Vol. 39 No. 3 March 2010 Journal of Pain and Symptom Management 515

Original Article

A Comparison of Two Spirituality Instrumentsand Their Relationship With Depression andQuality of Life in Chronic Heart FailureDavid B. Bekelman, MD, MPH, Carla Parry, PhD, Farr A. Curlin, MD,Traci E. Yamashita, MS, Diane L. Fairclough, DrPH, and Frederick S. Wamboldt, MDDenver VA Medical Center (D.B.B.), Denver, Colorado; Division of General Internal Medicine

(D.B.B., T.E.Y.), Division of Health Care Policy and Research (C.P.), Colorado Health Outcomes

Program (D.L.F.), and Department of Psychiatry (F.S.W.), University of Colorado Denver School of

Medicine, Aurora, Colorado; Section of General Internal Medicine (F.A.C.), MacLean Center for

Clinical Medical Ethics, University of Chicago, Chicago, Illinois; Department of Biostatistics and

Informatics (D.L.F.), Colorado School of Public Health, Aurora, Colorado; and Division of

Psychosocial Medicine (D.B.B., F.S.W.), National Jewish Health, Denver, Colorado, USA

Abstract

Spirituality is a multifaceted construct related to health outcomes that remains ill definedand difficult to measure. Spirituality in patients with advanced chronic illnesses, such aschronic heart failure, has received limited attention. We compared two widely usedspirituality instruments, the Functional Assessment of Chronic Illness TherapydSpiritualWell-Being (FACIT-Sp) and the Ironson-Woods Spirituality/Religiousness Index (IW), tobetter understand what they measure in 60 outpatients with chronic heart failure. Weexamined how these instruments related to each other and to measures of depression andquality of life using correlations and principal component analyses. The FACIT-Sp measuredaspects of spirituality related to feelings of peace and coping, whereas the IW measured beliefs,coping, and relational aspects of spirituality. Only the FACIT-Sp Meaning/Peace subscaleconsistently correlated with depression (r¼�0.50, P< 0.0001) and quality of life(r¼ 0.41, P¼ 0.001). Three items from the depression measure loaded onto the same factoras the FACIT-Sp Meaning/Peace subscale (r¼ 0.43, �0.43, and 0.71), whereas theremaining 12 items formed a separate factor (Cronbach’s alpha¼ 0.82) when combined withthe spirituality instruments in a principal component analysis. The results demonstrateseveral clinically useful constructs of spirituality in patients with heart failure and suggestthat psychological and spiritual well-being, despite some overlap, remain distinctphenomena. J Pain Symptom Manage 2010;39:515e526. � 2010 U.S. Cancer PainRelief Committee. Published by Elsevier Inc. All rights reserved.

This study was funded by the Johns Hopkins Centerfor Complementary and Alternative Medicine; theNational Center for Complementary and Alterna-tive Medicine, National Institutes of Health; andthe University of Colorado Denver Hartford/Jahni-gen Center of Excellence in Geriatric Medicine.The views in this article are those of the authors

and do not necessarily reflect the views of the De-partment of Veterans Affairs.

Address correspondence to: David Bekelman, MD, MPH,Denver VA Medical Center, Research (151), 1055Clermont Street, Denver, CO 80220, USA. E-mail:david.bekelman@ucdenver.edu

Accepted for publication: August 29, 2009.

� 2010 U.S. Cancer Pain Relief CommitteePublished by Elsevier Inc. All rights reserved.

0885-3924/10/$esee front matterdoi:10.1016/j.jpainsymman.2009.08.005

516 Vol. 39 No. 3 March 2010Bekelman et al.

Key Words

Spirituality, quality of life, depression, questionnaires, measurement, heart failure

IntroductionSpirituality is a complex, multifaceted con-

struct that is challenging to define and mea-sure.1 It may be defined broadly as ‘‘the wayin which people understand their lives inview of their ultimate meaning and value.’’2

Measuring spirituality is difficult, in part, be-cause of the multiple domains ‘‘spirituality’’encompasses and the varied ways in which peo-ple construct and interpret the concept of spir-ituality. For example, spirituality may includea system of beliefs or set of rituals as part ofan organized religion, namely, the strengthor comfort derived from one’s faith; thedegree to which a relationship with a higherpower has been internalized; existential well-being; or a sense of meaning, peace, or well-being. Spirituality is related to religion, whichmay be defined as ‘‘a set of beliefs, practices,and language that characterizes a communitythat is searching for transcendent meaning ina particular way, generally on the basis of beliefin a deity.’’3 Because spirituality has beendefined in many different ways, it has alsobeen measured using numerous differentapproaches.

Contemporary measures of spirituality havebeen associated with a variety of health out-comes and quality-of-life domains. Aspects ofspirituality have been associated with lowerlevels of depression, and spiritual well-beinghas been shown to be associated with overallquality of life, even after accounting for psy-chological and physical well-being.4e8 In a mul-tivariate model, spirituality, but not religiosity,was an important predictor of self-appraisedgood health, after adjusting for quality of life,physical functioning, age, race, and depres-sion.9 Religiosity is associated with less morbid-ity and mortality, but the mechanism of thisassociation and whether it is a causal relation-ship is unclear.10e14

Spirituality in outpatients with advancedchronic illnesses, such as chronic heart failure,has received limited attention. Chronic heartfailure is a leading cause of disability, hospital-ization, death, and health care costs in the

United States. Nearly 5 million Americanshave chronic heart failure, and the prevalencecontinues to increase as the population ages.15

Between 20% and 30% of patients with heartfailure are depressed, and depression in heartfailure is associated with worse physical func-tioning, poorer quality of life, and higher costsand mortality. Spirituality in patients withheart failure is related to adjustment,16 depres-sion,6 and quality of life.17 Thus, a deeper un-derstanding of the relevance of spirituality tooutpatients with heart failure, how to measurespirituality in this population, and of the rela-tionships between different measures of spiri-tuality and depression and quality of life isimportant.

