
Other factors impacting the relationship between stressful life events and physical health
Several demographic and psychosocial variables have been reported to affect the
relationship between spousal bereavement and physical health, and each will either be
considered as a covariate or moderator in the current study in order to examine its role in non-
spousal bereavement. Regarding ethnicity, for example, there is evidence of greater risk of
spousal bereavement-related mortality in White individuals than in Black individuals (Elwert &
Christakis, 2006). There is also some evidence that widowhood confers greater risk for mental
disorders in White men than in White women while Black women experienced worse effects
compared to Black men, with such discrepancies potentially being due to cultural differences in
social support (Williams, Takeuchi, & Adair, 1992). The role of education is somewhat unclear,
with some evidence that better educated adults experience fewer physical and psychological
symptoms in general and some evidence that they experience fewer yet more severe life stressors
(Grzywacz et al., 2004). In addition, there is evidence that individuals of lower socioeconomic
status are exposed to greater amounts of stress over the life course and thus may be more
vulnerable to the harmful physical consequences of stressful life events (Lantz et al., 2005; Baum
et al., 1999; van den Berg et al., 2011).
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 15

Several commonly studied psychosocial factors have demonstrated direct effects on
physical health as well as indirect influence on the relationship between stress and health. For
example, social support has been reported to buffer the negative impact of bereavement and
other stressful life events on physical health (Cohen & Janicki-Deverts, 2009; Cohen, 2004;
Callaghan & Morrissey, 1993; Cohen & McKay, 1984). In particular, these buffering effects
have been observed for perceived availability of functional support, rather than simply the size or
density of social networks (in other words, quality over quantity). Furthermore, the high rates of
depression observed after spousal bereavement (Chen et al., 1999; Bruce et al., 1990; Clayton,
1990), in addition to any pre-existing depressive symptoms, may heighten the physical health
consequences of this major life event. Given that depression also confers independent risk for
morbidity and mortality (Ganguli et al., 2002; Unutzer et al., 2002; Schulz et al., 2000; Penninx
et al, 1999), it is likely that stressful life events and depression may have additive or even
compounding detrimental effects on physical health. Finally, models of stress and coping suggest
that the impact of a life event such as bereavement depends on perceived stress, or the cognitive
process of appraisal regarding the situation. Distress results when the perceived demands of the
situation exceed or deplete the available resources and options for coping, and can lead to
detrimental health outcomes if prolonged (Lazarus & Folkman, 1994; Folkman, 2001; Stroebe et
al., 2006).
Study Aims
This study aimed to build upon the literature on widowhood by clarifying the longitudinal
relationship between non-spousal bereavement and physical health-related quality of life
(HRQOL) in a sample of community-dwelling middle-to-older aged women. Psychosocial
functioning, stressful life events, and physical HRQOL were assessed via self-report
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 16

questionnaires every six months for a period of two years. Non-bereaved women were not
expected to experience any significant changes in physical HRQOL over two years, while, based
on literature regarding spousal bereavement, bereaved women were predicted to have a
curvilinear effect in which declines in physical HRQOL would be observed after bereavement,
persist for six months and then gradually return to baseline levels by two years (Figure 1). The
negative impact of bereavement on physical HRQOL was predicted to weaken with increasing
age (Figure 2). In addition to the primary health-related outcomes, emotional and social
indicators (depression, perceived stress, social support) were tested as secondary outcomes.
Exploratory analyses were also conducted to examine whether these psychosocial factors
(depression, perceived stress, social support) and contextual factors surrounding the bereavement
event (e.g., type/quality of relationship with the deceased, timing and predictability of the death)
affected trajectories of physical HRQOL over two years.
Methods
Participants
Participants were recruited from an existing sample of women enrolled in the Early
versus Late Intervention Trial with Estradiol (ELITE) conducted at the Atherosclerosis Research
Unit at the Keck School of Medicine at the University of Southern California. ELITE
commenced in 2005 and is a randomized, double-blind, placebo-controlled clinical trial of oral
17β-estradiol in a sample of healthy community-dwelling postmenopausal women from the
greater Los Angeles area with no clinical evidence of cardiovascular disease at baseline.
Menopause was defined as the absence of menses for twelve consecutive months, and
postmenopausal status was confirmed by measuring estradiol levels. Participants returned to the
clinic for a major clinical evaluation every six months.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 17

Participation in the psychosocial substudy was voluntary and did not affect their
involvement in the larger ELITE study. Of the 643 women who were originally enrolled in
ELITE, 131 women had either dropped out or completed the ELITE study by the time the
psychosocial substudy was commenced in January 2009. Hence the current sample was recruited
from the remaining 512 women who were still active in ELITE at that time. A total of 448
women (87.5%) agreed to provide psychosocial data, while the remaining 64 declined. The most
commonly cited reasons for nonparticipation in the remaining participants were lack of time,
length of the questionnaire packet, and discomfort with the personal nature of the questions.
Participants were categorized as either bereaved or non-bereaved for the purpose of the
current analyses. Participants were considered to be bereaved if they endorsed any of the items
indicating the death of a spouse, child, boyfriend/girlfriend, parent, sibling, other family member,
or close friend on a bereavement measure (described below). For these participants, the
psychosocial baseline was defined as the ELITE clinic visit immediately preceding the first
reported bereavement event. Those who did not report any bereavement events during the course
of their participation in the psychosocial substudy were classified as non-bereaved, and their
baseline timepoint was simply defined as the ELITE clinic visit at which they completed their
first questionnaire packet for the psychosocial substudy. It is important to note that the
proportion of widows (6.2%) at the ELITE baseline remained unchanged at the psychosocial
baseline, ensuring that the non-bereaved group did not consist of recent widows who had lost a
spouse prior to starting the psychosocial study. The proportion of widows did not differ
significantly between the bereaved and non-bereaved groups. The 41 individuals who reported
multiple bereavements at consecutive timepoints and 14 individuals who did not specify the date
of their bereavement event were excluded from analyses. The 54 participants who provided
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 18

baseline but no follow-up data were also excluded from the current analyses. See Figure 3 for a
flowchart of participant exclusion and attrition. In the current study, there were only three cases
of spousal bereavement reported, one of which was excluded from analysis due to unavailability
of follow-up data, and two of which were excluded due to missing date of bereavement. Without
a sufficient number of women who had experienced spousal loss, a systematic comparison of
spousal and non-spousal bereavement was not possible in the current study. There were no
reported cases of the loss of a child. The final sample for analyses consisted of 339 women, with
135 in the bereaved group and 204 in the non-bereaved group. Of those who reported
bereavement, 25 (18.5%) lost a parent, 10 (7.4%) lost a sibling, 42 (31.1%) lost another relative,
and 58 (43.0%) lost a close friend.
Procedure
All ELITE participants completed a series of medical and health questionnaires,
including those that assessed HRQOL and depressive symptoms, at the clinic every six months
as a standard part of their routine study visits. All clinical data, including personal and family
medical history and current list of medications, were obtained via self-report. Participants were
informed of the purpose and details of the psychosocial substudy by a research nurse during a
clinic visit for the ELITE study. Those who agreed to participate in this substudy received a
packet of questionnaires which assessed their psychosocial characteristics and life events, and
were given instructions to mail the packet back to the clinic upon completion. Data for this
substudy were collected every six months between January 2009 and December 2012, resulting
in a total of five timepoints for each participant. This study protocol was approved by the
University and Hospital Institutional Review Boards. All participants provided written informed
consent. Participants were compensated $40 per packet they returned.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 19

Measures
Age. Participant age at baseline was normally distributed and was treated as both a
continuous and categorical variable in the following analyses. The middle-aged group included
individuals aged 64 and younger, while the older-aged group included individuals aged 65 and
older.
Time elapsed since bereavement. In addition to the fixed nominal measurement times
spaced 6 months apart, the number of months elapsed since bereavement was calculated by
subtracting the date of the death event from the date on which the participant completed the
questionnaire.
Physical HRQOL. The Women’s Health Questionnaire (WHQ; see Appendix A) is a 36-
item questionnaire assessing women’s physical and emotional HRQOL (Hunter, 2000; Hunter,
2003). This scale is widely used in clinical trials, intervention studies, and epidemiological
studies as a measure of quality of life in peri- and post-menopausal women. It consists of 9
subscales assessing physical and emotional functioning, including anxiety/fears, attractiveness,
somatic symptoms, memory/concentration, vasomotor symptoms, depressed mood, sleep
problems, sexual behavior, and menstrual symptoms. Responses are made on a 4-point Likert
scale (1= yes, definitely, 2 = yes, sometimes, 3 = no, not much, 4 = no, not at all), which are then
reduced to binary format (0 = “yes, definitely” or “yes, sometimes;” 1 = “no, not much” or “no,
not at all”) per the standard scoring procedure. These subscales can be analyzed separately to
evaluate individual dimensions of quality of life (Hunter, 2003). For each subscale, the scores of
the relevant items are averaged to produce a prorated subscale score ranging from 0 to 1. The
WHQ has been shown to be sensitive to change across all subscales, where a clinically
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 20

significant change on a subscale is typically defined as 0.10 to 0.20 (Hunter, 2003). It also
demonstrates excellent concurrent validity with the General Health Questionnaire and SF36, two
commonly used validated self-report measures of health. In order to assess physical health, the
vasomotor (2 items), sleep (3 items), and somatic (8 items) dimensions were extracted for
analysis (see Appendix A for specific items). The symptoms include both menopause-related
symptoms (e.g., hot flashes) and general somatic complaints (e.g., sleeping badly, feeling tired,
feeling restless and unable to sit still). Internal consistency for each subscale has been
demonstrated to be reasonably high in populations of middle-aged women, with Cronbach’s α
of .76 for somatic symptoms, .84 for vasomotor symptoms, and .73 for sleep problems (Hunter,
2003). Internal consistency of these subscales in the current sample were comparable for somatic
and vasomotor symptoms (Cronbach’s α of .71 and .80, respectively) but lower (Cronbach’s α
of .59) for sleep problems. The overall index physical HRQOL score for all three domains
demonstrated high internal consistency (Cronbach’s α of .77).
Life Events. The Psychiatric Epidemiological Research Institute (PERI) Life Events
Scale (see Appendix B) is a self-report questionnaire used to measure the occurrence and impact
of stressful life events (Dohrenwend et al., 1978). The scale consists of 110 major life events in
various categories (work, school, finances, health/safety, social activities, crime/legal matters,
housing, family, children, and relationships/marriage) that reflect chronic stress as they generally
require a shift in lifestyle and have long-lasting consequences. Participants were asked to
indicate which events they had experienced over the previous twelve months. For each event
they endorsed, they were also asked to rate their perceived undesirability of the event on a 7-
point Likert scale, with anchor points ranging from “Not undesirable at all (1)” to “Extremely
Undesirable (7).” The PERI was used to identify participants who experienced bereavement if
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 21

they endorsed any of the items indicating the death of a spouse, child, boyfriend/girlfriend,
mother, father, brother/sister, grandparent, or close friend. Since the PERI was administered
every six months in the current study, responses at each timepoint were adjusted to exclude any
events that had occurred more than six months prior to the date on which the participant
completed the questionnaire packet.
Context of Bereavement. The PERI allows the participant to rate her perceived
undesirability of the death, but this strategy does not adequately account for contextual
differences since it can be confounded with individual differences in emotional reactivity (Turner
& Wheaton, 1995; Zimmerman, 1983). It has been suggested that a more effective approach is to
examine objective information about the individual’s life situation at the time of, and the process
leading up to, the death event. For the bereaved group, the circumstances surrounding the death
event were assessed by a series of questions (see Appendix C) regarding (a) age of the
participant at the time of the deceased individual’s death, (b) amount of warning time available
before the death (Lee & Carr, 2007; Carr et al., 2001), (c) whether the participant was present at
the time of the individual’s death, (d) the quality of the relationship with the deceased individual,
(e) degree of communication with the individual regarding death, (f) the presence of serious
ongoing health problems in the deceased individual before death, (g) whether the participant had
caregiving responsibilities, and (h) whether the deceased individual was in a nursing home.
Moderators
The following three psychosocial variables were examined as potential moderators of the
relationship between bereavement and physical HRQOL.
1. Depression – The Center for Epidemiological Studies – Depression (CES-D) scale (see
Appendix D; Radloff, 1977) is a 20-item measure that assesses clinical manifestations of
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 22

depression, including sadness, feelings of hopelessness and worthlessness, lack of interest,
interpersonal difficulties, and somatic complaints. Response options occurred on a 4-
point scale ranging from “0 = Rarely or none of the time (less than 1 day this week)” to
“3 = Most or all of the time (5 to 7 days this week).” Factor analytic studies of this
instrument have identified four relatively invariant factors: (a) depressed affect, (b)
positive affect, (c) somatic complaints, and (d) interpersonal problems. The seven items
comprising the somatic symptoms of depression (poor appetite, restless sleep, inability to
get going, talking less, everything is effortful, trouble concentrating, and being bothered
by things; see Callahan & Wolinsky, 1994) were removed from the main analyses since
their overlap with the physical symptoms assessed in this study could lead to an
overestimation of depression severity. After reverse scoring the appropriate items, a mean
score was created for the 13 items, where higher scores reflect higher depressive
symptomatology. The CES-D full scale and the somatic subscale were considered for
secondary and exploratory analyses. In the current sample, the modified 13-item scale
(Cronbach’s α = .85), the 7-item somatic factor (Cronbach’s α = .72), and the original 20-
item scale (Cronbach’s α = .87) demonstrated high internal consistency.
2. Social Support was assessed using the Medical Outcomes Study (MOS) Social Support
Survey (see Appendix E; Sherbourne & Stewart, 1992), a 19-item questionnaire which
assesses the degree to which an individual perceives the availability of functional social
support. Response values occur on a 5-point Likert scale that ranges from “None of the
time (1)” to “All of the time (5).” The overall index was calculated by averaging scores
across all items, with higher scores indicating greater levels of social support. This scale
demonstrated high internal consistency in the current sample (Cronbach’s α = .97).
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 23

3. Perceived Stress was assessed using the Perceived Stress Scale (see Appendix F; Cohen
& Williamson, 1988), a 10-item questionnaire which assesses the degree to which one
perceives aspects of one’s life as uncontrollable, unpredictable, and overloading.
Responses values occur on a 5-point Likert scale that ranges from “None of the time (1)”
to “Very Often (5).” The overall index was calculated by summing the scores across all
items after reverse-coding the positively-worded items. Higher scores indicated greater
levels of perceived stress. This scale demonstrated high internal consistency in the
current sample (Cronbach’s α = .91).
Covariates
Several demographic, psychosocial, and clinical covariates were identified a priori due to
their theoretical and empirical significance.
1. Ethnicity was evaluated with a single multiple choice item which asks participants to
indicate with which ethnic group they most closely identify. Category choices were:
White (non-Hispanic), Black (non-Hispanic), Hispanic, Asian or Pacific Islander, and
Other.
2. Marital/Partner status was evaluated with a single multiple choice item which asks
participants to indicate whether they are married, living with a partner, widowed,
divorced, separated, or single at baseline (i.e., pre-bereavement). Participants were
assigned to a binary category such that those who indicated that they were married or
living with a partner were classified as “partnered,” while all others were classified as
“unpartnered.”
3. Socioeconomic status was evaluated with a single multiple choice item which asks
participants to indicate their total family income. Response options were given in $10,000
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 24

increments between $10,000 to $100,000, with “under $10,000” and “over $100,000” at
the lower and higher ends, respectively. These categories were treated as a continuous
scale for the purpose of the current analyses.
4. Education – The participant’s highest level of education was evaluated with a single
multiple choice item. Category choices were: 8th grade or less, some high school, high
school graduate, trade or business school after high school, some college, received
bachelor’s degree, and graduate or professional education. These categories were treated
as a continuous scale for the purpose of the current analyses.
5. ELITE Randomization Status – Each ELITE participant was randomized to either the
experimental (oral 17β-estradiol estradiol) or placebo group upon enrollment into the
clinical trial.
6. Total number of negative stressful life events was represented by the total frequency of
stressful life events that the participant reported as negative in the PERI between baseline
and each follow-up. The bereavement event itself was excluded from this count. Since
some items on the PERI are ambiguous with respect to desirability (e.g. “sharply reduced
workload,” “new person moved into the household,” “change in the frequency of family
get-togethers”), a negative stressful life event was defined as any item for which the
participant gave an undesirability rating higher than “Not undesirable at all” (see PERI in
Appendix B for the full range of response options; Dohrenwend et al., 1978).
7. Total perceived undesirability of negative stressful life events at each wave was rated on
a 7-point scale where 1 = “not undesirable at all” and 7 = “extremely undesirable.” A
total score was calculated by summing the undesirability ratings across all negative
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 25

