Abstract
Background:
The relationships between religiosity/spirituality (RS) and self-harm behavior, including non-suicidal self-harm behavior (NS-SHB) and suicide attempts/completions, remain of keen interest. Whereas the majority of studies strongly suggest that RS protects against suicide attempts/completions, relationships between RS and NS-SHB have been rarely studied.
Aim:
In this study, we examined RS in relationship to both NS-SHB (six explicit behaviors) and past history of suicide attempts.
Method:
In a cross-sectional sample of 306 consecutive primary care outpatients, we administered four self-report assessments for RS (extent participant considered self a religious person, extent participant considered self a spiritual person, extent religion is involved in understanding/dealing with stressful situations, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale (FACIT-Sp-12)) as well as examined seven items on the Self-Harm Inventory: six items reflecting NS-SHB and one item reflecting a past suicide attempt.
Results:
While two RS items yielded no significant findings (extent participant considered self a spiritual person, extent religion is involved in understanding/dealing with stressful situations), the remaining two items were associated with a lowered risk of self-harm behavior, particularly the FACIT-Sp-12.
Conclusions:
Some but not all aspects of RS are associated with lowered risk for self-harm. In this study, considering oneself a religious person and reporting a general sense of RS well-being offered the most protective effect to participants, particularly the latter.
Relationships between religiosity/spirituality (RS) and both non-suicidal self-harm behaviors (NS-SHB) and suicide attempts are meaningful not only from a psychiatric perspective but also from a humanistic perspective. Whereas there have been numerous studies on RS and suicide attempts/completions, relationships between RS and NS-SHB have been studied infrequently.
RS and NS-SHB
With regard to RS and NS-SHB, Young, Riordan, and Stark (2011) examined a cohort of over 2,000 children and adolescents and reported that, in comparison with Protestants, those with other religious affiliations or no religious affiliation, atheists and agnostics, the subsample consisting of Roman Catholics was less likely to engage in NS-SHB. Likewise, Rosmarin et al. (2013) examined 159 psychiatric outpatients in a day treatment program and found that greater religious belief was associated with greater reductions in NS-SHB during treatment. While the evidence is meager, these studies suggest that RS may in some forms protect against NS-SHB.
RS and suicide attempts/completions
With regard to RS and suicide attempts/completions, there are numerous studies in the extant literature. To summarize these studies, a number of recent literature reviews support the contention that RS provides a protective effect with regard to suicide attempts/completions. For example, Mueller, Plevak, and Rummans (2001) concluded that the majority of studies in this area support the proposition that RS is generally associated with better health outcomes in a number of areas (e.g. overall longevity, coping skills, health-related quality of life) as well as a reduced risk of completed suicide. Koenig (2001) examined 68 studies on the relationship between religious involvement and suicidal attitudes/suicide rates and found that 84% of studies demonstrated that higher levels of religious involvement were associated with more negative attitudes toward suicide as well as lower suicide rates. Colucci and Martin (2008) examined the literature on RS and suicide attempts/completions and concluded that although in general religious influences are associated with lower suicidal ideation as well as more negative attitudes toward the act of suicide, in a minority of studies RS was associated with a possible increased risk.
Zagozdzon (2012) reviewed the literature on religiosity and morbidity and mortality and found that religiosity demonstrated a favorable effect on survival, including the risk of suicide. Likewise, in a review of the literature between 1990 and 2010, Bonelli and Koenig (2013) reported that higher levels of RS were generally predictive of less psychological dysfunction, including completion of suicide. Specifically, these authors found that among the reviewed publications, 43 (72%) indicated a favorable influence of RS on mental disorders, including suicide; 19% reported mixed influences and 5% reported negative influences.
Virtually all the preceding studies have been retrospective in methodology. However, Kleiman and Liu (2014) prospectively examined the prediction of suicide in a nationally representative sample in the United States of more than 20,000 participants. Based on data collected at the outset regarding the frequency of religious-service attendance, subsequent mortality data indicated that frequent religious-service attendance demonstrated a long-term protective effect against suicide – even after accounting for other relevant risk factors.
Although the majority of the extant literature suggests a protective influence of RS with regard to suicide attempts/completions, we located two studies whose findings opposed this general trend. In the first, Rezaeian (2010) found a higher rate of completed suicide among young Middle Eastern Muslim females – a finding that was attributed to the method chosen for suicide, the presence of comorbid psychiatric disorders, marital status and the role of gender socialization in that culture. In a second study, Wang, Lightsey, Tran, and Bonaparte (2013) found that among black women college students, ‘religious awareness did not predict unique variance in suicidal thoughts and behaviors’ (p. 9). Despite these two studies, in summarizing this overall body of literature, Gearing and Lizardi (2009) concluded that ‘Research has established that degree of religiosity is directly related to degree of suicidality, with greater religiosity predicting decreased risk of suicidal behavior’ (p. 338).
