Duke Religion Index

1. Abstract

The Duke Religion Index (widely known by the acronym DUREL, or the Duke University Religion Index) is a brief, five-item psychometric instrument designed to assess three major dimensions of religious involvement: organizational religious activity (OR), non-organizational religious activity (NOR), and intrinsic religiosity (IR) or subjective religious motivation. Developed in 1997 by Harold G. Koenig, George R. Parkerson Jr., and Keith G. Meador at the Duke University Medical Center, the scale was created to address the acute need in psychiatric, epidemiological, and medical research for an instrument that is psychometrically robust, concise enough for large-scale clinical surveys, and culturally non-offensive across diverse religious traditions.

The scale consists of two single-item behavioral subscales—measuring frequency of attendance at public religious services (OR) and frequency of private devotional activities such as prayer, meditation, or scripture study (NOR)—and a three-item cognitive-motivational subscale measuring intrinsic religiosity adapted from Dean R. Hoge’s 10-item Intrinsic Religious Motivation Scale. The OR and NOR items use 6-point response scales, whereas the IR items utilize a 5-point Likert scale. Psychometric investigations consistently indicate strong internal consistency for the intrinsic subscale (α = .75 to .91 across diverse international cohorts), solid test-retest reliability (intraclass correlation coefficients ranging from .80 to .93), and robust construct, convergent, and predictive validity across psychiatric, oncological, geriatric, and cardiovascular populations. Rather than aggregating the items into a single composite score, methodological guidelines recommend analyzing each of the three dimensions separately to evaluate their unique, independent, and sometimes contrasting relationships with mental and physical health outcomes.

2. Keywords

Duke Religion Index, DUREL, Harold G. Koenig, intrinsic religiosity, organizational religious activity, non-organizational religious activity, religious coping, psychiatric epidemiology, psychometrics, health outcomes, religion and health, Allport intrinsic extrinsic

3. Authors

The Duke University Religion Index was authored by a multidisciplinary team of physician-researchers and psychiatrists based at Duke University Medical Center in Durham, North Carolina, USA:

• Harold G. Koenig, M.D., M.H.Sc. — Professor of Psychiatry and Behavioral Sciences, Associate Professor of Medicine, and Director of the Center for Spirituality, Theology and Health at Duke University Medical Center. Dr. Koenig is widely recognized as a pioneering authority on the intersection between religion, spirituality, and medicine.
• George R. Parkerson Jr., M.D., M.P.H. — Professor Emeritus and former Chair of the Department of Community and Family Medicine at Duke University Medical Center, specialized in health status measurement, functional assessment, and primary care research.
• Keith G. Meador, M.D., Th.M., M.P.H. — Professor of Psychiatry and Health Policy, Professor of Ethics and Society, and Director of the Center for Biomedical Ethics and Society at Vanderbilt University (formerly of Duke University Medical Center). Dr. Meador works at the intersection of theological ethics, public health, and social psychiatry.

4. Purpose

The primary purpose of the Duke Religion Index is to provide clinical researchers, psychiatric epidemiologists, and behavioral scientists with a rapid, psychometrically sound, multidimensional assessment of religious involvement that can be incorporated into comprehensive health outcome batteries without overburdening participants. Prior to the development of the DUREL, investigations into the epidemiology of religion and health were hindered by two major measurement problems:

1. Oversimplification: Many large-scale epidemiological studies relied on a single, global question—most frequently, religious affiliation (e.g., Catholic, Protestant, Jewish, None) or attendance frequency. This practice conflated sociocultural identity with active religious engagement and obscured distinct behavioral and motivational pathways.
2. Participant Burden: Existing validated multidimensional instruments, such as the Religious Orientation Scale (Allport & Ross, 1967) or the Intrinsic Religious Motivation Scale (Hoge, 1972), contained between 10 and 30 items. In acute medical, geriatric, or intensive psychiatric settings where patients experience cognitive fatigue, pain, or functional impairment, administering lengthy scales is practically unfeasible.

The DUREL was designed specifically to bridge this gap by capturing the most salient aspects of religious commitment in just five questions that can be completed in less than two minutes. The scale measures religiousness in a manner that is applicable across Western monotheistic traditions (Christianity, Judaism, Islam) as well as adaptable to other religious frameworks. By isolating organizational, non-organizational, and intrinsic domains, the instrument enables researchers to determine whether observed correlations with health parameters—such as depression, systemic inflammation, cardiovascular reactivity, or mortality—are driven by social capital and community networks (OR), personal cognitive coping mechanisms (NOR), or deeply held existential commitments and internal worldviews (IR).