stressful life events (as defined above) endorsed by the participant, excluding the
bereavement event itself.
8. Perceived undesirability of the bereavement event was rated on a 7-point scale where 1
= “not undesirable at all” and 7 = “extremely undesirable.”
Analyses
Preliminary and Descriptive Analyses. Independent samples t-tests or chi-square tests
were used to examine demographic, psychosocial, and clinical differences between participants
and nonparticipants, bereaved and nonbereaved participants, and middle-aged and older-aged
participants.
In addition to the a priori covariates listed above, other potential variables (i.e.. number
of medical conditions, medications, BMI, weight, waist-to-hip ratio, cholesterol) were tested for
significant associations with bereavement status and physical HRQOL, using a relaxed criterion
of p<.15. Any confounders identified through this method were included in the models for
analysis and examined for significant change over the two years. If their values remained stable
across timepoints, models were adjusted for baseline levels of these covariates; however, if they
did change significantly over time, they were considered as time-varying covariates.
Modeling trajectories of physical HRQOL. Longitudinal growth modeling with the
Mplus software package (version 5.1; Muthen & Muthen, 2007) was used to investigate the
trajectories of physical HRQOL (and later, other secondary psychosocial outcomes) over two
years. The first step was to examine a change model of only physical HRQOL as repeated
measures variables with no predictors, to which subsequent models would be compared. The
change model had 14 parameters, including: seven variances (of the intercept and slope factors
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 26

and five measurement errors), five covariances (one between the factors and four between
temporally adjacent measurement errors), and two factor means. The unstandardized loadings on
the intercept factor were all fixed to 1. Loadings on the slope factor were fixed to constraints that
correspond to times of measurement, beginning with 0 for the first measurement and ending with
-4 for the last. These weights (0, 1, 2, 3, 4) specify a linear trend, with the initial level set to the
first (or baseline) timepoint. The intercept and slope factors were specified to covary, with the
covariance indicating the degree to which initial levels of physical HRQOL predict rates of linear
change. The variance of the intercept factor reflects the range of individual differences around
the initial level, and the variance of the slope factor represents the range of individual differences
in the rate of linear increase in physical HRQOL per six months over time. Curvilinear trends
were also estimated to examine whether they would explain the data better than a linear model
alone. A quadratic growth factor was added to the model by specifying that loadings on the
repeated measures indicators on this factor equal the square of the corresponding loadings on the
linear slope factor (0, 1, 4, 9, 16), and the quadratic change factor is included in the mean
structure and covaries with the intercept and linear slope factors. See Figure 4 for a visual
representation of the proposed path model.
Before the primary hypothesis of the study was tested, a baseline model adding only the
identified covariates to the change model was examined. Bereavement was then introduced as a
predictor of the mean slope and intercept factors of physical HRQOL (Model 1). The second
study hypothesis, predicting an interaction between age and bereavement, was examined by
calculating the cross product of both predictors and including it directly into the growth models
(Model 2). The interactions between bereavement and depression (Model 3), bereavement and
perceived stress (Model 4), and bereavement and social support (Model 5) were also examined.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 27

Continuous moderators were centered (by subtracting the mean value) before calculating the
interaction term in order to reduce issues of multicollinearity. Full information maximum
likelihood (FIML) estimation was used to account for incomplete data due to attrition or varying
lengths of trajectories. This approach fully utilizes all available information for each participant
to estimate each parameter.
A number of additional analyses were carried out in order more fully to characterize the
data. Because the bereavement could have occurred at any time during the 6-month window
between baseline and the first follow-up, linear regression was used to examine the cross-
sectional relationship between reported length of time since bereavement and physical HRQOL
at 6 months in order to simulate a prospective examination of changes in physical HRQOL that
occur within the first 6 months of bereavement.
As follow-up analyses, linear regression models were used to examine physical HRQOL
at 6, 12, 18, and 24 months while controlling for baseline physical HRQOL and the demographic,
clinical, and psychosocial covariates. This method allowed for identification of significant
predictors of individual trajectories of physical HRQOL, even if the main analyses using
longitudinal growth models did not detect any changes in the mean slopes of physical HRQOL
over time.
As secondary analyses, the same models used to examine physical HRQOL were
repeated with depression, perceived stress, and social support as the outcomes in order to
examine if and how bereavement affected psychosocial functioning.

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 28

Results
Descriptive Analyses
Sample Characteristics. A detailed description of the demographic, clinical, and
psychosocial characteristics of the sample is reported in Table 1. The final sample consisted of
339 women with ages ranging from 49 to 87 (M = 63.10) at baseline. Just under half (44.0%)
were married or partnered at baseline. Nearly all participants had at least a high school education,
and approximately 65% had at least a college education. The sample was ethnically diverse,
consisting of 70.5% Caucasians, 6.5% African Americans, 9.1% Hispanics, 9.4% Asian
Americans, and 3.0% others. The majority of participants was in the middle class, with
approximately 15% reporting total annual family income below $30,000 and over 30% reporting
total annual income over $100,000.
Of this sample, 135 women were classified as bereaved and 204 were classified as non-
bereaved. A comparison using independent samples t-tests or chi-square tests using a relaxed
alpha level of .10 revealed that in terms of demographic characteristics, bereaved and non-
bereaved individuals did not differ significantly in their ethnicity, education, or total annual
family income. However, the bereaved (mean age = 63.89) group was slightly older than the
non-bereaved (mean age 62.58) group, and included a slightly higher proportion of women who
were married or partnered at baseline. The non-bereaved group had a higher proportion of
women with hypertension (defined as a blood pressure >= 140/90 mmHg or taking
antihypertensive medications) at baseline. In terms of psychosocial functioning at baseline,
bereaved and non-bereaved individuals did not differ in their baseline (i.e., pre-bereavement)
CES-D scores, baseline social support, or baseline perceived stress. See Table 2 for the mean
values of physical HRQOL and key psychosocial predictors across timepoints. At baseline,
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 29

women did not differ on their mean physical HRQOL (t = .572, p =.568) or on the somatic (t
= .822, p = .412), vasomotor (t = -.305, p = .760), and sleep (t = .355, p = .723) domains. See
Table 3 for the correlations between the key predictors and physical HRQOL at each timepoint.
Attrition Analysis. There was no access to the demographic data of the 64 women who
declined to participate in psychosocial study or the 103 who were not active in ELITE at the time
of recruitment for the psychosocial study. Therefore, it could be not confirmed whether the final
sample of 339 women in the psychosocial substudy was representative of the original group of
women enrolled in ELITE.
Of the 339 participants whose data were included at baseline, 339 (100%) completed
follow-up at 6 months [as those who did not complete the first follow-up had been excluded from
the sample], 305 (89.97%) completed follow-up at 12 months, 243 (71.68%) at 18 months, and
212 (62.54%) at 24 months. In the non-bereaved group, 204 (100%) completed follow-up at 6
months, 181 (88.7%) at 12 months, 148 (72.5%) at 18 months, and 139 (68.1%) at 24 months. In
the bereaved group, 135 (100%) completed follow-up at 6 months, 124 (91.9%) at 12 months, 95
(70.4%) at 18 months, and 73 (54.1%) at 24 months. There were no differences in attrition
between bereaved and non-bereaved women at 6 months, 12 months and 18 months. However,
the bereaved group experienced a significantly greater drop-out rate than the non-bereaved group
at 24 months, χ
2
(1) = 6.858, p = .009.
Those who dropped out after 6 months did not differ from continuing participants with
regards to overall physical HRQOL, somatic HRQOL, vasomotor HRQOL, sleep HRQOL,
perceived stress, depressive symptoms, or social support at the timepoint immediately preceding
dropout (full results not shown here). Similarly, those who dropped out after 12 months did not
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 30

differ from continuing participants with regards to any of these measures at their final timepoint
preceding dropout. Finally, those who dropped out after 18 months did not differ from
continuing participants with regards to any of these measures at the timepoint immediately
preceding dropout.
Chi-square or t tests were conducted to examine differences in demographic or clinical
characteristics between the 339 participants who were included in analyses and the 109 who
dropped out after baseline, or who were excluded from analysis due to missing date of
bereavement or multiple bereavements. These results showed that these two groups did not differ
significantly in their bereavement status, age, education, ethnicity, marital status, ELITE
randomization status (estradiol vs. placebo), or their baseline medical characteristics (full results
not shown here). Under a relaxed criterion of alpha = 0.10, participants reported slightly better
baseline physical HRQOL than those who were excluded from analyses (mean difference =
0.038, t = -1.687, p = .092.) This discrepancy appeared to be driven by difference in the somatic
domain of HRQOL (mean difference = .0499, t = -1.978, p = .049.) No differences were
observed between participants and nonparticipants in their vasomotor and sleep domains of
HRQOL or depressive symptoms, perceived stress, and social support at baseline.
Preliminary Analyses
Identification of Covariates. Of the demographic covariates, Hispanic ethnicity was
associated with better vasomotor HRQOL (β = .110, t = 2.030, p = .043) at baseline. Having a
college education or higher was associated with better somatic HRQOL (β = .115, t = 2.121, p
= .035). Higher total family income was associated with better overall physical HRQOL (β
= .158, t = 2.853, p = .005) and somatic (β = .164, t = 2.974, p = .003) and sleep (β = .150,
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 31

t = 2.698, p = .007), but not vasomotor, domains at baseline. Those who were married or
partnered had better somatic HRQOL (β = .133, t = 2.190, p = .029) at baseline. These
demographic covariates were included in the baseline model.
Greater number of stressful negative life events reported at baseline was associated with
lower physical HRQOL at baseline (β = -.127, t = -2.145, p = .033). This relationship was
observed in the sleep domain (β = -.128, t = -2.158, p = .032) but not the vasomotor or somatic
domains. Furthermore, those who rated higher total undesirability of these life events reported
worse physical HRQOL (β = -.156, t = -2.567, p = .011) at baseline, particularly in the sleep
domain (β = -.164, t = -2.703, p = .007). Therefore, number of stressful life events and
undesirability of stressful life events at baseline were included as psychosocial covariates.
Participants who had been randomized to the estradiol condition in the ELITE clinical
trial had significantly higher physical HRQOL at baseline of the psychosocial substudy
compared to those who had been randomized to the placebo condition, (t(337) = 3.609, p < .001).
These difference were apparent in the vasomotor (t = 3.717, p < .001), somatic (t = 2.028, p
= .043), and sleep (t = 2.711, p = .007) domains of physical HRQOL. Therefore, ELITE clinical
trial group was included as a clinical covariate. Length of time of exposure to either
treatment/placebo did not predict physical HRQOL at the psychosocial baseline (β = -.023, t = -
.431, p = .667).
Stressful life events. Participants reported an average of 3.25 negative stressful life
events at baseline (see Table 1). There was no significant difference between the number of
stressful negative life events reported by bereaved and non-bereaved groups at baseline.
However, non-bereaved participants reported a slightly greater number of events at 6 months
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 32

(t = 1.892, p = .060), while bereaved participants reported a greater number of events at 12
months (t = -2.188, p = .029). At baseline, bereaved participants’ self-rated total undesirability of
their life events was greater than those reported by their non-bereaved counterparts (t = -2.308, p
= .023). However, the undesirability ratings made by the two groups did not differ at any of the
follow-up timepoints.
Middle-aged women (<65) reported a greater number of negative stressful life events
than older women at baseline (t = 2.932, p = .004) and at each timepoint thereafter (t = 2.572 to
2.861, p = .005 to .011), although the difference at 24 months was not statistically significant.
Furthermore, the total rating of the undesirability of the events was significantly higher in the
middle-aged group than the older-aged group at baseline (t = 3.350, p = .001) and all follow-up
timepoints thereafter (t = 2.086 to 3.396, p < .000 to .038).
Modeling Trajectory of Physical HRQOL
Longitudinal growth modeling with the Mplus software package (version 5.1; Muthen &
Muthen, 2007) was used to examine the hypothesized models, as described in the analysis
section. Before turning to the main study aims, the first step was to examine a change model of
only physical HRQOL as repeated measures variables with no predictors. The estimated linear
model showed adequate fit to the data, χ
2
(10) = 15.221, p = .124 (CFI = .994, TLI = .994,
RMSEA = .039). The average slope (-.001, p = .583) was not significantly different from zero,
indicating that mean levels of physical HRQOL in the full sample did not change across
timepoints. However, there was significant variance in the intercept (p < .001) and slope factor (p
= .017) of physical HRQOL, suggesting that there were some inter-individual differences in
starting points and rates of change that may possibly be explained by the predictors examined in
the current study. Examination of the three domains of physical HRQOL revealed that although
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 33

there was no change over time in vasomotor or sleep HRQOL, somatic HRQOL declined slightly
over time (mean slope = -.005, p = .06). This was supported by a repeated measures ANOVA
using complete cases, F(4,644) = 2.374, p = .05, as well as post hoc tests with Bonferroni
correction which suggested that this downward trend in somatic HRQOL occurred not between
consecutive timepoints, but rather, as an overall reduction between baseline and 24 months (p
= .05). A quadratic term was added to test for the possibility that a curvilinear trajectory of
physical HRQOL was contributing to a model of no overall linear change. However, this did not
offer an improvement in the fit of the model, and the means of the quadratic factor (-.003, p
= .101) and slope factor (p = .468) were not significantly different from zero. Therefore, the
linear model was used for further analyses.
We considered the possibility that the standard scoring procedure of reducing the 4-point
(1= yes, definitely, 2 = yes, sometimes, 3 = no, not much, 4 = no, not at all) response system on
the WHQ to binary (yes/no) format may have underestimated changes in overall physical
HRQOL over time by limiting the range of actual reported symptoms. However, when the same
change model was examined using the mean of non-binary responses, there still appeared to be
no significant changes in mean physical HRQOL across the two years (full statistical results not
shown). Hence, the original WHQ mean scores produced from the binary responses was used for
all analyses, per standard WHQ scoring procedures.
Main analyses. When only the demographic, psychosocial and clinical covariates were
entered as predictors, the model showed adequate fit to the data, χ
2
(31) = 34.620, p = .299 (CFI
= .993, TLI = .989, RMSEA = .024). Being married or partnered predicted greater rate of decline
in physical HRQOL across timepoints (mean slope = -.004, p = .012). ELITE randomization
status also predicted mean slope of physical HRQOL (0.012, p = .025), such that participants
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 34

who received estradiol had slower rate of decline than those who received placebo over two
years.
First, bereavement status was entered into the model in order to examine whether it
predicted differential change in physical HRQOL over time (Model 1). See Table 4 for a
summary of the intercept and slope factor estimates for each model. The linear growth model of
overall physical HRQOL with bereavement group and covariates as predictors (Model 1) fit
adequately with the data, χ
2
(34) = 37.785, p=.300 (CFI = .992, TLI = .989, RMSEA = .024).
Bereavement predicted neither the mean intercept (.025, p = .241) nor mean slope (.000, p
= .807) of physical HRQOL across timepoints. See Figure 5 for the trajectories of physical
HRQOL in the bereaved and non-bereaved groups. Independent sample t-tests yielded consistent
results, as the mean physical HRQOL of bereaved and non-bereaved groups did not differ
significantly at baseline (t = .572, p = .568), 6 months (t= .658, p = .511), 12 months (t = .330, p
= .742), 18 months (t = .507, p = .612), or 24 months (t = .591, p = .555).
Short –term effects of bereavement. In order to investigate whether the frequency of
measurement (i.e. every 6 months) had failed to capture more short-term effects of bereavement,
the cross-sectional associations between the length of time elapsed since bereavement and
physical HRQOL at six months were examined in the bereaved group as a simulation of
longitudinal relationships within the first six months of bereavement. These regression analyses
showed that the number of months since the bereavement event (β = -.047, t = -.527, p = .599)
was not a significant predictor of physical HRQOL within the first six months after bereavement
(Figure 6). While the data appeared to suggest that women who had been assessed either less
than 1 month after bereavement or more than 3 months after bereavement had poorer physical
HRQOL compared to others, ANOVA with follow-up comparisons between pairs of timepoints
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 35

