Death Studies

ISSN: 0748-1187 (Print) 1091-7683 (Online) Journal homepage: https://www.tandfonline.com/loi/udst20

“Death from a broken heart”: A systematic review

of the relationship between spousal bereavement
and physical and physiological health outcomes

Jeffrey Ennis & Umair Majid

To cite this article: Jeffrey Ennis & Umair Majid (2019): “Death from a broken heart”: A systematic
review of the relationship between spousal bereavement and physical and physiological health
outcomes, Death Studies, DOI: 10.1080/07481187.2019.1661884

To link to this article: https://doi.org/10.1080/07481187.2019.1661884

View supplementary material

Published online: 19 Sep 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=udst20
DEATH STUDIES
https://doi.org/10.1080/07481187.2019.1661884

“Death from a broken heart”: A systematic review of the relationship

between spousal bereavement and physical and physiological
health outcomes
Jeffrey Ennisa,b and Umair Majida,c,d
a
Ennis Centre for Pain Management, Hamilton, Canada; bDepartment of Medicine, McMaster University, Hamilton, Canada; cInstitute
of Health Policy, Management, and Evaluation, University of Toronto, Toronto, Canada; dDivision of Clinical Decision-Making and
Healthcare, University Health Network, Toronto, Canada

ABSTRACT
The loss of a loved one is often associated with “death from a broken heart” for the sur-
vivor, and there is evidence that shows that widowers and widows are at risk for higher
morbidity and mortality than the general population. This systematic review will summarize
the physical and physiological health outcomes of spousal bereavement. A systematic data-
base search was conducted, and 38 studies were analyzed. The majority of studies found a
statistically significant and positive association between spousal bereavement and adverse
physical and physiological health outcomes such as inflammation, cardiovascular risk,
chronic pain, and mortality.

Introduction depressive symptoms and risk of mental illness (Burns,

Browning, & Kendig, 2015; Jadhav & Weir, 2017;
Most individuals experience the loss of a spouse, sib-
Siflinger, 2017). One study found that individuals who
ling, parent, and/or child at some point in their lives.
experienced grief had higher levels of intrusive
The experience of losing someone is often referred to
thoughts, emotional numbness, disbelief and denial,
as bereavement, grief, widowhood, and mourning.
yearning, and lower meaning in life (O’Connor,
Other terms include broken heart, tako-tsubo, stress
Wellisch, Stanton, Olmstead, & Irwin, 2012). Similarly,
cardiomyopathy, transient apical ballooning syndrome,
caregivers of patients with chronic medical conditions
and ampulla syndrome (Efferth, Banerjee, & Paul, experience psychological and physical distress before
2017). These events can have significant adverse experiencing the loss of a spouse (Hebert, Prigerson,
effects on the mental, psychological, and emotional Schulz, & Arnold, 2006). A recent theoretical review
functioning of individuals. Spousal loss can disrupt an posited that traumatic events, such as the loss of
individual’s way of living, financial security, and social spouse, can disrupt an individual’s meaning in life,
status (Stroebe, 2001), and also increase feelings of which may cause psychological and existential distress
social isolation (Bunker et al., 2003). Grief from the that motivates the individual to seek new meaning or
loss of spouse may be classified as either acute, where renew their existing meaning (Majid & Ennis, 2018).
it improves overtime, or complicated, where it is The resolution of search for meaning is characterized
prolonged and interferes with daily activities and by the Meaning in Life-Adjustment Framework, which
thoughts (Shear et al., 2011). In the literature, compli- conjectures that individuals adjust positively or nega-
cated grief has also been referred to as prolonged grief tively to traumatic events.
disorder (PGD) (Mason & Duffy, 2018). The psychological correlates of widowhood have
Research has found adverse psychological outcomes been investigated for almost a century. Less research,
associated with bereavement. For example, individuals however, has focused on the physical and physio-
who experience the loss of a relative have higher logical health effects of widowhood. Some research

CONTACT Umair Majid umair.majid@mail.utoronto.ca Institute of Health Policy, Management, and Evaluation, University of Toronto, 100 Elizabeth
Street, Toronto, Ontario M5G2C4, Canada.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/udst.
Supplemental data for this article can be accessed on the publisher’s website.
ß 2019 Taylor & Francis Group, LLC
2 J. ENNIS AND U. MAJID

has documented a widowhood effect as an increased Materials and methods

risk of mortality for spousally-bereaved individuals. In
Database searching and screening
1963, a seminal study found that widowers had a 40%
increased risk in mortality than married men A systematic search was conducted in Embase,
(Young, Benjamin, & Wallis, 1963). Since then, mul- MEDLINE, and PsychINFO (all via Ovid) for studies
tiple studies have confirmed the presence of a widow- on the physical, physiological, and clinical outcomes
hood effect, including a large meta-analysis that found associated with bereavement from loss of a spouse or
a statistically significant and positive association partner. The search strategy for MEDLINE is shown
between widowhood and an increased risk in mortality in Additional File 1. The database search was con-
(Moon, Kondo, Glymour, & Subramanian, 2011). The ducted from 2008 to November 19, 2018. The search
risk is highest immediately after experiencing spousal was limited to past 10 years to reflect more recent evi-
loss (Buckley, McKinley, Tofler, & Bartrop, 2010; dence that could answer the research questions.
Moon et al., 2011), and ranges from 15% in the long- Included articles were primary quantitative or mixed-
term to 90% immediately after bereavement (Schaefer, methods studies published in English and in peer-
Quesenberry, & Wi, 1995; Elwert & Christakis, 2006; reviewed journals. Primary qualitative studies and evi-
Martikainen & Valkonen, 1996). Other studies have dence syntheses were not included but these studies
also found a heightened cardiovascular risk in spou- informed the background and discussion sections of
sally-bereaved individuals (Buckley et al., 2010, 2012; this manuscript. The screening process looked for
studies that included spousally-bereaved individuals,
Mason & Duffy, 2018), which may be due to a disrup-
defined as individuals who lost a partner or spouse of
tion in health service access and utilization
any sex at any age. Studies that included multiple
(Simeonova, 2013). Despite some research in this area,
bereaved populations (e.g., loss of child, loss of parent,
there remain important knowledge gaps between the
loss of spouse) were excluded if the results specific to
experience of widowhood and physical and physio-
spousally-bereaved individuals could not be located.
logical health outcomes, and how patterns of health
Eligible articles must have included physical health,
service utilization may influence health outcomes.
clinical, and/or physiological outcomes. Studies that
In a meta-analysis of the relationship between
included mental health, emotional, psychological, and
widowhood and mortality, Moon et al. (2011) iden-
social outcomes were excluded unless these articles
tified many gaps in the literature including: (1) the
also described physical, clinical, and/or physiological
need for more longitudinal studies that investigate a
outcomes. Since this study focused on physical and
greater number of physical correlates, (2) studies
physiological outcomes, studies that were about the
that examine the physiological underpinnings of the
transmission of sexually-transmitted diseases were
widowhood effect, and (3) lack of clarity on the
excluded because these studies emphasized outcomes
factors that may predict the widowhood effect.
related to attitudes, beliefs, and behaviors. A full list
Since the publication of Moon and colleagues’ of eligibility criteria is depicted in Table 1.
meta-analysis, more longitudinal studies have been
published, some of which have explored the physio-
logical correlates of spousal bereavement and other Data extraction
factors that may predict the widowhood effect. This A standard data extraction form was used to extract
systematic review will examine these studies and details from included studies (Additional File 2). The
address the gaps identified by Moon et al. (2011) following details were extracted: last name of first
by exploring the relationships between bereavement author, year of publication, title, research objectives/
from the loss of a spouse and physical, clinical, or questions, country of publication, type of quantitative
physiological health outcomes. The primary research methodology, number and type of participants, defini-
question of this study was: What physical, clinical, tions of outcomes measured in the study, overall con-
and physiological health outcomes or correlates are clusions (positive relationships and outcomes) for
associated with spousal bereavement? The secondary primary and secondary outcomes, statistical conclu-
research questions were: Are there gender/sex differ- sions (positive and negative correlations), and any
ences in the relationship between bereavement and reported differences between outcomes and sex. One
physical, physiological, and clinical outcomes? What researcher (UM) extracted all data and prepared
are the effects of spousal bereavement on healthcare for analysis, which were checked by another
access, quality, and utilization? researcher (JE).
 DEATH STUDIES 3

Table 1. Eligibility (inclusion and exclusion) criteria.