A critical view of spirituality measures sug-gests that they are corrupted by measures ofpositive character traits and psychologicalstates, thereby confounding spiritual well-being with psychological well-being.18 Forexample, individuals who score high on theFunctional Assessment of Chronic IllnessTherapydSpiritual Well-Being (FACIT-Sp) in-strument, a popular instrument for measuringspirituality, have been found to have less de-pression6,19 and suicidality and to be less likelyto desire help in hastening their death.20 Koe-nig suggests that such findings are spuriousbecause they are tautological: ‘‘Research thatdocuments an association between spiritualitydefined this way and positive mental health,is meaningless since constructs measuredwith the same or similar items will always becorrelated with one another.’’18

Despite the widespread study of spiritualityand health outcomes, few studies beyond theoriginal psychometric validation studies havecompared different spirituality instruments tounderstand the specific aspects of spiritualitymeasured, and even fewer studies have explic-itly sought to explore the extent to which spir-ituality measures may overlap with measures ofpsychological constructs. In this study, we com-pared two spirituality instrumentsdthe FACIT-Sp and the Ironson-Woods Spirituality/Reli-giousness Index (IW)din 60 outpatients withchronic heart failure. We chose these

Vol. 39 No. 3 March 2010 517Measuring Spirituality in CHF

instruments because both have been widelyused and have demonstrated validity and reli-ability in previous psychometric testing.21,22

In addition, they are uniquely suited to thisinquiry because they appear to measure boththe same and different aspects of spirituality/religiosity. For example, both instrumentsmeasure ‘‘sense of peace’’ and ‘‘faith,’’ butonly the IW measures ‘‘religious behavior.’’Thus, it would be informative to see if theseconstructs indeed measure the same or differ-ent information when used in a single popula-tion. Furthermore, the FACIT-Sp has beencriticized for measuring psychological well-being rather than spiritual well-being.18 Thepurpose of this study was to compare theexplicit domains and latent constructs mea-sured by these two spirituality instrumentsand to examine their associations with mea-sures of depression and quality of life.

MethodsStudy Population and Design

A cross-sectional study in outpatients withheart failure was conducted between August2004 and April 2005. Participants wererecruited from cardiology clinics at an aca-demic-affiliated community hospital anda tertiary care academic referral hospital inBaltimore, Maryland. Eligible study partici-pants were 60 years or older and had been di-agnosed by a cardiologist as having congestiveheart failure (New York Heart Association(NYHA) Functional Class II, III, or IV). Pa-tients were excluded if they were diagnosedwith dementia, were unable to understandthe study protocol or provide informed con-sent, or were on a heart transplant list. At thestudy visit, participants completed surveys toassess spiritual well-being, heart failure-relatedhealth status, and depressive symptoms. De-tailed descriptions of the study have been pub-lished elsewhere.6 The study was approved bythe Institutional Review Board at the JohnsHopkins Bloomberg School of Public Health.

MeasurementsSpirituality was assessed using two instru-

ments, the FACIT-Sp and the IW (see tablesin Results for instrument item lists). The FA-CIT-Sp is a 12-item self-report measure that

assesses overall spiritual well-being (scale range0e48; higher scores signifying greater spiritualwell-being). The instrument includes two sub-scales, ‘‘Meaning/Peace’’ (Items 1e8) and‘‘Faith’’ (Items 9e12). The Meaning/Peace sub-scale measures a sense of meaning, peace andharmony, and purpose in life. The Faith sub-scale assesses the relationship between illness,faith, and spiritual beliefs, and how one finds so-lace in one’s faith.21 The instrument had highinternal consistency in the original validationstudy (Cronbach’s alpha for total scale: 0.87;for Meaning/Peace subscale: 0.81; for Faithsubscale: 0.88), which included 1,617 subjects,83.1% of whom had cancer and 16.9% ofwhom had human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS).21

Immediately after the FACIT-Sp, partici-pants completed the IW, a 25-item self-reportinstrument that measures spirituality in boththe traditional religious and private spiritualdimensions. The IW contains four subscales,described as ‘‘Sense of Peace’’ (Items 1e9),‘‘Faith in God’’ (Items 10e15), ‘‘ReligiousBehavior’’ (Items 16e20), and ‘‘Compassion-ate View of Others’’ (Items 21e25). Thisinstrument was validated in a study of 279 peo-ple living with HIV/AIDS (Cronbach’s alphafor Sense of Peace: 0.94; Faith in God: 0.93;Religious Behavior: 0.87; and CompassionateView of Others, two-item: 0.87).22

Depression was measured, as it is an impor-tant indicator of psychological well-being. Weused the Geriatric Depression Scale-ShortForm (GDS-SF), a self-report, reliable, andvalid screening tool for current depression inthis population.23,24 The GDS is widely usedin elderly, medically ill populations because itexcludes somatic symptoms of depressionwhile maintaining sensitivity.25 The scale rangeis from 0 to 15, with a higher score indicatinga greater number of depressive symptoms. Ascore greater than 4 is 60% sensitive and89% specific for a diagnosis of depressive dis-order.26 The scale has high internal consis-tency (Cronbach’s alpha: 0.80).26

Heart failure-related quality of life was mea-sured using the quality-of-life subscale of theself-reported Kansas City CardiomyopathyQuestionnaire (KCCQ). The quality-of-life sub-scale of the KCCQ has been validated and isscored from 0e100, with a higher score

518 Vol. 39 No. 3 March 2010Bekelman et al.

indicating better quality of life.27 It includesthe following items: 1) ‘‘Over the past 2 weeks,how much has your heart failure limited yourenjoyment of life?’’ Response items are as fol-lows: extremely, quite a bit, moderately,slightly, and not at all; and 2) ‘‘If you had tospend the rest of your life with your heart fail-ure the way it is right now, how would you feelabout this?’’ Response items are as follows: notat all satisfied, mostly dissatisfied, somewhatsatisfied, mostly satisfied, and completely satis-fied. A third item asks about depression symp-toms. To specifically ascertain quality of liferather than depression, we excluded the ques-tion on depression that is typically included inthis subscale.

