Psychological Science http://pss.sagepub.

com/

Death Without God: Religious Struggle, Death Concerns, and Depression in the Terminally III
Donald Edmondson, Crystal L. Park, Stephenie R. Chaudoir and Jennifer H. Wortmann
Psychological Science 2008 19: 754
DOI: 10.1111/j.1467-9280.2008.02152.x

The online version of this article can be found at:

http://pss.sagepub.com/content/19/8/754

Published by:

http://www.sagepublications.com

On behalf of:

Association for Psychological Science

Additional services and information for Psychological Science can be found at:

Email Alerts: http://pss.sagepub.com/cgi/alerts

Subscriptions: http://pss.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Aug 1, 2008

What is This?

Downloaded from pss.sagepub.com at EMORY UNIV on June 23, 2014

 P SY CH OL OG I C AL S CIE N CE

Research Report

Death Without God

Religious Struggle, Death Concerns, and Depression in the
Terminally Ill
Donald Edmondson, Crystal L. Park, Stephenie R. Chaudoir, and Jennifer H. Wortmann

University of Connecticut

ABSTRACT—Religious worldviews often provide comfort Empirical studies of religion’s role at the end of life are few, but
near the end of life, but they can cause distress if life cir- generally suggest that religious belief is beneficial and provides
cumstances are perceived as evidence of God’s disfavor. increasing comfort as death nears (Idler, Kasl, & Hays, 2001).
This study, the first to test terror management theory However, some people experience significant religious struggle as
(TMT) with terminally ill participants, examined the hy- they near life’s end, and this struggle increases anxiety, depression,
pothesis that concerns about death mediate the relation- and functional disability (Burker, Evon, Sedway, & Egan, 2004).
ship between religious struggle (and religious comfort) and We employed terror management theory (TMT; Solomon,
depression in the terminally ill. Ninety-eight patients with Greenberg, & Pyszczynski, 1991) to examine whether religious
end-stage congestive heart failure (CHF) completed mea- struggle is a mechanism through which religion influences psycho-
sures of religious comfort, religious struggle, belief in an logical well-being at the end of life. According to TMT, indi-
afterlife, concerns about death, and depression. In sepa- viduals neutralize the ubiquitous, paralyzing fear of death by
rate hierarchical linear regression models that controlled adopting cultural worldviews that imbue the world with meaning
for degree of belief in an afterlife, death concerns fully and order, lend meaning to human activity, and provide self-
mediated the relationships between religious struggle and esteem. Individuals invested in these cultural worldviews are
depression and between religious comfort and depression. rewarded with assurances that life is ultimately meaningful and
These findings suggest that religious struggle is a break- that personal mortality will be transcended, either symbolically
down in the terror management system that leaves the or literally (Solomon et al., 1991).
individual vulnerable to the terror of death, and that Religious worldviews are among the most common cultural
properly functioning religious worldviews offer comfort by worldviews; experimental manipulations have demonstrated that
buffering the individual against death concerns. functioning religious worldviews can reduce the accessibility of
death-related thoughts in situations in which one’s mortality is
Many individuals experience anxiety about death and dying, salient (Jonas & Fischer, 2006). When functioning effectively,
particularly near the end of their lives (Ardelt & Koenig, 2006). religious worldviews provide individuals with religious comfort, a
When one is faced with death, finding meaning and value in life generally pleasant and orderly view of the world and their place in
typically becomes a central focus of attention (Sulmasy, 2006); it, a promise of life’s eternal significance, and an ultimate source of
these concerns are often more prominent than complaints about self-worth (Baumeister, 1991). However, religious worldviews can
physical symptoms in terminally ill patients (Portenoy et al., become sources of self-esteem threat, uncertainty, or despair if they
1994). Many turn to religion because religion break down, such as is the case in religious struggle, when indi-
viduals perceive that they are being punished or abandoned by God
provides order to the world even in the presence of physical de- in the course of stressful life circumstances. In religious struggle,
cline, social losses, suffering, and impending death and offers an people perceive that God continues to exist and exert control but
existential meaning that provides a sense of peace and a recog-
does not provide them with care and comfort, and this perception
nition of one’s place in the broader cosmic context. (Ardelt &
impairs the effectiveness of their religious worldview in defending
Koenig, 2006, p. 189)
against death terror. Religious struggle of this kind has been shown
to increase depression (Pargament, Koenig, Tarakeshwar, & Hahn,
Address correspondence to Donald Edmondson, University of Con-
necticut, Box U-1020, Storrs, CT 06269, e-mail: donald.edmondson@ 2004) and mortality risk over the course of 1 year in medically ill
uconn.edu. older adults (Pargament, Koenig, Tarakeshwar, & Hahn, 2001).

