Introduction

Suicide is a leading cause of death in young people in the 15- to 29-years age bracket worldwide [1]. A recent meta-analysis of 634,662 students by Mortier et al. [2] found that the pooled prevalence of lifetime suicidal ideation and suicide attempts were 22.3% and 3.2%, respectively. In China, a study by Yang and colleagues [3] indicated that the pooled prevalence of lifetime suicide attempts among 88,225 college students was 2.8%, with the highest rates recorded among rural students (5.1%). The risk of lifetime suicidal ideation and suicide attempts among college students was higher than adults’ prevalence rates worldwide [4].

Religion and religiosity have been proposed to be protective factors for suicide since Emile Durkheim’s study, which revealed that Protestants reported higher suicide rates than Catholics in late 20^th century Western Europe [5]. The connection between religion and emotional well-being has spurred numerous empirical studies over the past 50 years to understand further the complex relationships between religion and suicidality [6]. Religious affiliation, commitment, and practice are among the few independent factors reported to contribute substantially to lower the likelihood of individual suicidality [7, 8]. Religion has also been related to the country and regional suicide rates [9, 10], with most studies focusing on Western countries. In a cross-national study of 22 European countries, the results suggested that religion is a protective factor in the relatively secularized European nations and regions after controlling for economic level, economic strain, and education [11]. A study on 124 countries using the WHO data on suicide rates indicated that Islamic countries had significantly lower suicide rates than non-Islamic countries [12]. A meta-analysis on religion and completed suicide also showed that religion serves as a protective factor in Western countries but loses its protectiveness in Eastern countries [13].

A number of studies on religious affiliation and suicide among university students have revealed inconsistent findings. For example, a study in Australia found that a greater proportion of students affiliated to Protestantism and Catholicism reported experiencing suicidal ideation compared to the group with no religion. However, the no-religion group had a larger proportion of individuals who had attempted suicide [14]. Among Ukrainian college students, those with a religious affiliation had a lower likelihood of reporting lifetime suicide attempts compared to individuals with no religion [15]. In terms of inter-religion differences, a large-scale study on LGBQ university students found that Jewish-affiliated participants were less likely to report lifetime suicidal ideation, but non-religious participants were more likely to wish they were dead compared to Christian LGBQ participants [16]. Eskin et al.’s [17] study of university students across 12 countries showed that Muslims had a higher risk of attempting suicide while Buddhists, Catholics, and individuals with no religion had lower risks. These findings indicate that the association between religious affiliation and suicidality is an important area for further investigation, especially in non-Western settings such as China.

In China, only an estimated 15% of the population report a religious affiliation [18]. Studies examining associations between religion and suicide have gained momentum, although existing studies indicated mixed findings. For example, the age-standardized rate of lifetime suicidal ideation and lifetime suicide planning was significantly higher in the mostly Muslim Hui ethnic group than in the mainly atheist Han ethnic group [19]. Similarly, Zhang et al. [20] found that religious believers reported higher depression and suicidal ideation, suggesting that religion may be a risk factor for suicide. Lew et al. [21] and He et al. [22] found that religion may be a protective factor in their study of college students in Shandong and Northwestern China. Wang et al. [23], however, found no significant correlations between religious involvement and suicidal ideation, plan, and/or attempt in either Muslims or those not following any religion.

We have chosen to measure the type of religious affiliation as a risk/protective factor for suicidal thoughts and behaviors due to a few reasons. First, doctrinal beliefs regarding suicide vary between different religious beliefs and may influence suicidal behavior. For example, devotees of Christianity and Islam believe that suicide is a sin, and therefore a doctrinal commitment to these religions may serve as a suicide deterrent [24]. Compared to other religions, Muslims have the lowest level of permissiveness towards suicide [25]. In contrast, it has been found that Hindus and Buddhists hold more permissive views of suicide [26]. As past studies have indicated a positive relationship between attitudes toward suicide and suicidality, an individual who is affiliated with Islam may have different levels of suicidality compared to an individual who is affiliated with Hinduism or Buddhism. Secondly, the association between religious affiliation and suicidality needs to be examined against the background of the predominant culture of a locality. Eskin et al. [17] proposed that religiosity may lose its protectiveness in a culture where religious affiliation is compulsory or is normative. On the other hand, according to Hu et al. [27], the protectiveness of a religion against psychological distress depends on the social status of the religion. In their study, Christians were found to exhibit worse depression symptoms compared to Buddhists and individuals with no religion. This led them to hypothesize that self-identification with a marginalized religion may be a psychological burden [27]. Therefore, we aim to examine whether different religious affiliations among college students in China are associated with overall suicidal behavior, suicide attempt, and suicide risk.

For this study, religious affiliation is defined as the self-reported identification as belonging to a specific religious denomination (such as Buddhism/Daoism, Christianity, Islam, and other religions).