Statistical AnalysesTo confirm the internal consistency of the

FACIT-Sp and IW subscales, we calculated thestandardized Cronbach’s coefficient alpha ofthe subscales. Cronbach’s alpha reflects the de-gree to which all of the items pertain to a uni-form concept. Pearson correlations were usedto assess the FACIT-Sp and IW between- andwithin-subscale associations as well as the asso-ciations between the spirituality subscales withdepression (GDS-SF) and quality of life(KCCQ subscale). To further clarify the dimen-sions of spirituality/religiosity measured by theFACIT-Sp and IW scales, a principal compo-nent analysis with varimax rotation was per-formed on the 37 combined items from theseinstruments. The number of factors was cho-sen based on the examination of the screeplots, percentage of variability explained, andinterpretability. We then examined the overlapbetween psychological well-being and spiritualwell-being, as measured by the FACIT-Sp, IW,and GDS-SF, using a principal componentanalysis with varimax rotation, which com-bined the 15 items from the GDS-SF instru-ment with the 37 spirituality/religiosity itemsfrom FACIT-Sp and IW. Factors were selectedbased on the same criteria described earlier.All analyses were implemented with SAS soft-ware (Version 9.1; SAS Institute, Cary, NC).

Regarding missing data, only one partici-pant had incomplete FACIT-Sp data, missingfour items out of 12. Sixteen (27%) of thestudy subjects had incomplete data on the IWsurvey, missing an average of three out of 25items per person. In all cases, missing data

was the result of nonresponse by the partici-pants on individual survey items. The twoitems from the Ironson-Woods survey withthe most missing data (8%) were the onlytwo questions that specifically used the term‘‘religious.’’ If fewer than half of the items com-prising the subscales for the FACIT-Sp or IWwere missing, we replaced the missing itemswith the average of the non-missing items inthe corresponding subscales. Otherwise, theitems and subscales were classified as missing.There were no missing data for the KCCQ sur-veys. Four (7%) of the 60 participants had in-complete GDS-SF data, missing an average of1.3 out of 15 items per person. For the missingitems, we used a weighted GDS score as de-scribed on the GDS Web site, which removesthe missing items from the denominator toavoid underestimation of the scale score.28

ResultsParticipants

The study sample is described in Table 1.The median age of the participants was 75years; 37% were females and 12% were non-whites. Most of them had at least a high schooleducation (58%). The median heart failurehealth status score (KCCQ) of 71 has been as-sociated with New York Heart Association ClassII.29 Thirty-two percent had current clinicallysignificant depression (GDS-SF> 4). Femaleswere more likely to have higher depressionscores (median GDS-SF for men¼ 2, for wom-en¼ 4, P¼ 0.04).

The Functional Assessment of Chronic IllnessTherapydSpiritual Well-Being and Ironson-Woods Spirituality/Religiousness Index inPatients With Heart Failure

Cronbach’s coefficient alphas for the FACIT-Sp were 0.86 (full scale), 0.82 (Faith subscale),and 0.82 (Meaning/Peace subscale). Thealpha coefficients for the IW were 0.98 (fullscale), 0.97 (Sense of Peace subscale), 0.97(Faith in God subscale), 0.90 (Religious Behav-ior subscale), and 0.95 (Compassionate Viewsubscale). Table 2 displays the intra- and inter-instrument correlations between the subscalesof FACIT-Sp and IW. Intra-instrument subscalecorrelations ranged from 0.53 to 0.89(unshaded cells), whereas the interinstrument

Table 1Study Population Characteristics (n¼ 60)

Characteristics n (%)a

Female 22 (36.7)Age, years, median (IQR) 75 (70, 81)Race, African American 7 (11.7)Married/significant other

or partner (n¼ 59)30 (50.9)

Highest grade completed in school (n¼ 59)Less than high school graduate 25 (42.4)High school graduate 20 (33.9)Additional education 14 (23.7)

Total household income (n¼ 56)<$20,000 20 (35.7)$20,000e30,000 19 (33.9)>$30,000 17 (30.4)

Comorbid medical illnessChronic obstructive pulmonary disease 11 (18.3)Cancer 8 (13.3)Stroke 11 (18.3)Prior myocardial infarction 28 (46.7)

MMSE score, median (IQR) 29 (28, 29)Health status score (KCCQ), median (IQR) 71 (48, 90)Quality-of-life subscale score

(KCCQ), median (IQR)75 (50, 88)

Depression score (GDS-SF),median (IQR)

2 (1, 5)

Probable depression (GDS-SF> 4) 19 (31.7)

IQR¼ interquartile range; MMSE¼Mini-Mental State Examina-tion; MSAS-SF¼Memorial Symptom Assessment ScaledShortForm.aUnless otherwise indicated.