Inclusion Exclusion
 Loss of a spouse/partner and “death from a broken heart”  Loss of non-spousal family members (parents and children)
 Physical, clinical, and/or physiological outcomes  Loss of family members associated with disasters, natural or otherwise
 Studies that included combination of different populations, but must  Psychological, emotional, and social outcomes
have included adults facing the loss of spouse/partner  Outcomes associated with the transmission of sexually-
 Studies published from 2008 to November 19, 2018 (last 10 years) transmitted infections
 Primary, empirical quantitative or mixed-methods studies published in  Editorial, commentaries, dissertations, case reports, letters to the editor
peer-reviewed journals  Work that had not been peer-reviewed, published in a journal, or did
 English full-text publications not include empirical quantitative data
 Work that was available in abstract or book chapter form only
 Primary empirical qualitative studies of any descriptive of interpretive
methodology
 Any reviews or syntheses of evidence or secondary analyses of
primary data

Data analysis Network since all studies were observational in nature

(Von Elm et al., 2007). Where applicable, we used the
Descriptive characteristics were analyzed, and sum-
cohort and cross-sectional versions of this checklist.
mary statistics were computed for country of publica-
The criteria were applied to each included study and
tion, study design, participants, and measured
appraised as yes, no, or not applicable. The results of
outcomes. We were primarily interested observational
studies since randomized trial designs are not possible the appraisal were then reported narratively based on
to study bereavement. Study design was differentiated the number of studies that fulfilled each criterion
between cohort (prospective and retrospective) or compared to the number of studies that did not.
cross-sectional designs. Types of participants were dif-
ferentiated into the following categories: couples, Results
spousally-bereaved individuals, total deaths deter-
mined from a registry, matched controls (non- The database search retrieved 2023 hits and after
bereaved individuals), and expectant widows. Studies removing duplicates, a total of 1395 references were
with expectant widows were only included if these screened for relevance and eligibility. After screening,
studies also included participants who were already we excluded 1324 citations if they were not relevant
widows. Studies that included only expectant widows to physical health, clinical, or physiological outcomes
were excluded because this experience is distinct from associated with bereavement from the loss of spouse.
the processes of dealing with the loss of spouse(i.e., We reviewed the full-text of 71 articles and excluded
the expectedness of losing a spouse may lead to differ- 33 for reasons identified in Figure 1. After screening,
ent outcomes and cognitive appraisals). This study we deemed 38 articles eligible for inclusion.
was primarily interested in the outcomes after the loss
of spouse, and as such, studies that only included Descriptive analysis
individuals expecting widowhood were excluded.
Measured outcomes were differentiated into two cate- Country of publication
gories. First, mortality-specific outcomes included the Table 2 summarizes studies according to the country
examination of associations between a variety of fac- of publication.
tors and the risk of death after widowhood, also
referred to as the “widowhood effect” in the literature. Study design
Second, disease-specific outcomes focused on the rela- Table 3 summarizes studies according to study design.
tionship between spousal bereavement and non-mor-
tality outcomes such as the recurrence and/or Participants
development of a disease, increased risk of a disease, There was limited information in most included stud-
etc. More details on these factors are provided in ies on the median age and sex ratio of research partic-
Additional Files 1 and 2. ipants. Where mentioned, there were multiple
participant populations, including couples, spousally-
bereaved individuals, total deaths determined from a
Quality appraisal
registry, matched controls (non-bereaved individuals),
We appraised the quality of included studies using the and expectant widows. Table 4 summarizes participant
STROBE checklist obtained from the Equator populations in included studies.
4 J. ENNIS AND U. MAJID

Records identified through

database searching 2008-2019

Identification
(n = 2023)

Records after duplicates removed

(n = 1395)
Screening

Records screened Records excluded

(n = 1395) (n = 1324)
Eligibility

Full-text articles excluded (n = 33)

• Mental health outcomes only
Full-text articles assessed
(26)
for eligibility (n = 71)
• Acquiring an infection (2)
• Abstract form only (5)
Included

Studies included in
synthesis
(n = 38)

Figure 1. PRISMA diagram.

Table 2. Studies according to country of publication. Table 4. Participant populations in included studies.
Country Count (% of Total) Participant population Count (% of Total)
United States 15 (39.5) Spousally-bereaved individuals 31 (81.6)
United Kingdom 9 (23.7) Matched (non-bereaved) controls 8 (21.1)
Australia 4 (10.5) Couples 6 (15.8)
Sweden 3 (7.9) Expectant widows and spousally-bereaved individuals 2 (5.3)
Denmark 3 (7.9) Deaths from a registry 1 (2.6)
Germany 1 (2.6)
Norway 1 (2.6)
Finland 1 (2.6) M€oller, Bj€
orkenstam, Ljung, & Åberg Yngwe, 2011;
Switzerland 1 (2.6)
Shah et al., 2013a, Tseng, Petrie, Wang, Macduff, &
Stephen, 2018). These studies analyzed the number
Table 3. Studies according to study design. and type of medication use, using psychiatric or pri-
Study design Count (% of Total)
mary care, and the quality of care received. Four
Cohort (prospective or retrospective) 35 (92.1)
Cross-sectional 3 (7.9) (16.7% of 24) studies examined general predictors of
death (Bowling, 2009; Boyle, Feng, & Raab, 2011;
Elwert & Christakis, 2008; Infurna et al., 2017), three
Measured outcomes (12.5% of 24) identified the relationship between time
Studies examined mortality (24 studies; 63.2%) or dis- since widowhood and risk of death (Berntsen &
ease-specific outcomes (14 studies; 36.8%). Among the Kravdal, 2012; King, Lodwick, Jones, Whitaker, &
studies that looked at mortality-specific outcomes, five Petersen, 2017; Vable, Subramanian, Rist, & Glymour,
(20.8% of 24) examined the relationship between qual- 2015), three (12.5% of 24) looked at the relationship
ity or access to health services and risk of death after between demographic factors such as body mass
widowhood (Jin & Chrisatakis, 2009; King et al., 2013; index, socioeconomic status, and gender and the risk
 DEATH STUDIES 5

of death after widowhood (Moon, Glymour, Vable, general risk of mortality after widowhood (Allegra
Liu, & Subramanian, 2014; Oliveira, Rostila, Saarela, et al., 2015), various predictors of death (Bowling,
& Lopes, 2014; Sullivan & Fenelon, 2014), three 2009), use of primary health services after widowhood
(12.5% of 24) examined the association between psy- (King et al., 2013), the relationship between body-
chological and social factors (social support, the level mass index and the risk of death after widowhood
of unexpectedness of bereavement, and levels of con- (Oliveira et al., 2014), and the relationship between
scientiousness and neuroticism) and the risk of death psychological factors (neuroticism and conscientious-
(Aoun et al., 2015; Bratt, Stenstr€ om, & Rennemark, ness) and the risk of death after widowhood (Bratt et
2016; Shah et al., 2013b). Other studies examined rela- al., 2016). Among studies that analyzed disease-spe-
tionship between disease characteristics (e.g., spouse cific outcomes, 12 (85.7% of 14) found statistically sig-
dying from suicide, spouse dying from dementia) and nificant and positive associations between primary
risk of death after widowhood (Erlangsen et al., 2017; 
variables of interest (Asgeirsd
ottir et al., 2013; Buckley
Shah et al., 2016), one (4.2% of 24) study looked at et al., 2010, 2011, 2012; Carey et al., 2014; d’Epinay
the general risk of death after widowhood without any et al. 2010; Eini€o et al., 2017; Fagundes et al., 2018,
specification on demographic or disease characteristics 2019; Holland et al., 2013; Schultze-Florey et al., 2012;
(Allegra, Ezeamama, Simpson, & Miles, 2015), and Stahl et al., 2016), and two (14.3% of 14) found statis-
one (4.2% of 24) analyzed the relationship between tically non-significant associations (Olsen et al., 2012;
the concentration of widows in neighborhood and Santacroce et al., 2018). All studies that measured dif-
the risk of death (Subramanian, Elwert, & ferent predictors of cardiovascular risk after spousal
Christakis, 2008). bereavement found a positive and statistically signifi-
Among the studies that looked at disease-specific cant association (Buckley et al., 2010, 2011, 2012;
outcomes, nine (64.3% of 14) looked at various indi- Carey et al., 2014; Eini€o et al., 2017; Fagundes et al.,
cators of cardiovascular risk such as blood pressure, 2018, 2019; Holland et al., 2013; Schultze-Florey et al.,
heart rate, cardiovascular events, and inflammation 2012; Stahl et al., 2016). The two studies that found
(Buckley et al., 2010, 2011, 2012; Carey et al., 2014; statistically non-significant associations focused on:
Eini€
o, Moustgaard, Martikainen, & Leinonen, 2017; breast cancer recurrence (Olsen et al., 2012), and
Fagundes et al., 2018, 2019; Schultze-Florey et al., immune system strength after experiencing the loss of
2012; Stahl, Arnold, Chen, Anderson, & Schulz, 2016). spouse (Santacroce et al., 2018). Overall, this finding
The remaining five (35.7% of 14) examined the fol- indicates that a higher proportion of studies that ana-

