Professional Documents
Culture Documents
Forgiveness, Ego-Integrity, and Depressive Symptoms in Community-Dwelling and Residential Elderly Adults
Forgiveness, Ego-Integrity, and Depressive Symptoms in Community-Dwelling and Residential Elderly Adults
Forgiveness, ego-integrity, and depressive symptoms in community-dwelling and residential elderly adults. Journals of Gerontology, Series B:
Psychological Sciences and Social Sciences, doi:10.1093/geronb/gbu146
1
Faculty of Psychology and Educational Sciences, KU Leuven University of Leuven, Belgium.
2
Department of Psychology, Luther College Decorah, Iowa.
Objectives. This study aimed first, to investigate the relation between forgivingness and depressive symptoms in
Methods. Community-dwelling elderly individuals (n = 280, M = 75.98 years) and residential elderly individuals
(n = 205, M = 83.20 years) in Belgium filled out questionnaires on forgivingness, depressive symptoms, and the develop-
mental task of integrity/despair. Structural equation modeling was used to test the mediational role of integrity/despair in
the relation between forgivingness and depression in both samples. Furthermore, multigroup analyses tested the conver-
gence of the mediational models in the community-dwelling elderly individuals and the residential elderly individuals.
Results. Forgivingness and depression were negatively associated in both residential elderly individuals and com-
munity-dwelling elderly individuals. The developmental task of finding integrity and avoiding despair showed to be a
significant mediator in this relationship.
Discussion. Framed within a life-span perspective, we showed that the developmental task of finding a balance
between integrity and despair is an important intrapersonal mechanism through which forgivingness is related with
depressed feelings for residential as well as community-dwelling elderly individuals.
is an additional key element of forgiveness (Brown, 2003; adults (Toussaint et al., 2008), and adult patients of a pri-
Lawler-Row, Karremans, Scott, Edlis-Matityahou, & mary care clinic (Nsamenang, Webb, Cukrowicz, & Hirsch,
Edwards, 2008; Worthington, 2005). Fincham (2000), for 2013).
example, labeled these two dimensions of forgiveness “a
negative dimension” (with focus on reduction in resent-
Integrity-Despair and Connections to Mental Health
ment) and “a positive dimension” (with focus on foster-
The idea of forgiveness being pivotal in late life can be
ing compassion). Lawler and colleagues (2008) stated that
framed within the life stage theory of Erikson (1982). In
letting go of one’s right of resentment can be expressed in
Erikson’s model, adults in the final stage of life must come
a reduction in negative emotions (e.g., anger, hostility),
to terms with their past and need to find meaning in their
negative behaviors (e.g., revenge), or negative thoughts
lives in the face of mortality. The crucial task is to come to
(e.g., rumination). Offering undeserved compassion, on the
ego-integrity, which involves a process whereby individu-
other hand, involves a more positive process of increasing
als must “come to terms with the lives they have lived and
anticipate that forgiveness becomes more salient in light of environmental control, and increased levels of formalized
one’s own finality. care might reveal a different pattern of findings. Because to
Within Erikson’s (1982) model of life-span develop- our knowledge there are no earlier empirical studies avail-
ment, we argue that being able to forgive, as formalized in able in this area, we do not have specific hypotheses with
dispositional forgiveness, might indeed facilitate the devel- regard to possible differences between living circumstances
opmental task of integrity and despair. Past transgressions and thus this part of the study is exploratory.
are often associated with anger and hurt serving as road-
blocks for the process of life review. The painful memo-
ries associated with the transgressions can create feelings Method
of “unresolvedness,” bitterness, and rumination, all of them
Participants and Procedure
blocking the process of accepting one’s life and leading,
Participants were recruited with the help of nine psychol-
ultimately, to despair. The ability to forgive, on the other
ogy students pursuing master’s degrees. Exclusion criteria
Vansteenkiste (2009). The measure has two subscales: ego- Items were rated in a 4-point Likert scale ranging from 1
integrity (8 items, e.g., “I can accept my faults and missed (totally disagree) to 4 (totally agree).