Tab

le2

son

Co

rrel

atio

ns

(P-V

alu

e)B

etw

een

IWan

dF

AC

IT-S

pS

ub

scal

eS

core

s

FAC

IT-S

p:

Mea

nin

g/P

eace

FAC

IT-S

p:

Fai

thIW

:Se

nse

of

Pea

ceIW

:F

aith

inG

od

IW:

Rel

igio

us

Beh

avio

rIW

:C

om

pas

sio

nat

eV

iew

6060

5959

5957

1.0

0.32

a0.

22b

0.11

c0.

03d

0.53

e1.

00.

84e

0.77

e0.

69e

0.54

e

1.0

0.89

e1.

00.

84e

0.79

e1.

00.

74e

0.63

e0.

81e

1.0

ded

area

sco

nta

inw

ith

in-s

cale

corr

elat

ion

s.

Vol. 39 No. 3 March 2010 519Measuring Spirituality in CHF

subscale correlations ranged widely from 0.03to 0.84 (shaded cells). The Faith subscale ofthe FACIT-Sp was significantly correlated witheach of the IW subscales (0.54e0.84;P< 0.001 for all), whereas the Meaning/Peacesubscale of the FACIT-Sp was only modestlycorrelated with the Sense of Peace subscale(0.32, P¼ 0.01) and was not correlated withthree of the four IW subscales.

Pea

r

Spir

itu

alit

yIn

stru

men

tSu

bsc

ale

n

n FAC

IT-S

p:

Mea

nin

g/P

eace

60FA

CIT

-Sp

:F

aith

60IW

:Se

nse

of

Pea

ce59

IW:

Fai

thin

Go

d59

IW:

Rel

igio

us

Beh

avio

r59

IW:

Co

mp

assi

on

ate

Vie

w57

Shad

edar

eas

con

tain

bet

wee

n-s

cale

corr

elat

ion

s,an

du

nsh

aap¼

0.01

.b p¼

0.09

.c p¼

0.43

.dp¼

0.80

.e p<

0.00

01.

Association of Functional Assessment ofChronic Illness TherapydSpiritual Well-Being and Ironson-Woods Spirituality/Religiousness Index Subscales With Depressionand Quality of Life

The Meaning/Peace subscale of the FACIT-Sp was strongly associated with both depres-sion and quality of life, whereas the Faithsubscale was associated with quality of life butnot significantly associated with depression(Table 3). As sense of meaning/peace or faithincreased, depression scores decreased, andquality-of-life scores increased. In contrast,none of the IW subscales was strongly

Table 3Pearson Correlation (P-Value) Between the GDS-SF and KCCQ-QoL and Subscales of the IW and the

FACIT-Sp Instruments

Spirituality Instrument Subscale n GDS-SF KCCQ-QoL

FACIT-Sp: Meaning/Peace 60 �0.50 (<0.0001) 0.41 (0.001)FACIT-Sp: Faith 60 �0.23 (0.07) 0.38 (0.003)IW: Sense of Peace 59 �0.07 (0.59) 0.21 (0.10)IW: Faith in God 59 �0.10 (0.46) 0.25 (0.05)IW: Religious Behavior 59 �0.00 (0.98) 0.09 (0.52)IW: Compassionate View 57 0.12 (0.36) �0.05 (0.73)

GDS-SF ¼ Geriatric Depression Scale-Short Form; KCCQ-QoL ¼ Kansas City Cardiomyopathy Questionnaire-Quality of Life Subscale.

520 Vol. 39 No. 3 March 2010Bekelman et al.

associated with either depression or quality oflife except for a modest association betweenthe Faith in God subscale and quality of life(r¼ 0.25, P¼ 0.05).

Domains of Spirituality/Religiosity MeasuredAfter Combining Scale Items

To further clarify the dimensions of spiritu-ality/religiosity measured by the FACIT-Spand IW scales, a principal component analysiswith varimax rotation was performed on the 37items that result when both scales are com-bined. Although six factors had eigenvaluesgreater than 1.0, which together accountedfor 81% of the variance (53%, 12%, 6%, 5%,3%, and 3%), a three-factor solution appearedreasonable after examination of the scree plot.We examined models for three to six factors.The models containing five or more factorshad one or more factors consisting of one ortwo items. Hence, a three-factor model wasdeemed the best based on interpretability.

The factor loadings for the three-factormodel are displayed in Table 4. Factor 1encompassed both beliefs and practicesrelated to God and religion and patients’ as-sessments of the impacts of those beliefs ontheir lives. Several items reflected how one’sspiritual/religious beliefs help one cope withillness in particular. Factor 1 contained all ofthe items from the IW ‘‘Sense of Peace’’ and‘‘Faith in God’’ subscales, three of the five ‘‘Re-ligious Behavior’’ subscale items, and the threeFACIT-Sp items that include the terms ‘‘faithor spiritual beliefs’’ (all from the ‘‘Faith’’ sub-scale). Factor 2 described how one practicesone’s spiritual beliefs and how one’s beliefsshape one’s interactions and relationshipswith others. This factor included all of theitems from the IW ‘‘Compassionate View’’ sub-scale and two items from the IW ‘‘Religious

Behavior’’ subscale. One item from the FA-CIT-Sp also loaded onto this factor, but theloading was weak (0.39), and this item alsodid not load well onto models containinga higher number of factors. Factor 3 capturedfeelings of inner peace and included seven ofthe eight FACIT-Sp ‘‘Meaning/Peace’’ subscaleitems.

Spirituality/Religiosity Compared WithPsychological Well-Being

To evaluate the overlap between measures ofspirituality/religiosity and psychological well-being, we conducted a principal componentanalysis with varimax rotation on combineditems from the FACIT-Sp, IW, and GDS-SF.Ten factors had eigenvalues greater than 1.0,which together accounted for 81% of the vari-ance (38%, 14%, 6%, 5%, 4%, 4%, 3%, 3%,3%, and 2%). The scree plot suggested that be-tween three and seven factors would be appro-priate, and we reviewed these models. Themodel with four factors is displayed in Table 5.