5. Psychological Construct

The DUREL assesses religiosity (or religious involvement), operationalized as a multidimensional construct comprising behavioral, social, and psychological components. Modern psychometrics explicitly distinguishes religiosity from broader contemporary definitions of spirituality. Whereas spirituality is often conceptualized as an individualistic, subjective quest for meaning, transcendence, or connectedness that may or may not involve institutional structures, religiosity denotes adherence to beliefs, practices, and community rituals rooted in an established religious tradition. The DUREL measures three distinct dimensions:

1. Organizational Religious Activity (OR)

Organizational religiosity refers to participation in public, institutional, and communal religious activities. Measured by Item 1 (“How often do you attend church or other religious meetings?”), OR assesses the degree to which an individual integrates into a formal religious community. This construct embodies several active mechanisms:

• Social Support and Community Integration: Attendance provides structured opportunities to receive emotional, informational, and tangible social support from fellow congregants.
• Normative Behavioral Reinforcement: Regular engagement in religious communities often reinforces prosocial behaviors and health-promoting lifestyle choices (e.g., lower rates of substance abuse, smoking, and risky behaviors).
• Cognitive and Ritual Structuring: Participation in liturgy, collective singing, and shared rituals offers regular emotional regulation and cognitive reappraisal within a community setting.

2. Non-Organizational Religious Activity (NOR)

Non-organizational religiosity encompasses private, individualized devotional practices performed in solitude or informal settings. Measured by Item 2 (“How often do you spend time in private religious activities, such as prayer, meditation, or Bible study?”), NOR reflects personal devotion independent of public supervision. Key psychological attributes include:

• Spontaneous Coping Strategy: Private prayer or contemplative meditation is frequently employed as an active, self-initiated coping mechanism during acute life stressors, medical crises, or depressive episodes.
• Internalized Emotional Regulation: Contemplative practices provide immediate personal sanctuary, stress mitigation, and affective stabilization.
• Independence from Physical Function: Unlike public attendance (which requires physical mobility and transportation), private devotion can be sustained by bedridden, chronically ill, or institutionalized individuals, making NOR uniquely informative in clinical populations.

3. Intrinsic Religiosity (IR)

Intrinsic religiosity operationalizes the degree to which an individual has internalized their religious faith as the primary master-motive of their life. Measured by Items 3, 4, and 5, IR reflects an experiential and existential orientation wherein religion is lived for its own sake rather than used as an instrumental means to secondary ends (such as social status or personal comfort). Its psychological dimensions include:

• Experiential Transcendence (Item 3): Experiencing the presence of the Divine in daily life, providing affective reassurance, emotional security, and a sense of divine companionship.
• Foundational Worldview (Item 4): Religious beliefs serving as the primary interpretive lens and cognitive architecture through which all life events, challenges, and moral choices are understood.
• Behavioral Integration across Domains (Item 5): Deliberately applying religious principles to interpersonal interactions, ethical decisions, and occupational responsibilities, fostering psychological coherence and integrity.

6. Theoretical Framework

The structural and conceptual foundation of the DUREL is grounded in several major theoretical developments in the psychology and sociology of religion, as well as the behavioral medicine literature:

Allport’s Religious Orientation Theory

The primary conceptual root of the intrinsic subscale is Gordon Allport’s seminal distinction between intrinsic and extrinsic religious orientations (Allport & Ross, 1967). Allport posited that:

“The extrinsically motivated person uses his religion, whereas the intrinsically motivated person lives his religion.”

Extrinsic religiosity is utilitarian, self-serving, and instrumental, used to provide comfort, status, social acceptance, or business contacts. In contrast, intrinsic religiosity represents an internalized master-motive that harmonizes desires, disciplines impulses, and provides an overarching purpose in life. Dean R. Hoge (1972) operationalized Allport’s concept into a psychometrically refined 10-item scale. In developing the DUREL, Koenig and colleagues extracted the three items with the highest factor loadings from Hoge’s scale that demonstrated the most consistent associations with health outcomes, preserving the theoretical purity of the intrinsic construct while eliminating psychometric redundancy.

Sociological Multidimensionality

From a sociological perspective, the DUREL incorporates insights from Charles Glock and Rodney Stark’s (1965) multidimensional framework of religiosity, which asserted that religious commitment cannot be captured by a unidimensional indicator. Glock identified ritualistic, ideological, intellectual, experiential, and consequential dimensions. By explicitly separating behavioral-institutional practice (ritualistic/organizational), private devotional behavior (individual practice), and cognitive-experiential dedication (intrinsic faith), the DUREL respects the theoretical reality that these domains do not necessarily co-vary synchronously.