did not show significant effects. Moreover, the magnitude of these differences fell slightly short
of what would be regarded as a clinically significant change on this measure.
Age moderation. When age was added as a main predictor in the linear growth model of
physical HRQOL (Model 2a), the model continued to show adequate fit to the data, χ
2
(37) =
41.377, p = .285 (CFI = .991, TLI = .987, RMSEA = .024). Age was not a significant predictor
of the mean intercept (.002, p = .296) of physical HRQOL. However, when individual domains
were considered separately, older age was associated with better baseline vasomotor HRQOL,
perhaps reflecting the fact that older women tend to be farther removed from the hormone-
mediated somatic symptoms of menopause. Age did not predict the mean rate of change (mean
slope = .000, p = .714) of physical HRQOL across two years. The interaction between age and
bereavement (Model 2b) did not predict mean slope of physical HRQOL (.000, p = .665) over
two years (see Figure 7). Although there appeared to be a trend such that bereaved women
scored worse than non-bereaved women in the middle-aged group while nonbereaved women
scored worse than bereaved women in the older aged group, none of the differences reached the
magnitude to be called clinically significant.
Other psychosocial moderators. Separate linear growth models were used to examine
the respective roles of depression, perceived stress, and social support as predictors of physical
HRQOL in addition to bereavement, age, and covariates.
Bereavement x Depression. Linear growth modeling showed that greater baseline
depressive symptomatology was associated with lower physical HRQOL at baseline (mean
intercept = -.135, p < .001) and was a marginally significant predictor of changes in mean
physical HRQOL (mean slope = .013, p = .057) over two years (Model 3a). Follow-up analyses
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 36

with linear regression to examine predictors of individual trajectories of physical HRQOL
suggested that higher levels of baseline depressive symptomatology predicted lower physical
HRQOL at 6, 12, 18, and 24 months (β = -.232 to -.151, t = -3.458 to -2.355, p = .019 to .001)
while controlling for baseline physical HRQOL and covariates. There was an interaction
between baseline CES-D scores and bereavement in predicting mean slope of physical HRQOL
(.025, p = .045) over two years (Model 3b). When moderation was tested in regression analyses
to examine whether it predicted individual trajectories of physical HRQOL, there was an
interaction between baseline CES-D and bereavement over 12 months only (β = .165, t = 2.518,
p = .012), such that depression affected physical HRQOL less negatively for bereaved
individuals compared to non-bereaved individuals (see Figure 8). This interaction was observed
for total physical HRQOL and in the somatic domain only, and there was no such interaction at 6,
18, or 24 months. Change in CES-D scores did not predict change in physical HRQOL over two
years, and did not moderate the effects of bereavement on physical HRQOL trajectories.
Bereavement x Social Support. Greater social support at baseline was associated with
higher physical HRQOL at baseline (mean intercept = .028, p = .012) but did not predict changes
in mean physical HRQOL (mean slope = .001, p = .643) over time (Model 4a). Baseline social
support did not moderate the relationship between bereavement and physical HRQOL
trajectories (mean slope = .004, p = .690) over two years (Model 4b); see Figure 9. Increase in
social support predicted increase in physical HRQOL over 6 months (β = .124, t = 2.365, p
= .019), 12 months (β = .123, t = 2.187, p = .030), and 18 months (β = .142, t = 2.059, p = .041),
but the effect faded by 24 months (β = .010, t =.120, p = .905). With respect to moderation, there
was no interaction between bereavement and changes in social support in predicting trajectories
of physical HRQOL across timepoints.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 37

Bereavement x Perceived Stress. There was a main effect of perceived stress such that
greater baseline stress was associated with lower baseline physical HRQOL (mean intercept = -
.008, p < .000) and steeper declines in physical HRQOL (mean slope = -.001, p < .001) over two
years (Model 5a). Follow-up analyses using linear regression confirmed that higher levels of
perceived stress at baseline predicted lower physical HRQOL at 6, 12, 18, and 24 months (β = -
.221 to -.160, t = -4.107 to -2.834, p < .000 to p =.005) while controlling for baseline physical
HRQOL. These patterns were observed overall and in the somatic and sleep domains,
specifically. However, baseline perceived stress did not moderate the influence of bereavement
on physical HRQOL (Model 5b) over time (mean slope = .001, p = .128; see Figure 10). Change
in perceived stress did not predict change in physical HRQOL over 6, 12, 18, or 24 months, and
did not moderate the relationship between bereavement and physical HRQOL trajectories. In
other words, the relationship between perceived stress and physical HRQOL was no different for
bereaved and non-bereaved.
When baseline values of perceived stress, depression, and social support were included in
the linear regression model together with bereavement and age while controlling for covariates
and baseline HRQOL, baseline perceived stress was the only variable that remained a significant
predictor of physical HRQOL at 6 months (β = -.162, t = -3.135, p = .002), 12 months (β = -.158,
t = -2.968, p = .003), 18 months (β = -.245, t = -3.790, p < .000), and 24 months (β = -.139, t = -
2.109, p = .036). Baseline depressive symptomatology and perceived social support were no
longer statistically significant in these models.

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 38

Types of Bereavement
Regression analyses indicated that the deceased individual’s relationship to the
participant, coded as family versus friend, did not significantly predict changes in physical
HRQOL (β = .005 to .098, t = .075 to 1.307, p = .194 to .941) over 6, 12, or 18 months. However,
losing a friend compared to losing a family member was associated with faster increase in
physical HRQOL over 24 months (β = .226, t =2.459, p = .016). This association appeared to be
driven by effects in the somatic domain, where those who had lost a friend experienced faster
increase in QOL compared to those who had lost a family member at 24 months (β = .238, t =
2.610, p = .011), but not at 6, 12, or 18 months (β = -.012 to .105, t = -.174 to 1.299, p = .197
to .862). There were no differences between participants who lost a family member vs. those
who lost a friend with regards to their trajectories of vasomotor and sleep HRQOL over 6, 12, 18,
and 24 months. Of the key psychosocial predictors, depressive symptomatology of those who
had lost a friend decreased more rapidly over 24 months compared to those who had lost a
family member (β = -.271, t = -2.602, p = .011), but not over 6, 12, or 18 months. Rates of
change in perceived stress and social support over 6, 12, 18, or 24 months did not differ
significantly between these two types of bereavement. Exploratory analyses of the group of
participants who had lost a family member revealed no differences in changes in physical
HRQOL, depression, perceived stress, or social support between those who had lost a parent,
sibling, or relative.
The perceived undesirability of the death event (rated on a scale from 0 = not undesirable
at all to 7 = extremely undesirable) did not predict changes in physical HRQOL, depression,
social support, or perceived stress over 6, 12, 18, or 24 months, controlling for baseline levels.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 39

Context of Bereavement
Further data concerning the contextual factors surrounding the bereavement event could
only be collected for 42 cases, as the Context of Bereavement questionnaire (Appendix C) was a
later addition to the study. A preliminary look at these limited data suggested that the age of the
deceased individual at the time of the death did not predict physical HRQOL at 6 months.
However, the older the age of the deceased individual, the lower the depressive symptomatology
(β = -.523, t = -2.858, p = .008) and perceived stress (β = -.623, t = -3.419, p = .002) of the
participant after 6 months. Furthermore, the deaths of those with ongoing serious health concerns
predicted better overall physical HRQOL in the bereaved participant after 6 months as compared
to deaths of healthy individuals (β = .604, t = 3.011, p = .005). These effects could be seen in the
somatic (β = .486, t = 2.268, p = .031) and sleep (β = .640, t = 3.085, p = .004) domains but not
in the vasomotor domain. Ongoing serious health concerns in the deceased individual also
predicted lower depressive symptomatology (β = -.492, t = -2.344, p = .026) in the bereaved
participant after 6 months. However, the amount of warning time available before the death, the
participant’s presence at the death, personal caregiving duties for the deceased individual, and
the deceased individual’s residence in a nursing home did not predict physical HRQOL,
perceived stress, or depression. Similarly, neither the quality of the relationship with the
deceased individual nor the amount of conversation with individual to plan for life without him
or her appeared to matter. Social support at 6 months was not predicted by any of these
contextual factors surrounding the bereavement event. Due to the limited data available at the
time of analysis, there was insufficient power to examine whether these contextual factors of
bereavement predict longitudinal trajectories of physical HRQOL and its domains over two years.

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 40

Secondary Psychosocial Outcomes
Given the significant correlation of depression, social support, and perceived stress with
physical HRQOL over time, secondary analyses examined these three psychosocial variables as
alternative outcomes, and tested whether their trajectories over time differed with bereavement
status, and whether there was age moderation.
Depression. When considering depression as an outcome variable, the full CES-D scale
(including the somatic symptoms that were previously removed) was used for analysis, since
there were no issues of confounding overlap between symptoms that were problematic when
examining depression as a predictor of physical HRQOL. A linear growth model of depression
including only the covariates as predictors showed adequate fit to the data, χ
2
(31) = 38.37, p
= .17 (RMSEA = .027, CFI = .986, TLI = .979). Greater total annual family income (mean
intercept = -.015, p = .014) and higher total undesirability of life events (mean intercept = .008, p
= .018) significantly predicted mean intercept of CES-D.
A linear growth model with bereavement added as a predictor of depression showed
adequate fit to the data, χ
2
(34) = 39.04, p=.25 (RMSEA = .02, CFI = .990, TLI = .985).
Bereavement did not predict mean intercept (.048, p = .204) or mean slope (.004, p = .714) of
CES-D scores over two years. See Figure 11 for the trajectories of CES-D scores in the bereaved
and nonbereaved groups. An examination of the cross-sectional association between the time
since bereavement and CES-D scores at 6 months, which simulated a short-term longitudinal
design of effects within the first 6 months after bereavement, revealed that greater number of
months since bereavement was not a significant predictor of depressive symptomatology (β
= .085, t = .948, p = .345) at six months.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 41

Model fit improved when age was added as a predictor, χ
2
(37) = 39.54, p=.36 (RMSEA
= .015, CFI = .995, TLI = .993). However, age also did not predict mean intercept (-.004, p
= .146) or mean slope (.000, p = .617) of CES-D scores. There was no interaction between
bereavement and age in predicting CES-D score trajectories (mean slope = .001, p = .752).
Social Support. When the covariates alone were examined as predictors in the linear
growth model of social support, not surprisingly, being married or partnered was associated with
higher levels of perceived social support at baseline (mean intercept = .446, p < .001) and higher
mean slope of change in perceived social support trajectories over time (mean slope = .048, p
= .029). Total income was positively associated with the mean intercept (.040, p = .008) of social
support. Higher education was a significant predictor of more rapid increase in social support
over two years (mean slope = .028, p = .017).
When bereavement was entered into the model, it did not significantly predict the mean
intercept (.059, p = .500) or mean slope (.008, p = .676) of social support over two years. See
Figure 12 for the trajectories of social support in the bereaved and nonbereaved groups. When
trajectories of social support within 6 months was simulated by examining cross-sectional
associations between the time since bereavement and social support, the number of months since
the bereavement event was not a significant predictor of social support (β = -.042, t = -.765, p
= .446) at 6 months while controlling for baseline social support
Older age was associated with higher social support at baseline (mean intercept = -.013, p
= .033) but did not predict the mean rate of change in social support across time (mean slope = -
.001, p = .405). Age did not moderate the effects of bereavement on social support (mean slope =
-.001, p = .822).
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 42

Perceived Stress. When covariates alone were entered as predictors, the estimated linear
growth model of perceived stress was a good fit to the data, χ
2
(31) = 36.65, p = .22 (RMSEA
= .024, CFI = .993, TLI = .990). Higher total rating of undesirability of stressful life events was
associated with greater perceived stress at baseline (mean intercept = .210, p < .000) but did not
predict the mean slope of perceived stress over two years.
A linear growth model with bereavement group included as a predictor of perceived
stress fit adequately with the data, χ
2
(34) = 41.057, p = .19 (RMSEA = .025, CFI = .991, TLI
= .987). Bereavement did not significantly predict mean intercept (-.424, p = .543) or mean slope
(-.196, p = .294) of perceived stress over two years. See Figure 13 for the trajectories of
perceived stress scores in the bereaved and non-bereaved groups. When longitudinal analyses
within the first 6 months were simulated by examining cross-sectional associations between the
time since bereavement and perceived stress, the number of months since the bereavement event
was not a significant predictor of perceived stress (β = .084, t = .944, p = .347) at 6 months.
When age was added, the model continued to show adequate fit to the data, χ
2
(37) =
43.334, p = .22 (RMSEA = .023, CFI = .992, TLI = .989). Age was a significant predictor of
mean intercept (-.177, p < .001), suggesting that older participants had lower perceived stress at
baseline. When age was examined as a categorical variable, older participants continued to report
significantly lower perceived stress than middle-aged adults at each follow-up timepoint (t =
2.421 to 4.086, p < .001 to p = .016) until differences dissipated after 18 months. Age was a
significant predictor of the mean slope (.029, p = .019) of perceived stress over time. Stress
levels declined over time for middle-aged women but increased over time for older women. A
repeated measures ANOVA confirmed significant differences in mean perceived stress across
time in the middle-aged (< 64) group, F(4, 260) = 3.875, p = .004, with Bonferroni post-hoc tests
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 43

revealing significant declines between baseline and 12 months (p = .003) and between 6 months
and 12 months (p = .013). Similarly, repeated measures ANOVA for the older-aged (>65) group
confirmed that mean perceived stress differed significantly across timepoints, F(4,176) = 2.904,
p = .023, with significant increase between baseline and 24 months (p = .043). However, age did
not moderate the relationship between bereavement and perceived stress over time.
Posthoc Analyses
Several possible reasons for failure to support the hypotheses were considered, in order
either to explain the findings or to eliminate alternative explanations.
Secondary Health Outcome. One possibility is that the measure of physical HRQOL
may not have captured the physical health consequences commonly experienced after
bereavement. The WHQ was a tool specifically developed to assess perceptions of physical and
emotional health in peri- and postmenopausal women, and accordingly, many of the items relate
to experiences secondary to changes in levels of estrogen. Selecting the measure was supported
by the overlap of menopausal and stress symptoms (e.g., flushing) and the evidence in the
literature that stress may exacerbate the somatic symptoms of menopause and that stressful life
events may interact with hormonal changes to create vulnerability (Alexander et al., 2007;
Dennerstein et al., 1999; Dennerstein et al., 2007). However, all of the findings regarding
relationships between psychosocial predictors and physical HRQOL occurred in the somatic and
sleep domains, not the vasomotor domain. Therefore we tested whether bereavement affected
another measure of general physical symptoms commonly associated with the experience of
stress: the somatic factor of CES-D. The somatic factor consisted of the seven CES-D items that

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 44

had previously been removed from analysis in order to minimize confounding overlap between
predictor and outcome when examining the relationship between depression and physical health.
When only the covariates were entered as predictors, a linear growth model of somatic
depression showed close fit to the data, χ
2
(31) = 27.259, p = .66 (RMSEA = .000, CFI = 1.000,
TLI = 1.011). Higher total income was associated with fewer somatic symptoms at baseline
(mean intercept = -.020, p = .010), and higher education (mean slope = .015, p = .05) predicted
faster rates of improvement in somatic depression symptoms over two years. Higher total
undesirability of stressful life events at baseline predicted higher levels of somatic depression
(mean intercept = .010, p = .013).
Overall, findings regarding somatic depression were similar to those obtained with the
primary physical HRQOL outcome. Bereavement did not predict mean intercept (.018, p=.712)
or mean slope (.005, p = .714) of somatic CES-D scores over two years. Follow-up regression
analyses in the bereaved group showed that the cross-sectional association between the number
of months since bereavement and somatic depressive symptomatology at six months was not
significant (β = .078, t = .870, p = .386), suggesting that notable changes in somatic symptoms
were unlikely to have occurred within the first six months of bereavement. There was no
interaction between bereavement and age in predicting somatic CES-D score trajectories (mean
slope = -.001, p = .695). There was no main effect of age in predicting the mean intercept (-.002,
p = .574) or mean slope (-.001, p = .338) of somatic CES-D scores and no interaction of age and
bereavement.
Length of ELITE Treatment. As described in the Methods, women in this study were
part of a clinical trial. Psychosocial baseline for the bereaved group was defined as the timepoint
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 45

at which a death event was first reported, while for the non-bereaved group it was defined as the
first data collection timepoint. Due to this study design, the bereaved group as a whole had
experienced greater and varied lengths of exposure to either estradiol or placebo than the non-
bereaved group. In the ELITE trial, it was hypothesized that estradiol exposure would have
beneficial effects on physical HRQOL. Therefore it was necessary to confirm that the findings of
the study were not an artifact of the length of treatment in the clinical trial on the part of the
bereaved women. When the bereaved group was examined separately, length of exposure to
estradiol/placebo, calculated as the number of months that a participant had been active in
ELITE following randomization, did not significantly predict baseline physical HRQOL (β = -
.009, t = -.099, p = .921) or changes in physical HRQOL over 6, 12, 18, or 24 months. Similar
patterns were observed for other psychosocial outcomes, as neither baseline nor trajectories of
depression or perceived stress were predicted by the length of one’s exposure to the randomized
treatment.
Stressful Life Events. Next we considered whether there may have been qualitative
differences in concurrent stressful life events between bereaved and non-bereaved participants
that contributed to the lack of difference in the physical HRQOL trajectories between the two
groups. The frequencies of each item on the PERI stressful life event scale were examined in
order to identify the top ten most commonly reported negative events in the sample at baseline
(see Table 5). As mentioned above, an event was defined as negative if the participant rated the
undesirability of the event as greater than 1 (not undesirable at all). While threats to health of self
and loved ones and financial loss comprised the most common life events across the two groups,
the top ten list of non-bereaved participants consisted of a greater number of work-related
negative events compared to that of bereaved participants.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 46