This model is presented because the firstthree factors remained similar and stablethroughout the models containing four toseven factors. Factors 5 and higher in thesemodels contained one to three items andwere difficult to interpret. Across thesemodels, Factor 1 encompassed a broad rangeof spiritual and religious beliefs, values, andbehaviors. It included all of the items fromthe IW as well as the three FACIT-Sp itemsthat include the terms ‘‘faith or spiritual be-liefs’’ (all from the ‘‘Faith’’ subscale). Factor 2described feelings of inner peace and content-ment and included most of the FACIT-Sp‘‘Meaning/Peace’’ items as well as a few itemsfrom the GDS-SF. Factor 3 captured thoughtsand feelings of depression, includingfear, helplessness, worthlessness, emptiness,

Table 4Factor Loadings for Combined Items From the Ironson-Woods Spirituality/Religiousness Index (IW) and the

Functional Assessment of Chronic Illness Therapy - Spiritual Well-Being Scale (FACIT-Sp)

Items from the Ironson-Woods and FACIT-Sp scales Factor 1 Factor 2 Factor 3

IW itemsMy beliefs give me a sense of peace 0.72 0.55 dMy beliefs help me to know everything will be fine 0.81 d dMy beliefs give meaning to my life 0.67 0.53 dMy beliefs help me to be relaxed 0.66 0.54 dMy beliefs help me to feel protected 0.73 d dMy beliefs help me feel I am not alone 0.83 d dMy beliefs help me feel I have a relationship or a connection with

a higher form of being0.72 0.46 d

My beliefs help me be less afraid of death 0.74 0.46 dI believe my soul will live on in some form after my body dies 0.80 d dI believe God created all things in the universe 0.90 d dGod will not turn his back on me no matter what I do 0.88 d dWhen I am ill, God gives me courage to cope with my illness 0.86 d dWhen I am ill, God will answer my prayers for a recovery 0.89 d dMy beliefs are very influential in my recovery when I am ill 0.90 d dWhen I am ill my faith gives me optimism that I will recover 0.93 d dI attend religious services 0.61 0.51 dI participate in religious rituals 0.48 0.70 dI pray or meditate to get in touch with God 0.73 d dI discuss my beliefs with others who share my beliefs 0.62 0.55 dMy beliefs give me a set of rules I must obey 0.57 0.62 dMy beliefs teach me to help other people who are in need 0.55 0.67 dMy beliefs help me feel compassion/love/respect for others 0.49 0.61 dI have a responsibility to help others 0.42 0.79 dMy beliefs increase my acceptance and tolerance of others d 0.84 dI feel I am connected to all humanity d 0.89 d

FACIT-Sp itemsI feel peaceful d d 0.69I have a reason for living d d 0.70My life has been productive d d 0.61I have trouble feeling peace of mind d 0.39 dI feel a sense of purpose in my life d d 0.78I am able to reach down deep into myself for comfort d d 0.72I feel a sense of harmony within myself d d 0.80My life lacks meaning or purpose d d L0.62I find comfort in my faith or spiritual beliefs 0.65 d 0.41I find strength in my faith or spiritual beliefs 0.53 0.42 dMy illness has strengthened my faith or spiritual beliefs 0.66 d dI know whatever happens with my illness, things will be okay 0.50 d 0.63

Percent of total variance 38.0 19.0 13.6Coefficient alpha 0.98 0.89 0.86

Primary factor loadings are presented in boldface, and any additional loadings above 0.40 are displayed.

Vol. 39 No. 3 March 2010 521Measuring Spirituality in CHF

hopelessness, and fatigue. Throughout modelswith four to seven factors, Factor 3 includednine or 10 of the 15 GDS-SF items. Factors 5and higher in these models were difficult to in-terpret because they contained one to threeitems that were weakly loaded and the factorshad low alpha coefficients.

DiscussionThis comparison of the FACIT-Sp and IW in-

struments in a sample of outpatients with heartfailure informs the ongoing effort to measurespirituality and to make sense of associations

between spirituality measures, mental health,and quality-of-life outcomes. The major find-ings are as follows: 1) the scales had adequateinternal consistency in patients with heart fail-ure; 2) subscales of the FACIT-Sp were gener-ally associated with depression and heartfailure-related quality of life, whereas the IWsubscales generally were not; 3) combiningitems from the two measures showed thata mix of items from each measure formedone factor that explained the largest amountof variance; and 4) several items measuringpsychological well-being overlapped withspiritual well-being, but most of them were

Table 5Factor Loadings for Combined Items From the Ironson-Woods Spirituality/Religiousness Index,

the Functional Assessment of Chronic Illness TherapydSpiritual Well-Being Scale, and the GeriatricDepression Scale

Factor 1 Factor 2 Factor 3 Factor 4

IW itemsMy beliefs give me a sense of peace 0.91 d d dMy beliefs help me to know everything will be fine 0.89 d d dMy beliefs give meaning to my life 0.87 d d dMy beliefs help me to be relaxed 0.86 d d dMy beliefs help me to feel protected 0.82 d d dMy beliefs help me feel I am not alone 0.86 d d dMy beliefs help me feel I have a relationship or a connection

with a higher form of being0.86 d d d

My beliefs help me be less afraid of death 0.87 d d dI believe my soul will live on in some form after my body dies 0.82 d d dI believe God created all things in the universe 0.77 d d 0.45God will not turn his back on me no matter what I do 0.83 d d dWhen I am ill, God gives me courage to cope with my illness 0.94 d d dWhen I am ill, God will answer my prayers for a recovery 0.89 d d dMy beliefs are very influential in my recovery when I am ill 0.92 d d dWhen I am ill my faith gives me optimism that I will recover 0.91 d d dI attend religious services 0.78 d d dI participate in religious rituals 0.76 d d dI pray or meditate to get in touch with God 0.79 d d dI discuss my beliefs with others who share my beliefs 0.83 d d dMy beliefs give me a set of rules I must obey 0.80 d d dMy beliefs teach me to help other people who are in need 0.81 d d dMy beliefs help me feel compassion/love/respect for others 0.74 d d dI have a responsibility to help others 0.76 d d �0.47My beliefs increase my acceptance and tolerance of others 0.73 d d �0.47I feel I am connected to all humanity 0.69 d d �0.52