Stress and Coping Theory

In behavioral medicine, the DUREL aligns with Richard Lazarus and Susan Folkman’s (1984) Transactional Model of Stress and Coping, as elaborated in Kenneth Pargament’s (1997) theory of religious coping. In this paradigm, religious beliefs and practices serve as primary cognitive appraisals (interpreting adversity as meaningful or benign within a cosmic order) and secondary coping resources (drawing on prayer or pastoral support to manage uncontrollable stressors). By isolating private devotion from public attendance, the DUREL allows researchers to observe how individuals shift coping modalities when external physical capacities diminish.

7. Validity

The psychometric validity of the DUREL has been thoroughly documented across clinical, community, and cross-cultural samples worldwide.

Construct and Convergent Validity

During the scale’s original validation in the Duke Hospital Study of 458 medically hospitalized patients aged 60 and older, Koenig et al. (1997) demonstrated that the 3-item intrinsic subscale correlated exceptionally highly (r = .85) with Hoge’s full 10-item Intrinsic Religious Motivation Scale, proving that reducing the item count did not compromise construct coverage. Furthermore, the intrinsic subscale correlated moderately with OR (r = .40) and NOR (r = .42), confirming that while these dimensions share conceptual commonality, they remain psychometrically distinct.

Subsequent international validation studies have reaffirmed strong convergent validity. In validation studies of the Portuguese, Spanish, German, Persian, and Chinese adaptations, DUREL subscales demonstrated strong, positive correlations with established measures of spirituality and religious coping, including the Brief RCOPE, the FACIT-Sp (Functional Assessment of Chronic Illness Therapy–Spiritual Well-Being), and the Santa Clara Strength of Religious Faith Questionnaire (SCSORFQ).

Discriminant Validity

The discriminant validity of the DUREL is demonstrated by the differential relationships that OR, NOR, and IR exhibit with psychosocial and clinical variables. In the initial cohort (Koenig et al., 1997; Koenig, Hays, et al., 1997):

• OR was significantly associated with greater perceived social support, lower self-rated and clinician-rated depressive symptoms, lower severity of medical illness, and less functional impairment.
• NOR showed a distinct pattern: it was associated with high social support, but often showed null or positive cross-sectional correlations with physical illness severity and functional impairment. This reflects a well-known clinical phenomenon: patients experiencing severe functional decline or chronic pain may increase private prayer even when unable to attend public religious gatherings.
• IR correlated robustly with cognitive resilience, psychological well-being, and accelerated recovery from major depression, while remaining distinct from generalized optimism or trait social desirability.

Predictive and Criterion Validity

Prospective epidemiological studies utilizing the DUREL have demonstrated predictive validity across several objective health endpoints:

• Depression Remission: In a landmark prospective study of medically ill hospitalized older adults with major depressive disorder, high scores on the intrinsic religiosity subscale independently predicted a shorter time to remission of depression (hazard ratio = 1.70, p = .01), controlling for baseline severity, social support, and physical functional status (Koenig, George, & Peterson, 1998).
• Biological Biomarkers: Higher DUREL scores have been prospectively associated with lower circulating levels of interleukin-6 (IL-6), improved immune function, reduced cardiovascular reactivity to acute laboratory stressors, and longer leukocyte telomere length in older adults.
• Mortality: In large-scale prospective cohort studies, organizational religious attendance (Item 1) has repeatedly predicted reduced all-cause mortality, even after rigorous adjustment for baseline health status, socioeconomic status, and health behaviors.

8. Reliability

The reliability of the DUREL has been extensively tested using classical test theory and internal consistency metrics across varied languages, demographic cohorts, and clinical conditions.

Internal Consistency

Because Items 1 and 2 are independent single-item behavioral measures assessing distinct frequencies of different activities, computing a global Cronbach’s alpha across all five items is psychometrically problematic and generally discouraged by the scale authors. Internal consistency is formally evaluated on the 3-item Intrinsic Religiosity (IR) subscale:

• Original Duke Validation Cohort (Koenig et al., 1997): The 3-item IR subscale demonstrated a Cronbach’s alpha of α = .75 in a clinical sample of older hospitalized medical patients.
• General Population and Cross-Cultural Validations: In subsequent community and clinical studies, the IR subscale has consistently demonstrated high internal consistency:
• Brazilian Portuguese version (Taunay et al., 2012): α = .88 in psychiatric outpatients and α = .86 in community controls.
• Spanish version (Alminana et al., 2013): α = .84 to .89 across university and community cohorts.
• Persian version (Hafizi et al., 2013): α = .86 to .91 among university students and medical patients.
• German version (Büssing et al., 2015): α = .82 to .86.
• Chinese version (Wang et al., 2016): α = .85 in primary healthcare settings.