Given that the bereaved group was slightly older than the non-bereaved group, these
differences may have been partially related to age and subsequent life stage. Indeed, similar
patterns of qualitative differences were found for middle-aged (<64) and older aged (>65)
participants, such that middle-aged participants experienced a greater number of negative life
events related to work/career while older adults, who were presumably less likely to be actively
working, appeared to have experienced a greater number of negative life events related to
family/household. Some studies have reported that work-related problems have more detrimental
effects on health than any other type of stressor such as family, relationship, and financial issues
(Karasek, Gardell, & Lindell, 1987) or may further exacerbate the effects of non-work stressors
(Klitzman et al., 1990; Greenhaus & Parasuraman, 1987). Despite the bereaved group rating
greater undesirability of life events at baseline compared to the non-bereaved group, this
qualitative imbalance of life events may have reduced any differences in health-related outcomes
between the two groups.
Discussion

This study examined the longitudinal relationships between non-spousal bereavement and
physical HRQOL over a period of two years in a sample of 135 women who reported having
been bereaved and 204 women who reported no bereavement. Non-bereaved women were not
expected to experience any significant changes in physical HRQOL over two years, while
bereaved women were hypothesized to have a curvilinear effect in which declines in physical
HRQOL would be observed immediately after bereavement and would gradually lessen over
time, in line with similar trends that have been reported in the widowhood literature regarding
new illness and disability (Buckley et al., 2010; Thompson et al., 1984) and mortality risk
(Christakis & Iwashyna, 2003; Cottington et al., 1980; Hart et al., 2007; Lichtenstein et al., 1998;
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 47

Martikainen & Valkonen, 1996; Mendes de Leon et al., 1993) after bereavement. Differences in
these patterns by age were also examined in order to clarify conflicting evidence in the literature
as to whether older age is associated with a more intense (Hansson & Stroebe, 2007; Kraajj,
Arensman, & Spinhoven, 2002; Moss, Moss, & Hansson, 2001; Parkes, 2001; Stroebe & Schut,
2001) or less intense (Martikainen & Valkonen, 1996; Schaefer et al., 1995; Lichtenstein et al.
1998) deleterious effect of bereavement on physical health.
Contrary to expectations, mean physical HRQOL did not change significantly across
timepoints for either group of women. In contrast to what was hypothesized, bereavement did not
predict changes in physical HRQOL, either when mean and individual trajectories were
examined. When the three domains of physical HRQOL were examined separately, somatic
HRQOL declined slightly over time in the sample as a whole, but not differentially for bereaved
and non-bereaved women. Vasomotor and sleep HRQOL did not change over time.
Differences between bereaved and non-bereaved women were also evaluated on several
psychosocial variables. Bereavement was not a significant predictor of changes in depressive
symptomatology, contrary to much prior literature (Stroebe, Schut, & Stroebe, 2007; Murrell,
Himmelfarb, & Phifer, 1989; Lee et al., 1998; Kraaj et al., 2002; Carr et al., 2001). However,
there was an interaction between bereavement and baseline CES-D scores such that depression
affected physical HRQOL less negatively for bereaved individuals compared to non-bereaved
individuals. This may suggest that pre-existing depressive symptoms have a more powerful
influence on women’s vulnerability to detrimental functional health outcomes than the
occurrence of a single non-spousal bereavement event. However, these effects only occurred at
12 and 18 months but not at 6 or 24 months, suggesting possible loss or power or Type I error
due to the large number of models and variables examined in the current analyses. Bereavement
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 48

was not a significant predictor of changes in perceived stress or changes in social support over
two years. There was no interaction between age of the participant and bereavement on any of
the physical or psychosocial outcomes.
Exploratory analyses suggested that perhaps physical HRQOL outcomes depend not on
simply whether bereavement occurred or not, but rather, the contextual factors surrounding the
death event. Over the course of 24 months, participants who lost a friend experienced faster
improvement in physical HRQOL (particularly in the somatic domain) and depressive
symptomatology compared to those who lost a family member. This is consistent with findings
in the literature suggesting that the death of those closer in kin confer greater detrimental effects
on emotional and physical health (Holland & Neimeyer, 2011). Due to the greater length of the
relationship and greater likelihood of dependency and shared lifestyle with a family member,
adjustment to life without the individual may pose greater challenges than adjusting to life
without a friend. When the category of family members was broken down further, there were no
differences in physical HRQOL, depression, or perceived stress among participants who lost a
parent, sibling, or relative. Older age of the deceased individual was associated with lower levels
of perceived stress and depression at 6 months, perhaps supporting the theory that death at older
age is more likely to be an “on-time” event or less tragic in nature (Sherbourne et al., 1992;
Hansson, Remondet, & Galusha, 1993; Wortman & Silver, 1992; Parkes, 2001; Stroebe & Schut,
2001). The presence of ongoing serious health concerns in the deceased individual prior to death
predicted higher physical HRQOL and lower depression in the participant at 6 months compared
to bereavement cases in which the deceased individual was in relatively good health. These
effects appeared to be independent of the warning time, amount of preparation, or caregiving
duties for the deceased individual prior to the death event. However, these findings regarding the
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 49

differential effects of varying scenarios of the death event are preliminary and must be
interpreted with caution, as data regarding the contextual factors of bereavement were provided
by only approximately 30% of the bereaved sample.
In the current group of participants, therefore, results do not support the hypothesis that
bereavement is an exceptional stressor with significant effects on physical HRQOL. Notably, all
of the bereavement in the current study was non-spousal. There were only three cases of spousal
bereavement reported in the current sample, all of which were excluded from analysis due to
unavailability of follow-up data, either because of attrition or because the death was reported at
the final wave of data collection. Therefore we could not form a group of bereaved spouses to
compare to findings in the widowhood literature, nor could we compare the functioning of
women who had lost a spouse to those experiencing non-spousal bereavement.
Furthermore, there were no cases of child death reported during the entire course of the
psychosocial study. Studies have consistently shown that the intensity of grief reaction and
depressive symptomatology are highest in individuals who lost a child, primarily due to
disruption in one’s beliefs about the “natural order” of mortality (Arbuckle & de Vries, 1995;
Leahy, 1992; Middleton et al., 1998). Consequently, the distribution of the types of bereavement
reported in the current study was devoid of the subgroup of individuals who may have been most
vulnerable to the detrimental consequences of bereavement on quality of life.
Clinicians and researchers have toyed with the implications of normal grief versus
complex bereavement disorder, the latter of which is listed in the DSM-5 as a condition for
further study and not yet official for clinical use. The vast majority (80-90%) of bereaved
individuals experience normal or uncomplicated grief reactions (Prigerson, 2004). However,
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 50

there may be subtypes of individuals who experience severe grief reactions with a specific and
cohesive cluster of symptoms independent of depression and anxiety that persists for at least 12
months after the death event. This pathological grief response may interfere with the individual’s
functioning and health-related quality of life in both emotional and physical domains. Although
this study did not have specific measures assessing for complicated grief, it seems likely that
most if not all of the women were experiencing uncomplicated grief.
Reexamining study design
Given that the findings were inconsistent with those reported in the existing bereavement
literature, a number of alternative explanations were considered, including measurement and
sample selection.
Measurement considerations. Some studies have suggested that there may be some
variation in the temporal impact of stressful life events on health-related outcomes. For example,
Sherbourne (1992) reported that occurrence of bereavement was not related to physical
functioning except at the two-year follow-up, at which point those who had lost somebody close
had significantly worse physical functioning than nonbereaved individuals. It is possible that
there was a delayed impact of bereavement on overall physical HRQOL in the current sample
which would have only started to become visible after the timeframe examined in the current
study, although the 24 month results provide minimal suggestion that this was the case.
Alternatively, it may be that some effects of bereavement on physical health, such as
impairments in immune functioning (Bartrop et al. 1977; Schleifer et al., 1983; Beem et al.,
1999), are acute and transient and may not have translated to tangible changes in physical
HRQOL at data collection timepoints spaced 6 months. Short-term effects evaluated by the
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 51

cross-sectional association between the number of months since bereavement and 6-month
physical HRQOL scores showed that bereavement did not significantly predict patterns of
physical HRQOL within the first 6 months after bereavement. Possibly even shorter time
intervals than one month are needed.
Furthermore, self-report questionnaires as a method of measuring negative life events
such as bereavement are subject to several concerns. For example, there is potential for recall
bias for life events, which has been demonstrated to increase with age and aging-related
cognitive and physical impairments (Dex, 1995; Schwarz et al., 1999). There was no way to
check whether participants reported inaccurate dates of bereavement, forgot to report
bereavement events, or deliberately chose not to report bereavement due to feelings of denial or
avoidance. Individuals with lower HRQOL may be more likely to report an increased number of
stressful life events compared to those with higher HRQOL. However, the self-report
questionnaire remains a gold standard in life event research, and has the advantages of brevity
and ease of administration than an in-depth standardized interview, which may be more sensitive
but still not immune to bias.
Finally, this study was not able to assess for the possible effect of bereavement overload,
i.e. multiple bereavements reported at the same timepoint and/or across consecutive timepoints.
One contributing issue is that due to the yes/no format of a life event checklist (i.e. “Check if this
event happened”), the PERI does not have the capacity to fully account for multiple occurrences
of a single stressful life event. As a result, the PERI may underestimate the cumulative stress
load in an individual’s life during a six month interval. While the deaths of different types of
individuals (e.g. friend, spouse, child, other family member) can be specified since they comprise
different items on the PERI, the loss of several close friends within the same time period, for
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 52

example, may not have been captured. Furthermore, the number of women who reported
bereavements at more than one timepoint was too small for analysis, as well as underestimating
the actual number with multiple bereavements. It would be important for future studies to
compare the health outcomes of individuals who have experienced varying numbers of
bereavements, and to examine possible differences in the effects of the accumulation of
bereavement on health outcomes over time.
Sample Considerations. By design, the present study is comprised entirely of women.
Yet, some previous work has found bereavement to have a less detrimental effect on women than
on men (Lee, Willets, & Seccombe, 1998; Matthews, 1991; Stroebe & Stroebe, 1993), or in
some cases, no effect (van den Berg et al., 2007; Moon et al., 2011; Helsing & Szklo, 1981) or a
positive effect on women in this age group (Lichtenstein, Gatz, & Berg, 1998; Mendes de Leon,
Kasl, & Jacobs, 1993).
The study sample only included postmenopausal women who committed to participate in
a longitudinal clinical trial and may not have been representative of the broader population of
middle-aged and older women. The perimenopausal transition is a stressful life event that is
marked with hormone instability and physiological changes, such as increased abdominal and
total body fat distribution (Franklin et al., 2009; Lovejoy, 2009; Lovejoy et al., 2008). Such
changes may challenge self-identity, which could potentially impact an individual’s perception
of health functioning, negatively influencing physical HRQOL for all of the women in the study.
Indeed, the literature shows that women who are in the midst of or have completed menopause
have lower quality of life than those who still conserve regular menstrual cycles (Blumel, 2000;
Williams, 2009). The literature reports that estradiol replacement therapy predicts better overall
physical HRQOL in postmenopausal women (Limouzin-Lamothe et al., 1994). The current study
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 53

found better physical HRQOL at the baseline for the psychosocial substudy in women in one arm
of the study compared to the other. This factor was included as a covariate. We also tested
whether length of exposure to estradiol or placebo affected physical HRQOL in bereaved women,
and found no effect.
Furthermore, this study was a sample of convenience comprised of healthy community-
dwelling individuals, and excluded institutionalized older adults or those with chronic illnesses
who may be more vulnerable to the health consequences of bereavement. Any examination of
age differences may have been limited by the fact that the sample did not consist of any
participants in the oldest old (age 85+) category. This fact may have also limited the rates of
spousal bereavement in the current sample. Participants were recruited from the ELITE study
(described above), and thus were individuals who had met the stringent criteria for enrollment in
a clinical trial. Thus the current sample may have been representative of a particularly high-
functioning or resilient population, and it is possible that bereavement, as a one-time event, may
not have conferred obvious negative consequences on their perceived health.
Selective attrition can also be problematic in longitudinal studies in that individuals who
are less healthy and have fewer protective factors (e.g. social and financial resources) may be
more likely to drop out. In the current sample, participants reported better baseline physical
HRQOL than those who did not provide follow-up data. Furthermore, the bereaved group
experienced a significantly greater drop-out rate than the nonbereaved group by the two-year
follow-up, raising the possibility of biases that may have underestimated the detrimental
consequences of bereavement on physical HRQOL, although attrition analyses did not suggest
differences between dropouts and those who continued to complete questionnaires.

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 54

Predictors of physical HRQOL
Despite limited change in mean slopes of physical HRQOL in the sample over 24
months, there appeared to be significant variability in the individual trajectories. This allowed for
analyses to identify specific predictors of differential changes in physical HRQOL across
timepoints.
Higher levels of baseline depressive symptomatology predicted lower baseline physical
HRQOL and greater declines in physical HRQOL over two years. These findings suggest that
baseline emotional dysregulation, independent of bereavement, is an important influence on
physical HRQOL. By using a CES-D that excluded somatic symptoms (e.g., lack of appetite,
fatigue) that may overlap and confound with the items on the WHQ, this study provided a more
precise evaluation of the relationships between psychosocial factors and physical HRQOL. For
the most part, changes in depressive symptomatology did not predict changes in physical
HRQOL, suggesting that perhaps depression serves as a pre-existing vulnerability to future
detrimental functional health outcomes rather than acting through a parallel process of change.
Higher social support at baseline was associated with better physical HRQOL at baseline,
but did not predict increases in physical HRQOL over two years. However, increase in social
support across timepoints was associated with increase in physical HRQOL, particularly in the
somatic domain. Although there were no differences in the rate at which the physical HRQOL of
bereaved and nonbereaved groups responded to increased social support, for the bereaved group,
this association may have at least partially reflected the beneficial effects of the outpouring of
support that one often receives from others following the death of a loved one. This is consistent
with the findings in a study of heart failure patients (Bennett et al., 2001) in which increases in
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 55

social support, perhaps due to the circumstances of severe chronic illness, significantly predicted
improvements in HRQOL, while baseline social support did not.
Older age was associated with better physical HRQOL at baseline, but in general did not
predict changes in physical HRQOL across timepoints. Age also did not predict changes in social
support or depressive symptoms over time. However, older age was associated with lower
perceived stress at baseline and follow-up timepoints in the current sample. This may be related
to indications in the literature that older adults experience fewer and different types of stressful
life events than those who are younger. For example, the stresses of older age may be more
chronic, less likely to dissipate with time, and more likely to involve threats and losses compared
to younger individuals (Sherbourne, 1992). In the current sample, while financial difficulties and
health problems in self and loved ones appeared to be shared concerns between both age groups,
middle-aged participants appeared to have experienced a greater number of negative life events
related to work/career while older adults appeared to have experienced a greater number of
negative life events related to family/household. Some studies have reported that work-related
problems have more detrimental effects on health than any other type of stressor such as family,
relationship, and financial issues (Karasek, Gardell, & Lindell, 1987) or may further exacerbate
the effects of non-work stressors (Klitzman et al., 1990; Greenhaus & Parasuraman, 1987). In the
current sample, older (≥ age 65) participants reported fewer negative stressful life events and
lower total ratings of undesirability of these events than younger (< age 65) participants at each
timepoint. Higher perceived stress at baseline was associated with lower physical HRQOL at
baseline and steeper declines in physical HRQOL over two years. However, the simple number
of negative stressful life events did not predict change in perceived stress; rather, what mattered
was the self-rated total undesirability of all such events. Taken together, these findings suggest
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 56

that cumulative appraisal of stress, rather than a single bereavement event, was most influential
on physical HRQOL.
Conclusion & Future Directions