lowing: development of chronic pain (Asgeirsd ottir, lyzed disease-specific compared to mortality-specific
Valdimarsd ottir, F€urst, Steineck, & Hauksd ottir, outcomes showed a statistically meaningful result.
2013), functional status (d’Epinay, Cavalli, & Guillet,
2010), levels of diurnal cortisol (Holland et al., 2013), Sex differences
breast cancer recurrence (Olsen et al., 2012), and Among the included studies, 21 (55.3%) looked at sex
immune system strength (Santacroce, Wastesson, differences in the outcomes measured (Bernstein 2012;
H€ohn, Christensen, & Oksuzyan, 2018). Bowling, 2009; Boyle et al., 2011; Eini€ o et al., 2017;
Elwert & Christakis, 2008; Erlangsen et al., 2017;
Significant vs. non-significant associations Holland et al., 2013; Jin & Chrisatakis, 2009; King
Among studies that analyzed mortality-specific out- et al., 2013; Moon et al. 2014; Santacroce et al., 2018;
comes, 19 (79.2% of 24) found statistically significant Seifter et al.; 2014; Shah et al., 2016; Stahl et al., 2016;
associations between primary variables of interest Subramanian et al., 2008; Sullivan & Fenelon 2014;
(Abel & Kruger, 2009; Aoun et al., 2015; Berntsen & Abel & Kruger, 2009; Carey et al., 2014; Shah et al.,
Kravdal 2012; Boyle et al., 2011; Elwert & Christakis, 2013a, 2013b; Tseng et al., 2018), and 17 (44.7%) did
2008; Erlangsen et al., 2017; Infurna et al., 2017; Jin & not look at these differences (Allegra et al., 2015;
Chrisatakis, 2009; King et al., 2017; M€ oller et al., 
Aoun et al., 2015; Asgeirsd ottir et al., 2013; Bratt et
2011; Moon et al., 2014; Shah et al., 2013a, 2013b, al., 2016; Buckley et al., 2010, 2011, 2012; d’Epinay
2016; Seifter et al., 2014; Subramanian et al., 2008; et al. 2010; Fagundes et al., 2018, 2019; Infurna et al.,
Sullivan & Fenelon, 2014; Tseng et al., 2018; Vable 2017; King et al., 2017; M€ oller et al., 2011; Oliveira
et al., 2015), and five (20.8% of 24) found statistically et al., 2014; Olsen et al., 2012; Schultze-Florey et al.,
non-significant associations (Allegra et al., 2015; 2012; Vable et al., 2015). In the 21 studies that exam-
Bowling, 2009; Bratt et al., 2016; King et al., 2013; ined sex differences, 16 (76.2% of 21) identified differ-
Oliveira et al., 2014). These five studies focused on: ences between males and females on variables
6 J. ENNIS AND U. MAJID

measured in the study (Berntsen & Kravdal, 2012; examine how their findings differ across this character-
Bowling, 2009; Boyle et al., 2011; Eini€ o et al., 2017; istic. This observation is a significant limitation of the
Elwert & Christakis, 2008; Erlangsen et al., 2017; literature captured in this study. Investigating how fac-
Holland et al., 2013; Jin & Chrisatakis, 2009; King tors associated with widowhood differ across demo-
et al., 2013; Moon et al. 2014; Santacroce et al., 2018; graphic groups, specifically racial-ethnic mix, can aid
Seifter et al. 2014; Shah et al., 2016, Stahl et al., 2016; in tailoring responses to widowhood that are unique
Subramanian et al., 2008; Sullivan & Fenelon 2014), to individuals’ social locations.
whereas 5 (23.8% of 21) studies did not identify any
differences (Abel & Kruger, 2009; Carey et al., 2014; Predictors of widowhood mortality
Shah et al., 2013a, 2013b; Tseng et al., 2018). Among Multiple studies investigated the relationship between
the 16 studies that reported on differences between various factors and mortality- and disease-specific out-
sexes, 10 (62.5% of 16) found that men have a higher comes. In general, studies found that individuals who
risk of mortality (Berntsen & Kravdal 2012; Bowling, experienced spousal loss were at a higher risk of mor-
2009; Eini€ o et al., 2017; Erlangsen et al., 2017; tality than the general population (Abel & Kruger,
Holland et al., 2013; King et al., 2013; Moon et al. 2009; Allegra et al., 2015; Aoun et al., 2015;
2014; Seifter et al. 2014; Stahl et al., 2016; Sullivan & 
Asgeirsdottir et al., 2013; Bowling, 2009; Buckley
Fenelon 2014), all of which showed statistically signifi- et al., 2011; Elwert & Christakis, 2008; Erlangsen
cant associations except for one which showed a trend et al., 2017; Infurna et al., 2017; M€oller et al., 2011;
towards significance (Moon et al. 2014). Four (25.0% Moon et al. 2014; Oliveira et al., 2014; Seifter et al.
of 16) studies found that women had a higher risk of 2014; Shah et al., 2013a, 2013b, 2016; Subramanian
mortality than men (Jin & Chrisatakis, 2009; et al., 2008; Sullivan & Fenelon 2014; Stahl et al.,
Santacroce et al., 2018; Shah et al., 2016; Subramanian 2016; Tseng et al., 2018). Some studies referred to this
et al., 2008), and two (12.5% of 16) found that men phenomenon as the widowhood effect; an increased
and women have higher risk of mortality at different risk of mortality that characterizes widowhood. The
ages (Boyle et al., 2011), and there are differences in increased risk of mortality ranged from 19.2% (Tseng
disease-specific deaths between the sexes (Elwert & et al., 2018), to 48% (Sullivan & Fenelon 2014), and
Christakis, 2008). some studies reported that the widowhood effect
remained regardless of the cause of death (Elwert &
Quality appraisal Christakis, 2008; Erlangsen et al., 2017; M€ oller et al.,
Using the STROBE checklist for cohort and cross-sec- 2011). The following sections summarize various pre-
tional designs, quality appraisal revealed many dictors and factors that included studies showed to be
strengths and limitations of included studies. The associated with an increased mortality (i.e., the
majority of studies (32 studies; 84.2%) provided a widowhood effect) after experiencing spousal loss.
detailed description of the background, context, and These sections are: Demographic Characteristics and
rationale for conducting their study. Moreover, 36 the Widowhood Effect; Age of Widows and Increased
studies (94.7%) clearly stated specific hypotheses of Risk of Mortality; Symptoms, Disorders, and
objectives of their investigation, and 36 studies (94.7%) Widowhood; Lifestyle Factors and Widowhood;
described their statistical analysis plan in sufficient Remarriage After Widowhood; and Time
detail. However, 13 studies (34.2%) did not provide a Since Widowhood.
balanced description of methods and results in the
abstract. This observation could possibly reflect differ- Demographic characteristics and the widowhood
ences in disciplines and journals in which included effect. The relationship between socioeconomic status
studies were published. All except for four studies and social class was investigated by some included
(10.5%) did not describe strategies to reduce sources studies. Whereas three studies found that these factors
of bias in their sample or analysis plan, and only three partially explained the widowhood effect (Sullivan &
studies (7.9%) discussed how they formulated their Fenelon, 2014; Subramanian et al., 2008; Tseng et al.,
sample size. Out of 38 studies, 10 (26.3%) did not pro- 2018), and one study concluded that one-third of the
vide a description of the participants’ demographic increase in mortality risk was explained by socioeco-
characteristics. The remaining 28 studies that described nomic disadvantage (Sullivan & Fenelon, 2014), two
these characteristics did not use them for subgroup studies did not report a statistically significant associ-
analyses. For example, even though some studies noted ation between these variables (Oliveira et al., 2014;
the racial-ethnic mix of their participants, they did not Moon et al., 2014). One non-significant study
 DEATH STUDIES 7

