Participation in Church or Religious Groups and its Association with Health: A National Study of Young Canadians

Sample. The 2009–2010 HBSC was the sixth cycle of this health study conducted in Canada. HBSC in turn is an international study conducted in collaboration with the World Health Organization that aims to increase understanding of health and its determinants in populations of young people in some 43 countries (Currie et al. 2008). It involves written health surveys conducted with students in classroom settings, with a focus on the early adolescent years. Surveys in participating countries are conducted every 4 years following a common protocol. The 2009–2010 Canadian HBSC was conducted from November 2009 to May 2010 and involved participants in eight provinces and the three territories. The national sample was stratified by province/territory, type of school board (public vs. separate), urban–rural geographic status, school population size, and language of instruction (French vs. English). Standardized population weights were generated to account for the oversampling in some jurisdictions and the stratification criteria. Children from private schools, home school situations, native reserves, street youth, incarcerated youth, and youth not providing informed consent were excluded.

Study variables of specific interest were items describing the following: (1) participation or not in the three key groups of interest; (2) engagement in risk-taking behaviors; (3) outward-looking “pro-social” behaviors; (4) internal feelings, both positive and negative; and (5) sociodemographic variables that were considered as key covariates in our models. These variables have been used in multiple surveys and publications that span the past 20 years, and information on their origins and validity/reliability is documented extensively elsewhere (e.g., Currie et al. 2010). We suggest that it is the composite relationship between the first three types of variables that is related to the theoretical concept of shalom and in turn to a holistic understanding of the health of children.

Analysis. We described the study population demographically. Next, we subdivided the national population based upon self-reported membership (yes or no) in a church or religious group, then again for self-reported involvement in a sports club or team (yes or no), and then in combinations defined by group membership (none, church or religious group only, sports club or team only, both types of groups in combination). Groups were compared with respect to engagement in the risk-taking behaviors, the outward-looking “prosocial” behaviors, and the indicators (positive and negative) of internal feelings. Statistical tests for between-group differences in proportion were based upon two-sided p values generated from bivariate logistic regression models generated using the SAS Procedure PROC GLIMMIX (SAS Institute, Cary, NC, 2010), applied in order to test for differences while accounting for the clustered sampling design.

We then conducted a series of multi-level logistic regression analyses using the same procedures to study possible associations between group membership and each item on our list of potential health outcomes. We assumed fixed effects but random intercepts for schools. Associations were initially examined in bivariate models, then multiple logistic models that adjusted for sex, grade level, urban–rural geography, socioeconomic status, family structure, and the family connectedness measures. These covariates had been selected as a standard set of control variables following exploratory analyses. Adjusted odds ratios and associated 95 % confidence intervals were again estimated with inflation of standard errors to account for clustering. Interpretively, we looked for general and consistent patterns in the potential effects as opposed to highlighting specific relationships on their own.

Demographic characteristics of the study population are described in Table 1. Of the 24,244 (weighted n) participants that responded to the church or religious group involvement item, there was a roughly equal split between boys and girls, most were in grades 6–8 (with a roughly equal division in the five grade levels), the majority were from large urban centers and from Central Canada, more reported average or better material wealth than not, and the majority were from families with both a mother and father (intact family structures).

Table 1 also summarizes group participation for the two major group activities under study. Overall, 16.4 % of young people reported involvement in a church or religious group, and 54.6 % reported involvement in a sports club or team. For church or religious groups, higher prevalence values were reported for: girls, the lower grades, those from rural or dissemination areas, Western Canada, and those perceiving themselves to be not very well off. For sports, higher reported engagement levels were observed for: boys, the lower grades, Eastern Canada, those perceiving themselves to be more economically advantaged, and those from intact families.

In Table 2, findings from four example covariate models are presented to summarize relationships between the six key demographic covariates used in our models, and reports of four different health indicators: two negative items indicative of risk-taking and therefore the absence of shalom (frequent physical fights, loneliness) and two positive indicators indicative of outward-looking behaviors and therefore the presence of shalom (do favors for people, help people). These covariates were always included in subsequent modeling of relationships between group involvement and health and are provided for illustrative purposes only. Higher odds of fighting were reported by the following groups of young people: boys, the lower grades, and those from remote areas, non-intact families, families with lower reported material wealth, and families that did not eat meals together as often. The same general pattern was observed for loneliness, except that girls reported loneliness more often than boys. For the positive indicators, higher relative odds were reported for both indicators among girls, those from intact families, those reporting families with higher material health, and those from families that ate meals together more frequently.