In summary, this study examined the impact of non-spousal bereavement and related
psychosocial factors on physical HRQOL outcomes using a relatively large sample of
community-dwelling postmenopausal women. Contrary to expectations, bereavement did not
directly predict changes in physical HRQOL. Within the bereaved group, women who had lost a
family member rather than a friend showed more negative effects on somatic HRQOL and on
non-somatic depressive symptoms. Perceived stress appeared to be a key factor that may explain
the link between stressful life events and declines in physical HRQOL. As bereavement occurs
against a backdrop of concurrent stressful life events, pre-existing emotional dysregulation, and
social circumstances, the baseline vulnerability of an individual may be more predictive of
fluctuations in HRQOL than the death event itself.
It would be important for future studies to examine contextual factors surrounding not
only the death event itself, but the pre-bereavement and post-bereavement processes as well. For
example, the trajectory of physical HRQOL in the months prior to bereavement may provide
information about declines that occurred in anticipation of a loved one’s death or due to
increased caregiving responsibilities, if applicable. Similarly, the many ways in which
individuals demonstrate grieving after a loss (e.g., finding meaning in the loss, focusing on
positive vs. negative memories and conversations) may have differential impacts on physical
well-being. The contextual risk factors that increase certain individuals’ vulnerability to the
detrimental effects of bereavement on HRQOL needs continued clarification.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 57

References
Aalto, A., Uutela, A., Aro, A.R. (1997). Health related quality of life among insulin-dependent
diabetics: disease-related and psychosocial correlates. Patient Education and Counseling, 30,
215-225.
Alexander, J.L., Dennerstein, L., Woods, N.F., McEwen, B.S., Halbreich, U., Kotz, K.,
Richardson, G. (2007). Role of stressful life events and menopausal stage in wellbeing and
health. Expert Review of Neurotherapeutics, 7(11), S93-S113.
Arbuckle, N.W., deVries, B. (1995). The long-term effects of later life spousal and parental
bereavement on personal functioning. The Gerontologist, 35(5), 637-647.
Ballinger, S.E. (1985). Psychosocial stress and symptoms of menopause: a comparative study of
menopause clinic patients and non-patients. Maturitas, 7, 315-327.
Bartrop, R.W., Lazarus, L., Luckhurst, E., Kiloh, L.G., Penny, R. (1977). Depressed lymphocyte
function after bereavement. The Lancet, 309(8016), 834-836.
Baum, A., Garofalo, J.P., Yali, A.M. (1999). Socioeconomic status and chronic stress: does
stress account for SES effects on health? Annals of New York Academy of Sciences, 896, 131-
144.
Baum, A., & Grunberg, N.E. (1991). Gender, stress, and health. Health Psychology, 10, 80-85.
Beem, E.E., Hooijkaas, H., Cleiren, M., Schut, H., Garssen, B., Croon, M.A., Jabaaij, L.,
Goodkin, K., Wind, H., de Vries, M.J. (1999). The immunological and psyclogical effects of
bereavement: does grief counseling really make a difference? A pilot study. Psychiatry
Research, 85, 81-93.
Bennett, S.J., Perkins, S.M., Lane, K.A., Deer, M., Brater, D.C., Murray, M.D. (2001). Social
support and health-related quality of life in chronic heart failure patients. Quality of Life
Research, 10, 671-682.
Blumel, J.E., Castelo-Branco, C., Binfa, L., Gramegna, G., Tacla, X., Aracena, B., Cumsille,
M.A., Sanjuan, A. (2000). Quality of life after the menopause: a population study. Maturitas,
34(1), 17-23.
Bonanno, G.A., Wortman, C.B., Nesse, R.M. (2004). Prospective patterns of resilience and
maladjustment during widowhood. Psychology and Aging, 19(2), 260-271.
Boyle, P.J., Feng, Z., Raab, G. (2011). Does widowhood increase mortality risk? Testing for
selection effects by comparing causes of spousal death. Epidemiology, 22, 1-5.
Bradbeer, M., Helme, R.D., Yong, H.H., Kendig, H.L., Gibson, S.J. (2003). Widowhood and
other demographic associations of pain in independent older people. Clin J Pain, 19, 247-54.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 58

Bruce, M.L., Kim, K., Leaf, P.J., Jacob, S. (1990). Depressive episodes and dysphoria resulting
from conjugal bereavement in a prospective community sample. American Journal of
Psychiatry, 147, 608-611.
Bryk, A., Raudenbush, S. (1992). Hierarchical linear models. Newbury Park, California: Sage
Publications.
Buckley, T., McKinley, S., Tofler, G., Bartrop, R. (2010). Cardiovascular risk in early
bereavement: a literature review and proposed mechanisms. International Journal of Nursing
Studies, 47, 229-238.
Callaghan, P., Morrissey, J. (1993). Social support and health: a review. Journal of Advanced
Nursing, 18(2), 203-210.
Callahan, C.M., Wolinsky, F.D. (1994). The effect of gender and race on the measurement
properties of the CES-D in older adults. Med Care, 32(4), 341-356.
Carnelley, K.B., Wortman, C.B., Bolger, N., Burke, C.T. (2006). The time course of grief
reactions to spousal loss: evidence from a national probability sample. Journal of Personality
and Social Psychology, 91(3), 476-492.
Carr, D., House, J.S., Wortman, C., Nesse, R., Kessler, R.C. (2001). Psychological adjustment to
sudden and anticipated spousal loss among older widowed persons. Journal of Gerontology:
Social Sciences, 56B(4), S237-S248.
Carr, D. (2003). A “good death” for whom? Quality of spouse’s death and psychological distress
among older widowed persons. Journal of Health and Social Behavior, 44, 215-232.
Carroll, D., Phillips, A.C., Ring, C., et al. (2005). Life events and hemodynamic stress reactivity
in the middle-aged and elderly. Psychophysiology, 42, 269-276.
Carstensen, L.L., Pasupathi, M., Mayr, U., Nesselroade, J.R. (2000). Emotional experience in
everyday life across the adult life span. Journal of Personality and Social Psychology, 79(4),
644-655.
Chen, J.H., Bierhals, A.J., Prigerson, H.G., Kasl, S.V., Mazure, C.M., Jacobs, S. (1999). Gender
differences in the effects of bereavement-related psychological distress in health outcomes.
Psychol Med, 29, 367-80.
Christakis, N.A., Iwashyna, T.J. (2003). The health impact of health care on families: a matched
cohort study of hospice use by decedents and mortality outcomes in surviving widowed
spouses. Social Science and Medicine, 57(3), 465-475.
Clayton, P.J. (1990). Bereavement and depression. Journal of Clinical Psychiatry, 51, 34-38.
Cohen, S. (2004). Social relationships and health. American Psychologist, 59, 676-684.
Cohen, S., Janicki-Deverts. (2009). Can we improve our physical health by altering our social
networks? Perspectives on Psychological Science, 4(4), 375-378.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 59

Cohen, S., McKay, G. (1984). Social support, stress, and the buffering hypothesis: a theoretical
analysis. In A. Baum, S.E. Taylor, & J.E. Singer (Eds.), Handbook of Psychology and Health.
Hillsdale, NJ: Erlbaum.
Cohen, S., & Williamson, G. (1988). Perceived stress in a probability sample of the United
States. In S. Spacapan & S. Oskamp (Eds.), The social psychology of health: Claremont
Symposium on Applied Social Psychology (pp. 31-67). Newbury Park, CA: Sage.
Cooke, D.J., Greene, J.G. (1981). Types of life events in relation to symptom at the climacterium.
Journal of Psychosomatic Research, 25(1), 5-11.
Cottington, E.M., Matthews, K.A., Talbott, E., Kuller, L.H. (1980). Environmental events
preceding sudden death in women. Psychosomatic Medicine, 42(6), 567-574.
Dennerstein, L., Lehert, P., Burger, H., Dudley, E. (1999). Mood and the menopausal transition.
Journal of Nervous Mental Disease, 187(11), 685-691.
Dennerstein, L., Lehert, P., Guthrie, J., Burger, H. (2007). Modeling women’s health during the
menopausal transition: a longitudinal analysis. Menopause, 14(1), 53-62.
D’Epinay, C.J.L., Cavalli, S., Guillet, L.A. (2010). Bereavement in very old age: impact on
health and relationships of the loss of a spouse, a child, a sibling, or a close friend. OMEGA,
60(4), 201-325.
Dex, S. (1995). Reliability of recall data: a literature review. Bulletin de Methodologie
Sociologique, 49, 58-89.
Dohrenwend, B.S., Krasnoff, L., Askenasy, A.R., et al. (1978). Exemplification of a method for
scaling life events: the PERI Life Events Scale. J Health Soc Behav,19, 205–229.
Elwert, F., Christakis, N.A. (2008). Wives and ex-wives: a new test for homogamy bias in the
widowhood effect. Demography, 45, 851-873.
Elwert, F., Christakis, N.A. (2006). Widowhood and race. Am Sociol Rev, 71, 16-41.
Espinosa, J., Evans, W.N. (2008). Heightened mortality after the death of a spouse: marriage
protection of marriage selection? Journal of Health Economics, 27, 1326-1342.
Faul, F., Erdfelder, E., Lang, A.G., Buchner, A. (2007). G*Power 3: a flexible statistical power
analysis program for the social, behavioral, and biomedical sciences. Behavior Research
Methods, 39, 175-191.
Folkman, S. (2001). Revised coping theory and the process of bereavement. In M. Stroebe, R.O.,
Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences,
coping and care (pp. 563-584). Washington, DC: American Psychological Association.
Franklin, R.M., Ploutz-Snyder, L., Kanaley, J.A. (2009). Longitudinal changes in abdominal fat
distribution with menopause. Metabolism Clinical and Experimental, 58, 311-315.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 60

Fry, P.S. (1998). Spousal loss in late life: a 1-year follow-up of perceived changes in life
meaning and psychosocial functioning following bereavement. Journal of Personal and
Interpersonal Loss, 3, 369-391.
Ganguli, M., Dodge, H.H., Mulsant, B.H. (2002). Rates and predictors of mortality in an aging,
rural, community-based cohort: the role of depression. Arch Gen Psychiatry, 59, 1046-1052.
Glaser, R. (2005). Stress-associated immune dysregulation and its importance for human health:
a personal history of psychoneuroimmunology. Brain, Behavior, and Immunity, 19, 3-11.
Goldberg, D. (1978). Manual of the general health questionnaire. Windsor, England: NFER
Publishing.
Greene, J.G., Cooke, D.J. (1980). Life stress and symptoms at the climacterium. British Journal
of Psychiatry, 136, 486-491.
Greenhaus, J.H., Parasuraman, S. (1987). A work-nonwork interactive perspective of stress and
its consequences. Journal of Organizational Behavior Management, 8(2), 37-60.
Grzywacz, J.G., Almeida, D.M., Neupert, S.D., Ettner, S.L. (2004). Socioeconomic status and
health: a micro-level analysis of exposure and vulnerability to daily stressors. Journal of
Health and Social Behavior, 45(1), 1-16.
Hansson, R.O., Remondet, J.H., Galusha, M. (1993). Handbook of bereavement: theory,
research, and intervention. New York, NY: Cambridge University Press.
Hansson, R.O., Stroebe, M.S. (2007). Bereavement in late life: coping, adaptation, and
developmental influences. Washington, DC: American Psychological Association.
Hart, C.L., Hole, D.J., Lawlor, D.A., Smith, G.D., Lever, T.F. (2007). Effect of conjugal
bereavement on mortality of the bereaved spouse in participants of the Renfrew/Paisley
Study. Journal of Epidemiology and Community Health, 61(5), 455-460.
Helsing, K.J., Szklo, M. (1981). Mortality after bereavement. Am J Epidemiol, 114, 41-52.
Holland, J.M., Neimeyer, R.A. (2011). Separation and traumatic distress in prolonged grief: the
role of causes of death and relationship to the deceased. J Psychopathol Behav Assess, 33,
254-263.
Holmes, T.J., Rahe, R.H. (1967). The social readjustment rating scale. Journal of Psychosomatic
Research, 11, 213-218.
Hunter, M.S. (2000). The Women’s Health Questionnaire (WHQ): The development,
standardization and application of a measure of mid-aged women’s emotional and physical
health. Quality of Life Research, 9, 733-738.
Hunter, M.S. (2003). The Women’s Health Questionnaire: Frequently Asked Questions (FAQ).
Health Qual Life Outcomes, 1, 41.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 61

Karasek, R., Gardell, B., Lindell, J. (1987). Work and non-work correlates of illness and
behavior in male and female Swedish white collar workers. Journal of Occupational
Behavior, 8, 187-207.
Kessler, R.C., McLeod, J.D. (1984). Sex differences in vulnerability to undesirable life events.
American Sociological Review, 49(5), 620-631.
Klitzman, S., House, J.S., Israel, B.A., Mero, R.P. (1990). Work stress, nonwork stress, and
health. Journal of Behavioral Medicine, 13(3), 221-243.
Kraajj, V.E., Arensman, E., Spinhoven, P. (2002). Negative life events and depression in elderly
persons: a meta-analysis. Journal of Gerontology: Medical Sciences, 57A, M87-M94.
Lantz, P.M., House, J.S., Mero, R.P., Williams, D.R. (2005). Stress, life events, and
socioeconomic disparities in health: results from the Americans’ Changing Lives Study.
Journal of Health and Social Behavior, 46, 274-288.
Lazarus, R.S., Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.
Leahy, J.M. (1992). A comparison of depression in women bereaved of a spouse, child or a
parent. OMEGA: Journal of Death and Dying, 26(3), 207-217.
Lee, M., Carr, D. (2007). Does the context of spousal loss affect the physical functioning of older
widowed persons? A longitudinal analysis. Research on Aging, 29(5), 457-487.
Lee, G.R., Willets, M.C., Seccombe, K. (1998). Widowhood and depression: gender differences.
Research on Aging, 20, 611-630.
Li, J., Precht, D.H., Mortensen, P.B, Olsen, J. (2003). Mortality in parents after death of a child
in Denmark: a nationwide follow-up study. The Lancet, 361, 363-367.
Lichtenstein, P., Gatz, M., Berg, S. (1998). A twin study of mortality after spousal bereavement.
Psychol Med, 28, 635-643.
Lieberman, M.A. (1989). All family losses are not equal. Journal of Family Psychology, 2, 368-
372.
Limouzin-Lamothe, M.A., Mairon, N., Joyce, C.R.B., Le Gal, M. (1994). Quality of life after
the menopause: influence of hormonal replacement therapy. American Journal of Obstetrics
and Gynecology, 170(2), 618-624.
Littlefield, C.H., Rushton, J.P. (1986). When a child dies: the sociobiology of bereavement.
Journal of Personality and Social Psychology, 51(4), 797-802.
Lovejoy, J.C., Champagne, C., DeJonge, L., Xie, H., Smith, S.R. (2008). Decreased physical
activity and increased visceral fat during the menopausal transition. Int J Obes, 32, 949-958.
Lovejoy, J.C. (2009). Weight gain in women at midlife: the influence of menopause. Obesity
Management, 5(20), 52-56.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 62