reported a trend towards significance between losing a same risk of mortality than those who were married
spouse and higher body-mass index that was mediated for the first time and did not experience spousal loss
by social class (Oliveira et al., 2014). On the other (Berntsen & Kravdal, 2012).
hand, one study found that the risk of death increases
by up to 6% for individuals who lost a spouse and Time since widowhood. Multiple studies explored the
live in a low-poverty neighborhood (Subramanian relationship between different variables and the time
et al., 2008). Surviving spouses who smoked had a since individuals experienced the loss of spouse.
higher neutrophil count than the general population, Generally, some studies found that individuals experi-
indicating a more active immune system that may sig- enced the highest risk of mortality immediately after
nal a higher risk of adverse health consequences as a widowhood up until three months (King et al., 2017;
result of widowhood (Buckley et al., 2011). Moon et al., 2014), until six months (Bowling, 2009;
Boyle et al., 2011; Jin & Chrisatakis, 2009; Seifter
Age of widows and increased risk of mortality. et al., 2014; Sullivan & Fenelon, 2014), or until
Among included studies, there was uncertainty 10 years after widowhood (Buckley et al., 2010). Other
regarding whether younger (Aoun et al., 2015; studies found that the increased risk of mortality that

Asgeirsdottir et al., 2013; Seifter et al., 2014; Fagundes widows experienced immediately after spousal loss
et al., 2019; King et al., 2013), or older widows decreases overtime (Berntsen & Kravdal, 2012;
(Bowling, 2009; Infurna et al., 2017) had a higher risk Buckley et al., 2010; King et al., 2017; Seifter et al.,
of mortality. Two studies found that younger widows 2014; Sullivan & Fenelon 2014). For some partici-
experienced more adverse symptoms (muscle tension, pants, the risk decreased to normal levels 24 months
headache, burnout, and inflammation) (Asgeirsd ottir after widowhood (King et al., 2017; Sullivan &
et al., 2013; Fagundes et al., 2019). One study Fenelon, 2014). For other participants, the risk of
hypothesized that a stronger association existed mortality remained high at six months but disap-
between grief and inflammation for younger com- peared 28 years after experiencing the loss of spouse
pared to older widows (Fagundes et al., 2019). (Bowling, 2009). However, one study found that sur-
Therefore, younger widows may be at an increased viving spouses had a 10% less risk of death at 18 years
risk of higher inflammation after experiencing spousal post-widowhood than individuals who did not experi-
loss, and accordingly, have a higher risk of developing ence spousal loss in the first place (Tseng et al., 2018).
cardiovascular disease (Fagundes et al., 2019). On the This decrease in risk of mortality overtime was not
other hand, one study found that older widows may explained by psychosocial factors, social circumstan-
have worse adaptation to life-changing circumstances ces, social network, support, and participation, prob-
as a result of spousal loss (Infurna et al., 2017). lems with mental health, coming to terms with the
Interestingly, one of the studies that found younger death of spouse, and social class, but was explained by
widows to have a higher risk of mortality also found male sex, older age, poorer physical health, and
that mortality risk increased with age (King et al., expressed relief at the loss of spouse (Bowling, 2009).
2013), indicating that the relationship between the age Studies also found a difference between disease-spe-
of widows and risk of mortality warrants further cific outcomes and time since widowhood. For cardio-
investigation. vascular disease risk and related variables, multiple
studies found an increased risk of higher systolic
Remarriage after widowhood. One study found that blood pressure, heart rate, prothrombotic changes
individuals who remarried after spousal loss had a (higher neutrophil count, von Willebrand factor anti-
higher risk of mortality than those who were first- gen, Factor VIII, and platelet/granulocyte aggregates),
time married (Berntsen & Kravdal, 2012). However, heart rate variability, cardiovascular events, dysregu-
the authors noted that this difference could be attrib- lated levels of cortisol, blood pressure, and cholesterol
uted to the experience of losing a spouse rather than levels immediately after experiencing spousal loss,
remarrying after widowhood. On the other hand, a which returned to normal levels between 30 days to
different study found that individuals who remarried one year after widowhood (Buckley et al., 2010, 2011,
after widowhood had lower risk of mortality (Seifter 2012; Carey et al., 2014; Holland et al., 2013; Eini€ o
et al., 2014) and experienced less symptoms of chronic et al., 2017; Shah et al., 2013a). One study observed
pain compared to widows who did not remarry that widows experienced a decrease in functional sta-

(Asgeirsd ottir et al., 2013). Widows who remarried tus after spousal loss, which increased overtime as
within 10 years after spousal loss experienced the widows adjusted to their new lives without spouse
8 J. ENNIS AND U. MAJID

(d’Epinay et al., 2010). Despite these findings, one disease had difficulty maintaining their health imme-
study did not find a statistically significant association diately after experiencing spousal loss, which could
between time since widowhood and the following var- explain the increased risk of adverse health outcomes
iables: cytokine production, heart rate variability, and (Shah et al., 2013a; Stahl et al., 2016). On the other
inflammatory markers (Fagundes et al., 2018). hand, disease characteristics may also explain the
presence or absence of the widowhood effect. For
Symptoms, disorders, and widowhood. Widows gen- example, one study found interaction effects between
erally reported poorer health and a greater number of the nature of a preexisting illness and an increased
symptoms than the general population (Aoun et al., risk of mortality after spousal loss (Bowling, 2009); a

2015; Asgeirsd ottir et al., 2013; Infurna et al., 2017). second study found a trend towards less severe
Most studies that examined the relationship between widowhood effect for surviving spouses who experi-
experiencing widowhood and symptoms or disorders enced bereavement as a result of dementia (Shah et
found a positive and statistically significant association al., 2016); a third study found that individuals were at

(Asgeirsd ottir et al., 2013; Buckley et al., 2010, 2011, higher risk of death from "non-life limiting illnesses"
2012; Carey et al., 2014; d’Epinay et al. 2010; (Aoun et al., 2015); and a fourth study found an
Erlangsen et al., 2017; Fagundes et al., 2018, 2019; increased risk of mortality from a known chronic dis-
Holland et al., 2013; Oliveira et al., 2014; Schultze- ease (Shah et al., 2013a).
Florey et al., 2012; Shah et al., 2013a). Many studies
focused on quantifying the magnitude of increased Lifestyle factors and widowhood. Interestingly,
risk for developing cardiovascular disease after experi- expectedness of spousal loss seemed to influence the
encing spousal loss (Buckley et al., 2010, 2011, 2012; extent to which widows experienced adverse physical
Carey et al., 2014; Fagundes et al., 2018, 2019; health outcomes. Although both expectedness and
Holland et al., 2013; Schultze-Florey et al., 2012; Shah unexpectedness of spousal loss resulted in the widow-
et al., 2013a). These studies found that widows experi- hood effect, unexpectedness had a more pronounced
enced: increased systolic blood pressure (Buckley effect (Shah et al., 2013b). Similarly, another study
et al., 2010; Shah et al., 2013a), heart rate (Buckley found that higher levels of pre-spousal loss life satis-
et al., 2010, 2012; Fagundes et al., 2018), inflammation faction, possibly indicating an orientation towards
(Buckley et al., 2011; Fagundes et al., 2018, 2019; acceptance, may result in a lower widowhood effect
Schultze-Florey et al., 2012), ex vivo cytokine produc- (Infurna et al., 2017). Religion was also a predictor of
tion (Fagundes et al., 2018, 2019), levels of dysregu- spousal survival that could help individuals attain a
lated cortisol (Holland et al., 2013), autonomic level of expectedness or pre-loss satisfaction that
function (Buckley et al., 2012), prothrombotic changes reduces the widowhood effect (Abel & Kruger, 2009).
(Buckley et al., 2011), risk of cardiovascular events Another factor that reduced the widowhood effect was
(Carey et al., 2014), and cholesterol levels (Shah et al., physical activity or exercise (Allegra et al., 2015;
2013a). One study specified an interaction between a 
Asgeirsd ottir et al., 2013). One study reported that
pro-inflammatory genotype and environment spousally-bereaved individuals who were more physic-
(Schultze-Florey et al., 2012). In other words, some ally-active observed an 85% reduction in mortality
individuals may be more susceptible to a higher risk risk and fewer symptoms of chronic pain compared to
of cardiovascular disease after widowhood because of spousally-bereaved individuals who were less physic-
their genotype, which is activated by experiencing 
ally-active (Allegra et al., 2015; Asgeirsd ottir
spousal loss (Schultze-Florey et al., 2012). et al., 2013).
Apart from cardiovascular risk, studies also found Although not a primary topic of this study, social
that widows experienced more symptoms of chronic cohesion and support were mentioned as a predictor