opportunities” and “I’m able to accept the fine and the dif-
ficult moments of my life”) and despair (10 items, e.g., “I
look back upon my life with a feeling of discontent and Data Analysis
regret” and “I’m still feeling some bitterness on past rela- Descriptive analyses and structural equation modeling
tions”). Previous studies confirmed the factorial structure were used as the analytic techniques. Descriptive analyses
and showed that both dimensions were related in meaning- were first conducted using SPSS 20.0. Amos 22 was used to
ful ways to late adults’ well-being, death attitudes, self- conduct confirmatory factor analyses and assess measure-
esteem, and depressive symptoms (Dezutter, Wiesmann, ment invariance across the residential elderly and commu-
Apers, & Luyckx, 2013; Soenens & Duriez, 2012). To nity-dwelling participants. Structural equation models were
reduce the participation burden for the elderly individuals, tested to evaluate the extent to which associations between
Table 1. Correlations Between the Study Variables for the Participants in Elderly Homes (Above the Diagonal) and the Participants
Living Independently (Below the Diagonal)
Measures
Depressive
symptoms Integrity Despair Forgiveness
[1.67/.56] [3.36/.67] [1.69/.71] [2.93/.74] Age Cronbach’s α
Depressive symptoms [1.67/.48] — −.38** .33** −.24** .02 .77
Integrity [3.35/.53] −.31** — −.49** .40** .08 .83
Despair [1.98/.66] .31** −.28** — −.42** −.20** .76
Forgiveness [2.59/.56] −.26** .16* −.18** — .18** .62
Age .00 −.15* .11 .05
Cronbach’s α .79 .72 .74 .58 —
Subsequent mean-level analyses, using a set of multivari- Metric invariance indicates that the factor loadings are con-
ate analysis of variance’s followed by Tukey’s Honestly sistent across groups. Factor variance/covariance invariance
Significant Difference tests, showed no significant dif- tests whether the variances and covariances of the latent
ferences between men and women on our study variables factors are equivalent across groups. Residual invariance
(forgiveness, integrity, despair, and depressive symptoms) adds, further still, the requirement that all item residuals are
in the community-dwelling elderly (F(4, 245) = 1.84, invariant across groups. Because of the complexity of mul-
ns) or in the residential elderly sample (F(4, 193) = 2.40, tiple group confirmatory factor analysis models, root mean
ns). Significant differences between the marital status square error of approximation (RMSEA) is the preferred
were found on depressive symptoms (F(4, 244) = 3.02, choice in assessing fit due to it known distributional prop-
p < .05, ή2 = 0.05) and despair (F(4,244) = 3.45, p < .01, erties and because it remains unaffected by model com-
ή2 = 0.05) in the community-dwelling elderly population plexity or sample size (Browne & Cudeck, 1989; Wu, Li,
and on forgiveness (F(4,193) = 3.86, p < .01, ή2 = 0.07) and & Zumbo, 2007). Browne and Cudeck (1989) recommend
Table 3. Configural, Metric, Variance/Covariance, and Residual Invariant Factor Loadings for Confirmatory Factor Model of Forgiveness,
Integrity and Despair, and Depressive Symptoms
Variance/
Configural Metric covariance Residual
Construct/item UL SL UL SL UL SL UL SL UL SL
Forgiveness
1. Get over it quickly 1.00 0.51 1.00 0.59 1.00 0.53 1.00 0.59 1.00 0.57
2. Harbor grudgesa 0.80 0.39 1.11 0.61 0.99 0.49 0.99 0.55 1.02 0.54
3. Think about it a lot afterwardsa 1.22 0.57 0.92 0.64 0.99 0.49 1.10 0.61 1.00 0.60
4. Just to forgive and forget 0.74 0.42 0.58 0.45 0.62 0.37 0.67 0.46 0.63 0.44
Integrity
1. Accept fine and difficult moments of life 1.00 0.45 1.00 0.72 1.00 0.56 1.00 0.62 1.00 0.60
2. Cope with negative experiences from past 1.32 0.48 1.17 0.63 1.20 0.56 1.20 0.62 1.27 0.59
(see Table 3), and factor correlations ranged from r = .48 through integrity and despair on depression had an accepta-
to r = −.59. All loadings (p < .001) and factor correlations ble fit, RMSEA = .05, 90% CI = 0.05 – 0.06, and accounted
(p < .05) for all models were statistically significant. for 33% of the variability in depressive symptoms (see
Table 5). Higher levels of forgiveness were associated with
Structural Models higher levels of integrity (B = 0.94, β = 0.49, p < .001) and