In Table 3, models describing relationships between church or religious group involvement and risk-taking indicative of the absence of shalom are summarized. Church or religious group involvement was associated with lower relative odds of: smoking (3 indicators), alcohol misuse (3 indicators), drug use (3 indicators), ever having sexual intercourse, and four of five physical health measures, all indicative of protective effects. Sports group or team involvement was associated with lower relative odds of: smoking (3 indicators), drug use (3 indicators), and five of the five physical health measures. Sports team involvement was associated with higher relative odds of drunkenness and binge drinking, ever having sexual intercourse, and physical fighting.

Table 4 (top of table) summarizes the results of the modeling of relationships between group membership and factors indicative of the presence of shalom in terms of participation in prosocial behaviors that involve looking outward toward others. Participation in church or religious groups was related to higher relative odds of five of the five prosocial behaviors. The same general pattern was observed for participation in sports groups or teams, suggesting that both group involvements were associated with positive outward-looking behavioral tendencies. Findings for the internal feelings indicators (bottom of table) were different. For church or religious group participation, while increases were observed for feelings of confidence and reporting “the best possible life,” there was no statistically significant (p < 0.05) relationship between group involvement and the outcomes of loneliness, wishing they were someone else, helplessness, and feeling low or depressed. Increases in the odds of reporting regret (sorry for the things I have done) were observed in those reporting religious affiliations. For sports, young people reporting group involvement reported higher relative odds for the two positive indicators of internal feelings and lower relative odds (all p < 0.05) for each of the negative indicators, showing a clear protective effect.

The final analysis (Table 5) summarizes the prevalence of each of the health indicators among young people who reported combinations of group involvement (neither group, sports only, church or religious group only, and both groups). Findings for the “sports only” and “church or religious only” groups are fairly consistent with those patterns reported earlier in Tables 3 and 4, with strong evidence that church or religious group involvement is protective for some common risk behaviors (smoking, drinking, cannabis use, sexual intercourse) and the outward-looking prosocial behaviors (e.g., helping and sharing), but a risk factor for some of the physical health measures (e.g., toothbrushing, physical inactivity). Church or religious group vs. sports involvement was also related to deficits in internal feelings (e.g., confidence in oneself and feelings about the best possible life) but also each of the negative indicators of internal feelings (e.g., loneliness, wishing they were someone else).

However, in addition, analysis of Table 5 also looks at the potential influence on health of adding sports involvement to church or religious group participation. Findings suggested that the relative odds of reporting smoking or drug use was not associated with the combined group versus church or religious group participation only, but the combined group reported increases in the odds of drinking, sexual behaviors, and fighting. Relative to the “only church or religious” group, young people reporting involvement in both groups also reported better physical health, more outward-looking behaviors, and increased positive feelings of confidence in oneself and having the “best possible life” as well as decreased odds of loneliness, wishing they were someone else, regret, and depression. Sports involvement, while not protective for some overt risk-taking among church-involved children, also had many beneficial effects in terms of physical health indicators, outward-looking prosocial behaviors, and several negative indicators of the absence of shalom.

Our observation of positive relationships between religious group involvement and some aspects of health is consistent with those documented historically (Strawbridge et al. 2001; Holder et al. 2010; Pearce et al. 2003; Ovwigho and Cole 2010; George et al. 2000). Findings from the most established study on church involvement and its impact on Canadian children, albeit 15–19-year-olds, mirror our own in the area of religious involvement and lower participation in sexual behavior and drug use (Bibby 2009). Similarly, findings from a large US-based study, the National Study of Youth and Religion, suggest that religious teenagers report acting differently than non-religious teens, particularly in areas such as lower involvement in risk behaviors and quality of relations (Smith and Denton 2005). Hence, our findings support the existence of some general social patterns that seem to be consistent between countries and cultures. In contrast, our study findings suggesting negative or nonexistent associations between religious group attendance and some more holistic internal feelings are somewhat novel. Few studies have tested for the existence of such associations in large populations of children. Theologically, such holistic measures of health and their relations with group membership are important as they take into account the way that children experience shalom within themselves in their lives.

We believe that our study and our collaboration have several strengths. Our sample of children was contemporary, representative of Canadian children, and sufficiently large to permit meaningful subgroup analyses. It permitted a variety of analyses of health behaviors conducted across grade levels, by sex, and by group involvement. Second, to our best understanding our particular focus on church/religious involvement and this more composite, holistic approach toward health is novel in young adolescents. If these associations are indeed valid, there is potential to intervene with policies, educational practice, and organizational structures. In short, findings from this initial study provide basic guidance for theological discussions about the role of church and religious groups in nurturing wholeness in children.