pain (Asgeirsd ottir et al., 2013); lower functional sta- of the widowhood effect. One study concluded that
tus (d’Epinay et al., 2010); an increased risk of phys- individuals who lacked adequate social support were
ical and mental disorders, and suicidal behaviors at a higher risk of mortality (Aoun et al., 2015).
(Erlangsen et al., 2017); and an increased risk for can- Interestingly, however, one study found that the
cer (Erlangsen et al., 2017). One study found that widowhood effect still existed in socially-cohesive
higher levels of depression in widows may mediate communities where social support was high (e.g., the
the adverse physical health consequences reported Amish), and a higher number of children to support
after widowhood (Fagundes et al., 2019). Another is associated with an increased risk for mortality
study found that widows with existing cardiovascular (Seifter et al., 2014). Another study found that a lower
 DEATH STUDIES 9

concentration of widowed individuals in a neighbor- Differences between sexes

hood may explain up to 22% of the widowhood effect In general, seven studies found that males experienced
(Subramanian et al., 2008), perhaps underlining the a higher risk of mortality after widowhood than
relationship between social support and an increased females (Berntsen & Kravdal, 2012; Bowling, 2009;
risk in mortality. The factors that may explaingd these Boyle et al., 2011; King et al., 2013; Moon et al., 2014;
findings are unclear. However, another study found a Seifter et al., 2014; Stahl et al., 2016), particularly,
more active immune system in individuals who were within 12 months after experiencing the loss of a
living alone (Buckley et al., 2011), possibly indicating spouse (Moon et al., 2014). However, two studies
that individuals who lacked social support are at a found that females experienced a higher risk of mor-
higher risk of adverse physical health effects that tality (Jin & Chrisatakis, 2009; Shah et al., 2016). It
accompany the widowhood effect. was unclear what factors may have caused different
conclusions between studies.
There were also differences in the predictors of
mortality between sexes. For example, males had a
Use and access of health services higher risk of cardiac problems (Eini€ o et al., 2017),
Multiple studies found that individuals who experi- and males who had an existing cardiovascular disease
enced the loss of spouse were more likely to access observed an increased risk of mortality; the same
the healthcare system than the general population effect did not exist for women (Stahl et al., 2016). For
(Erlangsen et al., 2017; King et al., 2013; M€ oller et al., men, spousal death increased the risk of mortality by
2011; Shah et al., 2013a, Tseng et al., 2018). These more than 20% for chronic obstructive pulmonary
individuals also reported higher utilization of mental disease, diabetes, accidents or serious fractures, infec-
health services or psychiatric care (Erlangsen et al., tions or sepsis, lung cancer, and unknown causes of
2017; M€ oller et al., 2011), primary care services (King death; 10% for colon cancer, heart disease, heart fail-
ure, kidney disease, stroke, vascular diseases, and
et al., 2013), and hospitals (Tseng et al., 2018).
other cancers; and not significant associations between
Similarly, spousally-bereaved individuals reported a
the widowhood effect and flu, pneumonia,
higher likelihood of using various medications (anti-
Alzheimer’s disease, Parkinson’s disease (Elwert &
depressants, hypnotic medications, and psychotropic)
Christakis, 2008). The increased mortality in men may
than the general population (King et al., 2013; M€ oller
be explained by higher levels of depression among
et al., 2011; Shah, ND). Psychotropic medication use,
males compared to females (Stahl et al., 2016).
in particular, was higher for widows who experienced
Similarly, unexpected spousal loss affects men more
spousal loss as a result of dementia or related disor-
severely than women (54% increased risk for mortality
ders (Shah ND). On the other hand, individuals who
in men and no increase risk in women) (Sullivan &
lost a spouse from suicide reported higher use of
Fenelon 2014). Increased risk of mortality in men
municipal support, sick leave benefits, and disability may also be explained by a decline in the quality of
pension funds than the general population (Erlangsen care, lack of adequate coordination between the differ-
et al., 2017). These studies support the conclusion that ent levels of healthcare, and inability to advocate and
spousally-bereaved individuals’ access and use more communicate appropriately in formal medical consul-
health services than the general population. However, tations (Jin & Chrisatakis, 2009). However, one study
other research does not support this conclusion. For found that the presence of other widowed individuals
example, one study found that there was a short-term in the neighborhood protected men more than
disruption in health service utilization immediately women from the widowhood effect (Subramanian
after experiencing spousal loss (Jin & Chrisatakis, et al., 2008). Moreover, the rate of mortality increased
2009). Moreover, another study found that individuals for women 10 years post-widowhood but decreased
who were taking cardiovascular medications before for men (Berntsen & Kravdal, 2012).
spousal loss experienced challenges with maintaining For women, spousal loss increases the risk for mor-
use in the year following spousal loss (Shah et al., tality by more than 20% for chronic obstructive pul-
2013a). These studies highlight the uncertainty in how monary disease, colon cancer, accidents or serious
the experience of losing a spouse increases (or fractures, lung cancer, and unknown causes; 10% for
decreases) health service utilization, or the mecha- infections, flu, kidney disease, diabetes, vascular dis-
nisms that cause widows to access the healthcare sys- eases, and heart failure; and non-significant associa-
tem more frequently than the general population. tions between the widowhood effect and rapidly fatal
10 J. ENNIS AND U. MAJID