Materials and methods

Results

Suicidality by religion

Suicidality (i.e., suicidal thoughts and behaviors) which was measured using the SBQ-R total score, was the highest in the Shanghai province (5.38±2.80) and lowest in the Shandong province (4.32±2.28). Similarly, the percentage of participants reporting suicide risk was the highest in Shanghai province (28.1%) and the lowest in Shandong province (14.9%), χ² ₍₅₎ = 121.41, p<0.001 (Refer Table 1). Christians reported significantly higher suicidality score (5.71±3.30) compared to those affiliated with Buddhism/Daoism (4.87±2.68), other religions (4.72±2.61), individuals reporting not following a religion (4.67±2.44), and Islam (4.43±2.39). At the same time, Muslims had lower levels of suicidality than Buddhists/Daoists and those with no religion, F_{(4, 9601)} = 17.82, p < 0.001, η² = 0.007. Similarly, a higher percentage of Christians were at risk of suicide (30.4%; χ² ₍₄₎ = 39.34, p < .001) and had attempted suicide (7.5%; χ² ₍₄₎ = 48.95, p<0.001) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Suicidality measures by province and religious denomination.

https://doi.org/10.1371/journal.pone.0251698.t001

Religion and risk/protective factors for suicidality

Results of the one-way ANOVAs indicated that risk factors for suicide such as the level of hopelessness were significantly higher among all religions compared to those with no religion, except for Islam (F _{(4, 9446)} = 28.45, p<0.001, η² = 0.011). Psychache (F _{(4, 9460)} = 29.05, p<0.001, η² = 0.012) and depression (F _{(4, 9517)} = 29.34, p<0.001, η² = 0.012) were significantly higher for Christians, Buddhists/Daoists, and those with other religions. Compared to all other religions in this study, Christians and Buddhists/Daoists reported lower protective factors against suicidality, specifically in terms of self-esteem (F _{(4, 9479)} = 16.09, p<0.001, η² = 0.007), social support (F _{(4, 9482)} = 20.64, p<0.001, η² = 0.009), and purpose in life (F _{(4, 9542)} = 33.67, p<0.001, η² = 0.014) (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. One-way ANOVA analysis comparing hopelessness, psychache, depression, self-esteem, social support, and purpose in life by religious denomination.

https://doi.org/10.1371/journal.pone.0251698.t002

The results of the multiple linear regression suggested that, after adjusting for age, gender, province, hopelessness, depression, psychache, self-esteem, social support, and purpose in life, those with a Christian religious affiliation reported higher suicidality (suicidal thoughts and behavior) levels (B = 0.105, p < 0.001), whereas Muslims had lower suicidality levels (B = -0.034, p = 0.031). Together, the independent variables accounted for 19.2% of the variance, R² = 0.192, adjusted R² = 0.190, F _{(17, 9046)} = 126.43, p<0.001 (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Multiple linear regression analysis for factors predicting suicidality among college students.

https://doi.org/10.1371/journal.pone.0251698.t003

A multiple logistic regression analysis was conducted with suicide risk as the dependent. The full model was statistically significant, (Chi-square ₍₁₇₎ = 1344.68, p<0.001). This model accounted for 22.2% of the variance in suicide risk (Nagelkerke R² = 0.222). Together, the model correctly identified 81.7% of the participants with suicide risk. The results indicated that participants affiliated with a Christian denomination had 1.5 times higher odds of having an elevated suicide risk (OR: 1.51, 95% CI: 1.14, 1.99, p = 0.004) (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Multiple logistic regression analysis for factors predicting suicide risk (based on a total score of 7 and above on the Suicidal Behaviors Questionnaire-Revised) among college students.

https://doi.org/10.1371/journal.pone.0251698.t004

Another multiple logistic regression analysis was conducted with lifetime suicide attempts as the dependent. The full model was statistically significant, (Chi-square ₍₁₇₎ = 218.36, p<0.001). This model accounted for 12.6% of the variance in suicide attempts (Nagelkerke R² = 0.126). Together, the model correctly identified 97.8% of the participants with suicide attempts. The results indicated that Christian participants had 3.1-times higher odds of reporting a previous suicide attempt (OR: 3.09, 95% CI: 1.90, 5.04, p<0.001; Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Multiple logistic regression analysis for factors predicting suicide attempt among college students.