Our study also is subject to limitations. First, selection bias is a possibility. It is conceivable, for example, that children with certain social characteristics would be more likely to attend church or other religious events. While efforts were made to control for such social factors in the regression analysis, findings may be residually confounded and the observed patterns may in part be attributable to some sort of social selection in who attends religious functions, rather than what the church or other religious group is doing per se. Second, our analysis was also limited by the absence of a direct measure of spirituality. A third potential limitation surrounds the children who are not included in the HBSC sample. While our sampling frame includes Canadian children who attend the publicly funded Public School and Separate School Board Schools, approximately 7 % of Canadian children attend private religious-based schools, academic enrichment schools, alternative private schools, are homeschooled, or do not attend school. While this does mean that children in some religious-based schooling settings are excluded from the survey results, it is unlikely that within our Canadian context, removing this 7 % from the sample group substantially affects the representativeness of our results. This is further evidenced by a subanalysis (data not shown) that we performed solely with the young people from the Separate School Board systems. The latter likely included youth with mainly Catholic, Protestant Christian and other religious youth affiliations. This subanalysis yielded very similar findings to the larger analysis that was presented. Fourth, while our data sample included responses from a broad spectrum of children involved in a church or religious group, from very conservative Christians, to very liberal Christians, to Roman Catholics to religious groups other than Christianity, there were no controls for denominational or theological differences in this study and we were unable to examine trends within these more specific groups. Such analyses would be of obvious importance for future study. Fifth and finally, it is always difficult to attach deep and consistent meaning to key phrases and constructs used in a written survey. We do not know, for example, what words like “go to church” or “have feelings of regret” actually mean to the individual children who answered the survey questions, and whether there is consistency in meaning. Additional insight into such factors would assist interpretively. Analogously, we did not have any measures of the quality of the experience (religious or sports-oriented) of these children, which may be very important in regard to their impact on health.

Self-reported measures obtained by written survey are always subject to some measurement errors. For example, self-reports of substance abuse may be more influenced by cultural norms and group identity than by actual participation in these behaviors. With respect to our own Christian traditions, it was challenging to identify children within HBSC who are actively involved in a church community due to the lack of specificity in our questionnaire item. The HBSC item is not limited to Christianity, and there may be variations in how children perceive their involvement in existing groups and organizations across different faiths. Finally, as this is a cross-sectional study that provides information on children for a “snapshot” of time, it is difficult to establish the temporal sequence between group membership and the health indicators under study. Some caution therefore must be taken in interpreting all associations as causal, and our findings require confirmation in longitudinal analyses. That being said, we do view our analyses as an important starting point in understanding potential relations between church involvement and a whole-person approach to child health.

Findings from this study raise important questions and concerns with respect to ministry to children, both within and outside of the Christian tradition. These pertain to the role of ministry in the health of children and especially the integrated experience of shalom in their lives. While involvement in a church or religious group offers some protection with respect to engagement in risk behaviors and improved, outward-looking prosocial behaviors, one is left to wonder why this trend does not extend into these other areas of life that are vital to a more composite experience of holistic health in children, including a protective trend around internal feelings. Indeed, involvement in sports clubs appeared to have more potential impact on some important positive aspects of holistic health than church involvement. This finding is sobering. In our own religious traditions, given that a central claim of the Christian faith is a promise of wholeness or shalom, the absence of positive findings for these emotional indicators that are an integral part of holistic health is cause for concern. Regardless of the positive story that church or religious group involvement offers a protective trend around overt risk-taking behaviors and outward-looking prosocial behaviors, the notable absence of a continued protective trend in areas of physical and emotional health is problematic. It is important to note that the church operates within the confines of a broken world, and in its fullest sense, this promise of shalom is an eschatological hope rather than a present reality. While it is futile to idealize the church in the present as an absolute actualization of all of God’s promises, it nevertheless has the potential to be an agent of shalom, offering a foretaste in the present of God’s ultimate recreation of all things.

Our comparisons of findings for the church or religious group with those from the group involved with a sports club or team were also valuable. Involvement in sports appeared to have many positive relations with overall health, particularly around lower-risk behaviors in the area of physical health and higher-outward-looking prosocial behaviors. Our most striking observation was the high positive correlation between involvement in sports and increased positive perceptions of internal feelings. This could be connected to the children feeling a part of a group with a common purpose. It could also be connected with greater physical activity, a greater sense of accomplishment within a team or social context, and a higher sense of group identity. However, there was very little correlation between sports group involvement and reduced engagement in serious overt risk-taking behaviors, and the drinking, sexual, and substance use behaviors were even higher in the sports-involved group. Thus, while sports involvement clearly nurtures some important components of shalom, it is only when combined with the protective effect of church involvement that it offers protection in terms of overt risk-taking behaviors. While sports involvement on its own offers a striking protective trend in many areas that are in keeping with shalom, it is when it is combined with church involvement, as seen in Table 5, that the full spectrum of health behaviors indicative of shalom are observed.