754 Downloaded fromr

Copyright 2008 Association
pss.sagepub.com for UNIV
at EMORY Psychological
on June 23, Science
2014 Volume 19—Number 8
 D. Edmondson et al.

We propose that under circumstances of increasing awareness Association (NYHA) Classification III or IV (Criteria Committee
of mortality, religious worldviews may break down. The psycho- of the NYHA, 1994), indicating end-stage CHF, were recruited
logical consequences of a terror management system under from a regional cardiology practice in Cincinnati, Ohio. The
chronic siege have not been addressed empirically, as TMT has sample was European American (55%), African American
never been tested in terminally ill participants. One way to test (38%), Native American (5%), and biracial (1%). Participants
the validity of TMT in the face of chronic and intense mortality included Protestants (74%), Catholics (17%), Jews (1%), and
salience is to examine worldview functioning in the terminally ill. individuals with no religious affiliation (8%). Because this study
Individuals with terminal conditions, such as severe congestive focused on the function of explicitly religious worldviews,
heart failure (CHF), live in a chronic state of heightened mortality participants who reported being ‘‘not at all’’ religious (n 5 6)
salience, characterized by uncertainty regarding their time and were excluded from analyses. In addition, participants with
mode of death (Gottlieb, 2003) and symptoms of physical dete- missing data for any measure (n 5 7) were excluded. Therefore,
rioration (e.g., difficult and painful breathing) that are not easily the final sample consisted of 98 participants.
kept out of consciousness (Rohrbaugh et al., 2002). Thus,
individuals with terminal illnesses, including CHF, should be
particularly susceptible to the terror accompanying increased Procedure
accessibility of death-related thoughts (Winters, 1999). At their regularly scheduled cardiology appointments, partici-
TMT holds that cultural worldviews allow individuals to avoid pants completed a demographic questionnaire and the measures
thinking about their own mortality in day-to-day functioning, by for this study. Filler items assessing quality of life were included
offering a means for transcending death (Solomon et al., 1991). following the instruments measuring belief in an afterlife and
We propose that to the extent that religious worldviews are crip- concerns about death.
pled under the weight of chronic mortality salience (as reflected Religious comfort was assessed using the three-item Daily
in religious struggle), CHF patients are likely to experience Spiritual Experiences Scale (Fetzer Institute & National Insti-
depression (e.g., Simon, Arndt, Greenberg, Pyszczynski, & Solo- tute on Aging, 1999). Participants rated how often they felt the
mon, 1998). Because religious struggle would indicate a poorly presence of God, found strength or comfort in their religion or
functioning terror management system that is incapable of spirituality, and felt God’s love directly or through other people,
buffering death-related thoughts, we hypothesized that increased using a scale from 0 (never or almost never) to 6 (many times a
death concerns would mediate the relationship between religious day; a 5 .80). The sum of these ratings was the religious-comfort
struggle and increased depression. Conversely, because religious score (M 5 14.09, SD 5 2.94).
comfort would indicate a properly functioning terror management Religious identification was measured by the item, ‘‘To what
system that is capable of buffering death-related thoughts, we extent do you consider yourself a religious person?’’ (Fetzer
hypothesized that decreased death concerns would mediate the Institute & National Institute on Aging, 1999). The scale for
relationship between religious comfort and decreased depression. responding to this item ranged from 0 (not at all) to 4 (extremely).
If, indeed, fewer death-related thoughts were the mechanism Religious struggle was measured using the six-item Spiritual
through which religious comfort influenced depression, that fact Strain Scale (Exline & Yali, 2000; a 5 .88), which assesses per-
would suggest that the influence of religious struggle on depres- ception of a ruptured relationship with God (e.g., ‘‘To what extent do
sion through increased death-related thoughts is not merely a you currently feel abandoned by God?’’). Participants responded to
function of global negative affect producing morbid cognitions each question on a scale from 0 (not at all) to 10 (extremely). The
and mood disruption. Finally, we hypothesized that patients’ sum of these ratings was the religious-struggle score (M 5 3.44,
degree of belief in the existence of a literal life after death would SD 5 6.65); 42% of the sample had scores greater than 0.
not account for the relationships among religious struggle or Belief in an afterlife was measured using a single item, ‘‘Do
comfort, death concerns, and depression. Although the prospect you believe there is life after death?’’ Response options were 0
of literal immortality has been posited as the aspect of religious (no), 1 (uncertain), and 2 (yes). Most participants (78%) believed
worldviews most likely to serve anxiety-buffering functions, the in an afterlife (14% undecided, 8% no).
critical function of religious worldviews appears to be the pro- Concerns about death were assessed as a proxy for accessi-
vision of self-esteem (Dechesne et al., 2003). That is, although the bility of death-related thoughts. The three items for this measure
beliefs in an afterlife afforded by many religious worldviews may came from the Death and Dying subscale (a 5 .83) of the World
be beneficial, they are insufficient to keep terror at bay. Health Organization Quality of Life measure (Harper & Power,
1998). Individuals rated the degree to which they worried about
METHOD (a) death, (b) ‘‘the thought of not being able to die the way
you would want to,’’ and (c) ‘‘how and where you will die,’’
Participants using scales from 0 (not at all) to 4 (extremely). Ratings were
One hundred eleven individuals (mean age 5 66.85 years, summed to calculate a total score. Death concerns varied widely
SD 5 11.65; 61% male, 39% female) with a New York Heart (M 5 5.38, SD 5 2.77).