https://doi.org/10.1371/journal.pone.0251698.t005

Discussion

Our key finding suggests that religious denomination may be either a risk or a protective factor for suicidality. It should be noted, however, that suicide risk per se was low amongst this sample. Specifically, being of the Christian faith appeared to be associated with increased suicidality, suicide risk, and suicide attempt. These results are not consistent with past findings among college students in Western countries, where affiliation to Catholic and Protestant Christianity was protective against suicidality [14], and those who professed a religious affiliation were less likely to have attempted suicide [15]. Compared to other religions, being of Christian faith appeared to be associated with higher levels of hopelessness, psychache, and depression, as well as lower levels of self-esteem, social support, and purpose in life, all of which are known risk and protective factors for suicidality. The higher depression level among the Christian participants is comparable to the findings of a study by Hu et al. [27], where it was also found that Christians in China’s general population were more likely to be depressed compared to Buddhists and atheists. Other studies in China have also revealed that increased depression and hopelessness, and decreased social capital were predictive of increased suicidality [44, 45]. Taking these findings into consideration, the greater suicide risk among the Christian participants in our study could originate from their worse psychological health. This may be because adherence to the Christian faith places an individual in a minority group in China, as the Christian population in China is small and loosely spread across China [46]. For example, in this sample, most Christians were Han Chinese, which made them a minority (3.7%) compared to the majority of Han Chinese who did not profess any religious affiliation (83.0%). Also, Christians reported the lowest level of social support in this sample, indicating feelings of exclusion.

In contrast, being Muslim was found to be protective against suicidality, given the lowest scores in hopelessness. Unlike Christians, who are a minority within their nation and race, Muslims are a majority within the Hui race in this sample (85.8%). Even though Islam is not the majority religion in China, the concentration of Muslims in Northwestern China (e.g., Ningxia, Qinghai, and Gansu) is high [46].

The role of religion as a protective or risk factor for suicidality may be explained by Rodney Stark’s moral community hypothesis, which postulates that the aggregate level of religiousness in a group (e.g., neighborhood, community, city, country, state, region, and nation) will affect the attitudes and behaviours of its members, whether or not they are affiliated with a religion [47]. The moral community hypothesis found mixed support in a study on Chinese college students, where provinces that had a higher number of mosques (indicating a higher level of aggregate religiosity among Muslims) were associated with lower odds of deviant behaviour, but the relationship was not significant where Christianity and Buddhism were examined [46]. Similarly, in a community or province with higher religiousness, the likelihood of religion being a protective factor against suicidality will be higher. The finding also means a country or community with low religious rates, such as China may have lower religiosity, and this may lead to religion having a lower or no protective effect against suicidality. This hypothesis, however, remains tentative and requires further evaluation in the future.

However, given the low rate of religious affiliation in China, and also its repressive political and social policy towards religion [18], another alternative hypothesis we propose is the “religious minority hypothesis”. This hypothesis can characterize an environment where religious affiliation is low in the community, province, and nation, and the practice of religion is reported to be tightly controlled by the government. China, a communist country, has the largest population in the world, with 1.4 billion people, but it has the world’s lowest religion rate estimated to be about 15% [18]. The Chinese government has been reported to restrict the practice of religion and does not include religious elements into its public policy and decision-making processes [18]. Religious freedom is actively managed/controlled. In such an environment, it may be detrimental or even stressful to profess too strongly in a religious belief and/or to practice religion openly. This may be a root cause leading to a sense of discrimination, decreased social integration and connectivity, low belongingness to the society they interact with, disturbance in family harmony, social isolation, and clashes in value systems and aspirations. The restrictions imposed may have led to religion losing its protective effect against suicidality, and in our sample, Christians and Buddhists/Daoists even reported higher levels of suicidality and/or risk factors. Thus, the religious minority hypothesis may help to explain why certain religions are not protective against suicidality. Islamic adherents, however, are more likely to form close social support networks, offering a stronger sense of community and cohesion [47], which could serve as a protective factor, thus lowering suicide risk and negating other negative psychological factors. However, as stated in the moral community hypothesis, there is a potential for these suicidal risks to be stemming from regional differences, which is also tied to the religious demographics of the region. Therefore, this hypothesis should be subjected to further research.

The study has a number of limitations. First, with the cross-sectional design, inferences of causation cannot be drawn. Second, this study is limited to undergraduate students in urban provinces we sampled and is not generalizable to all universities in China. The religion rate reported in our study should not be assumed to represent the religion rate of any particular province or the general population in China. In addition, we did not measure the strength of the participants’ religious affiliation. A number of factors which has been shown to influence the association between religion and suicide were not included in our analysis, such as the acceptability of suicide [48]. We also did not examine Chinese values and philosophies such as Confucianism, which has been associated with suicide among young people in rural China [49]. Next, given the larger number of participants who reported not following a religion, the statistical methods employed may suffer from the problem of unequal variances.

Many previous studies of suicidality have been conducted in the US and Western European countries, as well as countries with a relatively higher rate of religious affiliation and a dominant religion. Future studies should involve other provinces in China, other age groups, and other relevant subgroups of the population to determine reasons why only Islam appears to be a protective factor, and whether this finding is due to social integration, specific beliefs, religious network, or moral community. Future studies could test our “religious minority hypothesis” in other countries with unique social and political atmospheres, and also to determine whether this pattern is generalizable throughout China. Finally, other aspects of religious affiliation which have been found to be related to suicidality, such as religious importance [50] could also be taken into account in future studies.

Supporting information