cancers, Alzheimer’s disease, and Parkinson’ disease increased risk of mortality before the loss of spouse
(Elwert & Christakis, 2008). Women experienced a (Vable et al., 2015; Eini€ o et al., 2017). From this
higher risk of cardiac problems due to a preexisting research, the risk of mortality rises around 30 days
chronic disease (Eini€ o et al., 2017), and use more before spousal loss, continues to increase after loss,
anti-infective drugs before and after the loss of spouse and peaks at some time after, which could range from
(Santacroce et al., 2018). Women also reported a trend a few weeks to several years post-bereavement. Once
towards higher baseline levels of dysregulated cortisol the risk of mortality peaks, it appears to decrease over
levels compared to men, indicating a higher risk of time. Factors such as remarriage and social support
developing cardiovascular disease after spousal loss can act as buffers by reducing the risk of mortality or
(Holland et al., 2013). Women were also more likely decreasing the time ittakes for an individual to reach
than men to die after widowhood if they lost a spouse their peak for mortality risk. However, other research
to dementia (Shah et al., 2016). This analysis indicates has found that mortality risk only increases after loss
that there is uncertainty regarding the predictors of (Dunlay et al., 2017). There is some confusion sur-
the widowhood effect between sexes, and the mecha- rounding the timeline of the widowhood effect and
nisms or factors that may explain the sex differences. future research should explore this aspect of widow-
hood longitudinally.
It also appears that surviving the widowhood effect
Discussion
may result in an overall reduction in an individual’s
This review described the relationships between spou- mortality risk. This finding may indicate that individ-
sal bereavement and physical and physiological health uals become cognitively stronger once they have
outcomes. Included studies revealed many variables adjusted to their life circumstances. The Meaning in
that clarify the possible physical health consequences Life-Adjustment Framework – describes how individ-
of bereavement. This review discussed the following uals adjust positively or negatively to a traumatic
categories of variables: Demographic Characteristics event such as spousal loss – illustrates why mortality
and the Widowhood Effect; Age of Widows and risk in the long-term may reduce for certain individu-
Increased Risk of Mortality; Remarriage After als (Majid & Ennis, 2018). According to this frame-
Widowhood; Time Since Widowhood; Symptoms, work, an individual experiencing a traumatic event
Disorders, and Widowhood; and Lifestyle Factors also experiences significant psychological and existen-
and Widowhood. tial distress, which may disrupt their meaning in life
Most studies found a widowhood effect, which and sense of purpose. This distress motivates them to
appears to indicate that an increased risk of mortality find new or renew their existing meaning in life in
is an expected outcome of widowhood. However, the order to make sense of their circumstances. These
relationship between widowhood and physical health individuals, depending on the personal and relational
outcomes is complicated by environmental and per- resources (e.g., social support) available to them, will
sonal factors. This review identified multiple, potential either adjust positively or negatively to the traumatic
buffers that can reduce the widowhood effect such as event. In the case of positive adjustment, individuals
physical activity and exercise, expectedness of spousal may become cognitively, physically, and emotionally
loss, remarriage, more time since widowhood, social stronger than before the traumatic event. On the other
support, personality factors, higher socioeconomic sta- hand, for individuals who adjust negatively, they may
tus, and higher frequency of interaction experience more negative health outcomes such as
with neighbors. suicide, depression, anxiety, and more (Majid &
In particular, it appears that time since widowhood Ennis, 2018).
is an important factor that clarified why widowhood Many studies included in this review looked at car-
accompanied with an increased risk of mortality. diovascular indicators of health and their relationship
Comparing the findings across studies, the risk of to spousal bereavement. A previously published litera-
mortality increases immediately prior to spousal loss ture review analyzed studies on this topic and pro-
in anticipation of bereavement. This observation sug- posed mechanisms of action between cardiovascular
gests that the cognitive appraisal of anticipated risk indicators and bereavement (Buckley et al., 2010).
widowhood may contribute to physical and physio- This review found many gaps in the literature, such as
logical health outcomes. Although this review did not limited studies on heart rate, heart rate variability,
focus on studies about pre-bereavement, two studies immune system function, and prothrombotic changes,
that included expectant widows confirmed an which the present review addressed by including a
 DEATH STUDIES 11

wider range of studies and more recent publications. particular, it appears that individuals who are gener-
Moreover, Buckley et al., (2010) found no effect ally healthy (i.e., without a known chronic medical
between the expectedness of spousal loss and reduced condition that requires daily self-management)
mortality risk, whereas the present review found two reported a lower widowhood effect if they accessed
studies that highlighted this relationship (Shah et al., basic primary and psychiatric care services after spou-
2013a, Infurna et al., 2017). Therefore, this review sal loss. This finding is intuitive since these services
adds to the current understanding of spousal bereave- may buffer some of the habits that adversely influence
ment with regards to cardiovascular risk outcomes. physical health and the cognitive effects of a traumatic
Similar to Buckley’s review, more research is needed event. On the other hand, individuals who are already
on blood pressure, inflammatory changes, and accessing health services to manage a chronic medical
immune system function, specifically, the physiological condition, losing a spouse may disrupt their self-man-
mechanisms of action. Moreover, a more recent agement activities and reduce routine use of health
review by Mason and Duffy (2018) looked at compli- services. As a result of coping with the personal and
cated grief and the cortisol response and found that emotional consequences of bereavement, physical
individuals with prolonged grief reported higher levels health needs may be neglected by these individuals,
of dysregulated cortisol levels. This finding was con- which may explain their increased risk of mortality.
firmed by the present review. The findings presented in this review found important
With regards to sex differences, a higher number of relationships between demographic and disease char-
studies seemed to indicate that men report more acteristics and health service utilization.These relation-
adverse physical and physiological outcomes from ships are unclear as other research indicates that
widowhood than women. However, some studies sug- recently bereaved individuals may use more psycho-
gest the opposite. There are also sex differences with tropic medications than those who have been
regards to the nature of spousal death, for example, bereaved for longer periods of time (M€ oller et al.,
men are more likely to have a higher risk of mortality 2011). More research is needed to link the widowhood
if the death was unexpected. Moreover, women may effect and health service utilization patterns.
have higher quality and quantity of social support
available to them, which may improve their likelihood
Strengths and limitations of this study
of buffering the negative health consequences of
bereavement. On the other hand, none of the studies This review has multiple strengths such as a system-
included in this review analyzed differences in out- atic database search and the inclusion of studies from
comes across racial, ethnic, or cultural groups, despite a variety of countries, contexts, and participant popu-
some studies displayed these characteristics in the lations. Moreover, this review classified outcomes if
manuscripts. This observation is a significant limita- they were mortality- or disease-specific. Within each
tion of the literature in this area, and it may be classification, there were multiple predictors, corre-
because the majority of studies analyzed registry or lates, and factors that were associated with bereave-
database data which may not contain information ment. As such, this analysis provides a breadth of the
about race or ethnicity. For example, Shah et al. literature on the physical, clinical, and physiological
(2013b) noted that due to the homogeneity of the health outcomes associated with spousal bereavement.
population, they were unable to include ethnicity as a On the other hand, this review has many limita-
covariate in their analyses because this information tions that provide opportunities for future research.
was unavailable from the databases they used. Future First, this study looked at spousally-bereaved individu-
research should investigate how physical and physio- als only. We recognize that studies on other forms of
logical outcomes of spousal bereavement differ bereavement, such as the loss of child, parent, and
between racial, ethnic, or cultural groups, or at the sibling, may support a greater and more nuanced
minimum conduct rigorous subgroup analyses. There understanding of the physical health effects after loss.
also needs to be concerted effort to report the various A comparative analysis between these types of losses
demographic characteristics of participants in manu- and their physical and physiological outcomes may
scripts and have authors comment on the link elucidate their differences and increase depth of
between these characteristics and the objectives of understanding in this area. For example, comparing
their study. these types of losses may clarify how closeness to an
Finally, some studies discussed medical care access individual may result in different combinations of
and utilization and the widowhood effect. In physical outcomes.
12 J. ENNIS AND U. MAJID