Volume 19—Number 8 Downloaded from pss.sagepub.com at EMORY UNIV on June 23, 2014 755
 Religious Struggle

Depression was measured using the Center for Epidemio-

logical Studies Depression Scale (a 5 .73; Radloff, 1977), on Death
which respondents rate their experience of 20 depressive Concerns .27** (.34***)
(.39**)
symptoms in the past week. The rating scale ranges from 0
(rarely or none of the time) to 4 (most or all of the time). The mean
in this study’s sample (15.95, SD 5 9.76) approximated the Religious Depressive
.17 (.27**)
Struggle Symptoms
suggested screening cutoff for clinical depression (i.e., 16).

RESULTS Death
Concerns
(−.30**) .30** (.34***)
Bivariate correlations (see Table 1) indicated that mediation
tests were appropriate, as the predictors (religious struggle and
comfort) were related to the criterion (depression) and the pro- Religious −.13 (−.22*) Depressive
posed mediator (death concerns), and the proposed mediator Comfort Symptoms
was related to the criterion. To test whether death concerns Fig. 1. Results of hierarchical regressions testing death concerns as a
mediated the relationship between religious struggle and de- mediator of the effect of (a) religious struggle and (b) religious comfort on
pression, we entered religious struggle in Step 1 of a hierarchical depressive symptoms. Numbers in parentheses represent the zero-order
correlations between two variables. Numbers outside parentheses rep-
regression with depression as the criterion. In Step 2, death resent the standardized regression coefficients (b) for a model with re-
concerns was entered as a covariate. As expected, death con- ligious struggle or comfort as the independent variable, death concerns as
cerns mediated the effect of religious struggle on depression the mediator variable, and depressive symptoms as the criterion.
(Fig. 1a). Entering death concerns significantly improved the
model, DR2 5 .06, p < .01, which explained 14% (adjusted R2) entered in Step 1, and religious comfort was entered in Step 2.
of the variance in depressive symptoms, F(2, 95) 5 7.59, The entry of religious comfort did not improve the model,
p < .001. To see whether the hypothesized mediation model best DR2 5 .01, p > .05; this result provides further support for the
represented the relationships among the variables, we tested an original, hypothesized model.
alternative model with religious struggle specified as a mediator Conceivably, religious struggle and comfort could merely
of the relationship between death concerns and depression. reflect belief in an afterlife. To test whether this belief accounted
When death concerns was entered in Step 1, the entry of reli- for the relationships among religious struggle or comfort, death
gious struggle in Step 2 did not significantly improve the model, concerns, and depression, we entered it as a covariate in both
DR2 5 .02, p > .05; this result provides further support for the models. When belief in an afterlife was entered as a covariate
original, hypothesized model. simultaneously with religious struggle in Step 1 of the first
We conducted a parallel analysis to test whether death con- model, the relationships among religious struggle, death con-
cerns mediated the relationship between religious comfort and cerns, and depression were largely unchanged. In Step 1, the
depression. As expected, death concerns fully mediated the effect model was significant, adjusted R2 5 .09, F(2, 95) 5 4.27, p < .05,
of religious comfort on depression (Fig. 1b). Entering death and religious struggle was positively related to depression, b 5
concerns significantly improved the model, DR2 5 .08, p < .01, .27, p < .01, whereas belief in an afterlife was unrelated to de-
which explained 13% (adjusted R2) of the variance in depressive pression, b 5 .09, p > .05. In Step 2, the entry of death
symptoms, F(2, 95) 5 7.05, p < .01. In an alternative model concerns improved the model, DR2 5 .06, p < .05; death
testing religious comfort as a mediator of the relationship concerns was positively related to depression, b 5 .27, p < .05;
between death concerns and depression, death concerns was and the relationship between religious struggle and depression
became statistically nonsignificant, b 5 .16, p > .05. The final
model, F(3, 94) 5 5.28, p < .01, explained 14% (adjusted R2) of
TABLE 1 the variance in depressive symptoms.
Bivariate Correlations Among the Variables When entered as a covariate in the second model, belief in an
Religious Religious Death afterlife had some impact on the model. In Step 1, the simul-
Variable Depression struggle comfort concerns taneous entry of religious comfort and belief in an afterlife
Depression yielded a nonsignificant model, adjusted R2 5 .03, F(2, 95) 5 2.36,
Religious struggle .27nn p 5 .10, though the relationship between religious comfort and
Religious comfort .22n .36nn depression was unaltered, b 5 .22, p 5 .06. The entry of death
Death concerns .34nn .39nn .30nn concerns in Step 2 significantly improved the model, DR2 5 .10,
Belief in an afterlife .10 .05 .51nn .06 p < .01. Again, death concerns was positively related to
Note. N 5 98. depression, b 5 .30, p < .01, and fully mediated the relationship
n
p < .05. nnp < .01. between religious comfort and depression. Neither religious