As a caveat to the preceding findings, it is important to note at this juncture that studies on the effects of RS on suicide attempts/completions are fraught with a number of potential confounds. At the outset, explicit assessment of RS – a complex and multi-faceted concept – is problematic. For example, RS assessments oftentime focus on a very limited number of study variables (e.g. church attendance, church affiliation/membership, participation in religious activities, specific religious beliefs such as life after death) and/or examine spirituality only in its religious form (e.g. religiosity scales). To underscore this concern, Colucci and Martin (2008) emphasized that RS is a ‘complex and multifaceted construct which requires measures that reflect its complexity and multidimensionality’ (p. 239). Studies may also fail to control for mediating and/or moderating variables, such as comorbid psychiatric illness or maltreatment by representatives of the church. In addition, many studies are based on samples from the United States and do not necessarily reflect a world experience. Likewise, while RS may function as a powerful source of support, it may also become entangled with neurotic and psychotic disorders and function as a liability in some cases (Koenig, 2009). Finally, examination of suicide attempts is oftentimes undertaken through a self-report methodology, and some participants may frankly be too embarrassed or ashamed to admit such behavior, particularly in an RS context (i.e. a number of religions actively proscribe suicide).
When RS functions as a positive resource and as a potential deterrent to suicide attempts/completions, it may do so in a number of different ways. For example, Koenig (2001) postulated that RS facilitates a positive world view, thereby providing a broader meaning to one’s life experience. In addition, Koenig stated that religious/spiritual beliefs and practices (e.g. meditation, prayer, communal worship) tend to evoke positive and supportive emotions from participants. Likewise, some religious practices specifically demarcate life transitions (e.g. baptism, bar/bat mitzvah, religious confirmation), easing and facilitating life passages within a community setting. RS also provides a distinctive level of life structure and guidance through shared subcultural values. Moreover, in a very broad sense, the concept of a ‘higher power’ may facilitate a sense of security and cosmic order among practitioners. A religious/spiritual community consisting of clergy and congregation members may also offer a very tangible and accessible support system. In the case of the Catholicism, Colucci and Martin (2008) proffered that there exists a father substitute who provides distinct therapeutic functions such as confession, eradication of guilt and conscious self-analysis. Finally, Gearing and Lizardi (2009) pointed out that RS is associated with lower levels of anger and hostility as well as the proscription of substance abuse – factors that may mitigate suicide attempts/completions. These are but a few examples of the potential mental health benefits of RS.
In accordance with previous research, we hypothesized that RS would provide a protective effect against both NS-SHB and suicide attempts. To explore this issue, we examined four variables of RS in relationship to lifetime engagement in any of six types of NS-SHB as well as a history of past suicide attempts.
Method
Participants
Participants in this study were males and females, ages 18 years or older, who were being seen at an internal medicine outpatient clinic for non-emergent medical care. The outpatient clinic is staffed by both faculty and residents in the Department of Internal Medicine in a community hospital and is located in a mid-sized mid-western city in the United States. The majority of patients who were recruited for this study were seen by resident providers. The recruiter excluded individuals with compromising medical (e.g. pain), intellectual (e.g. mental retardation), cognitive (e.g. dementia) or psychiatric symptoms (e.g. psychotic) of a severity to preclude the candidate’s ability to successfully complete a survey.
At the outset, 381 individuals were approached and 345 agreed to participate, for a participation rate of 91%. Of these, 306 completed the relevant study measures, 228 (74.5%) of whom were female and 78 (25.5%) of whom were male. Participants ranged in age from 18 to 92 years (M = 43.11 years, standard deviation (SD) = 12.45 years). Most participants were White/Caucasian (85.9%), followed by African-American (7.2%); 6.9% indicated some other ethnicity/race. With regard to educational attainment, all but 4.6% had at least graduated high school, whereas 31.7% had earned a 4-year college degree or higher.
Procedure
During clinic hours, the recruiter positioned in the lobby of the internal medicine outpatient clinic, approached consecutive incoming patients after registration, and informally assessed exclusion criteria. Exclusions were informal as surveys needed to be completed prior to participants’ appointments with primary care providers (time constraints). With potential candidates, the recruiter reviewed the focus of the project (i.e. a study exploring religious beliefs and one’s personal history) and then invited each to participate. Each participant was asked to complete a six-page survey, which took about 10 minutes. Participants were asked to complete surveys onsite in the lobby, and then to place completed surveys into sealed envelopes and into a collection box that was located in the lobby.