Having established the invariant measurement model, lower levels of despair (B = −0.99, β = −0.46, p < .001) and
we next tested a series of latent structural models exam- forgiveness, along with the other nonsignificant sociodemo-
ining direct effects, indirect effects, and the invariance of graphic predictors, accounted for 25% of the variability in
these parameters. All structural models showed accept- each variable. Higher levels in despair were associated with
able fit (RMSEA < .08) with measurement and structural higher levels of depressive symptoms (B = 0.15, β = 0.19,
weights, factor variance/covariances, and structural and p < .05), and higher levels of integrity were associated with
measurement residuals constrained across groups. Hence, lower levels of depressive symptoms (B = −0.24, β = −0.28,
all subsequent coefficients are reported from these fully p < .001). The total indirect effect of forgiveness on depres-
constrained models. The first structural model exam- sive symptoms through integrity and despair was significant
ined the associations between sociodemographic vari- (B = −0.38, β = −0.23, p < .01). The direct effect of forgive-
ables and depression. This model was an acceptable fit, ness on depression in this model was noticeably reduced
RMSEA = .07, 90% CI = 0.06 – 0.08, but accounted for from the direct effect only model but remained statistically
only 5% of the variance in depressive symptoms and none significant (B = −0.32, β = −0.19, p < .05).
of the variables emerged as unique predictors (see Table 4). Single mediator models can be used to investigate the indi-
The direct effect model of forgiveness on depression had an vidual indirect effects of forgiveness on depressive symp-
acceptable fit, RMSEA = .06, 90% CI = 0.05 – 0.07, and toms through integrity only and despair only. Tests of the
accounted for 20% of the variability in depressive symp- individual indirect effects in separate models allow one to
toms. Higher levels of forgiveness were associated with determine the extent to which each variable is responsible for
lower levels of depression (B = −0.45, β = −0.41, p < .001), the overall indirect effect. The indirect effect model of for-
(see Table 4). The indirect effects model of forgiveness giveness through integrity on depression had an acceptable
FORGIVENESS, EGO-INTEGRITY, AND DEPRESSIVE SYMPTOMS Page 7 of 11
fit, RMSEA = .05, 90% CI = 0.05 – 0.06, and accounted for levels of integrity (B = 0.93, β = 0.49, p < .001) and for-
31% of the variability in depressive symptoms (see Table 6). giveness accounted for 25% of the variability in integrity.
Higher levels of forgiveness were associated with higher Higher levels of integrity were associated with lower levels
Table 4. Unstandardized and Standardized Coefficients for Sociodemographic Variables and Forgiveness, Controlling for Sociodemographics,
Predicting Depressive Symptoms
Dependent variable = Depressive symptoms Dependent variable = Depressive symptoms
B β B β
Women 0.07 0.06 0.02 0.02
Men (referent) — — — —
Age 0.00 −0.02 0.00 0.03
Table 5. Unstandardized and Standardized Coefficients for Forgiveness Predicting Integrity, Despair, and Depressive Symptoms and Partial
Effects of Forgiveness, Integrity, and Despair on Depressive Symptoms, Controlling Sociodemographics
Dependent variable
(DV) = Integrity DV = Despair DV = Depression
B β B β B β
Female −0.02 −0.02 −0.02 −0.01 0.02 0.01
Male (referent) — — — — — —
Age −0.01 −.04 −0.01 −0.04 0.00 0.03
Education 0.09 0.10 −0.07 −0.07 −0.03 −0.04
Single −0.07 −0.03 −0.17 −0.06 0.05 0.02
Divorced −0.29 −0.07 0.27 0.06 −0.17 −0.05
Widowed −0.13 −0.09 0.14 0.08 0.17* 0.13
Married/cohabitating (referent) — — — — — —
Forgiveness direct effect 0.94*** 0.49 −0.99*** −0.46 −0.32* −0.19
Integrity direct effect −0.24*** −0.28
Despair direct effect 0.15* 0.19
Total integrity/despair indirect effect −0.38** −0.23
R2 .25 .25 .33
Table 6. Unstandardized and Standardized Coefficients for Forgiveness Predicting Integrity and Depressive Symptoms and Partial Effects of
Forgiveness and Integrity on Depressive Symptoms, Controlling Sociodemographics
Dependent variable (DV) = Integrity DV = Depression
B β
Women −0.02 −0.01 0.01 0.01
Men (referent) — — — —
Age −0.01 −0.04 0.00 0.01
