Religion, Spirituality, and Health Outcomes in Cancer: A Case for a Meta-Analytic Investigation

Introduction

National surveys routinely highlight the importance of religion and spirituality (R/S) to many individuals in the US. Gallup polls reveal that 65% of Americans report that religion is important in their daily lives, and 41% report attending religious services “almost every week” or more.¹ Sixty-nine percent of Americans report praying several times a week or more² and 43% pray specifically for their own health.³ The importance of R/S is further evidenced by the fact that 57% of Americans believe religion can answer all or most of today’s problems.⁴

In the context of cancer, R/S beliefs, behaviors, and experiences may be important resources and are beneficial for managing the physical, mental, and social challenges of the cancer experience. Daily spiritual experiences were associated with more self-assurance and less fear of cancer recurrence in a mixed group of cancer survivors.^{5, 6} In a sample of patients with cancer receiving active treatment, higher levels of religious beliefs were associated with greater perceived social support.⁷ R/S beliefs provide a context for patients to integrate their difficult experiences into their lives in ways that may help promote greater well-being and better quality of life, but this relationship is not uniformly positive, as negative aspects of R/S such as religious struggle may be associated with greater distress and poorer subjective health. For example, in patients receiving a bone marrow transplant, negative religious coping was correlated with more physical symptoms and poorer physical quality of life.⁸ Thus, while there does appear to be a relationship between R/S and health outcomes, this relationship is embedded in a complex matrix of varying dimensions of R/S and a variety of physical, mental, and social health outcomes.

Multiple professional organizations and accrediting bodies are formally recognizing the importance of R/S assessment and the potential role it may serve in delivering quality patient care. The Joint Commission on Accreditation of Healthcare Organizations, the largest accrediting body of health care organizations and programs in the United States, requires a spiritual assessment of all patients.⁹ The National Comprehensive Cancer Network’s Clinical Practice Guidelines on “Distress Management” identifies spiritual or religious concerns as a specific source of distress¹⁰; and the National Consensus Project for Quality Palliative Care’s Clinical Practice Guidelines recognize spiritual, religious, and existential aspects of care as one of eight core domains for optimal palliative care.¹¹ Similarly, cancer care providers are recognizing the importance of acknowledging the role of R/S in the care of their patients. In a recent survey of international members of the Multinational Association of Supportive Care in Cancer, nearly 40% of providers said it is their role to explore the spiritual concerns of their patients.¹²

Although religion, religiousness, and spirituality are often used interchangeably or even as overlapping terms, a growing segment of the US population (currently 18%) considers themselves spiritual but not religious.¹³ In the health literature, some have argued for distinct conceptualizations of the terms.^{14, 15} Others suggest R/S is best conceptualized as multidimensional, comprising related constructs.¹⁶ A discussion of multiple competing viewpoints on the unique and overlapping aspects of R/S is beyond the scope of this manuscript, but several excellent works on this topic are available for interested readers.^16–21 Moreover, we believe R/S includes “both the personal and the institutional, the traditional and the progressive, the helpful and the harmful” (p. 563).²² R/S collectively represents a constellation of important affective, behavioral, and cognitive variables that may impact the feelings, actions, and thoughts of people with cancer in ways that influence important health-related outcomes.

Emerging Trends in Research

Interest in R/S factors in cancer is continuing to expand. In a recent search of the PubMed database on the terms “religio*” or “spiritual* and cancer, a total of 3,778 abstracts were retrieved for the search period through December 31, 2013. This is a significant number of studies in this area, but what underscores the burgeoning interest in this topic is the publication pattern for five-year intervals over the last twenty years (Figure 1). In just the past decade, the number of abstracts concerning R/S and cancer have increased 2.5 times over the previous decade. During that same period, a small number of articles and book chapters have reviewed the literature on R/S in cancer. The scope of these projects has varied, ranging from broad overviews of R/S research across the cancer control continuum^{23, 24} to more recently-focused examinations of particular aspects of the R/S-health relationship within cancer such as the role of R/S coping strategies and psychological adjustment^{25, 26} or the relationship between spirituality and emotional well-being.

Figure 1.

Number of Abstracts in PubMed for the Search: Religio* or Spiritual* & Cancer

These reviews have yielded mixed findings. For example, out of 17 papers reviewed by Thune-Boyle et al.²⁵, 7 reported beneficial effects, 7 reported no effect, and 3 reported harmful effects. The authors urged caution in drawing firm conclusions about the potential beneficial or even harmful effects of R/S coping due to methodological shortcomings of the reviewed studies. Similarly, among the 27 studies reviewed by Visser et al.²⁶ only 4 employed a longitudinal design and they reported mixed results. They also suggested that no definitive conclusions could be drawn regarding the relationship between R/S and emotional well-being due to methodological shortcomings of the studies reviewed, including overlap in the items used to assess R/S and emotional well-being. In Sherman and Simonton’s overview²⁴, they acknowledged the variable findings, but noted there appeared to be a salutary association between certain dimensions of R/S and health (e.g., more adaptive psychological resources among those who are ill, improved screening behaviors), suggesting that some but not all aspects of R/S may be beneficial for some but not all aspects of health. Notably, Stefanek et al.²³ advocated a “wait and see” approach, given the stage of the research in R/S and health: “The study of religion and spirituality and cancer is in its infancy. It is too early to determine what role the [R/S] constructs play in cancer outcome either related to the disease itself, or to quality of life and adjustment measures” (p. 459). Almost a decade has passed since that comment was written and the field is going through a “growth spurt”, justifying and perhaps necessitating a closer as well as more comprehensive examination of these important relationships.

Before proceeding further, it is important to recognize that different subgroups of people express R/S in myriad and sometimes unique ways. This differential involvement across subgroups is well established. For example, older individuals often report greater R/S beliefs, including a greater belief in the existence of God²⁷ and R/S activities and beliefs are particularly important among African Americans.^{28, 29} When coping with a stressful life event such as cancer,³⁰ African-Americans often report relying on faith or spirituality more than do Whites and African American women engage in religious activities more frequently and score higher on measures of religiousness and commitment than do African American men.³¹ As a group, women are more likely than men to be religious, to engage more frequently in private prayer, and to report more frequent spiritual experiences.³² These group differences are increasingly being demonstrated among patients with cancer and survivors. However, the differential impact of R/S on health for these subgroups is less clearly established and is an important emerging focus in an era of growing interest in personalized medicine and patient-centered outcomes.^33–36

A Meta-Analytic Contribution?

Reviews that have been published to date provided helpful summaries of some of the existing findings in R/S and cancer, but no one has taken the “next step” and conducted a comprehensive yet focused meta-analysis to summarize and synthesize these diverse findings. Meta-analysis provides a particularly illuminating approach for synthesizing the results across studies to determine the strength of relationships between variables in a specific field. The lack of meta-analyses on this topic may be due to the time-intensive nature of meta-analyses, in general, and the specific challenges of summarizing research in the R/S and cancer field. These challenges include heterogeneity with respect to measurement of both R/S and cancer-related health, as well as the number of demographic and disease variables that may modify relationships between R/S and cancer.

From a measurement perspective, some of the mixed findings noted in the above reviews may be due to the challenging nature of R/S measurement. Well over a hundred measures of R/S have been used in research, many of which have poor or unestablished psychometric properties.³⁷ In addition, R/S encompasses a host of related but distinct constructs ranging from coping strategies to spiritual experiences to religious practices to beliefs about God and the world. It is not unreasonable to expect that these R/S dimensions would be associated with health outcomes in cancer in a somewhat dissimilar fashion. Accordingly, greater clarity in understanding the role of R/S in cancer may be achieved through intentional and thoughtful efforts to harmonize study findings such that similar R/S predictors are clustered together. In short, this should permit greater comparisons of “apples to apples” and “oranges to oranges,” and yield a clearer and more empirically-informed understanding of the relationship between R/S and health.

An R/S Taxonomy

In an effort to improve our understanding of the relationship between R/S and health in the cancer context, and thereby increase the potential interpretability and generalizability of study findings, we began by developing an organizing taxonomy centered on the assumption that most R/S constructs could be understood and categorized within one of three primary dimensions of R/S – affective, behavioral, or cognitive. This taxonomy has obvious roots in psychological perspectives on human experience in terms of feelings (affect), actions (behaviors), and thoughts (cognitions), but also highlights the functional or experiential nature of R/S independent of content or dispositional aspects of R/S by including measures with clear implications for religious and spiritual functioning as well as for health status (e.g., closeness to God, religious and spiritual struggle).^{38, 39} A distinct advantage to incorporating measures of functional or experiential R/S is that these measures may be more proximal to physical, mental, and social health and thus, may delineate the R/S and health connection more accurately than common R/S variables such as R/S affiliation or a global indices of R/S involvement.^{39, 40} Furthermore, this R/S taxonomy is broadly supported by empirical work on the multi-dimensionality of R/S which has identified, through factor analyses, the distinctiveness of several dimensions of R/S, including affective (e.g., religious struggle, guilt, spiritual well-being, existential well-being), behavioral (e.g., religious practices and organizational religiousness, religious/spiritual involvement) and cognitive (e.g., meaning, cognitive orientation towards spirituality, paranormal beliefs) dimensions.^41–43 Finally, the taxonomy has ties to specific operationalizations of the R/S dimensions (described below), which in structural equation modeling terminology, are latent variables expressed by manifest variables.

Affective R/S

The affective dimension of R/S refers to those aspects having to do with subjective emotional experience, such as a sense of transcendence, meaning, purpose, or connection to a source larger than oneself (Spiritual Meaning⁴⁴). Many emotions are specific to R/S, such as awe or reverence, as well as finding comfort and peace within one’s religious beliefs or spirituality (e.g., FACIT-Sp⁴⁵) or feeling struggle or anger towards God (e.g., as assessed with the Attitudes Toward God Scale^{46, 47}). Spiritual distress or feeling disconnected from God or one’s religious community are also examples of affective R/S constructs.^48–51

Behavioral R/S

Behavioral R/S refers to R/S practices or behaviors. Often these practices form part of the routine of daily life, such as public (e.g., service attendance) or private (e.g., prayer, meditation) R/S activities, which are commonly measured in R/S studies.¹⁵ Sometimes, however, these behaviors are engaged in to manage stress, illness demands, and life events that are specifically related to cancer and cancer treatments, with the goal of reducing or accommodating negative outcomes such as depression, anxiety, worry, poor quality of life, disrupted relations with family and friends, or inactivity.⁵² Other examples of R/S behaviors include coping by seeking a closer connection with the divine, pursuing specific R/S problem-solving strategies, “letting go of active striving” or trusting that God (supreme being, Nature) will resolve the stressor.⁵³

Cognitive R/S

Cognitive R/S typically focuses on the concept of religious or spiritual beliefs. Park writes, “Lay dictionaries typically define beliefs as statements that one accepts or is convicted of as true. Similarly, religious beliefs have been defined as, ‘Propositional statements a person considers to be true about religion’⁵⁴” (p.970).⁵⁵ Often, religious beliefs are grounded in doctrinal foundations or tenets of a particular faith tradition. In considering these beliefs, several different aspects might be kept in mind, including the specific content³⁷ and strength of the beliefs (e.g., Certainty in Religious belief)⁵⁶ as well as how literally or non-literally the beliefs are held.⁵⁷ Other types of beliefs (less bound to particular doctrine) involve religiously-based illness perceptions, including beliefs about God as the locus of control of one’s health or disease status,⁵⁸ fatalistic beliefs that health outcomes are inevitable and/or determined by God⁵⁹, beliefs about God’s role in healing, including miracles⁶⁰, and beliefs about how God answers prayer. The cognitive R/S dimension also includes religious doubts, salience or commitment, images of the divine, and perceptions of spiritual growth (e.g., positive spiritual change secondary to the experience of being diagnosed with and treated for cancer).

‘Other’ R/S

Despite our useful theoretically- and empirically-informed taxonomy, not all R/S constructs are easily categorized into an affective, behavioral, or cognitive dimensional structure. For example, religious affiliation is often a marker for a host of attitudes, beliefs, and behaviors that defy easy categorization. Similarly, religious social support may speak to the sense of comfort derived from attending R/S services, which may have both affective and behavioral components. Other measures may be a composite of multiple dimensions of R/S (e.g., service attendance, prayer, importance) or an index of general religiousness.^{61, 62} We developed the ‘other’ R/S category to account for the subset of R/S measures that are relatively frequently used and that may have important relationships with health outcomes, but that do not squarely fit into the affective, behavioral, and cognitive dimensions of R/S. Table 1 provides examples of R/S constructs and measures within each of the four R/S dimensions.

Table 1.

R/S Taxonomy

Dimension	Illustrative Constructs	Common Measures
Affective	Spiritual well-being	Functional Assessment of Chronic Illness Therapy – Spiritual Well-Being, Spiritual Well-Being45; Spiritual Well-Being Scale64
	Spiritual distress	Negative RCOPE53; Spiritual Strain Scale65
	R/S experiences	Daily Spiritual Experiences Scale66
Behavioral	R/S coping	Positive RCOPE53; Brief COPE – religious coping subscale67
	R/S activities - private	Frequency of prayer15
	R/S activities - public	Frequency of service attendance15
Cognitive	God image	Image of God Scale68
	R/S beliefs	Spiritual Involvement and Beliefs Scale – existential/meditative beliefs and internal beliefs subscales69
	Spiritual growth	PTGI – spiritual change subscale70
	Locus of control	Cancer Locus of Control Scale – religious control subscale71
	Importance	Single item – perceived importance of R/S15
	R/S problem-solving	Religious Problem-Solving Styles52
	Religious orientation	DURel – intrinsic religiousness subscale62
‘Other’	Composite Index	Santa-Clara Strength of Religious Faith Inventory61, System of Beliefs Inventory 15R – beliefs and practices subscale72
	Social support	System of Beliefs Inventory 15R – social support subscale72
	Affiliation	affiliation with R/S group(s)

Specific Aims

In the succeeding articles, we describe the results of three meta-analyses of the relationship between measures of R/S and patient-reported health outcomes (e.g., symptom burden, quality of life). Consistent with the World Health Organization’s view of health as wellness⁶³, we organized our meta-analyses by type of health outcome – physical, mental, and social – which comprises positive (i.e., well-being) and negative (i.e., dysfunction, infirmity) constructs within each type of health outcome. Our meta-analyses included two broad aims: 1) to identify the degree of association between measures of R/S and health outcomes; and 2) to compare the strength of the R/S dimensions relative to each other across the health domains. For exploratory purposes, we were also interested in identifying the conditions under which the relations between R/S and health were strengthened or attenuated. We anticipated that relations between R/S and health might vary as a function of gender, age, race, cancer type, stage, and phase of treatment and recovery or disease progression, as well as dimensions of R/S (affective, behavioral, cognitive, ‘other’) and health outcomes. In the papers that follow, we focus first on physical health outcomes such as fatigue, pain, and physical well-being. Second, we focus on mental health outcomes such as depression, anxiety, and emotional well-being. Third, we focus on social health outcomes such as social well-being, social distress, and social support. Finally, we close with a synthesis of the findings across all three meta-analyses and discussion of the implications for future research. We expect the results of this methodologically rigorous effort will help clarify current findings regarding associations between R/S variables and health outcomes in the cancer context. Additionally, it should help focus future work in this area by identifying key variables for inclusion in studies of R/S and cancer and highlighting particular subgroups for whom dimensions of R/S are particularly important to their health.