Lund, D.A., Caserta, M.S., & Dimond, M.F. (1993). The course of spousal bereavement in later
life. In M.S. Stroebe, W. Stroebe, & R.O. Hansson (Eds.), Handbook of bereavement: Theory,
research, and intervention (pp.240-254). Cambridge, England: Cambridge University Press.
Manzoli, L., Villari, P., Pirone, G.M., Boccia, A. (2007). Marital status and mortality in the
elderly: a systematic review and meta-analysis. Soc Sci Med, 64, 77-94.
Martikainen, P., Valkonen, R. (1996). Mortality after the death of a spouse: rates and causes of
death in a large Finnish cohort. Am J Pub Health, 86, 1087-93.
Matthews, A.M. (1991). Widowhood in late life. Toronto, Canada: Buttersworth.
Mendes de Leon, C., Kasl, S.V, Jacobs, S. (1993). Widowhood and mortality risk in a
community sample of the elderly: a prospective study. J Clin Epidemiol, 46, 519-527.
Middleton, W., Raphael, B., Martinek, N., Burnett, P. (1998). A longitudinal study comparing
bereavement phenomena in recently bereaved spouses, adult children, and parents.
Australian and New Zealand Journal of Psychiatry, 32, 235.
Miles, M. S. (1985). Emotional symptoms and physical health in a bereaved parents. Nursing
Research, 34(2), 76-81.
Moon, J.R., Kondo, N., Glymour, M.M., Subramanian, S.V. (2011). Widowhood and mortality:
a meta-analysis. PLoS ONE, 6(8), e23465.
Moss, M. S., Moss, S.Z. (1983). The impact of parental death on middle aged children. OMEGA,
14, 65-75.
Moss, M.S., Moss, S.Z., Hansson, R.O. (2001). Handbook of bereavement research:
consequences, coping, and care. Washington , DC: American Psychological Association.
Murrell, S.A., Himmelfarb, S., Phifer, J.F. (1989). Effects of attachment bereavement and pre-
event conditions on subsequent depressive symptoms in older adults. Psychology and Aging,
4(2), 166-172.
Murrell, S.A., Himmelfarb, S., Phifer, J.F. (1988). Effects of bereavement/loss and pre-event
status on subsequent physical health in older adults. International Journal of Aging and
Human Development, 27(2), 89-107.
Ott, C.H., Lueger, R.J. (2002). Patterns of change in mental health status during the first two
years of spousal bereavement. Death Studies, 26, 387-411.
Owens, J.F., Stoney, C.M., Matthews, K.A. (1993). Menopausal status influences ambulatory
blood pressure levels and blood pressure changes in during mental stress. Circulation, 88(6),
2794-2802.
Parkes, C.M. (2001). Bereavement: studies of grief in adult life (3
rd
ed.). New York:
Routledge/Taylor & Francis.
Pearlin, L.I., Menaghan, E.G., Lieberman, M.A., & Mullan, J.T. (1981). The stress process.
Journal of Health and Social Behavior, 22, 337-356.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 63

Penninx, B.W., Geerlings, S.W., Deeg, D.J., et al. (1999). Minor and major depression and the
risk of death in older persons. Arch Gen Psychiatry, 56, 889-895.
Perkins, H.W., Harris, L.B. (1990). Familial bereavement and health in adult life course
perspective. Journal of Marriage and Family, 52(1), 233-241.
Phillips, A.C., Carroll, D., Ring, C., et al. (2005). Life events and acute cardiovascular reactions
to mental stress: a cohort study. Psychosomatic Medicine, 67, 384-392.
Prigerson, H.G. (2004). Complicated grief: when the path of adjustment leads to a dead-end.
Bereavement Care, 23, 38-40.
Radloff, L.S. (1977). The CES-D scale: a self-report depression scale for research in the general
population. Appl Psychol Meas, 1, 385-401.
Rozanski, A., Blumenthal, J.A., Davidson, K.W., et al. (2005). The epidemiology,
pathophysiology and management of psychosocial risk factors in cardiac practice. Journal of
the American College of Cardiology, 45, 637-651.
Sanders, C.M. (1979). A comparison of adult bereavement in the death of a spouse, child, and
parent. OMEGA – Journal of Death and Dying, 10(4), 303-322.
Schaefer, C., Quesenberry, C.P., Soora, W. (1995). Mortality following conjugal bereavement
and the effects of a shared environment. Am J Epidemiol, 141, 1142-1152.
Schleifer, S.J., Keller, S.E., Camerino, M., Thornton, J.C., Stein, M. (1983). Suppression of
lymphocyte stimulation following bereavement. JAMA, 250(3), 374-377.
Schulz, R., O’Brien, A.T., Bookwala, J., Fleissner, K. (1995). Psychiatric and physical morbidity
effects of dementia caregiving: prevalence, correlates, and causes. The Gerontologist, 35,
771-91.
Schulz, R., Beach, S.R., Ives, D.G., et al. (2000). Association between depression and mortality
in older adults: the Cardiovascular Health Study. Arch Intern Med, 160, 1761-1768.
Schwarz, N., Park, D., Knauper, B., Sudman, S. (Eds.). (1999). Cognition, aging, and self-
reports. Philadelphia: Psychology Press.
Seeman, T.E., Singer, B., Wilkinson, C.W., McEwen, B. (2001). Gender differences in age-
related changes in HPA axis reactivity. Psychoneuroendocrinology, 26(3), 225-240.
Shahar, D., Schultz, R., Shahar, A., Wing, R. (2001). The effect of widowhood on weight change,
dietary intake, and eating behavior in the elderly population. J Aging Health, 13, 186-99.
Sherbourne, C.D., Meredith, L.S., Rogers, W., Ware, J.E. (1992). Social support and stressful
life events: age differences in their effects on health-related quality of life among the
chronically ill. Quality of Life Research, 1, 235-246.
Sherbourne, C.D., Stewart, A.L. (1992) The MOS social support survey. Soc Sci Med, 32(6),
705-714.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 64

Stroebe, M.S., Schut, H. (2001). Meaning making in the dual process model. In R. Neimeyer
(Ed.), Meaning reconstruction and the experience of loss (pp. 55-73). Washington, DC:
American Psychological Association.
Stroebe, M.S., Stroebe, W. (1993). The mortality of bereavement. In Handbook of Bereavement:
Theory, Research, and Intervention (ed. M.S. Stroebe, W. Stroebe and R.O. Hansson), pp.
175-195. Cambridge University Press: New York.
Stroebe, M.S., Schut, H., Stroebe, W. (2007). Health outcomes of bereavement. Lancet, 370,
1960-1973.
Stroebe, M.S., Folkman, S., Hansson, R.O., Schut, H. (2006). The prediction of bereavement
outcome: development of an integrative risk factor framework. Social Science & Medicine,
63, 2440-2451.
Stroebe, M.S., Hansson, R.O., Schut, H., Stroebe, W. (2008). Handbook of bereavement
research and practice: advances in theory and intervention. (1
st
Ed). Washington, D.C.:
American Psychological Association.
Stroebe, M.S., Stroebe, W., Schut, H. (2001). Gender differences in adjustment to bereavement:
an empirical and theoretical review. Review of General Psychology, 5(1), 62-83.
Sutin, A.R., Costa, P.T., Wethington, E., Eaton, W. (2010). Perceptions of stressful life events as
turning points are associated with self-rated health and psychological distress. Anxiety, Stress,
and Coping, 23(5), 479-492.
Thompson, L.W., Breckenridge, J.N., Gallagher, D., Peterson, J.A. (1984). Effects of
bereavement on self-perceptions of physical health in elderly widows and widowers. J
Gerontol, 39, 309-14.
Tosevski, D. L., Milovancevic, M. P. (2006). Stressful life events and physical health. Current
Opinion in Psychiatry, 19(2), 184-189.
Turner, R.J., Wheaton, B. (1995). Checklist measurement of stressful life events. In S. Cohen,
R.C. Kessler, L.U. Gordon (Eds.), Measuring stress: a guide for health and social scientists
(pp. 29-58). New York, NY: Oxford University Press.
Unutzer, J., Patrick, D.L., Marmon, T., et al. (2002). Depressive symptoms and mortality in a
prospective study of 2558 older adults. Am J Geriatr Psychiatry, 10, 521-530.
Vachon, M.L., Lyall, W.A., Rogers, J., et al. (1980). A controlled study of self-help intervention
for widows. Am J Psychiatry, 137, 1380-1384.
Van den Berg, G.J., Lindeboom, M., Lundborg, P. (2008). Bereavement, health, and mortality
among the elderly. Manuscript.
Van den Berg, G.J., Lindeboom, M., Portrait, F. (2011). Conjugal bereavement effects on health
and mortality at advanced ages. Journal of Health Economics, 30, 774-794.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 65

Wexler, D.J., Grant, R.W., Wittenberg, E., Bosch, J.L., Cagliero, E., Delahanty, L, Blais, M.A.,
Meigs, J.B. (2006). Correlates of health-related quality of life in type 2 diabetes.
Diabetologia, 49(7), 1489-1497.
Williams, R.E., Levine, K.B., Kalilani, L., Lewis, J., Clark, R.V. (2009). Menopause-specific
questionnaire assessment in US population-based study shows negative impact on health-
related quality of life. Maturitas, 62(2), 153-159.
Williams, D.R., Takeuchi, D.T., & Adair, R.K. (1992). Marital status and psychiatric disorders
among Blacks and Whites. Journal of Health and Social Behavior, 33, 140-157.
Wilson, I.B., Cleary, PD. (1995). Linking clinical variables with health-related quality of life: a
conceptual model of patient outcomes. JAMA, 273(1), 59-65.
Wortman, C.B., Silver, R.C.(1992). Reconsidering assumptions about coping with loss: An
overview of current research. In L. Montada, S.H. Filipp, & M.J. Lerner (Eds.), Life crises
and experiences of loss in adulthood (pp. 341-365). Hillsdale, NJ: Erlbaum.
Zimmerman, M. (1983). Methodological issues in the assessment of life events: a review of
issues and research. Clinical Psychology Review, 3(3), 339-370.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 66

Table 1. Baseline demographic, clinical, and psychosocial characteristics of participants,
expressed as mean (standard deviation) or frequency (%)

Participant characteristic Full sample
(N=339)
Bereaved
(N=135)
Nonbereaved
(N=204)
t test or Χ
2
test of
difference (bereaved vs.
nonbereaved)
Demographic
Age (years) 63.10 (6.90) 63.89 (6.84) 62.58 (6.92) t = -1.716, p = .087
Ethnicity χ
2
(4) = 4.263, p = .372
Caucasian
African American
Hispanic
Asian American
Other
239 (70.5)
22 (6.5)
36 (9.1)
32 (9.4)
10 (3.0)
95 (70.3)
12 (8.9)
14 (10.4)
9 (6.7)
5 (3.7)
144 (70.5)
10 (4.9)
22 (10.8)
23 (11.3)
5 (2.5)

Marital Status
Married/Partnered
Separated/Divorced
Widowed
Single
149 (44.0)
116 (34.2)
21 (6.2)
53 (15.7)
67 (49.6)
40 (29.5)
9 (6.7)
19 (14.2)
82 (40.2)
76 (37.2)
12 (5.9)
34 (16.7)
χ
2
(3) = 3.455, p = .327
Education χ
2
(3) = 3.589, p = .309
< High school
High school
College
>College
3 (1.0)
113 (33.3)
90 (26.5)
133 (39.2)
1 (0.7)
51 (37.8)
38 (28.2)
45 (33.3)
2 (1.0)
62 (30.4)
52 (25.5)
88 (42.1)

Family Income χ
2
(10) = 14.819, p = .139
< $10,000
$10,000 to $19,999
$20,000 to $29,999
$30,000 to $39,999
$40,000 to $49,999
$50,000 to $59,999
$60,000 to $69,999
$70,000 to $79,999
$80,000 to $89,999
$90,000 to $99,999
> $100,000
10 (3.1)
21 (6.6)
16 (5.0)
23 (7.2)
24 (7.5)
33 (10.3)
18 (5.6)
21 (6.6)
22 (6.9)
25 (7.8)
107 (33.4)
5 (3.9)
10 (7.8)
7 (5.5)
16 (12.5)
12 (9.4)
14 (10.9)
6 (4.7)
8 (6.3)
8 (6.3)
9 (7.0)
33 (25.8)
5 (2.6)
11 (5.8)
9 (4.7)
6 (3.1)
12 (6.3)
19 (9.9)
12 (6.3)
13 (6.8)
14 (7.3)
16 (8.4)
74 (38.7)

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 67

Table 1, continued

Participant Characteristic Total Sample
(N=339)
Bereaved
(N=135)
Non-bereaved
(N=204)
t test or Χ
2
test of
difference (bereaved vs.
nonbereaved)
Clinical
Randomized ELITE treatment χ
2
(1) = 2.210, p = .137
Estradiol
Placebo
170 (50.1)
169 (49.9)
61 (45.1)
74 (54.8)
109 (53.4)
95 (46.6)

Hypertension 88 (26.0) 25 (18.5) 63 (30.9) χ
2
(1) = 6.461, p = .011
Family hx of CHD 53 (15.6) 21 (15.6) 32 (15.7) χ
2
(1) = .001, p = .974
Current cigarette smoker 10 (2.9) 4 (3.0) 6 (2.9) χ
2
(1) = .000, p = .991
Cancer 19 (5.6) 7 (5.2) 12 (5.9) χ
2
(1) = .075, p = .785
Thyroid disease 49 (14.5) 22 (16.3) 27 (13.2) χ
2
(1) = 3.770, p = .152
Arthritis 83 (24.5) 38 (28.1) 45 (22.1) χ
2
(1) = .1.711, p = .425
Psychosocial
Depression (CES-D) 1.271 (.398) 1.299 (.482) 1.251 (.331) t = -1.006, p = .316
Perceived stress 20.16 (6.29) 20.29 (6.69) 20.09 (6.08) t = -.264, p = .792
Social support 4.083 (.785) 4.029 (.826) 4.110 (.764) t = .856, p = .393
Number of stressful negative
life events
3.25 (2.33) 3.47 (2.91) 3.13 (1.97) t = -1.175, p = .241
Total perceived
undesirability of stressful
negative life events
13.44 (13.43) 16.41 (17.21) 11.87 (10.67) t = -2.308. p = .023
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 68

Table 2. Means (SD) of physical HRQOL domains and other key psychosocial variables.
TOTAL
(N=339)
BEREAVED
(N=135)
NON-
BEREAVED
(N=204)
AGE < 65
(N=215)
AGE ≥ 65
(N=124)
Physical WHQ
Baseline
6 months
12 months
18 months
24 months
.801 (.176)
.804 (.168)
.806 (.176)
.809 (.171)
.793 (.185)
.794 (.175)
.796 (.176)
.802 (.173)
.802 (.157)
.782 (.174)
.805 (.177)
.808 (.162)
.809 (.178)
.813 (.180)
.798 (.190)
.796 (.179)
.805 (.165)
.807 (.179)
.806 (.177)
.793 (.190)
.810 (.171)
.802 (.172)
.805 (.171)
.810 (.163)
.793 (.177)
Somatic WHQ .
Baseline
6 months
12 months
18 months
24 months
.821 (.196)
.824 (.183)
.820 (.187)
.820 (.191)
.803 (.203)
.810 (.195)
.816 (.193)
.814 (.183)
.814 (.182)
.789 (.207)
.828 (.197)
.829 (.175)
.824 (.190)
.824 (.198)
.809 (.201)
.817 (.202)
.837 (.174)
.824 (.190)
.819 (.199)
.810 (.201)
.827 (.185)
.801 (.195)
.812 (.182)
.821 (.180)
.790 (.207)
Vasomotor WHQ
Baseline
6 months
12 months
18 months
24 months
.816 (.323)
.804 (.353)
.838 (.307)
.833 (.318)
.821 (.346)
.822 (.320)
.800 (.357)
.827 (.344)
.811 (.328)
.801 (.370)
.811 (.325)
.806 (.352)
.845 (.281)
.848 (.312)
.831 (.333)
.798 (.335)
.788 (.363)
.808 (.331)
.829 (.312)
.778 (.371)
.847 (.300)
.831 (.337)
.887 (.257)
.840 (.330)
.892 (.285)
Sleep WHQ
Baseline
6 months
12 months
18 months
24 months
.744 (.314)
.750 (.310)
.751 (.322)
.764 (.308)
.767 (.431)
.736 (.314)
.744 (.318)
.758 (.305)
.769 (.284)
.808 (.596)
.749 (.314)
.755 (.305)
.746 (.334)
.761 (.323)
.746 (.312)
.745 (.313)
.730 (.325)
.762 (.308)
.768 (.314)
.757 (.321)
.742 (.315)
.785 (.280)
.733 (.345)
.759 (.299)
.785 (.573)
Depression (CES-D)
Baseline
6 months
12 months
18 months
24 months
1.271 (.398)
1.279 (.411)
1.268 (.420)
1.258 (.396)
1.276 (.430)
1.299 (.482)
1.297 (.447)
1.293 (.428)
1.279 (.407)
1.288 (.374)
1.251 (.331)
1.268 (.386)
1.251 (.415)
1.245 (.389)
1.270 (.458)
1.284 (.384)
1.290 (.416)
1.286 (.466)
1.281 (.427)
1.320 (.490)
1.248 (.423)
1.260 (.403)
1.240 (.329)
1.222 (.340)
1.201 (.290)
Social support
Baseline
6 months
12 months
18 months
24 months
4.083 (.785)
4.081 (.861)
4.039 (.888)
4.055 (.867)
4.093 (.871)
4.029 (.826)
4.045 (.846)
4.023 (.903)
4.018 (.859)
4.144 (.878)
4.110 (.764)
4.108 (.874)
4.053 (.878)
4.080 (.875)
4.057 (.870)
4.099 (.790)
4.055 (.890)
3.999 (.922)
4.017 (.853)
4.071 (.869)
4.057 (.779)
4.122 (.816)
4.116 (.828)
4.115 (.892)
4.131 (.882)
(Continued on next page)
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 69