Education 0.09 0.10 −0.03 −0.05
Single −0.07 −0.03 0.01 0.00
Divorced −0.29 −0.07 −0.15 −0.04
Widowed −0.13 −0.09 0.17* 0.13
Married/cohabitating (referent) — — — —
Forgiveness direct effect 0.93*** 0.49 −0.37** −0.23
Integrity direct effect −0.32*** −0.38
Integrity indirect effect −0.30** −0.18
R2 .25 .31
of depressive symptoms (B = −0.32, β = −0.38, p < .001). higher levels of forgivingness suffer less of depressive
The indirect effect of forgiveness on depressive symptoms symptoms, and this relationship seems to be prevalent in
through integrity was significant (B = −0.30, β = −0.18, both residential elderly as in community-dwelling elderly
p < .01). The direct effect of forgiveness on depression in individuals. However, contrary to the existing literature,
this model was modestly reduced from the direct effect only we did not find higher levels of depressive symptoms in
model and remained statistically significant (B = −0.37, residential elderly individuals than in community-dwelling
β = −0.23, p = .01). elderly individuals, but differences did appear with regard
The indirect effect model of forgiveness through despair to feelings of despair as well as to levels of forgivingness.
on depression had an acceptable fit, RMSEA = .05, 90% In line with Erikson’s theory that later life brings the issue
CI = 0.05 – 0.06, and accounted for 28% of the variabil- of integrity/despair as a central focus, we assumed that this
ity in depressive symptoms (see Table 7). Higher levels of developmental process might play a crucial role in the rela-
forgiveness were associated with lower levels of despair tionship between forgivingness and depressive symptoms.
Table 7. Unstandardized and Standardized Coefficients for Forgiveness Predicting Despair and Depressive Symptoms and Partial Effects of
Forgiveness and Despair on Depressive Symptoms, Controlling Sociodemographics
Dependent variable (DV) = Despair DV = Depression
B β B β
Female −0.02 −0.01 0.03 0.02
Male (referent) — — — —
Age −0.01 −0.04 0.01 0.05
Education −0.07 −0.07 −0.04 −0.06
Single −0.16 −0.06 0.09 0.04
Divorced 0.14 0.08 0.19* 0.14
Widowed 0.27 0.06 −0.13 −0.04
Married/cohabitating (referent) — — — —
Forgiveness direct effect −1.09*** −0.47 −0.49** −0.27
Despair direct effect 00.25*** 0.32
Despair indirect effect −0.27** −0.15
R2 .26 .28
mechanisms. High levels of forgivingness might improve Our findings on the forgivingness–depressive symp-
healthy intrapersonal functioning which, in turn, has ben- toms association are consistent with previous work using
eficial associations with mental health. This is also in line samples of students and adults, but expand it to the final
with the study of Toussaint and colleagues (2008) show- stage of life where the topic might be of pivotal impor-
ing that hopelessness is an explanatory mechanism in the tance. Framed within a life-span perspective, we showed
forgiveness–depression association in a large sample of that the developmental task of finding a balance between
U.S. adults. Given the similarities between despair and integrity and despair is an important intrapersonal mecha-
hopelessness, a similar reasoning might be applicable. nism through which forgivingness is related with depressed
However, in this study, we focused in on a specific life stage feelings for residential as well as community-dwelling
where the feelings of hopelessness and despair might be of elderly population. Future work should be aimed at testing
pivotal importance. In this study, we focused explicitly on a this model in specific groups of elderly population, such as
specific type of forgiveness—forgiveness of others. Future elderly individual with severe medical illness or with cog-
Butler, R., & Lewis, M. (1982). Aging and mental health: Positive psycho- Lawler-Row, K. A., Scott, C. A., Raines, R. L., Edlis-Matityahou,
social and biomedical approaches. St Louis, MO: Mosby. M., & Moore, E. W. (2007). The varieties of forgiveness experi-
Chen, F., Curran, P. J., Bollen, K. A., Kirby, J., & Paxton, P. (2008). An ence: Working toward a comprehensive definition of forgive-
empirical evaluation of the use of fixed cutoff points in RMSEA ness. Journal of Religion & Health, 46, 233–248. doi:10.1016/j.