756 Downloaded from pss.sagepub.com at EMORY UNIV on June 23, 2014 Volume 19—Number 8
 D. Edmondson et al.

comfort, b 5 .11, p > .05, nor belief in an afterlife, b 5 .03, generally, provide external validation for the fundamental as-
p > .05, was a significant predictor in the final model, F(3, 94) 5 sertions of TMT. Our findings suggest that properly functioning
4.68, p < .01, which explained 10% (adjusted R2) of the vari- religious worldviews offer comfort by buffering the individual
ance in depressive symptoms. against death concerns near life’s end, but that the breakdown
of these worldviews (i.e., the experience of religious struggle)
DISCUSSION leaves the individual vulnerable to the terror of death.
However, the study’s limitations must be considered. The
In this study, we used the TMT framework to examine the impact cross-sectional design allows only the inference of temporal
of religious belief on psychological well-being in a sample relationships based on TMT and relevant experimental work.
of terminally ill CHF patients, a group for whom the inevitabil- Also, our death-concerns measure may not have served well as a
ity of death is a constant reality (Winters, 1999), as opposed proxy for death-thought accessibility, but may have reflected
to a philosophical abstraction (Arndt, Cook, Goldenberg, & some other process or trait. Further, our sample was predomi-
Cox, 2007). We proposed that cultural worldviews and their nantly Christian, and most participants believed in an afterlife;
anxiety-buffering functions may be compromised in some such generalizability to individuals of other faith traditions (e.g.,
patients. Specifically, we hypothesized that the effects of reli- Buddhism) remains to be demonstrated. Despite these limita-
gious struggle on depression previously noted in clinical studies tions, these findings provide valuable insight into the influence
of medically ill older adults (Pargament et al., 2004) would of religious belief near the end of life and are among the first to
be explained by the increase in death-related thoughts that highlight the workings of the terror management system under
has been shown, in the laboratory, to follow mortality-salience conditions of chronic and extreme mortality salience.
inductions (Arndt, Greenberg, Solomon, & Pyszczynski, 1997).
Our findings suggest that, indeed, a primary mechanism by
which religious struggle affects depression in terminally ill CHF Acknowledgments—The authors thank The Fetzer Institute
patients is through increased concerns about personal mortality. for funding and Jamie Arndt and Mary Alice Mills for indispens-
To show that increased death concerns accompanying greater able assistance with earlier drafts.
degrees of religious struggle truly reflected the terror manage-
ment functions of the religious aspect of participants’ cultural
REFERENCES
worldviews, and were not simply a by-product of negative affect,
we examined whether the converse was also true, that is, whether Ardelt, M., & Koenig, C.S. (2006). The role of religion for hospice
death concerns mediated the relationship between religious patients and relatively healthy older adults. Research on Aging,