Test-Retest Stability

The temporal stability of the DUREL has been verified across various retest intervals. Storch et al. (2004) examined the DUREL in a sample of young adults over a two-week interval and reported an overall test-retest reliability coefficient of r = .91. The Intrinsic Religiosity subscale yielded an intraclass correlation coefficient (ICC) of .92, while the single-item OR and NOR questions demonstrated ICC values of .88 and .85, respectively. Similar longitudinal stability has been documented over 1- to 3-month intervals in stable outpatients, confirming that the DUREL captures enduring religious orientations and behavioral patterns rather than transient affective states.

9. Factor Analysis

Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) have confirmed the underlying structural topology of the DUREL across diverse populations.

Exploratory Factor Structure

In the original developmental analyses conducted on data from the Duke Hospital Study (N = 458) and subsequent epidemiological cohorts involving over 7,000 community-dwelling adults, principal components and common factor analyses revealed a two- or three-factor structure depending on whether single behavioral items were forced into individual factors or allowed to load independently:

• When all 5 items are analyzed together, Items 3, 4, and 5 consistently cluster on a single dominant factor accounting for 60% to 75% of the common variance, with factor loadings ranging from .78 to .89.
• Items 1 (OR) and 2 (NOR) load on separate, secondary behavioral factors, confirming their divergence from pure cognitive/motivational intrinsic faith.

Confirmatory Factor Analysis (CFA)

Multiple CFA investigations (e.g., Storch et al., 2004; Taunay et al., 2012; Hafizi et al., 2013) have tested alternative structural models:

1. Unidimensional Model (1 Factor): All 5 items loading onto a single general “religiosity” factor. This model consistently demonstrates poor fit across all standard psychometric criteria (χ²/df > 15.0, RMSEA > .14, CFI < .85, TLI < .80).
2. Two-Factor Model: Factor 1 comprising the behavioral items (OR and NOR) and Factor 2 comprising the intrinsic items (Items 3, 4, 5). This model achieves moderate, but suboptimal fit.
3. Three-Factor Correlated Model: Factor 1 (OR: Item 1), Factor 2 (NOR: Item 2), and Factor 3 (IR: Items 3, 4, 5). This theoretical three-dimensional model consistently demonstrates excellent fit indices across cultures:

Standardized factor loadings for the three intrinsic items in the three-factor model are consistently high: Item 3 (.78–.88), Item 4 (.82–.92), and Item 5 (.76–.86). The correlations between the latent factors typically range between .35 and .55, providing empirical justification for treating them as related yet distinct dimensions.

10. Instrument / Measurement Tool

• Instrument Name: Duke University Religion Index (DUREL) / Duke Religion Index
• Target Population: Adults (aged 18 and older) across general, psychiatric, geriatric, and medically ill clinical populations.
• Administration Format: Self-administered questionnaire, structured clinical interview, or telephone survey.
• Administration Time: Approximately 1 to 2 minutes.
• Number of Items: 5 items.
• Subscale Breakdown:
  • Subscale 1: Organizational Religiosity (OR): Item 1 (public attendance).
  • Subscale 2: Non-Organizational Religiosity (NOR): Item 2 (private religious practices).
  • Subscale 3: Intrinsic Religiosity (IR): Items 3, 4, and 5 (subjective belief and lived faith).
• Response Scales:
  • Item 1 (OR): 6-point ordinal frequency scale (1 = Never; 2 = Once a year or less; 3 = A few times a year; 4 = A few times a month; 5 = Once a week; 6 = More than once a week).
  • Item 2 (NOR): 6-point ordinal frequency scale (1 = Rarely or never; 2 = A few times a month; 3 = Once a week; 4 = Two or more times/week; 5 = Daily; 6 = More than once a day).
  • Items 3, 4, 5 (IR): 5-point Likert scale (1 = Definitely not true; 2 = Tends not to be true; 3 = Unsure; 4 = Tends to be true; 5 = Definitely true of me).
• Scoring and Analytic Guidelines:
  • Recommended Scoring Approach: Analyze the scale as three separate subscales: OR (range: 1 to 6), NOR (range: 1 to 6), and IR (sum of items 3–5; range: 3 to 15). Higher scores denote higher levels of religious involvement.
  • Total Score Caution: Although some researchers compute a single total composite score (range: 5 to 27, where 5 = lowest religiousness and 27 = highest religiousness, or vice versa depending on direction of coding), the scale authors explicitly discourage summing all 5 items. Combining OR, NOR, and IR into a single index collapses distinct social, behavioral, and cognitive phenomena and can obscure divergent or opposing clinical effects.