Second, although there was some research on dis- Berntsen, K. N., & Kravdal, Ø. (2012). The relationship
ease-specific outcomes of spousal bereavement, most between mortality and time since divorce, widowhood or
of the research reviewed focused on cardiovascular remarriage in Norway. Social Science & Medicine (1982),
75(12), 2267–2274. doi:10.1016/j.socscimed.2012.08.028
risk indicators. More research is needed on the phys- Bowling, A. (2009). Predictors of mortality among a
ical health effects from bereavement associated with national sample of elderly widowed people: Analysis of
loss from a variety of medical conditions with differ- 28-year mortality rates. Age and Ageing, 38(5), 527–530.
ent levels of severity and unexpectedness. For doi:10.1093/ageing/afp108
example, we envision that the surviving spouse may Boyle, P. J., Feng, Z., & Raab, G. M. (2011). Does widow-
experience different physical health outcomes if their hood increase mortality risk? Testing for selection effects
by comparing causes of spousal death. Epidemiology,
spouse died from a car crash versus a debilitating,
22(1), 1–5. doi:10.1097/EDE.0b013e3181fdcc0b
chronic medical condition such as amyotrophic lateral Bratt, A. S., Stenstr€om, U., & Rennemark, M. (2016). The
sclerosis, or through the recurrence of potentially role of neuroticism and conscientiousness on mortality
severe medical condition such as breast cancer. risk in older adults after child and spouse bereavement.
Finally, since this study focused on the quantitative Aging & Mental Health, 20(6), 559–566. doi:10.1080/
aspects of the bereavement literature, there is a major 13607863.2015.1031638
Buckley, T., McKinley, S., Tofler, G., & Bartrop, R. (2010).
gap in understanding the cognitive and relational
Cardiovascular risk in early bereavement: A literature
processes that may influence physical health out- review and proposed mechanisms. International Journal
comes. We recognize that qualitative studies on the of Nursing Studies, 47(2), 229–238. doi:10.1016/j.ijnurstu.
experiences of bereavement can augment the current 2009.06.010
review by examining the perspectives, perceptions, Buckley, T., Mihailidou, A. S., Bartrop, R., McKinley, S.,
and experiences of widows after the loss of a spouse. Ward, C., Morel-Kopp, M.-C., … Tofler, G. H. (2011).
Although this review adds to the current evidence Haemodynamic changes during early bereavement:
Potential contribution to increased cardiovascular risk.
base on the physical and physiological outcomes of Heart, Lung and Circulation, 20(2), 91–98. doi:10.1016/j.
bereavement, there is a need to explore the experience hlc.2010.10.073
of widowhood, possibly through a qualitative evi- Buckley, T., Morel-Kopp, M.-C., Ward, C., Bartrop, R.,
dence synthesis. McKinley, S., Mihailidou, A. S., … Tofler, G. (2012).
Inflammatory and thrombotic changes in early bereave-
ment: A prospective evaluation. European Journal of
Disclosure statement Preventive Cardiology, 19(5), 1145–1152. doi:10.1177/
1741826711421686
No potential conflict of interest was reported by
Buckley, T., Stannard, A., Bartrop, R., McKinley, S., Ward,
the authors.
C., Mihailidou, A. S., … Tofler, G. (2012). Effect of early
bereavement on heart rate and heart rate variability. The
ORCID American Journal of Cardiology, 110(9), 1378–1383. doi:
10.1016/j.amjcard.2012.06.045
Umair Majid http://orcid.org/0000-0002-4581-7714 Buckley, T., Sunari, D., Marshall, A., Bartrop, R., McKinley,
S., & Tofler, G. (2012). Physiological correlates of
bereavement and the impact of bereavement interven-
References tions. Dialogues in Clinical Neuroscience, 14(2), 129.
Abel, E. L., & Kruger, M. L. (2009). The widowhood effect: Bunker, S. J., Colquhoun, D. M., Esler, M. D., Hickie, I. B.,
A comparison of Jews and Catholics. Omega, 59(4), Hunt, D., Jelinek, V. M., … Tonkin, A. M. (2003). Stress
325–337. doi:10.2190/OM.59.4.c and coronary heart disease: Psychosocial risk factors.
Allegra, J., Ezeamama, A., Simpson, C., & Miles, T. (2015). Medical Journal of Australia, 178(6), 272–276.
Population-level impact of loss on survivor mortality risk. Burns, R. A., Browning, C. J., & Kendig, H. L. (2015).
Quality of Life Research, 24(12), 2959–2961. doi:10.1007/ Examining the 16-year trajectories of mental health and
s11136-015-1048-x wellbeing through the transition into widowhood.
Aoun, S. M., Breen, L. J., Howting, D. A., Rumbold, B., International Psychogeriatrics, 27(12), 1979–1986. doi:10.
McNamara, B., & Hegney, D. (2015). Who needs 1017/S1041610215000472
bereavement support? A population based survey of Carey, I. M., Shah, S. M., DeWilde, S., Harris, T., Victor,
bereavement risk and support need. PloS One, 10(3), C. R., & Cook, D. G. (2014). Increased risk of acute car-
e0121101. doi:10.1371/journal.pone.0121101 diovascular events after partner bereavement: A matched

Asgeirsdottir, H. G., Valdimarsd ottir, U., F€
urst, C. J., cohort study. JAMA Internal Medicine, 174(4), 598–605.
Steineck, G., & Hauksdottir, A. (2013). Low preparedness doi:10.1001/jamainternmed.2013.14558
before the loss of a wife to cancer and the widower’s d’Epinay, C. J. L., Cavalli, S., & Guillet, L. A. (2010).
chronic pain 4–5 years later—a population-based study. Bereavement in very old age: Impact on health and rela-
Psycho-Oncology, 22(12), 2763–2770. doi:10.1002/pon. tionships of the loss of a spouse, a child, a sibling, or a
3345 close friend. Omega, 60(4), 301–325.
 DEATH STUDIES 13

Dunlay, S. M., Roger, V. L., Weston, S. A., Bangerter, L. R., Jin, L., & Chrisatakis, N. A. (2009). Investigating the mech-
Killian, J. M., & Griffin, J. M. (2017). Patient and spousal anism of marital mortality reduction: The transition to
health and outcomes in heart failure. Circulation: Heart widowhood and quality of health care. Demography,
Failure, 10(10), e004088. 46(3), 605–625. doi:10.1353/dem.0.0066
Efferth, T., Banerjee, M., & Paul, N. W. (2017). Broken King, M., Lodwick, R., Jones, R., Whitaker, H., & Petersen,
heart, tako-tsubo or stress cardiomyopathy? Metaphors, I. (2017). Death following partner bereavement: A self-
meanings and their medical impact. International Journal controlled case series analysis. PLoS One, 12(3),
of Cardiology, 230, 262–268. doi:10.1016/j.ijcard.2016.12. e0173870. doi:10.1371/journal.pone.0173870
129 King, M., Vasanthan, M., Petersen, I., Jones, L., Marston, L.,
Eini€o, E., Moustgaard, H., Martikainen, P., & Leinonen, T. & Nazareth, I. (2013). Mortality and medical care after
(2017). Does the risk of hospitalisation for ischaemic bereavement: A general practice cohort study. PLoS One,
heart disease rise already before widowhood? Journal of 8(1), e52561. doi:10.1371/journal.pone.0052561
Epidemiology and Community Health, 71(6), 599–605. Majid, U., & Ennis, J. (2018). The Role of Meaning in Life
doi:10.1136/jech-2016-207987 in Adjustment to a Chronic Medical Condition: A review.
Elwert, F., & Christakis, N. A. (2006). Widowhood and EC Psychology and Psychiatry, 7(12), 1023–1030.
race. American Sociological Review, 71(1), 16–41. doi:10. Retrieved from: https://www.ecronicon.com/ecpp/pdf/
1177/000312240607100102 ECPP-07-00371.pdf
Elwert, F., & Christakis, N. A. (2008). The effect of widow- Martikainen, P., & Valkonen, T. (1996). Mortality after the
hood on mortality by the causes of death of both spouses. death of a spouse: Rates and causes of death in a large
American Journal of Public Health, 98(11), 2092–2098. Finnish cohort. American Journal of Public Health,
doi:10.2105/AJPH.2007.114348 86(8_Pt_1), 1087–1093. doi:10.2105/AJPH.86.8_Pt_1.1087
Erlangsen, A., Runeson, B., Bolton, J. M., Wilcox, H. C., Mason, T. M., & Duffy, A. R. (2018). Complicated grief and
Forman, J. L., Krogh, J., … Conwell, Y. (2017). cortisol response: An integrative review of the literature.
Association between spousal suicide and mental, physical, Journal of the American Psychiatric Nurses Association,
and social health outcomes: A longitudinal and nation- 25, 181–188. doi:10.1177/1078390318807966
wide register-based study. JAMA Psychiatry, 74(5), M€oller, J., Bj€
orkenstam, E., Ljung, R., & Åberg Yngwe, M.
(2011). Widowhood and the risk of psychiatric care, psy-
456–464. doi:10.1001/jamapsychiatry.2017.0226
Fagundes, C. P., Brown, R. L., Chen, M. A., Murdock, chotropic medication and all-cause mortality: A cohort
study of 658,022 elderly people in Sweden. Aging &
K. W., Saucedo, L., LeRoy, A., … Heijnen, C. (2019).
Mental Health, 15(2), 259–266. doi:10.1080/13607863.
Grief, depressive symptoms, and inflammation in the
2010.513041
spousally bereaved. Psychoneuroendocrinology, 100,
Moon, J. R., Glymour, M. M., Vable, A. M., Liu, S. Y., &
190–197. doi:10.1016/j.psyneuen.2018.10.006
Subramanian, S. V. (2014). Short-and long-term associa-
Fagundes, C. P., Murdock, K. W., LeRoy, A., Baameur, F.,
tions between widowhood and mortality in the United
Thayer, J. F., & Heijnen, C. (2018). Spousal bereavement
States: Longitudinal analyses. Journal of Public Health,
is associated with more pronounced ex vivo cytokine pro-
36(3), 382–389. doi:10.1093/pubmed/fdt101
duction and lower heart rate variability: Mechanisms
Moon, J. R., Kondo, N., Glymour, M. M., & Subramanian,
underlying cardiovascular risk? Psychoneuroendocrinology, S. V. (2011). Widowhood and mortality: A meta-analysis.
93, 65–71. doi:10.1016/j.psyneuen.2018.04.010 PLoS One, 6(8), e23465doi:10.1371/journal.pone.0023465
Hebert, R. S., Prigerson, H. G., Schulz, R., & Arnold, R. M. O’Connor, M. F., Wellisch, D. K., Stanton, A. L., Olmstead,
(2006). Preparing caregivers for the death of a loved one: R., & Irwin, M. R. (2012). Diurnal cortisol in complicated
A theoretical framework and suggestions for future and non-complicated grief: Slope differences across the
research. Journal of Palliative Medicine, 9(5), 1164–1171. day. Psychoneuroendocrinology, 37(5), 725–728. doi:10.
Holland, J. M., Rozalski, V., Thompson, K. L., Tiongson, 1016/j.psyneuen.2011.08.009
R. J., Schatzberg, A. F., O’hara, R., & Gallagher- Oliveira, A. J., Rostila, M., Saarela, J., & Lopes, C. S. (2014).
Thompson, D. (2013). The unique impact of late-life The influence of bereavement on body mass index:
bereavement and prolonged grief on diurnal cortisol. The Results from a national Swedish survey. PLoS One, 9(4),
Journals of Gerontology Series B: Psychological Sciences e95201doi:10.1371/journal.pone.0095201
and Social Sciences, 69(1), 4–11. doi:10.1093/geronb/ Olsen, M. H., Bidstrup, P. E., Frederiksen, K., Rod, N. H.,
gbt051 Grønbaek, M., Dalton, S. O., & Johansen, C. (2012). Loss
Infurna, F. J., Wiest, M., Gerstorf, D., Ram, N., Schupp, J., of partner and breast cancer prognosis—a population-
Wagner, G. G., & Heckhausen, J. (2017). Changes in life based study, Denmark, 1994–2010. British Journal of
satisfaction when losing one’s spouse: Individual differen- Cancer, 106(9), 1560. doi:10.1038/bjc.2012.96
ces in anticipation, reaction, adaptation and longevity in Santacroce, A., Wastesson, J. W., H€ ohn, A., Christensen, K.,
the German Socio-economic Panel Study (SOEP). Ageing & Oksuzyan, A. (2018). Gender differences in the use of
and Society, 37(5), 899–934. doi:10.1017/S0144686X1 anti-infective medications before and after widowhood: A
5001543 register-based study. Journal of Epidemiology and
Jadhav, A., & Weir, D. (2017). Widowhood and depression Community Health, 72(6), 526–531. doi:10.1136/jech-
in a cross-national perspective: Evidence from the United 2017-210114
States, Europe, Korea, and China. The Journals of Schaefer, C., Quesenberry, C. P., & Wi, S. (1995). Mortality
Gerontology: Series B, 73, e143–e153. doi:10.1093/geronb/ following conjugal bereavement and the effects of a
gbx021 shared environment. American Journal of Epidemiology,
14 J. ENNIS AND U. MAJID