comfort and depression. We found that they did. Interestingly, 28, 184–215.
belief in an afterlife was related only to religious comfort (see Arndt, J., Cook, A., Goldenberg, J.L., & Cox, C.R. (2007). Cancer and
Table 1), which suggests that the expectation of literal immor- the threat of death: The cognitive dynamics of death-thought
suppression and its impact on behavioral health intentions. Jour-
tality is a significant aspect of the comfort that religious nal of Personality and Social Psychology, 92, 12–29.
worldviews provide individuals near death. Arndt, J., Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Sup-
Given that religious struggle follows stressful events (Par- pression, accessibility of death-related thoughts, and cultural
gament et al., 2001), we were compelled to test, but found no worldview defense: Exploring the psychodynamics of terror man-
empirical support for, the alternative notion that religious agement. Journal of Personality and Social Psychology, 73, 5–18.
struggle (or comfort) might mediate the relationship between Baumeister, R.F. (1991). Meanings of life. New York: Guilford.
Burker, E.J., Evon, D.M., Sedway, J.A., & Egan, T. (2004). Religious
death concerns and depression. Our conclusions regarding the coping, psychological distress and disability among patients with
relationships among these variables are also supported by end-stage pulmonary disease. Journal of Clinical Psychology in
converging evidence from experimental studies suggesting that Medical Settings, 11, 179–193.
people may rely on religious worldviews in the face of mortality Criteria Committee of the New York Heart Association. (1994). No-
salience (Norenzayan & Hansen, 2006), but are buffered from menclature and criteria for diagnosis of diseases of the heart and
great vessels (9th ed.). Boston: Little, Brown, & Co.
death concerns only if these worldviews are fully functioning
Dechesne, M., Pyszczynski, T., Arndt, J., Ransom, S., Sheldon, K.M.,
(i.e., provide meaning and value; Jonas & Fischer, 2006). van Knippenberg, A., & Janssen, J. (2003). Literal and symbolic
Further, mortality salience has been shown to cause individuals immortality: The effect of evidence of literal immortality on self-
to automatically orient ‘‘toward emotionally positive, pleasant esteem striving in response to mortality salience. Journal of
information and associations’’ (DeWall & Baumeister, 2007, Personality and Social Psychology, 84, 722–737.
p. 988). This suggests that increased death concerns would in- DeWall, C.N., & Baumeister, R.F. (2007). From terror to joy: Automatic
tuning to positive affective information following mortality sa-
crease religious comfort, not struggle, for individuals with intact
lience. Psychological Science, 18, 984–990.
worldviews. Exline, J.J., & Yali, A.M. (2000). Guilt, discord and alienation: The
These findings lend further support to the idea that religious role of religious strain in depression and suicidality. Journal
worldviews serve a terror management function and, more of Clinical Psychology, 56, 1481–1496.