11. Permissions & Fee and Test Year

The Duke Religion Index was officially published in 1997 in the American Journal of Psychiatry (Koenig, Parkerson, & Meador, 1997). The instrument was placed in the public domain by its authors to encourage research on spirituality, mental health, and medical outcomes.

• Usage Fees: There are no licensing fees, copyright royalties, or purchase costs associated with administering the DUREL.
• Permission Requirements: Researchers and clinicians do not need formal written permission to utilize the instrument in academic investigations, epidemiological surveys, or clinical practices. The authors require only that proper academic attribution and citation of the primary validation articles be maintained in any subsequent publication or presentation.
• Translations: Validated translations exist in over 20 languages (including Spanish, Portuguese, German, French, Italian, Persian, Arabic, Chinese, Japanese, and Korean) through independent academic validation studies worldwide.

12. References

• Allport, G. W., & Ross, J. M. (1967). Personal religious orientation and prejudice. Journal of Personality and Social Psychology, 5(4), 432–443. https://doi.org/10.1037/h0021212
• Alminana, R., Loewenthal, K. M., & Balbuena, F. (2013). Spanish adaptation of the Duke University Religion Index (DUREL). The Spanish Journal of Psychology, 16, E46. https://doi.org/10.1017/sjp.2013.43
• Büssing, A., Koenig, H. G., & Föller-Mancini, A. (2015). Validation of the German version of the Duke Health and Religion Index (DUREL) and its relation to emotional exhaustion in health care professionals. Religions, 6(4), 1319–1330. https://doi.org/10.3390/rel6041319
• Glock, C. Y., & Stark, R. (1965). Religion and society in tension. Rand McNally.
• Hafizi, S., Koenig, H. G., Arbabi, M., Pakrah, M., & Malakouti, S. K. (2013). Duke University Religion Index: Persian adaptation and psychometric properties. Psychological Reports, 112(1), 146–158. https://doi.org/10.2466/08.02.PR0.112.1.146-158
• Hoge, D. R. (1972). A validated intrinsic religious motivation scale. Journal for the Scientific Study of Religion, 11(4), 369–376. https://doi.org/10.2307/1384677
• Koenig, H. G. (1997). Is religion good for your health? The effects of religion on physical and mental health. Haworth Press.
• Koenig, H. G., George, L. K., & Peterson, B. L. (1998). Religiosity and remission of depression in medically ill older patients. American Journal of Psychiatry, 155(4), 536–542. https://doi.org/10.1176/ajp.155.4.536
• Koenig, H. G., Hays, J. C., George, L. K., & Blazer, D. G. (1997). Modeling the cross-sectional relationships between religion, physical health, social support, and depressive symptoms. American Journal of Geriatric Psychiatry, 5(2), 131–143. https://doi.org/10.1097/00019442-199700520-00006
• Koenig, H. G., Parkerson, G. R., & Meador, K. G. (1997). Religion index for psychiatric research: A 5-item measure for use in health outcome studies. American Journal of Psychiatry, 154(6), 885–886. https://doi.org/10.1176/ajp.154.6.885
• Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
• Pargament, K. I. (1997). The psychology of religion and coping: Theory, research, practice. Guilford Press.
• Storch, E. A., Roberti, J. W., Heidgerken, A. D., Storch, J. B., Lewin, A. B., Killiany, E. M., Baumeister, A. L., & Geffken, G. R. (2004). The Duke Religion Index: A psychometric investigation. Depression and Anxiety, 20(2), 97–99. https://doi.org/10.1002/da.20038
• Taunay, T. C., Gondim, F. A., Macêdo, D. S., Moreira-Almeida, A., Gurgel, L. A., Andrade, L. M., & Carvalho, A. F. (2012). Validity of the Duke Religion Index in a Brazilian sample. Revista Brasileira de Psiquiatria, 34(3), 327–335. https://doi.org/10.1016/j.rbp.2012.02.003
• Wang, Z., Koenig, H. G., Tong, Y., & Al Shohaib, S. (2016). Psychometric properties of the Chinese version of Duke University Religion Index. Annals of Depression and Anxiety, 3(1), 1073.

13. Items of the Scale (Questionnaire)