141(12), 1142–1152. doi:10.1093/oxfordjournals.aje. Economics, 32(1), 33–50. doi:10.1016/j.jhealeco.2012.10.

a117387 010
Schultze-Florey, C. R., Martınez-Maza, O., Magpantay, L., Stahl, S. T., Arnold, A. M., Chen, J. Y., Anderson, S., &
Breen, E. C., Irwin, M. R., G€ undel, H., & O’Connor, Schulz, R. (2016). Mortality after bereavement: The role
M. F. (2012). When grief makes you sick: Bereavement of cardiovascular disease and depression. Psychosomatic
induced systemic inflammation is a question of genotype. Medicine, 78(6), 697. doi:10.1097/PSY.0000000000000317
Brain, Behavior, and Immunity, 26(7), 1066–1071. doi:10. Stroebe, M. S. (2001). Handbook of bereavement research:
1016/j.bbi.2012.06.009 Consequences, coping, and care. Washington, DC:
Seifter, A., Singh, S., McArdle, P. F., Ryan, K. A., Shuldiner, American Psychological Association.
A. R., Mitchell, B. D., & Sch€affer, A. A. (2014). Analysis Subramanian, S. V., Elwert, F., & Christakis, N. (2008).
of the bereavement effect after the death of a spouse in Widowhood and mortality among the elderly: The modi-
the Amish: A population-based retrospective cohort fying role of neighborhood concentration of widowed
study. BMJ Open, 4(1), e003670. doi:10.1136/bmjopen- individuals. Social Science & Medicine, 66(4), 873–884.
2013-003670 doi:10.1016/j.socscimed.2007.11.029
Shah, S. M., Carey, I. M., Harris, T., DeWilde, S., Victor, Sullivan, A. R., & Fenelon, A. (2014). Patterns of widow-
C. R., & Cook, D. G. (2013a). The effect of unexpected hood mortality. The Journals of Gerontology Series B:
bereavement on mortality in older couples. American Psychological Sciences and Social Sciences, 69(1), 53–62.
Journal of Public Health, 103(6), 1140–1145. doi:10.2105/ doi:10.1093/geronb/gbt079
AJPH.2012.301050 Tseng, F. M., Petrie, D., Wang, S., Macduff, C., & Stephen,
Shah, S. M., Carey, I. M., Harris, T., DeWilde, S., Victor, A. I. (2018). The impact of spousal bereavement on hos-
C. R., & Cook, D. G. (2013b). Impact of partner bereave- pitalisations: Evidence from the Scottish Longitudinal
ment on quality of cardiovascular disease management. Study. Health Economics, 27(2), e120–e138. doi:10.1002/
Circulation, 128(25), 2745–2753. doi:10.1161/ hec.3573
CIRCULATIONAHA.113.004122 Vable, A. M., Subramanian, S. V., Rist, P. M., & Glymour,
Shah, S. M., Carey, I. M., Harris, T., DeWilde, S., Victor, M. M. (2015). Does the “widowhood effect” precede
C. R., & Cook, D. G. (2016). The mental health and mor- spousal bereavement? Results from a nationally represen-
tality impact of death of a partner with dementia. tative sample of older adults. The American Journal of
International Journal of Geriatric Psychiatry, 31(8), Geriatric Psychiatry, 23(3), 283–292. doi:10.1016/j.jagp.
929–937. doi:10.1002/gps.4411 2014.05.004
Shear, M. K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Von Elm, E., Altman, D. G., Egger, M., Pocock, S. J.,
Duan, N., … Keshaviah, A. (2011). Complicated grief Gøtzsche, P. C., & Vandenbroucke, J. P. (2007). The
and related bereavement issues for DSM-5. Depression Strengthening the Reporting of Observational Studies in
and Anxiety, 28(2), 103–117. doi:10.1002/da.20780 Epidemiology (STROBE) statement: Guidelines for
Siflinger, B. (2017). The effect of widowhood on mental reporting observational studies. Annals of Internal
health-an analysis of anticipation patterns surrounding Medicine, 147(8), 573–577. doi:10.7326/0003-4819-147-8-
the death of a spouse. Health Economics, 26(12), 200710160-00010
1505–1523. doi:10.1002/hec.3443 Young, M., Benjamin, B., & Wallis, C. (1963). The mortality
Simeonova, E. (2013). Marriage, bereavement and mortality: of widowers. Lancet (London, England)), 2(7305),
The role of health care utilization. Journal of Health 454–457. doi:10.1016/s0140-6736(63)92193-7