TOTAL
(N=339)
BEREAVED
(N=135)
NON-
BEREAVED
(N=204)
AGE < 65
(N=215)
AGE ≥ 65
(N=124)

Perceived Stress
Baseline
6 months
12 months
18 months
24 months
20.16 (6.29)
20.10 (6.24)
19.69 (6.83)
19.68 (6.52)
20.01 (6.72)
20.29 (6.69)
20.47 (6.29)
19.95 (6.85)
19.81 (6.68)
19.41 (5.46)
20.09 (6.08)
19.83 (6.20)
19.47 (6.83)
19.58 (6.43)
20.46 (7.52)
21.27 (6.47)
20.95 (6.37)
20.62 (7.22)
20.47 (6.93)
19.94 (6.61)
18.33 (5.53)
18.76 (5.79)
18.13 (5.83)
18.38 (5.58)
20.13 (6.96)
Number of stressful
negative life events

Baseline
6 months
12 months
18 months
24 months
3.25 (2.33)
3.72 (3.67)
3.33 (2.07)
3.94 (3.94)
3.60 (2.39)
3.47 (2.91)
3.28 (2.21)
3.68 (2.20)
3.76 (2.76)
3.50 (2.23)
3.13 (1.97)
4.03 (4.38)
3.13 (1.97)
4.03 (4.38)
3.64 (2.46)
3.55 (2.58)
4.20 (4.12)
3.55 (2.10)
4.50 (4.40)
3.80 (2.35)
2.72 (1.71)
2.99 (2.69)
2.90 (1.94)
3.04 (2.84)
3.23 (2.44)
Total undesirability of
stressful negative life events

Baseline
6 months
12 months
18 months
24 months
13.44 (13.43)
13.06 (12.91)
12.56 (10.96)
13.17 (13.88)
11.99 (10.64)
16.41 (17.21)
14.08 (11.03)
13.70 (11.37)
14.87 (13.46)
12.93 (10.35)
11.87 (10.67)
12.36 (14.05)
11.87 (10.67)
12.36 (14.05)
11.57 (10.77)
15.16 (15.33)
14.88 (14.65)
13.51 (11.34)
15.27 (15.80)
13.32 (11.55)
10.33 (8.23)
10.08 (8.66)
10.62 (9.91)
9.61 (8.77)
9.48 (8.15)
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 70

Table 3. Correlations between physical HRQOL and key study variables. (* p < .05, ** p < .01, ***p < .001)
1 2 3 4 5 6 7 8 9
1. Baseline physical
HRQOL

2. 6 mos physical
HRQOL
R = .660***
p = .000

3. 12 mos physical
HRQOL
R = .704***
p = .000
R = .682***
p = .000

4. 18 mos physical
HRQOL
R = .632***
p = .000
R = .689***
p = .000
R = .675***
p = .000

5. 24 mos physical
HRQOL
R = .630***
p = .000
R = .662***
p = .000
R = .666***
p = .000
R = .751***
p = .000

6. Baseline Age R = .017
p = .749
R = -.023
p= .674
R = .006
p = 923
R = .008
p = 901
R = - .034
p = .622

7. Baseline CES-D R = -.335***
p = .000
R = -.384***
p = .000
R = -.332***
p = .000
R = -.271***
p = .000
R = -.315***
p = .000
R = -.037
p = .498

8. Baseline Social
support
R = .218***
p = .000
R = .175**
p = .002
R = .195**
p = .002
R = .217**
p = .001
R = .169*
p = .019
R = -.123*
p = .031
R = -.242***
p = .000

9. Baseline
Perceived Stress
R = -.331***
p = .000
R = -.353***
p = .000
R = -.387***
p = .000
R = -.430***
p = .000
R = -.358***
p = .000
R = -.207***
p = .000
R = .415***
p = .000
R = -.327***
p = .000

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 71

Table 4. Results of Longitudinal Growth Modeling

Predictors Physical HRQOL
I S
Model 1 Model 1 Bereavement .025 (p = .241) .000 (p = .807)

Model 2a Bereavement
Age
.016 (p = .445)
.002 (p = .296)
.000 (p = .619)
.000 (p = .714)

Model 2b Bereavement
Age
Age x Bereavement
.015 (p = .465)
-.002 (p = .377)
.006 (p = .033)
-.002 (p = .673)
.000 (p = .515)
.000 (p = .665)

Model 3a Bereavement
Age
Depression
.022 (p = .257)
.000 (p = .868)
-.135 (p < .001)
-.003 (p = .608)
.000 (p = .781)
.013 (p = .057)

Model 3b Bereavement
Age
Depression
Depression x Bereavement
.022 (p = .257)
.000 (p = .878)
-.145 (p < .001)
.017 (p = .726)
-.002 (p = .665)
.000 (p = .818)
.003 (p = .749)
.025 (p = .045)

Model 4a Bereavement
Age
Social Support
.014 (p = .484)
.001 (p = .614)
.028 (p = .021)
-.002 (p = .656)
.000 (p = .619)
.001 (p = .643)

Model 4b Bereavement
Age
Social Support
Social Support x Bereavement
.015 (p = .476)
.001 (p = .613)
.024 (p = .099)
.012 (p = .634)
-.003 (p = .628)
.000 (p = .621)
.001 (p = .850)
.004 (p = .690)

Model 5a Bereavement
Age
Perceived Stress
.012 (p = .555)
-.001 (p = .563)
-.008 (p < .001)
-.003 (p = .590)
.000 (p = .501)
-.001 (p < .001)

Model 5b Bereavement
Age
Perceived Stress
Perceived Stress x Bereavement
.011 (p = .560)
-.001 (p = .554)
-.008 (p < .001)
-.001 (p = .671)
-.003 (p = .618)
.000 (p = .532)
-.001 (p = .018)
.001 (p = .128)
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 72

Table 5. Top ten most commonly reported stressful negative life events at baseline.
Rank TOTAL SAMPLE
(N = 339)
BEREAVED
(N=135)
NON-BEREAVED
(N=204)
MIDDLE-AGED
(N=215)
OLDER-AGED
(N= 124)
1
Serious illness or life-
threatening accident in
friends/family (N=74)
Serious illness or life-
threatening accident in
friends/family (N=39)
Serious illness or life-
threatening accident in
friends/family (N=35)
Serious illness or life-
threatening accident in
friends/family (N=46)
Serious illness or life-
threatening accident in
friends/family (N=22)
2
Financial loss or loss of
property (N=46)
Financial loss or loss of
property (N=16)
Physical illness (N=31) Financial loss or loss of
property (tied, N=28)
Physical illness (N=21)
3
Physical illness (N=46) Physical illness (N=15) Financial loss or loss or
property (N=30)
Increased work load (tied,
N=28)
Financial loss or loss or
property (N=18)
4
Death or serious illness of a
pet (N=33)
Serious family argument
(N=14)
Increased work load
(N=22)
Trouble with a boss (N=27) Death or serious illness of
pet (N=10)
5
Taking on greatly increased
work load (N=32)
Not able to take a planned
vacation (N=13)
Death or serious illness of
pet (N=21)
Person moving out of the
household (N=25)
Serious family argument
(N=9)
6
Person moving out of the
household (N=31)
Death or serious illness of
pet (N=12)
Person moved out of
household (N=20)
Physical illness (N=25) Loss of property due to theft
(N=9)
7
Trouble with a boss (N=28) Trouble with a boss (N=12) Stopped working for an
extended period of time
(N=19)
Death or serious illness of
pet (N=23)
Not being able to take a
planned vacation (N=8)
8
Stopped working for an
extended period of time
(N=27)
Person moved out of
household (N=11)
Trouble with a boss (N=16) Stopped working for an
extended period of time
(N=22)
Change in frequency of
family get-togethers (N=7)
9
New person moving into
the household (N=26)
Increased work load
(N=10)
Personal injury (N=14) Taking a pay cut (tied,
N=15)
Person moving out of the
household (N=6)
10
Serious family argument
(N=24)
Change in frequency of
family get-togethers (N=9)
Taking a pay cut (N=13) Serious family argument
(tied, N=15)
Personal injury (N=6)
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 73

Figure 1. Hypothesized physical HRQOL trajectories of bereaved and nonbereaved
participants.
It was expected that the nonbereaved group would not experience significant change in physical
HRQOL over time, while the bereaved group would experience a decline in physical HRQOL
within 6 months after the death event and gradually return to baseline functioning.

death
event
Physical HRQOL
Time (months)
0 6 12 18 24
Bereaved
Nonbereaved
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 74

Figure 2. Hypothesized interaction between age and bereavement.
Older age is expected to be associated with less deterioration in physical HRQOL through 6
months as a result of bereavement.

death
event
Physical HRQOL
Time (months)
0 6 12 18 24
Middle-aged Bereaved
Older Bereaved
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 75

Figure 3. Flowchart of Participant Exclusion and Attrition

512 active participants in original
ELITE study
448 agreed to participate in
psychosocial study
64 declined to participate in
psychosocial study
339 participants in final sample
54 completed baseline but did not
provide follow-up data
14 excluded because date of
bereavement was unspecified
41 excluded because reported
multiple bereavement events at
consecutive timepoints
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 76

Figure 4. Path model.

Physical HRQOL
0

I
initial WHQ

0
mos

6
mos

12
mos

18
mos

24
mos

1 1
1
1
1
1
2
3
4

Bereavement

0
1
4
9
16
Age
Covariates

Q
quadratic change
S
linear change
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 77

Figure 5. Mean physical HRQOL for bereaved (group = 1) vs. non-bereaved (group = 0)

Mean scores of physical HRQOL
0.7000
0.7500
0.8000
0.8500
0.9000
Months
Group 0 1
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 78

Figure 6. The cross-sectional relationship between physical HRQOL at 6 months and the
number of months elapsed since bereavement, while controlling for baseline physical HRQOL.
This was a simulation of longitudinal processes occurring within the first 6 months of the study.

Mean scores of physical HRQOL
0.7000
0.7500
0.8000
0.8500
0.9000
Months since Bereavement
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 79

Figure 7. Age x Bereavement
Age did not moderate the relationship between bereavement and physical HRQOL.
1 = middle aged + nonbereaved
2 = middle aged + bereaved
3 = older aged + nonbereaved
4 = older aged + bereaved

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 80

Figure 8. Depression x Bereavement
Baseline depression and bereavement interacted over 12 months only, such that bereavement
affected physical HRQOL less negatively for those with higher baseline depression compared to
those with lower baseline depression.

1 = low depression + nonbereaved
2 = low depression + bereaved
3 = high depression + nonbereaved
4 = high depression + bereaved

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 81

Figure 9. Social support x Bereavement
There was no interaction between social support and bereavement in predicting physical HRQOL
trajectories over two years.
1 = low social support + nonbereaved
2 = low social support + bereaved
3 = high social support + nonbereaved
4 = high social support + bereaved

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 82

Figure 10. Perceived stress x Bereavement
There was no interaction between baseline perceived stress and bereavement in predicting
physical HRQOL trajectories over two years.
1 = low perceived stress + nonbereaved
2 = low perceived stress + bereaved
3 = high perceived stress + nonbereaved
4 = high perceived stress + bereaved

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 83

Figure 11. Mean CES-D scores for bereaved (group = 1) vs. nonbereaved (group = 0)

Mean scores of depression
1.0000
1.1000
1.2000
1.3000
1.4000
1.5000
Months
0 6
Group 0 1
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 84

Figure 12. Mean social support scores for bereaved (group = 1) vs. nonbereaved (group = 0)

Mean scores of social support
3.5000
3.6000
3.7000
3.8000
3.9000
4.0000
4.1000
4.2000
4.3000
4.4000
4.5000
4.6000
4.7000
4.8000
4.9000
5.0000
Months
0 6
Group 0 1
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 85

Figure 13. Mean perceived stress scores for bereaved (group = 1) vs. nonbereaved (group = 0)

Mean scores of perceived stress
18.00
18.50
19.00
19.50
20.00
20.50
21.00
Months
0 6
Group 0 1
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 86

Appendix A
The Women's Health Questionnaire

Please indicate how you are feeling now, or how you have been feeling THE LAST FEW
DAYS, by putting a tick in the correct box in the answer to each of the following items:

Yes,
definitely
Yes,
sometimes
No,
not
much
No,
not at
all
1. I wake early and then sleep badly for the rest
of the night (Sl)

2. I get very frightened or panic feelings for
apparently no reason at all

3. I feel miserable and sad
4. I feel anxious when I go out of the house on my
own.

5. I have lost interest in things
6. I get palpitations or a sensation of `butterflies' in
my stomach or chest

7. I still enjoy the things I used to
8. I feel life is not worth living
9. I feel tense or `wound up'
10. I have a good appetite
11. I am restless and can't keep still (Sl)

12. I am more irritable than usual
13. I worry about growing old
14. I have headaches (So)

15. I feel more tired than usual (So)

16. I have dizzy spells (So)

17. My breasts feel tender or uncomfortable
18. I suffer from backache or pain in my limbs
(So)

19. I have hot flushes (V)

20. I am more clumsy than usual (So)

21. I feel rather lively and excitable
22. I have abdominal cramps or discomfort
23. I feel sick or nauseous (So)

24. I have lost interest in sexual activity
25. I have feelings of well-being
26. I have heavy periods (please omit if no periods
at all)

27. I suffer from night sweats (V)

28. My stomach feels bloated
29. I have difficulty in getting off to sleep (Sl)

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 87

30. I often notice pins and needles in my hands
and feet (So)

31. I am satisfied with my current sexual
relationship (please omit if not sexually active)

32. I feel physically attractive
33. I have difficulty in concentrating
34. As a result of vaginal dryness sexual
intercourse has become uncomfortable (please
omit if not sexually active)

35. I need to pass urine/water more frequently
than usual (So)

36. My memory is poor

Note: Questions in bold font represent items that were included in the analysis.
So = somatic, Sl = sleep, V = vasomotor.

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 88

Appendix B
PERI Life Events Scale

Listed below are a number of events that sometimes happen in life. Please let us know if any of these
happened to you in the past 12 months.
(1). Check only the item(s) that happened to you in the past 12 months (leave others blank)
(2). When did it happen? (month/year)
(3). Rate how undesirable do you consider the item was to you:
1 = Not undesirable at all
2 = A little undesirable
3 = Somewhat undesirable
4 = Moderately undesirable
5 = Quite undesirable
6 = Very undesirable
7 = Extremely undesirable

Check if
happened
When did it
happen
(month/year)
How
Undesir-
able?