FACIT-Sp itemsI feel peaceful d 0.60 d dI have a reason for living d 0.69 d dMy life has been productive d 0.60 d dI have trouble feeling peace of mind d d d L0.35I feel a sense of purpose in my life d 0.78 d dI am able to reach down deep into myself for comfort d 0.58 d dI feel a sense of harmony within myself d 0.72 d dMy life lacks meaning or purpose d L0.69 d dI find comfort in my faith or spiritual beliefs 0.64 0.42 d dI find strength in my faith or spiritual beliefs 0.73 d d dMy illness has strengthened my faith or spiritual beliefs 0.67 d d dI know whatever happens with my illness, things will be okay 0.47 0.58 d d

GDS-SFAre you basically satisfied with your life? d 0.43 d dHave you dropped many of your activities and interests? d L0.43 d dDo you feel happy most of the time? d 0.71 d dDo you prefer to stay at home rather than going out

and doing new things?d d d 0.31

Do you feel that life is empty? d d 0.53 dDo you often get bored? d �0.46 0.56 dAre you in good spirits most of the time? d d L0.45 dAre you afraid that something bad is going to happen to you? d d 0.75 dDo you feel helpless? d d 0.64 dDo you feel that you have more problems with memory than most? d d 0.54 dDo you think it is wonderful to be alive? d d d 0.67Do you feel pretty worthless the way you are now? d d 0.57 dDo you feel full of energy? d d L0.48 dDo you feel that your situation is hopeless? d d 0.63 dDo you think that most people are better off than you are? d d 0.44 d

Percent of total variance 37.3 11.9 7.5 6.3Coefficient alpha 0.98 0.87 0.82 0.24

Primary factor loadings are presented in boldface, and any additional loadings above 0.40 are displayed.

522 Vol. 39 No. 3 March 2010Bekelman et al.

Vol. 39 No. 3 March 2010 523Measuring Spirituality in CHF

conceptually and statistically distinct from spir-itual well-being.

The FACIT-Sp and the IW generally mea-sured different constructs, although the FA-CIT-Sp Faith subscale overlapped with itemsfrom the IW. The IW measured self-reportedbeliefs and behaviors in explicitly spiritualand religious language, whereas the FACIT-Splargely measured a sense of inner peace andpurpose. The overlap in the instruments onlyincluded items from the FACIT-Sp Faith sub-scale. Indeed, the strongest correlationbetween the subscales from the two instru-ments was between the FACIT-Sp Faith sub-scale and the IW Sense of Peace subscale.The items from the FACIT-Sp Faith subscalethat included the words ‘‘faith or spiritual be-liefs’’ loaded onto the construct measured bythe IW in the principal component analysis.The association of the FACIT-Sp subscaleswith depression and quality of life supportedthe notion that the FACIT-Sp, particularly theMeaning/Peace subscale, measured a differentconstruct from the IW.

Our research demonstrated that instrumentsthat claim to be measuring the same spiritualdomain or construct may actually be measuringdifferent concepts altogether. We found thatthe subscales from two spirituality instrumentsthat purported to measure sense of peacewere weakly correlated, and the items fromthese subscales formed different factors ina principal component analysis. Although thereare more than 100 spirituality/religiosity mea-sures available, little research has comparedthe available instruments in the same popula-tion. However, articles have provided guidanceby classifying these instruments according totheir measurement domains.3,30,31 For exam-ple, Sulmasy3 described four measurement do-mains: religiosity, spiritual/religious copingand support, spiritual well-being, and spiritualneeds. He emphasized the need to better un-derstand relationships between these domainsand how they, in turn, relate to other variables.Thus, analyses such as this one are a necessarystep to gain a better understanding of constructvalidity, and by extension, to gain a betterunderstanding of the relationship betweenspiritual and psychological factors and theimplications for clinical practice.

The FACIT-Sp Meaning/Peace subscale wasassociated with depression and several

depression items loaded onto the same factorwith the FACIT-Sp Meaning/Peace subscale.However, most of the depression items formeda separate factor when combined with theFACIT-Sp and IW items. This finding revealsthat there can be some measurement overlapbetween depression, as an indicator of psycho-logical well-being, and spiritual well-being,although there is also substantial distinctnessof psychological and spiritual well-being.Although this conclusion should be tentativeand deserves replication, given the small sam-ple size, several implications arise. First, itadvocates for asking about spiritual well-beingin addition to psychological well-being inpatients with chronic heart failure who are inthe chronic care setting. Although some clini-cians may already recognize the importanceof doing this, our research provides evidencethat the spiritual domain is indeed separatefrom the psychological domain and is associ-ated with quality of life. Second, one cannot as-sume that depressed patients have low spiritualwell-being or that nondepressed patients havehigh spiritual well-being. Asking specificallyabout spiritual well-being in addition to psy-chological well-being could allow clinicians inthe chronic care setting to help patients fosteran important coping resource, discover a con-tributor to psychological well-being, or identifyan area of need separate from psychologicalwell-being but critically related to quality oflife. For example, a patient with heart failurewho describes a sense of meaninglessness andhopelessness without loss of interest in activi-ties or depressed mood may be sufferingfrom a spiritual crisis. Identifying this withthe patient could assist in more appropriatelyfocused therapy and/or referral of the mostappropriate resource, such as chaplain, pastor,community group, or counselor.