These findings then beg the following question: Irrespective of the findings around overt risk-taking, what is it about sport club involvement that appears to influence measures about internal feelings that are essential to our composite understanding of health that is lacking in church/religious involvement? This may be connected with the reality that fewer children attend church than are involved with sports, and with smaller groups and less time of contact, there is more limited potential for comparable peer interaction. It is also possible that in the church, while attention has been given to providing a safe and nurturing environment, insufficient attention has been given to challenging children with tasks and issues that offer a sense of teamwork, achievement, and accomplishment.

Our finding that the protective effects of church/religious group participation did not extend into the particular internal feelings (measures of loneliness, wishing they were someone else, feelings of helplessness, and feelings of depression) is striking and problematic. For example, one theological assumption of this study is that the Christian faith is inherently one of community and that central to the notion of shalom is reconciliation not only with God but with one another. Our core finding, that children in church are just as lonely as those outside the church, reveals a disparity between the promise of the Christian life and its reality, at least for young people, and moreover suggests that the church may not be as effective as it should be in integrating children into the full community of church life. While there is no way of understanding the quality of the church or religious experience of participants reporting involvement from the measures we have available, it is likely that the youth who self-reported their involvement represent those with a meaningful connection to the group rather than a nominal attendance. Our findings were in keeping with those from other studies (Bibby 2009; Bibby and Penner 2010), which report similar (or slightly higher) numbers of youth who are actively involved in church, with meaningful engagement. This suggests that beyond not influencing feelings of loneliness in those who have nominal attendance, church involvement is failing to provide a meaningful intervention in terms of feelings of loneliness in some of its potentially most involved young people. This could be connected to a potential lack of peer connection for these children within the church context; however, peer connection is only one factor in terms of mediating feelings of loneliness, because the church generally has no shortage of caring, older adults.

These findings issue a challenge to those involved in church leadership to take seriously the church’s mission of being an intergenerational community of God’s people seriously, and to be more intentional (while being conscious of appropriate and safe boundaries) about integrating young people into the life of the church family. Previous research has shown that this approach to church ministry has radically transformed church experiences of young people and that the church indeed has the potential for this sort of meaningful intervention in the lives of its young people (Devries 2004).

The shalom central to the Christian faith also takes seriously matters of one’s identity and unique belovedness, and one’s gifts with which to engage the world. That many church-connected children reported an increased sense of “wishing they were someone else,” “feel depressed or low,” and feelings of helplessness again affirms that children in the Christian community are often not experiencing the fullness of the promise of shalom. A meaningful connection with a church should equip and empower children to experience improvements in these areas of emotional life.

The measure regret was a bit of an anomaly in terms of its relations as well as their meaning and interpretation. While church-connected children did not report higher prevalence values with respect to the other measures around internal feelings, they did report more heightened feelings of regret. This raises questions around how these children understand regret, how it reflects their theological worldview, and what it looks like in their lives. For example, it is important to understand whether regret is understood to be positive or negative: It is unclear whether feelings of regret can be interpreted as having feelings of guilt, having ownership and accountability for one’s actions, or something different altogether.

Areas of life such as physical health, emotional health, and communal life are fundamental to the concept of shalom. Thus, our findings, from our Christian perspective, suggest that the essential mission of the church may have become distorted by disintegrative theologies that offer a moral code to be followed rather than an invitation into the new and holistic way of being in the world offered to us by Jesus. One possible reason for our findings may be a shallow understanding of human physicality and the body that emphasizes the spiritual over the physical and reflects a narrow or distorted understanding of the incarnation. This dualism becomes more significant since in recent years there has a been a renewed emphasis on the importance of the relationship between human beings and the world in which we live. This in turn naturally raises interest in a more holistic approach both to the human body and to overall health. Any suggestions regarding the significance of our findings and their interpretation must take into account a complex and deeper issue for theology. It would be misleading to suggest that the Christian life carries with it the straightforward promise of a better life with a corresponding sense of overall well-being. While the promise of shalom remains a central affirmation of the Christian faith, there are a number of qualifiers as to how and in what way that shalom is to be realized.

Inherent within the assumption that the incarnation informs a Christian understanding of shalom is the understanding that the incarnation resulted in conflict with societal norms and values, conflict which eventually led to the crucifixion and death of Jesus. Jesus himself asserted that to follow him was to take up one’s cross daily. In other words, there are ways in which the Christian life inevitably results in feelings of loneliness, helplessness, and regret. This is perhaps more evident for children who will experience conflict and uncertainty as they work to integrate their relationships in society at large with their relationships in the church. While acknowledging this complexity and seeking to allow for it in the assumptions of this study, we would continue to affirm that the promise of shalom is a real promise even though it may not be the promise of a “better life” in the way it is understood culturally.

This national study investigated the involvement of younger adolescent Canadians in church or religious groups, and relations between this involvement and various aspects of their health. It was based theoretically on the idea that regardless of whether or not the terminology of shalom is explicitly taught within a church context, church involvement should lead to fuller and more holistic health, as encompassed in the concept of shalom.