by Michael D. Anestis, M.S.
A substantial amount of research has indicated that a higher level of religiousness - a term that refers to a general tendency to adhere to a set of religious beliefs or practices shared by a group (National Institute on Aging Workgroup, 1999) - is associated with lower levels of suicide risk (Hilton, Fellingham, & Lyon, 2002). This reduced risk includes suicidal ideation, suicide attempts, and death by suicide (Burr, McCall, & Powell-Griner, 1994; King, Hampton, Berstein, & Schichor, 1996; Walker & Bishop, 2005), meaning that every meaningful component of suicidality is diminished in severity for individuals with higher levels of religiousness. This association, which has been replicated independently many times, is robust. What remains open to question, however, is why the association between these two variables exists. Some individuals propose that the relationship is mediated (explained) by certain religious beliefs that may offer protection from suicide (Dervic et al., 2004). Others, however, believe that the relationship is mediated by the social support inherent in religiousness (e.g., Durkheim, 1897/1951; Joiner, 2005; Stack & Wasserman, 1992). In an effort to test these two competing theories, Alee Robins and Amy Fiske (2009) of West Virginia University conducted a study recently published in Suicide and Life Threatening Behavior in which they were able to test this precise question.
The link between religiousness and diminished suicide risk
Robins and Fiske (2009) opened their article with a compelling summary of the literature detailing the link between higher levels of religiousness and lower levels of suicidality. Amongst the findings they mentioned were the following:
- Clinically depressed psychiatric inpatients affiliated with a religion exhibited lower levels of suicidal ideation and fewer suicide attempts than did depressed psychiatric inpatients with no such affiliation (Dervic et al., 2004).
- Within the Church of Latter Day Saints in Utah, active members exhibited lower suicide rates than did nonactive members and nonmembers (Hilton et al., 2002).
The current study
In order to test whether specific religious beliefs (e.g., suicide is a sin) or social support explained the link between religiousness and suicidality, Robins and Fiske (2009) recruited 454 undergraduates ranging in age from 18-21, 74% of whom were female. 72.7% of the students reported that they were raised in a religious tradition and, amongst those individuals, 81.5% were still practicing the same religion they practiced as children. 76.2% of individuals who endorsed a religious affiliation indicated that they were affiliated with Christianity, with 2.9% indicating an affiliation with Judaism, 0.7% Islam, and 2% other.
35% of the sample indicated that they had experienced suicidal ideation within the past year. 10% of the sample indicated that they had attempted suicide within the past year. As such, despite the use of an undergraduate sample, the authors were able to collect a substantial amount of data on individuals with elevated suicide risk.
The authors found that public religiousness (e.g., attending services) but not private religiousness (e.g., studying the Bible, meditating) was associated with lower levels of suicidal ideation and suicide attempts. In other words, individuals who partook in religious activities that involved other people were at lower risk for thoughts of suicide and actual suicide attempts, whereas engaging in private religious activities did not demonstrate any relationship to suicide related variables. Similarly, public religiousness but not private religiousness was associated with social support, such that higher levels of engagement in public religious activities predicted higher levels of general social support. These findings thus indicate that public religiousness is associated with greater social support and lower suicide risk, but what about religious beliefs?
Robins and Fiske (2009) found that religious beliefs were not significantly related to suicidal ideation or suicide attempts. In other words, the degree to which an individual maintains specific religious beliefs that some think might offer protection from suicidal behavior (e.g., suicide is a sin) does not impact the degree to which they experience suicidal ideation or attempt suicide. So, as it turns out, individuals with no religious beliefs related to suicide are no more likely to think about or attempt suicide than are individuals whose beliefs specifically discourage such behaviors, at least according to the data collected in this particular sample.
The authors were not done here though. The central aim of the paper was to test whether religious beliefs or social support mediate the relationship between religiousness and suicidality (ideation and attempts). Because religious beliefs were not related to suicidality, it was impossible for religious beliefs to mediate the relationship between religiousness and suicidality, so that hypothesis did not even need to be fully tested.
Before going any further, let me clarify what "mediate" means in the context of statistical analyses. Essentially, the word "mediate" means "explain." Baron and Kenny (1986) suggested the framework upon which most mediational analyses are based and that framework requires the following things:
- The independent variable (religiousness) must be related to the dependent variable (ideation and attempts). This establishes that there is a relationship to explain in the first place.
- The mediator (social support) must be related to both the independent (religiousness) and dependent (ideation and attempts) variables. This establishes that the variable we think explains the relationship is actually related to both sides of the relationship.
- The relationship between the independent variable (religiousness) and dependent variable (ideation and attempts) changes from significant (p < .05) to non-significant (p > .05) when the mediator (social support) is added to the equation. In other words, once you factor in the mediator, the independent and dependent variables are not related, so the mediator explained the relationship between those two variables.
So far, the authors had established the first two points, but needed to test the third. As it turns out, social support mediated the relationship between public religiousness and suicidal ideation as well as the relationship between public religiousness and suicide attempts. So, statistically speaking, the relationship between public religiousness and suicidality is better explained by social support. What do I mean by that? Essentially, public religiousness, which involves activities such as attending services, involves membership in a community and, as such, provides a natural source of social support. Additionally, social support has been shown to protect against suicidal thoughts and behaviors (remember our discussion of thwarted belongingness in Joiner's interpersonal-psychological theory of suicidal behavior). It thus follows that the relationship between religiousness and diminished suicidality is explained by the fact that public religiousness - which is the only type of religiousness that demonstrated a significant relationship with suicide outcomes - involves substantial social support.
You might find yourself wondering why a study like this is important. After all, if we have established that religiousness is associated with less suicidality, isn't that sufficient information? To be blunt, the answer is no. Just knowing that two variables are related to one another does not tell us why or how they are related. Because of this, it is easy to arrive at faulty conclusions regarding the rationale for a relationship. In this case, many people believe that religiousness is associated with less suicidal ideation and fewer suicide attempts because specific religious beliefs directly address such outcomes; however, when we test that hypothesis, the data simply do not support that this is the case. In fact, whether or not a person maintains such beliefs is unrelated to suicide outcomes. The religious beliefs hypothesis was not crazy - it makes intuitive sense - but it simply is not supported by facts. Without running mediational analyses like this, interesting but false beliefs can be perpetuated and, as a result, we can position ourselves to actually be less effective at preventing outcomes like suicide attempts.
This study was not perfect and no single study should be used to draw conclusions about a broad phenomenon, but these findings are rather interesting. They represent the first attempt to actually test the reasons why religiousness is related to suicidality and should be a good foundation for future work that examines clinical populations and diverse religious and cultural backgrounds.
*************
If you are experiencing thoughts of suicide, please dial 1-800-273-TALK for help. Alternatively, you can find help by dialing 911 or finding somebody to take you to the emergency room (if you can not get there yourself). If you would like to learn more about the nature and treatment of suicide, we recommend the following items, all of which are available through our online store of scientifically-based psychological resources:
- Why People Die by Suicide
by Thomas Joiner
- The Interpersonal Theory of Suicide: Guidance for Working With Suicidal Clients
by Thomas Joiner, Kim Van Orden, Tracy Witte, and David Rudd
- Treating Suicidal Behavior: An Effective, Time-Limited Approach
by David Rudd, Thomas Joiner, and Hasan Rajab
Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University
Subscribe in a reader