In the past 12 months…
About School
_____ ________ _____ 1. Started school or a training program after not going to school for a long
time.
_____ ________ _____ 2. Changed schools or training program.
_____ ________ _____ 3. Graduated from school or training program.
_____ ________ _____ 4. Had problems in school or in training program.
_____ ________ _____ 5. Failed school or training program.
_____ ________ _____ 6. Did not graduate from school or training program.
About Work
_____ ________ _____ 7. Started work for the first time.
_____ ________ _____ 8. Returned to work after not working for a long time.
_____ ________ _____ 9. Changed jobs for a better one.
_____ ________ _____ 10. Changed jobs for a worse one.
_____ ________ _____ 11. Changed jobs for one that was no better and no worse than the last one.
_____ ________ _____ 12. Had trouble with a boss.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 89

_____ ________ _____ 13. Demoted at work.
_____ ________ _____ 14. Found out not going to be promoted at work.
_____ ________ _____ 15. Conditions at work got worse, other than demotion or trouble with the
boss.
_____ ________ _____ 16. Promoted.
_____ ________ _____ 17. Had significant success at work.
_____ ________ _____ 18. Conditions at work improved, not counting promotion or other personal
success.
_____ ________ _____ 19. Laid off.
_____ ________ _____ 20. Fired.
_____ ________ _____ 21. Started a business or profession.
_____ ________ _____ 22. Expanded business or professional practice.
_____ ________ _____ 23. Took on a greatly increased work load.
_____ ________ _____ 24. Suffered a business loss or failure.
_____ ________ _____ 25. Sharply reduced work load.
_____ ________ _____ 26. Retired.
_____ ________ _____ 27. Stopped working, not retirement, for an extended period.
About Relationships or Marriage
_____ ________ _____ 28. Became engaged.
_____ ________ _____ 29. Engagement was broken.
_____ ________ _____ 30. Married.
_____ ________ _____ 31. Started a love affair.
_____ ________ _____ 32. Relationship with partner/significant other changed for the worse,
without separation or divorce.
_____ ________ _____ 33. Couple separated.
_____ ________ _____ 34. Termination of love relationship.
_____ ________ _____ 35. Relations with spouse/significant other changed for the better.
_____ ________ _____ 36. Couple reunited after separation.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 90

_____ ________ _____ 37. Infidelity on the behalf of spouse/significant other.
_____ ________ _____ 38. Trouble with in-laws.
_____ ________ _____ 39. Spouse died.
_____ ________ _____ 40. Boyfriend/girlfriend died.
About Children
_____ ________ _____ 41. Became pregnant.
_____ ________ _____ 42. Birth of a first child or grandchild (please circle one).
_____ ________ _____ 43. Birth of a second or later child or grandchild (please circle one).
_____ ________ _____ 44. Abortion.
_____ ________ _____ 45. Miscarriage or stillbirth.
_____ ________ _____ 46. Found out that cannot have children.
_____ ________ _____ 47. Child died.
_____ ________ _____ 48. Adopted a child.
About Family
_____ ________ _____ 50. New person moved into the household.
_____ ________ _____ 51. Person moved out of the household.
_____ ________ _____ 52. Someone stayed on in the household after he or she was expected to
leave.
_____ ________ _____ 53. Serious family argument other than with spouse.
_____ ________ _____ 54. A change in the frequency of family get-togethers.
_____ ________ _____ 55. Family member other than spouse or child dies:
_____ ________ _____ _____ Mother
_____ ________ _____ _____ Father
_____ ________ _____ _____ Brother or sister.
_____ ________ _____ _____ Grandparent.
_____ ________ _____ _____ Other (please list: _______________________________)
About Housing
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 91

_____ ________ _____ 56. Moved to a better residence or neighborhood.
_____ ________ _____ 57. Moved to a worse residence or neighborhood.
_____ ________ _____ 58. Moved to a residence or neighborhood no better or no worse than the last
one.
_____ ________ _____ 59. Unable to move after expecting to be able to move on.
_____ ________ _____ 60. Built a home or had one built.
_____ ________ _____ 61. Remodeled a home.
_____ ________ _____ 62. Suffered severe damage to a home.
_____ ________ _____ 63. Lost a home through fire, flood or other disaster.
About Crimes, Legal Matters
_____ ________ _____ 64. Lost property due to theft.
_____ ________ _____ 65. Was robbed.
_____ ________ _____ 66. Accident in which there were no injuries.
_____ ________ _____ 67. Involved in a law suit.
_____ ________ _____ 68. Accused of something for which a person could be sent to jail.
_____ ________ _____ 69. Lost driver’s license.
_____ ________ _____ 70. Arrested.
_____ ________ _____ 71. Went to jail.
_____ ________ _____ 72. Got involved in a court case.
_____ ________ _____ 73. Convicted of a crime.
_____ ________ _____ 74. Acquitted of a crime.
_____ ________ _____ 76. Didn’t get out of jail when expected.
About Finances
_____ ________ _____ 77. Took out a mortgage.
_____ ________ _____ 78. Started buying a car, furniture, or other large purchase on the installment
plan.
_____ ________ _____ 79. Foreclosure of a mortgage or loan.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 92

_____ ________ _____ 80. Repossession of a car, furniture or other items bought on the installment
plan.
_____ ________ _____ 81. Took a cut in wage or salary without a demotion.
_____ ________ _____ 82. Suffered a financial loss or loss of property not related to work.
_____ ________ _____ 83. Went on welfare.
_____ ________ _____ 84. Went off welfare.
_____ ________ _____ 85. Got a substantial increase in wage or salary without a promotion.
_____ ________ _____ 86. Did not get an expected wage or salary increase.
_____ ________ _____ 87. Had financial improvement not related to work.
About Social Activities
_____ ________ _____ 88. Increased church or synagogue, club, neighborhood, or other
organizational activities.
_____ ________ _____ 89. Took a vacation.
_____ ________ _____ 90. Was not able to take a planned vacation.
_____ ________ _____ 91. Took up a new hobby, sport, craft, or recreational activity.
_____ ________ _____ 92. Dropped a hobby, sport, craft, or recreational activity.
_____ ________ _____ 93. Acquired a pet.
_____ ________ _____ 94. Pet died or was seriously ill (circle one).
_____ ________ _____ 95. Made new friends.
_____ ________ _____ 96. Broke up with a friend.
_____ ________ _____ 97. Close friend died.
Miscellaneous
_____ ________ _____ 98. Enter the Armed Services.
_____ ________ _____ 99. Left the Armed Services.
_____ ________ _____ 100. Took a trip other than a vacation.

About Health and Safety
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 93

_____ ________ _____ 101. Physical health improved.
_____ ________ _____ 102. Physical illness.
_____ ________ _____ 103. Injury.
_____ ________ _____ 104. Unable to get treatment for an illness or injury.
_____ ________ _____ 105. Serious illness or accident (life-threatening) that happens to
friends/family:
_____ ________ _____ _____ Spouse
_____ ________ _____ _____ Child
_____ ________ _____ _____ Boyfriend/girlfriend
_____ ________ _____ _____ Close friend
_____ ________ _____ _____ Close family member
_____ ________ _____ _____ Distant family member
_____ ________ _____ 106. Sexual assault or forced/pressured sexual contact by someone other than
a spouse or partner
_____ ________ _____ 107. Sexual assault or forced/pressured sexual contact by a spouse or partner
_____ ________ _____ 108. Physical assault or unwanted physical contact (hitting, kicking, pushing,
slapping) by someone other than a spouse or partner
_____ ________ _____ 109. Physical assault or unwanted physical contact by a spouse or partner
_____ ________ _____ 110. Experienced a natural disaster
_____ ________ _____ 111. Other: ___________________________________________
_____ ________ _____ 112. Other: ___________________________________________
_____ ________ _____ 113. Other: ___________________________________________

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 94

Appendix C

Context of Bereavement
If you have lost a significant person or pet to death in the past 2 years, please answer the questions below.
If you have not recently lost anybody to death, please skip to the next page.

Regarding the loss of a significant person:
1. When did the individual die? _________________ (month/date/year)

2. How old was the individual when he/she died? ______ years

3. Were you there with the individual at the moment when she/he died?

Yes No

4. How long before the individual’s death did you realize that he/she was going to die?

1 2 3 4 5 6
Sudden Several hours Several days Several
weeks
Several
months
Several years

5. Did the individual have any serious, ongoing health problems before she/he died?
No Yes (please describe: ____________________________)

6. What was your relationship to the individual who died? _______________________

7. Did you yourself have to provide physical care to the individual in the 6 months before he/she
died?

No Yes

8. Was the individual residing in a nursing home prior to death?

No Yes

9. How often did you and the individual talk about how you would deal with life without him/her?

1 2 3 4 5
None of the time Almost Never Some of the time Fairly Often Very Often
10. Please indicate how much you agree with the following statement:
In the weeks before his/her death, things between the individual and me were going well.
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 95

1 2 3 4 5
Mostly or
completely
disagree
Somewhat
disagree
Neither agree nor
disagree
Somewhat agree Mostly or
completely agree

Any other context you wish to provide:
_____________________________________________________________________________________
_____________________________________________________________________________________
__________________________________________________________

Regarding the loss of a pet:
1. When did the pet die? _________________ (month/date/year)

2. How old was the pet when he/she died? ______ years

3. Were you there with the pet at the moment when she/he died? Yes No

4. How long before the pet’s death did you realize that he/she was going to die?

1 2 3 4 5 6
Sudden Several hours Several days Several
weeks
Several
months
Several years
5. Did the pet have any serious, ongoing health problems before she/he died?
No Yes (please describe: ____________________________)
6. Did you yourself have to provide physical care to the pet in the 6 months before he/she died?
No Yes

Any other context you wish to provide:
_____________________________________________________________________________________
___________________________________________________________________________________
_____________________________________________________________

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 96

Appendix D
CES-D
Below is a list of the ways you may have been feeling or behaving. Please check the boxes to
indicate how often you have felt this way in the past week.

Rarely or
none of the
time (less
than 1 day
this week)
Some or a
little of the
time (1 to 2
days this
week)
Occasionally
or a moderate
amount of
time (3 to 4
days this
week)
Most or all of
the time (5 to
7 days this
week)
1. I was bothered by things that usually don’t
bother me.
0 1 2 3
2. I did not feel like eating; my appetite was
poor.
0 1 2 3
3. I felt that I could not shake off the blues even
with help from my family or friends.
0 1 2 3
4. I felt I was just as good as other people. 0 1 2 3
5. I had trouble keeping my mind on what I was
doing.
0 1 2 3
6. I felt depressed. 0 1 2 3
7. I felt that everything I did was an effort. 0 1 2 3
8. I felt hopeful about the future. 0 1 2 3
9. I thought my life had been a failure. 0 1 2 3
10. I felt fearful. 0 1 2 3
11. My sleep was restless. 0 1 2 3
12. I was happy. 0 1 2 3
13. I talked less than usual. 0 1 2 3
14. I felt lonely. 0 1 2 3
15. People were unfriendly. 0 1 2 3
16. I enjoyed life. 0 1 2 3
17. I had crying spells. 0 1 2 3
18. I felt sad. 0 1 2 3
19. I felt that people dislike me. 0 1 2 3
20. I could not get “going.” 0 1 2 3
AGE, BEREAVEMENT, AND HEALTH IN WOMEN 97

Appendix E
Medical Outcomes Study – Social Support Scale
People sometimes look to others for companionship, assistance, or other types of support. How often is
each of the following kinds of support available to you if you need it?

None of
the
time
A little
of the
time
Some
of the
time
Most of
the
time
All of
the
time
2. Someone to help you if you were confined to bed 1 2 3 4 5
3. Someone you can count on to listen to you when you need to
talk
1 2 3 4 5
4. Someone to give you good advice about a crisis 1 2 3 4 5
5. Someone to take you to the doctor if you need it 1 2 3 4 5
6. Someone who shows you love and affection 1 2 3 4 5
7. Someone to have a good time with 1 2 3 4 5
8. Someone to give you information to help you understand a
situation
1 2 3 4 5
9. Someone to confide in or talk to about yourself or your
problems
1 2 3 4 5
10. Someone who hugs you 1 2 3 4 5
11. Someone to get together with for relaxation 1 2 3 4 5
12. Someone to prepare your meals if you were unable to do it
yourself
1 2 3 4 5
13. Someone whose advice you really want 1 2 3 4 5
14. Someone to do things with to help you get your mind off
things
1 2 3 4 5
15. Someone to help with daily chores if you were sick 1 2 3 4 5
16. Someone to share your most private worries and fears with 1 2 3 4 5
17. Someone to turn to for suggestions about how to deal with a
personal problem
1 2 3 4 5
18. Someone to do something enjoyable with 1 2 3 4 5
19. Someone who understands your problems 1 2 3 4 5
20. Someone to love and make you feel wanted 1 2 3 4 5
About how many close friends and close relatives do you have (people who you feel at
ease with and can talk to about what is on your mind)?
___________

AGE, BEREAVEMENT, AND HEALTH IN WOMEN 98

Appendix F
Perceived Stress Scale

The questions in this scale ask you about your feelings and thoughts during the past 2
weeks. For each item, please answer by circling how often you felt or thought a certain way.

In the past 2 weeks… None
of the
time
Almost
Never
Some
of the
time
Fairly
Often
Very
Often
1. How often have you been upset because of something
that happened unexpectedly?
1 2 3 4 5
2. How often have you felt that you were unable to
control the important things in your life?
1 2 3 4 5
3. How often have you felt nervous and “stressed?” 1 2 3 4 5
4. How often have you found that you could not cope
with all the things that you had to do?
1 2 3 4 5
5. How often have you been angered because of things
(that happened) that were outside your control?
1 2 3 4 5
6. How often have you felt difficulties were piling up so
high that you could not overcome them?
1 2 3 4 5
7. How often have you felt confident (about your ability
to handle your personal problems)?
1 2 3 4 5
8. How often have you felt that things were going your
way?
1 2 3 4 5
9. How often have you felt that you were on top of
things?
1 2 3 4 5
10
.
How often have you been able to control irritations in
your life?
1 2 3 4 5

Abstract (if available)

Abstract Although the detrimental effects of bereavement on structural aspects of physical health (e.g., mortality, disease) have been well documented in the literature, its impact on functional health patterns has been less studied. This study examined the longitudinal relationship between non‐spousal bereavement and physical health‐related quality of life (HRQOL) in a large sample of community‐dwelling women over a period of two years. Each participant had a pre‐bereavement baseline (if applicable) and was followed every 6 months thereafter. Potential moderators of the relationship between bereavement and physical HRQOL, including age, depression, perceived stress, social support, and context of bereavement, were explored. Contrary to expectations, the occurrence of a non‐spousal bereavement event did not predict change over time in physical HRQOL and there was no difference in the pattern of physical HRQOL over time between bereaved and non‐bereaved women. Within the bereaved group, those who had lost a family member had slower recovery of physical HRQOL over 24 months compared to those who had lost a friend. Descriptive data suggested that participants had better physical HRQOL and emotional functioning 6 months after bereavement if the deceased individual was of older age or had ongoing serious health concerns prior to death. The age of the participant did not moderate the relationship between bereavement and physical HRQOL. Depression, perceived stress, and social support were each significant predictors of physical HRQOL trajectories but themselves were unaffected by bereavement, and they did not moderate the effects of bereavement on physical HRQOL. Overall, these findings were inconsistent with the existing bereavement literature, which largely focuses on spousal bereavement. Potential explanations for this discrepancy, including frequency or length of measurement, choice of physical HRQOL measure, and biases of the obtained sample were systematically explored.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Social support, self-efficacy, and gender in treatment adherence of heart failure patients

PDF

Social support, constraints, and protective buffering in prostate cancer patients and their partners

PDF

The influences of anxiety, coping, and social support on physical functioning among heart failure patients

PDF

Cognitive functioning following ovarian removal before or after natural menopause

PDF

Health-related quality of life correlates of the treatment decision making process of newly diagnosed prostate cancer patients

PDF

Role of age and comorbidities in the quality of life of newly diagnosed prostate cancer patients

PDF

Relationships between lifetime chronic stress exposure, vascular risk, cognition, and brain structure in HIV

PDF

Pregnancy in the time of COVID-19: effects on perinatal mental health, birth, and infant development

PDF

Personality and cognitive aging: are positive traits protective?

PDF

Underlying neural mechanisms of depression and dementia

PDF

Psychosocial adjustment among patients with metastatic colorectal cancer

PDF

Neuropsychological functioning in older adult long-term cancer survivors

PDF

Spouse aggression, depression, and physical health: a multivariate longitudinal study of midlife couples

PDF

Effectiveness of a heart plus brain health-focused nutrition intervention among Latinas: a randomized controlled trial of Buenos Hábitos Alimenticios Para Una Buena Salud (Good Eating Habits for ...

PDF

Perceived discrimination and Latino youth adjustment: examining the influence of relinquished control and immigration status

PDF

The effects of familism and cultural justification on the mental and physical health of family caregivers

PDF

The effects of health-related spousal influence on couples coping with chronic heart failure: an application of the actor-partner interdependence model

PDF

A longitudinal study of sexual minority stress and behavioral health in sexual minority adolescents

PDF

Evaluating a cultural process model of depression and suicidality in Latino adolescents

PDF

Subclinical carotid atherosclerosis, psychosocial measures, and cognitive function in middle- to older-aged adults

Asset Metadata

Creator Maeda, Uta (author)

Core Title Longitudinal relationships between bereavement and physical health‐related quality of life in middle‐ to older‐aged women

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Psychology

Publication Date 08/13/2014

Defense Date 05/06/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bereavement,health,OAI-PMH Harvest,Women

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Gatz, Margaret (committee chair), Mack, Wendy Jean (committee member), Meyerowitz, Beth E. (committee member), Read, Stephen (committee member), Shen, Biing-Jiun (committee member)

Creator Email utamaeda@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-459814

Unique identifier UC11287085

Identifier etd-MaedaUta-2816.pdf (filename),usctheses-c3-459814 (legacy record id)

Legacy Identifier etd-MaedaUta-2816.pdf

Dmrecord 459814

Document Type Dissertation

Format application/pdf (imt)

Rights Maeda, Uta

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA