Introduction

Over the last few decades, an important increase in the elderly population has been recorded, with the worldwide proportion of people aged 65 years and older increasing from 6.1% to 8.8% from 1990 to 2017 (Cheng et al. 2020). The elderly population is concentrated in more developed countries, which consequently face increasing prevalence of chronic degenerative diseases (United Nations 2020). In fact, older people are more affected by life-threatening diseases such as neoplastic pathologies (White et al. 2014) and heart failure (Bosch et al. 2019), which can severely reduce their quality of life and may lead to hopelessness, depression, and suicidal ideation (Uncapher et al. 1998). Thus, it is important to invest more resources to gain a deeper understanding of the health determinants that may help improve their quality of life and life expectancy by providing them with a sense of well-being, hope, and wholeness. In this context, religion and spirituality are two resources that can not only help improve the quality of life of the elderly population but also provide them with significant health benefits (Litalien et al. 2021).

Religiosity is a complex phenomenon characterized by two main dimensions (Allport and Ross 1967): intrinsic religiosity, which consists of the personal commitment to faith driven by the involvement of people in the principles of their religion, and extrinsic religiosity, which denotes the adherence to religion driven by the pursuit of personal benefits and social approval. On the other hand, spiritual well-being is defined as a sense of wholeness promoted by a fulfilling relationship with God, self, the community, or the environment (Ellison 1983). A study of the Pew Research Center's Forum on Religion and Public Life (Hackett et al. 2012) estimated that 84% of the world population had a religious affiliation. According to the bio-psychosocial model of health promoted by the World Health Organization since its foundation, and given the positive correlation between age group and religiosity/spirituality (R/S) habits (Stearns et al. 2018), we can underline the importance of further studies on its determinants of health, as it involves a large number of individuals. Interest in the health determinants of the elderly population has increased in recent years (Litalien et al. 2021). The aim of this systematic review was to estimate the association between people's exposure to R/S and longevity among older adults.

Methods

Search strategy and data extraction

For the present study, a systematic search was conducted on the PubMed, PsycINFO, and CINAHL databases to find observational studies that evaluated the association between R/S and survival, life expectancy, or all-cause mortality. Search strings were created by combining the keywords “Religiosity,” “Spirituality,” “Survival,” “Mortality,” “Life expectancy,” “Old people,” and “Older adult*” and their respective synonyms using Boolean operators. For the PubMed database, a combination of Medical Subject Headings (MeSH) terms and text words was used. The search strings are reported in Appendix A.

The retrieved studies were independently reviewed for eligibility by two authors (DV and GG) in a two-step process: an initial screening was performed according to titles and abstracts, and a second screening was then performed to assess the full texts. At both stages, disagreements between the two reviewers were resolved by consensus or, if the agreement was not reached, by consulting another author (AB). The reference lists of the included studies were also checked by the authors for any relevant articles not already considered.

A data extraction form based on the research question was created using Microsoft Excel. Data on the following study characteristics were collected: (1) author name, year, and country of publication; (2) study design and data sources; (3) sample size and characteristics; (4) measure of exposure; (5) measure of outcome; (6) results; (7) confounders; and (8) conclusions. We performed a descriptive analysis to report the characteristics of the included studies (Table 1).

Table 1 Characteristics of the included studies
Full size table

Inclusion criteria

The following inclusion criteria were considered: studies that involved older adults (65+ years old or mean age of the sample of 65+ years), studies that reported measures of exposure (religiosity/spirituality) and outcome (survival, all-cause mortality, or life expectancy), and a measure of the association between exposure and outcome. Articles were excluded when not available in the English language. If multiple papers used data from the same study, only the most comprehensive was considered.

Methodological assessment

The methodological quality of the included studies was evaluated using the Newcastle–Ottawa Scale (NOS) for cohort studies (Wells et al. 2000). A score of at least 6 out of 9 indicated high quality, and the cutoff value for an adequate follow-up period was set a priori to 60 months (5 years), with a follow-up adequacy rate of 70%. All 13 included cohort studies were of high quality, with a mean NOS score of 7.8. The quality assessment results are reported in Appendix B.

Results

Identified studies

The search strategy returned a total of 238 citations from electronic databases. After removal of duplicates, 218 titles and abstracts were screened, and 32 full texts were assessed for eligibility, of which 13 met the inclusion criteria. None of the publications in the bibliographic references of the articles met the inclusion criteria. The literature search is shown in detail in Fig. 1.

Fig. 1
figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram of the article search

Full size image

Measures of exposure and outcome assessment

To quantify the exposure to religiosity, 11 studies asked for information about the subjects' religious attendance frequency (i.e., “How often do you usually attend religious services?”) (Oman and Reed 1998; Hart 2001; Hill et al. 2005; Bagiella et al. 2005; Teinonen et al. 2005; Dupre et al. 2006; Zhang 2008; Schnall et al. 2010; Zeng et al. 2011; McDougle et al. 2016; Park et al. 2016), one asked whether they were part of a religious community (Fraser et al. 2020), two assessed their involvement in private religious activities such as prayer, meditation, or Bible study (Helm et al. 2000; McDougle et al. 2016), and one used a validated questionnaire to evaluate both religious attendance and spiritual peace (Park et al. 2016), namely the Brief Multidimensional Measure of Religiousness and Spirituality (BMMRS), which was first proposed by the Fetzer Institute in 1999 (Fetzer Institute/National Institute on Aging Working Group 1999).

Characteristics of the included studies

All the included publications were prospective cohort studies. Most of them (10/13) were conducted in the United States, one study was conducted in Finland (Teinonen et al. 2005), and two studies were conducted in China (Zhang 2008; Zeng et al. 2011). All the studies conducted in the United States focused mainly on a Caucasian sample, with two exceptions: the study by Dupre et al. (2006) involved whites and African Americans in an almost 1:1 proportion, and the study by Hill et al. (2005) included a population of Mexican origin. Overall, a small percentage of other ethnicities were studied. Fourteen of 15 studies involved men and women almost equally, whereas the publication by Schnall et al. (2010) included only women. One publication focused on patients with congestive heart failure (Park et al. 2016).

The sample sizes in the included studies ranged from 191 (Park et al. 2016) to 22,451 individuals (Schnall et al. 2010) older than 65 years. As three articles also examined younger individuals, only a subgroup analysis that focused on older adults was considered in the systematic review. Oman and Reed (1998) recruited subjects with ages starting from 55 years, and only the subgroup data for those older than 65 years were taken into account. In the study by Fraser et al. (2020), people aged 30 years and older were recruited, and only the subgroup of individuals older than 70 years were examined in this review. Schnall et al. (2010) enrolled individuals aged 50 years and older. Thus, a subgroup analysis of subjects aged 70 to 79 years was considered for the present study.

Study results

Table 1 reports the estimates of exposure and effect measures of the included studies. Among the 13 included studies, seven found a negative association between religiosity and mortality (Oman and Reed 1998; Hart 2001; Hill et al. 2005; Dupre et al. 2006; Schnall et al. 2010; Zeng et al. 2011; Fraser et al. 2020), and six showed mixed results (Helm et al. 2000; Bagiella et al. 2005; Teinonen et al. 2005; Zhang 2008; McDougle et al. 2016; Park et al. 2016). Of the seven articles that reported a negative association between religiosity and mortality, six were conducted in the United States and one was conducted in China (Zeng et al. 2011). Most study subjects described in the US publications were of Caucasian ethnicity (non-Hispanic whites), with the exception of those in one article (Hill et al. 2005), which included Caucasians and African Americans at a proportion of almost 1:1. In general, a high prevalence of Christian affiliation was reported. The study from China (Zhang 2008; Zeng et al. 2011) enrolled subjects from 22 provinces where Han ethnicity represented the vast majority (more than 90% of the total population). No data concerning the religious affiliation of these subjects were provided, although the authors indicated Buddhism and Taoism as the two most important religious affiliations in the country.

Four of the six studies that showed mixed results were conducted in the United States (Helm et al. 2000; Bagiella et al. 2005; McDougle et al. 2016; Park et al. 2016); the remaining two took place in Finland (Teinonen et al. 2005) and China (Zhang 2008). Five of the six studies were conducted mainly with a Caucasian (non-Hispanic white) population; two involved samples whose main religious group was protestant (Helm et al. 2000; Teinonen et al. 2005), whereas four gave no information about religious affiliations (Bagiella et al. 2005; Zhang 2008; McDougle et al. 2016; Park et al. 2016), although three of these studies were conducted in the United States, which had a protestant majority (Bureau 2021), and the other one was conducted in China.

In the study by McDougle et al. (2016), religious attendance showed a negative association with mortality, although praying more to cope with stress was found to be positively associated with increased mortality. Park et al. (2016) found that subjects who experienced spiritual peace showed a lower mortality rate than those who did not, but no association was found between religiosity and mortality. Bagiella et al. (2005) found no consistent results, as only people from two of the four different samples showed a negative association between religiosity and mortality after adjusting for confounders. Helm et al. (2000) found that religious attendance was associated with a reduced mortality rate only in subjects without impairments in activities of daily living (ADL). Lastly, Teinonen et al. (2005) and Zhang (2008) demonstrated a negative association between religious attendance and mortality only among women. In addition, Zhang (2008) found similar results for individuals in poor health but not for those in good health.

Discussion

Overall, more than half of the included studies showed a negative association between religiosity and mortality. The other studies reported mixed results; that is, they found a significant negative association only for specific subgroups of the population (e.g., women) or components of the spiritual activities.

Several mechanisms have been hypothesized to explain the link between religious attendance and longevity. Among these mechanisms, the potential mediating factors of social support, health behavior, and mental health were the most studied (Hill et al. 2005). For instance, religious attendance may reduce the risk of mortality, partly through the promotion of social contacts and social resources (Ellison and Levin 1998; George et al. 2002). In fact, people who regularly attend a religious community may benefit from greater social support, community involvement, and access to material and psychological help (Fraser et al. 2020), which can reduce stress and provide options for assistance that may also affect mortality (Ellwardt et al. 2015; Olaya et al. 2017).

Moreover, people who are more religious are likely to engage in healthier lifestyles (Ellison and Levin 1998; George et al. 2002). For example, religious involvement may deter drinking and smoking by increasing exposure to anti-abuse norms and peers and by reducing contact with deviant networks (Gorsuch 1995; Ellison and Levin 1998; Hill et al. 2005). Previous literature reviews also suggested that attendance at religious services is associated with better mental health and psychological well-being (Hackney and Sanders 2003), as religiously motivated expressions such as hope, forgiveness, altruism, and love have been proposed as psychological factors that may strengthen host resistance (Levin 1996) and satisfy the need for social contact and meaning in life (Oman and Reed 1998).

Although social support, mental health, and health behavior differences have often been suggested as the primary mechanisms of health benefit among religious people, several religious and mortality studies found that adjustments for those mediating factors did not fully account for the survival benefit that the religious participants received (Oman and Reed 1998; Koenig et al. 1999; Helm et al. 2000). This suggests that other important mechanisms that may connect religious involvement to reduced mortality are not yet understood.

Concerning the included studies that showed mixed results, two revealed gender-related differences, with religiosity being associated with better survival among women but not among men (Teinonen et al. 2005; Zhang 2008). The two studies were conducted in two countries, China and Finland, where the social component of religion is weaker than in other states. In particular, the fairly regularized religious lives typical of Western religions, characterized by attending weekly religious services and other activities, do not exist in China (Zeng et al. 2011), and Finland presents a large discrepancy between engagement in private prayer and public worship, with almost half of all Finns usually praying at least once a month, but only 14% of them attending religious services as often (Teinonen et al. 2005). Moreover, a previous meta-analysis revealed that the impact of religiosity on longevity was weaker in studies that used measures of private religious involvement (McCullough et al. 2000) and that the association between religious attendance and mortality was weaker among Finnish women than among the older population in the United States (Koenig et al. 1999), where the social component of religious activity is stronger. In addition, praying has been described as more helpful to females than males (Pargament 1997).

An interesting finding on the impacts of social religious activities was reported by McDougle et al. (2016) in their investigation of the effects of various coping strategies, including social (e.g., church attendance) and individual religious activities (e.g., prayer), on mortality risk. The participants were asked how often they typically sought comfort through praying or church attendance when they had problems or difficulties in their family, work, or personal life. As a result, a reduction in mortality risk was observed among people who attended church more frequently to cope with stress, whereas an increase in mortality risk was found among those who used prayer more frequently (McDougle et al. 2016). In other words, social approaches to religious coping appear to be more protective than individual approaches. This finding seems to support the protective role of social approaches to religion that may be a way for individuals to not only relieve their anxiety but also allow them an opportunity to obtain the relevant affirmations needed that will enable them to build their own coping abilities (McDougle et al. 2016). In general, it confirms the impacts that social integration and support can buffer against negative health outcomes (Thoits and Hewitt 2001).

On the other hand, Parks and colleagues reported that spiritual peace and not religious attendance was associated with lower mortality in a sample of patients with congestive heart failure (Park et al. 2016). This result is in line with a previous study that showed that the association between religiousness, particularly service attendance, and reduced risk of mortality was usually found in healthy subjects but not in populations already diagnosed with a serious disease (Chida et al. 2009). Similar results were obtained by Helm et al., who specifically investigated private religious activities and concluded that they had a protective effect only among participants with good functional ability and not among those with impaired performance in activities of daily living (Helm et al. 2000). One explanation proposed for this phenomenon is that religion may be more important in resisting disease than in helping people already diagnosed with a disease and undergoing treatment (Powell et al. 2003; Chida et al. 2009). However, spirituality was confirmed to be related to a lower mortality risk, even after considering many other variables (Park et al. 2016). The authors defined spirituality in terms of a sense of inner peace and harmony and focused specifically on the spiritual component of deep peacefulness (Park et al. 2016), which has been shown to be critically important to individuals with serious and life-limiting illnesses (Ironson et al. 2002; Steinhauser et al. 2006; Canada et al. 2008; Whitford and Olver 2012). This finding is consistent with those of other studies that linked spirituality and mortality in patients with serious illness, in which this inner experience of a sense of peace may matter most in terms of survival and exert the strongest protective effects on mortality risk (Ironson et al. 2002; Whitford and Olver 2012).

Limitations

This systematic review has several limitations. First, the number of studies that assessed the association between spirituality and mortality among older adults was small compared with the number of those that focused on religious attendance. Further studies are needed to better understand the possible link between spirituality and health status. Second, most of the included studies were conducted in the United States, where the Christian religion is predominantly practiced. Studies from other parts of the world and on other religious beliefs would broaden the generalizability of the results of the present study.

Conclusion

In conclusion, most studies conducted among older adults supported the protective role of religiosity or spirituality on longevity, particularly for religious activities with an active social component. The linkage between religiosity and longevity might be mediated by the beneficial effects of social support and better health behavior and mental health that often characterize religious people. However, because most studies found a reduction in all-cause mortality even after adjusting for the abovementioned confounding variables, other important mechanisms not yet understood might be involved.