Religion/spirituality assessment
In addition to a demographic query, the survey contained assessments for RS as well as NS-SHB and suicide attempts. RS was measured through four assessments. The first three assessments consisted of singular queries that were each scored from 1 to 4: (a) to what extent do you consider yourself a religious person? (response options: very religious, moderately religious, slightly religious, not religious at all), (b) to what extent do you consider yourself a spiritual person? (response options: very spiritual, moderately spiritual, slightly spiritual, not spiritual at all) and (c) to what extent is your religion involved in understanding or dealing with stressful situations in any way? (response options: very involved, somewhat involved, not very involved, not involved at all). As a caveat to this approach to measuring RS, religiosity and spirituality are patently elusive terms, and although there are many proposed definitions, these phenomena represent experiences that words generally cannot describe (Tokayer, 2002). Therefore, in approaching these phenomena in this study, we did not provide any definitions or constructs for these terms, but rather left participants to define them.
The fourth assessment for RS was the Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale (FACIT-Sp-12) (Peterman, Fitchett, Brady, Hernandez, & Cella, 2002), a 12-item scale that assesses a sense of meaning in life, harmony, peacefulness and a sense of strength and comfort from one’s faith. The FACIT-Sp-12 has a 5-point Likert-style response scale (0 = not at all, 1 = a little bit, 2 = somewhat, 3 = quite a bit, 4 = very much) and culminates in a total summed score (after reverse-scoring two items). The FACIT-Sp-12 was developed from a larger measure, the Functional Assessment of Chronic Illness TherapySp (FACIT-Sp), which was originally designed to assess various facets of well-being in individuals with chronic illness (Peterman et al., 2002). We elected this particular scale for use in this study because of its prior empirical exposure, availability without cost, self-report format and brief nature (12 items). The FACIT-Sp-12 correlates moderately with a number of related religion/spirituality scales, with the Faith subscale demonstrating moderate-to-high correlations. Individuals who are Jewish or who report no religious affiliation have the lowest scores on the FACIT-Sp-12 (Peterman et al., 2002).
In using the FACIT-Sp-12, we assessed current (‘… over the past year’) religious/spiritual well-being. In addition, to accommodate the psychological focus of the study (i.e. the examination of religious/spiritual well-being in terms of emotional health), we replaced the word ‘illness’ in items 11 and 12 of the FACIT-Sp-12 with ‘emotional difficulties’. At the time we made this decision, we were unaware that there was an available modification of this scale in which the developers replaced in items 11 and 12 the word illness with ‘difficult times’ – a very close approximation of the modifications that we made for this study. Permission was obtained to use both the FACIT-Sp-12 scale and the modifications of the scale that we describe above (personal communication). Cronbach’s alpha was .91 in this study.
Assessment of NS-SHB and suicide attempts
The last section of the survey explored past histories of self-harm behavior through the Self-Harm Inventory (SHI; Sansone, Wiederman, & Sansone, 1998). The SHI is a 22-item, yes/no, self-report inventory that explores participants’ histories of self-harm behavior. Each item in the inventory is preceded by the stem, ‘Have you ever intentionally, or on purpose …’. Individual items include ‘overdosed, cut yourself on purpose, burned yourself on purpose’ and ‘hit yourself’. Each endorsement is in the pathological direction. The SHI total score is the summation of ‘yes’ responses. For this study, we elected to focus on six graphic behaviors (i.e. overdosed, cut self on purpose, burned self on purpose, hit self, banged head on purpose and scratched self on purpose) and attempted suicide.
Results
For comparison, we categorized respondents into three groups: (a) those who denied any of the self-harm behaviors (n = 216), (b) those who reported at least one of the six NS-SHBs but denied having attempted suicide (i.e. the NS-SHB subsample; n = 35) and (c) those who reported ever having attempted suicide (n = 55). Among those 35 respondents in the NS-SHB group, the number of such behaviors reported ranged from 1 to 6 (M = 2.40, SD = 1.52).
Responses to the three one-item measures of RS spanned the range of response options and demonstrated a fair degree of variance: To what extent do you consider yourself a religious person? (M = 2.28, SD = 0.95), to what extent do you consider yourself a spiritual person? (M = 1.97, SD = 0.90) and to what extent is your religion involved in understanding or dealing with stressful situations in any way? (M = 2.03, SD = 1.00). Scores on the FACIT-Sp-12 ranged from 0 to 44 (M = 29.37, SD = 9.46).