Volume 19—Number 8 Downloaded from pss.sagepub.com at EMORY UNIV on June 23, 2014 757
 Religious Struggle

Fetzer Institute & National Institute on Aging. (1999). Multidimen- Assessment Scale: An instrument for the evaluation of symptom
sional measurement of religiousness/spirituality for use in health prevalence, characteristics and distress. European Journal of
research. Kalamazoo, MI: Fetzer Institute. Cancer, 30, 1326–1336.
Gottlieb, S. (2003). Palliative care in heart failure. Advanced Studies in Radloff, L.S. (1977). The CES-D scale: A self-report depression scale
Medicine, 3, 456–463. for research in the general population. Applied Psychological
Harper, A., & Power, M. (1998). Development of the World Health Measurement, 1, 385–401.
Organization WHOQOL-BREF Quality of Life Assessment. Psy- Rohrbaugh, M.J., Cranford, J.A., Shoham, V., Nicklas, J.M., Sonnega,
chological Medicine, 28, 551–558. J.S., & Coyne, J.C. (2002). Couples coping with congestive heart
Idler, E.L., Kasl, S.V., & Hays, J.C. (2001). Patterns of religious failure: Role and gender differences in psychological distress.
practice and belief in the last year of life. Journals of Gerontology Journal of Family Psychology, 16, 3–13.
Series B: Psychological Sciences and Social Sciences, 56, 326– Simon, L., Arndt, J., Greenberg, J., Pyszczynski, T., & Solomon, S. (1998).
334. Terror management and meaning: Evidence that the opportunity to
Jonas, E., & Fischer, P. (2006). Terror management and religion: defend the worldview in response to mortality salience increases
Evidence that intrinsic religiousness mitigates worldview defense the meaningfulness of life in the mildly depressed. Journal of
following mortality salience. Journal of Personality and Social Personality, 66, 359–382.
Psychology, 91, 553–567. Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). A terror
Norenzayan, A., & Hansen, I.G. (2006). Belief in supernatural agents management theory of social behavior: The psychological func-
in the face of death. Personality and Social Psychology Bulletin, tions of self-esteem and cultural worldviews. In M. Zanna (Ed.),
32, 174–187. Advances in experimental social psychology (Vol. 24, pp. 93–159).
Pargament, K.I., Koenig, H.G., Tarakeshwar, N., & Hahn, J. (2001). San Diego, CA: Academic Press.
Religious struggle as a predictor of mortality among medically ill Sulmasy, D.P. (2006). Spiritual issues in the care of dying patients: ‘‘It’s
elderly patients: A 2-year longitudinal study. Archives of Internal okay between me and God.’’ Journal of the American Medical
Medicine, 161, 1881–1885. Association, 296, 1385–1392.
Pargament, K.I., Koenig, H.G., Tarakeshwar, N., & Hahn, J. (2004). Winters, C.A. (1999). Heart failure: Living with uncertainty. Pro-
Religious coping methods as predictors of psychological, physi- gressive Cardiovascular Nursing, 14, 85–91.
cal and spiritual outcomes among medically ill elderly patients: A
two-year longitudinal study. Journal of Health Psychology, 9,
713–730.
Portenoy, R.K., Thaler, H.T., Kornblith, A.B., Lepore, J.M., Fried-
lander-Klar, H., Kiyasu, E., et al. (1994). The Memorial Symptom (RECEIVED 8/7/07; REVISION ACCEPTED 2/14/08)

758 Downloaded from pss.sagepub.com at EMORY UNIV on June 23, 2014 Volume 19—Number 8