This study supports an effort to create assess-ment tools that would help clinicians with a spir-itual ‘‘differential diagnosis.’’ Combining theitems from the FACIT-Sp and IW generated sev-eral factors or domains that may help cliniciansbetter understand the spiritual or religiousneeds of patients. When patients describe dis-tress related to their spiritual or religious beliefs(Table 4; Factor 1), clinicians might consider re-ferral to a spiritual authority most relevant to thepatient, such as a priest, minister, rabbi, or coun-selor. Distress relating to the relational aspects of

524 Vol. 39 No. 3 March 2010Bekelman et al.

spirituality might prompt facilitated involve-ment of the patient’s family or spiritual commu-nity (Table 4; Factor 2). Distress relating tofeelings or lack of inner peace (Table 4; Factor3) may be helped by providers comfortable inaddressing patient needs that overlap betweenpsychological and spiritual needs, includingthe doctors or other providers caring for the pa-tient, a chaplain, or a mental health profes-sional. Although some clinicians mayrecognize these domains and their implicationsfrom good clinical judgment, our study supportsthe possibility that different dimensions of spiri-tuality can be measured and can be clinically rel-evant in outpatients with chronic heart failure.Further work on developing valid and reliableinstruments that guide clinicians in this type ofspiritual ‘‘differential diagnosis’’ would be veryhelpful.

The results reveal the complexity in concep-tualizing and measuring spirituality and sup-port the effort to create conceptual modelsthat inform measurement of this complex con-struct.3,18 Lessons can be drawn from attemptsto understand and measure another complexconstruct: psychological well-being. We mea-sured depression in this study as an indicatorof psychological well-being because it is wellresearched in basic science, health services,and outcomes research.32 However, patientswith heart failure and other chronic or ad-vanced illnesses may develop demoralization,grief, and anxiety, among other experiences.Prior conceptual and empirical research hasdistinguished some of these clinically impor-tant states.33e36 In our study, each spiritualityinstrument captured only a part of the rangeof patients’ spiritual issues. To answer the ques-tion ‘‘What are the most important or ‘core’domains of spirituality?’’ subsequent steps nec-essarily involve studies like this one to developand test measures to identify ‘‘core’’ spiritualdomains. This work should inform cohesive,testable, and clinically relevant conceptualmodels of spirituality.

This study has several limitations. The smallsample size limited our ability to examineinstrument performance in subgroups ofpatients with different demographic or clinicalcharacteristics. The participants were froma single geographic area, and thus, the resultsmay not be generalizable. Longitudinal datawould have been helpful to understand how

the measures change and interrelate overtime. Given the small sample size, the principalcomponent analyses should be consideredexploratory. Principal component analysisinvolves some subjective interpretation, andthus, replication of these findings wouldstrengthen their validity. The use of thismethod to understand domains of spiritualityor the relationship between psychological andspiritual well-being is but one way of investigat-ing complex issues that require multiplemethods and further in-depth study. Finally,we only examined two spirituality measures.Although the FACIT-Sp and IW are widelyused, there are numerous other measures avail-able that may assess different domains of spiri-tuality not identified by the measures weexamined. The overlap found in our study be-tween spiritual and psychological well-beingcould have been related to measurement erroror sample bias because of the small sample size.This overlap also could have been caused byanother unmeasured construct related toboth spiritual and psychological well-being.For example, a measure of a sense of innercalm may explain the overlap and clarify thedistinctness of feelings of peace and depres-sion. Further research is needed to better un-derstand whether there is measurementoverlap between spiritual and psychological in-struments and to describe the nature of theoverlap.

In conclusion, in patients with chronic heartfailure, the FACIT-Sp measured aspects of spir-ituality related to feelings of peace and copingand was correlated with depression and qualityof life. The IW measured aspects of spiritualityrelated to beliefs, coping, and relationship, andwas generally not correlated with depression orquality of life. The results suggested that psy-chological and spiritual well-being, despitesome overlap, remain distinct phenomena.

References1. Zinnbauer BJ, Pargament KI, Scott AB. The

emerging meanings of religiousness and spirituality:problems and prospects. J Pers 1999;67(6):889e919.

2. Muldoon M, King N. Spirituality, health care,and bioethics. J Relig Health 1995;34(4):329e349.

3. Sulmasy DP. A biopsychosocial-spiritual modelfor the care of patients at the end of life. Gerontol-ogist 2002;42(Spec No 3):24e33.

Vol. 39 No. 3 March 2010 525Measuring Spirituality in CHF

4. Yi MS, Mrus JM, Wade TJ, et al. Religion, spiritu-ality, and depressive symptoms in patients withHIV/AIDS. J Gen Intern Med 2006;21(Suppl 5):S21eS27.

5. Yoon DP, Lee EK. The impact of religiousness,spirituality, and social support on psychologicalwell-being among older adults in rural areas. J Ger-ontol Soc Work 2007;48(3e4):281e298.

6. Bekelman DB, Dy SM, Becker DM, et al. Spiri-tual well-being and depression in patients with heartfailure. J Gen Intern Med 2007;22(4):470e477.

7. Dunn LL, Shelton MM. Spiritual well-being,anxiety, and depression in antepartal women onbedrest. Issues Ment Health Nurs 2007;28(11):1235e1246.