One-way analyses of variance (ANOVAs) were performed to examine RS as a function of group membership (see Table 1). We took a statistically conservative approach to individual group comparisons by conducting Bonferroni post hoc tests. Note that two RS assessments, the extent that one considers oneself a religious person and the FACIT-Sp-12 summed score, evidenced protective effects against self-harm, with the extent of being a religious person being protective against suicide and the FACIT-Sp-12 summed score being protective against both NS-SHB and suicide attempts.
RS scores as a function of ever having engaged in NS-SHB or suicide attempts (N = 308).
RS: religiosity/spirituality; NS-SHB: non-suicidal self-harm behavior; SD: standard deviation; FACIT-Sp-12: Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale (Peterman, Fitchett, Brady, Hernandez, & Cella, 2002).
For comparison groups, controls = those with no reported self-harm behavior, NS-SHB = those with at least one reported self-harm behavior but not attempted suicide, suicide attempts = those reporting attempted suicide. With the first three assessments, lower scores indicate higher levels of RS; with the FACIT-Sp-12, higher scores indicate higher levels of RS. Within each row, means with differing superscripts are statistically significant different (p < .05) according to Bonferroni post hoc tests.
Discussion
There are a number of findings in this study. First, perceiving oneself as a religious person appears to be associated with a lesser likelihood of attempting suicide, which is reflective of our literature review – but not with a lesser likelihood of NS-SHB. To our knowledge, this is the first study to examine within a single study population relationships between religiosity and both NS-SHB and suicide attempts. It may be that religious proscription of suicide is patently clear, whereas the proscription of NS-SHB is somewhat ambiguous or implied, leaving open the door of possibility.
As for a second finding, being a spiritual person does not appear to confer any protection against either NS-SHB or suicide attempts. This finding may relate to the multitudinous variations in spirituality, the ambiguity of the term, the general absence of spiritual dogma and therefore proscriptions against such behavior and/or the oftentimes highly personalized nature of spirituality, which is customized to one’s beliefs. The observed differences in this study between RS and their relationships to NS-SHB and suicide attempts also reinforce the concept that the religion and spirituality are likely different – a difference emphasized by King and Koenig (2009).
As for a third finding, using religion to understand and deal with stressful life situations does not appear to confer any protection against either NS-SHB or suicide attempts. This finding is a bit perplexing as the use of religion for enhanced coping seems to intuitively predict for less stress-generated self-harm behavior, yet apparently this is either inaccurate or not fully effective under extreme stress.
As for the last finding, RS well-being as measured by the FACIT-Sp-12 demonstrated protective effects against both NS-SHB and suicide attempts. Based on the construct of FACIT-Sp-12 as reflected by the individual items, this finding alludes to the conclusion that overall spiritual well-being regardless of context provides statistically significant protection against both NS-SHB and suicide attempts. In fact, the FACIT-Sp-12 demonstrated the most impressive statistical differences among the groups, suggesting that RS, itself, is perhaps somewhat different than the well-being it may potentially provide.
This study has a number of potential limitations. First, all data were self-report in nature and subject to the vicissitudes of this method of data collection (e.g. recollection difficulties, denial, deceit, suppression, repression). Second, the comparison subgroups were relatively small. Third, the gender distribution in the sample was uneven, with a predominance of women (a reflection of general utilization patterns by gender in this clinic); gender may have exerted an influence on these findings. Fourth, enquiries about a multi-faceted and complex phenomenon such as religious/spiritual beliefs will always be hampered by the limitations of the associated measurement tools as well as the unique and endless number of belief systems of participants. Finally, some participants may not have admitted to the item enquiring about suicide because of shame, embarrassment, disclosure discomfort, denial or interpretation (e.g. ‘During that overdose, I wasn’t really trying to kill myself’).
Despite the potential limitations, this study offers a number of unique contributions to the literature, including the use of multiple measures for RS, assessment of multiple NS-SHB (a total of six behaviors) and suicide attempts and use of a unique study sample – primary care outpatients. With regard to NS-SHB, findings indicate that RS well-being appears to protect against these behaviors. With regard to suicide attempts, findings indicate that both religious identification and RS well-being appear to be protective. The overall clinical message seems to be that some aspects of RS protect against self-harming behavior, suicidal or otherwise, but not all. Further research is warranted to tease out the underlying facets of RS that offer protection.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.