test statistic in structural equation models. Sociological Methods & ijpsycho.2008.01.001
Research, 36, 462–494. Lewis, J. S. (1996). Sense of coherence and the strengths perspective with
Chimich, W. T., & Nekolaichuk, C. L. (2004). Exploring the links between older persons. Journal of Gerontological Social Work, 26, 99–112.
depression, integrity, and hope in the elderly. Canadian Journal of Luppa, M., Sikorski, C., Luck, T., Weyerer, S., Villringer, A., Koenig, H.-
Psychiatry, 49, 428–433. H., & Riedel-Heller, S. G. (2011). Epidemiology of depression in old
Cuijpers, P., Vogelzangs, N., Twisk, J., Kleiboer, A., Li, J., & Penninx, age: Results of the Leipzig longitudinal study of the aged (leila 75+).
B. W. (2013). Differential mortality rates in major and subthresh- Journal of Epidemiology and Community Health, 65, A127–A127.
old depression: Meta-analysis of studies that measured both. British doi:10.1136/jech.2011.142976e.14
Journal of Psychiatry, 202, 22–27. McCullough, M. E. (2000). Forgiveness as human strength: Theory, meas-
Dezutter, J., Wiesmann, U., Apers, S., & Luyckx, K. (2013). Sense of urement, and links to well-being. Journal of Social and Clinical
of Research in Personality, 42, 1004–1019. doi:10.1016/j. van’t Veer-Tazelaar, P. J., van Marwijk, H. W., Jansen, A. P., Rijmen, F., Kostense,
jrp.2008.02.006 P. J., van Oppen, P., … Beekman, A. T. (2008). Depression in old age
Toussaint, L. L., & Webb, J. R. (2005). Theoretical and empirical connec- (75+), the PIKO study. Journal of Affective Disorders, 106, 295–299.
tions between forgiveness, mental health, and well-being. In J. E. L. Westerhof, G. J., Bohlmeijer, E. T., van Beljouw, I. M. J., & Margriet, A.
Worthington (Ed.), Handbook of forgiveness (pp. 349–362). New (2010). Improvement in personal meaning mediates the effects of a
York, NY: Routledge. life review: Intervention on depressive symptoms in a randomized con-
Toussaint, L. L., Williams, D. R., Musick, M. A., & Everson, S. A. trolled trial. Gerontologist, 50, 541–549. doi:10.1093/geront/gnp168
(2001). Forgiveness and health: Age differences in a US prob- Worthington, E. L. (2005). More questions about forgiveness: Research
ability sample. Journal of Adult Development, 8, 249–257. agenda for 2005–2015. In E. L. Worthington, Jr. (Ed.), Handbook of
doi:10.1023/A:1011394629736 forgiveness (pp. 557–575). New York, NY: Routledge.
Toussaint, L. L., Williams, D. R., Musick, M. A., & Everson-Rose, Wu, A. D., Li, Z., & Zumbo, B. D. (2007). Decoding the meaning of facto-
S. A. (2008). Why forgiveness may protect against depression: rial invariance and updating the practice of multi-group confirma-
Hopelessness as an explanatory mechanism. Personality and Mental tory factor analysis: A demonstration with TIMSS data. Practical
Health, 2, 89–103. doi:10.1002/pmh.35 Assessment Research & Evaluation, 12, 1–26.