8. Cohen SR, Mount BM, Bruera E, et al. Validityof the McGill Quality of Life Questionnaire in thepalliative care setting: a multi-centre Canadian studydemonstrating the importance of the existentialdomain. Palliat Med 1997;11(1):3e20.

9. Daaleman TP, Perera S, Studenski SA. Religion,spirituality, and health status in geriatric outpa-tients. Ann Fam Med 2004;2(1):49e53.

10. Hummer RA, Rogers RG, Nam CB, Ellison CG.Religious involvement and U.S. adult mortality.Demography 1999;36(2):273e285.

11. Strawbridge WJ, Cohen RD, Shema SJ,Kaplan GA. Frequent attendance at religiousservices and mortality over 28 years. Am J PublicHealth 1997;87(6):957e961.

12. Oman D, Reed D. Religion and mortalityamong the community-dwelling elderly. Am J PublicHealth 1998;88(10):1469e1475.

13. Koenig HG, Hays JC, Larson DB, et al. Doesreligious attendance prolong survival? A six-year fol-low-up study of 3,968 older adults. J Gerontol A BiolSci Med Sci 1999;54(7):M370eM376.

14. Gillum RF, King DE, Obisesan TO, Koenig HG.Frequency of attendance at religious services andmortality in a U.S. national cohort. Ann Epidemiol2008;18(2):124e129.

15. Miller LW, Missov ED. Epidemiology of heartfailure. Cardiol Clin 2001;19(4):547e555.

16. Westlake C, Dracup K. Role of spirituality inadjustment of patients with advanced heart failure.Prog Cardiovasc Nurs 2001;16(3):119e125.

17. Beery TA, Baas LS, Fowler C, Allen G. Spiritual-ity in persons with heart failure. J Holist Nurs 2002;20(1):5e25.

18. Koenig HG. Concerns about measuring ‘‘spiri-tuality’’ in research. J Nerv Ment Dis 2008;196(5):349e355.

19. Nelson CJ, Rosenfeld B, Breitbart W, Galietta M.Spirituality, religion, and depression in the termi-nally ill. Psychosomatics 2002;43(3):213e220.

20. McClain CS, Rosenfeld B, Breitbart W. Effect ofspiritual well-being on end-of-life despair in termi-nally-ill cancer patients. Lancet 2003;361(9369):1603e1607.

21. Peterman AH, Fitchett G, Brady MJ,Hernandez L, Cella D. Measuring spiritual well--being in people with cancer: the Functional Assess-ment of Chronic Illness Therapy-SpiritualWell-being Scale (FACIT-Sp). Ann Behav Med2002;24(1):49e58.

22. Ironson G, Solomon GF, Balbin EG, et al. TheIronson-Woods Spirituality/Religiousness Index isassociated with long survival, health behaviors, lessdistress, and low cortisol in people with HIV/AIDS. Ann Behav Med 2002;24(1):34e48.

23. Yesavage JA. Geriatric Depression Scale. Psycho-pharmacol Bull 1988;24(4):709e711.

24. Sheikh JI, Yesavage JA. Geriatric DepressionScale (GDS): recent evidence and development ofa shorter version. In: Brink TL, ed. Clinical geron-tology: A guide to assessment and intervention.New York, NY: The Hayworth Press, 1986: 165e173.

25. Whooley MA, Browner WS. Association betweendepressive symptoms and mortality in older women.Study of Osteoporotic Fractures Research Group.Arch Intern Med 1998;158(19):2129e2135.

26. Arthur A, Jagger C, Lindesay J, Graham C,Clarke M. Using an annual over-75 health checkto screen for depression: validation of the short Ge-riatric Depression Scale (GDS15) within generalpractice. Int J Geriatr Psychiatry 1999;14(6):431e439.

27. Green CP, Porter CB, Bresnahan DR,Spertus JA. Development and evaluation of theKansas City Cardiomyopathy Questionnaire: a newhealth status measure for heart failure. J Am CollCardiol 2000;35(5):1245e1255.

28. Geriatric Depression Scale. Available fromhttp://www.stanford.edu/wyesavage/GDS.html 2008;Accessed September 8, 2008.

29. Heidenreich PA, Spertus JA, Jones PG, et al.Health status identifies heart failure outpatients atrisk for hospitalization or death. J Am Coll Cardiol2006;47(4):752e756.

30. Hill PC, Pargament KI. Advances in the concep-tualization and measurement of religion and spiri-tuality. Implications for physical and mental healthresearch. Am Psychol 2003;58(1):64e74.

31. Egbert N, Mickley J, Coeling H. A review and ap-plication of social scientific measures of religiosityand spirituality: assessing a missing component inhealth communication research. Health Commun2004;16(1):7e27.

32. Katon WJ. Clinical and health services relation-ships between major depression, depressive symp-toms, and general medical illness. Biol Psychiatry2003;54(3):216e226.

526 Vol. 39 No. 3 March 2010Bekelman et al.

33. Prigerson HG, Frank E, Kasl SV, et al. Compli-cated grief and bereavement-related depression asdistinct disorders: preliminary empirical validationin elderly bereaved spouses. Am J Psychiatry 1995;152(1):22e30.

34. Prigerson HG, Shear MK, Newsom JT, et al.Anxiety among widowed elders: is it distinct fromdepression and grief? Anxiety 1996;2(1):1e12.

35. Boelen PA, van den BJ. Complicated grief,depression, and anxiety as distinct postloss syn-dromes: a confirmatory factor analysis study. Am JPsychiatry 2005;162(11):2175e2177.

36. Lichtenthal WG, Cruess DG, Prigerson HG. Acase for establishing complicated grief as a distinctmental disorder in DSM-V. Clin Psychol Rev 2004;24(6):637e662.