Is Religiosity a Protective Factor Against Attempted Suicide: A Cross-Cultural Case-Control Study

Elsevier Editorial System(tm) for Social Science & Medicine Manuscript Draft

Manuscript Number: Title: Religiosity as a protective factor against attempted suicide: a case-control study within WHO SUPREMISS Article Type: Article Section/Category: Medical Sociology Keywords: WHO SUPRE-MISS; Case-control study; Attempted suicide; Religion; Religious denomination; Subjective religiosity Corresponding Author: Mrs Merike Sisask, MSc Corresponding Author's Institution: Estonian-Swedish Mental Health and Suicidology Institute; Estonian Centre of Behavioural and Health Sciences; Tallinn University First Author: Merike Sisask, MSc Order of Authors: Merike Sisask, MSc; Airi Värnik, MD PhD; Kairi Kõlves, PhD; José M Bertolote, MD; Jafar Bolhari; Neury J Botega; Alexandra Fleischmann, PhD; Lakshmi Vijayakumar; Danuta Wasserman, MD PhD Manuscript Region of Origin:

Manuscript

Religiosity as a protective factor against attempted suicide: a case-control study within WHO SUPRE-MISS

Abstract The aim of the current study was to investigate from a cross-cultural perspective the hypothesis whether religiosity assessed in two dimensions (religious denomination and subjective religiosity) is a protective factor against attempted suicide. 2819 suicide attempters seen in emergency settings and 5484 controls from community survey in the same catchment area in certain regions of 7 countries (Brazil, Estonia, India, Islamic Republic of Iran, South Africa, Sri Lanka, Viet Nam) participated in WHO Suicide Prevention – Multisite Intervention Study on Suicidal Behaviors (SUPRE-MISS). The results of the study revealed large diversity of religious denominations across participating sites. Predominant religions were Catholicism and Protestantism in Brazil, Protestantism and Orthodox in Estonia, Hinduism in India, Islam (Shi'ite) in Islamic Republic of Iran, various denominations without any of them prevailing in South Africa, and Buddhism in Sri Lanka. In Viet Nam over 90% of the respondents reported no religious denomination. In India, Sri Lanka and Islamic Republic of Iran both controls and/or suicide attempters reported some kind of religious denomination, therefore the odds ratio was not calculable. In Estonia religious denomination had protective effect (OR = 0.53; 95% CI = 0.38-0.73) and in South Africa risk effect (OR = 6.19; 95% CI = 3.49-10.95). No effect was found in Brazil and Viet Nam. Subjective religiosity was protective factor against suicide attempt in Brazil (OR = 0.20; 95% CI = 0.12-0.32), Estonia (OR = 0.55; 95% CI = 0.39-0.78), Islamic Republic of Iran (OR = 0.62; 95% CI = 0.45-0.84) and Sri Lanka (OR = 0.29; 95% CI = 0.15-0.57).

1

In South Africa it turned out to be a risk factor (OR = 2.88; 95% CI = 2.01-4.02). In India and Viet Nam the subjective religiosity had no effect.

Introduction Since Durkheim (1897/2002) the general tendency of research findings on the impact of religiosity on suicidal behavior is in favor of inverse association and protective effect – higher level of religiosity implicates lower level of suicidality – although exceptional and controversial findings in this issue cannot be denied. As a matter of fact, most of the studies performed so far are hardly comparable. The study designs vary from ecological level to individual, target groups are chosen from different populations, the range of suicidality investigated is wide (completed suicides, attempted suicides, suicidal ideation, attitude towards suicidal behavior), the aspects of religiosity and the questions asked about it differ substantially. Majority of the studies have been ecological by the design and relatively few individual-level findings have been reported on this matter. Furthermore, the majority of studies have been conducted in developed countries and less work has been done in developing countries (Vijayakumar, John, Pirkis, & Whiteford, 2005). Bertolote et al (2005) have suggested, that the religiosity of a person might serve as a protective factor against suicidal behavior and the WHO Suicide Prevention – Multisite Intervention Study on Suicidal Behaviors (SUPRE-MISS) will allow a more thorough exploration of this issue.

Social integration Durkheim pointed out the protective effect of religious denomination via social integration it provides (shared beliefs and shared practices). He concluded this is the reason why Catholic countries compared to Protestant countries report lower suicide rates (Durkheim, 1897/2002). 2

There are results supporting the Durkheim’s classical findings, that show higher suicide rates in predominantly Protestant regions and report persons with religious affiliation less likely to have a history of suicide attempt and having less suicidal ideation (Dervic, Oquendo, Grunebaum, Ellis, Burke, & Mann, 2004; Faria, Victora, Meneghel, de Carvalho, & Falk, 2006). However, the findings from early studies were usually inconsistent (Moreira-Almeida, Neto, & Koenig, 2006). A study by Neeleman et al (2004) suggested that religious affiliation had no independent effects on suicidality as the protective effects were no longer apparent once other socio-economic variables had been taken into account. Several others doubt in the effect of religious affiliation as a measure of religious integration and regulation in the contemporary world. This type of measure is currently facing serious challenge due to the growing convergence of Catholicism and Protestantism (Stack, 1983). Furthermore, religious affiliation tells us little about what is religiosity and how important it is in someone’s life (Moreira-Almeida et al., 2006).

Religious networks Traditional Durkheimian perspective on suicide has been questioned and reformulated in the religious networks perspective (Pescosolido & Georgianna, 1989; Stack, 2000). Pescosolido and Georgianna (1989) claimed either religious or other network ties to have both integrative and regulative aspects. If a church structure facilitates friendship ties among members of their congregations, these ties act as important sources of social and emotional support. Regulation guides action through advice and behavior monitoring. The strength of the network ties is expressed in their ability to carry out either function in a balanced way, which, in turn, reduces suicide risk. Exactly which elements of religious participation reduce the risk of suicide cannot be discerned. Analyses showed that visiting or talking with friends or relatives did not

3

reduce the likelihood of suicide relative to death by natural causes, but frequent participation in religious activities did. This suggests that it may not be the social contact inherent in some forms of religious participation that decreases suicide risk, but something more specifically intrinsic in religiousness (Nisbet, Duberstein, Conwell, & Seidlitz, 2000).

Religious commitment and core beliefs Stack (1983; 1992) has argued that a new theory of the effect of religion on suicide based on alternative religious concepts, such as religious commitment, can explain the phenomenon. In contrast to the classic integration view, the theory of religious commitment argues that the mere number of shared beliefs and practices is less important than the content of religious beliefs. The degree of commitment to a few life saving core aspects of religious beliefs (such as belief in an afterlife) may be critical in lowering suicide risk. A finding by Greening and Stoppelbein (2002) supports Stack’s theory that commitment to a few, core beliefs of traditional Christian doctrines explain the protective power that religion can have over self-destructive tendencies.

Religious practice and church attendance Actual church attendance regardless of denomination can be used as an indirect indicator of religious commitment, and in turn can be considered protective against suicide (Kelleher, Chambers, Corcoran, Williamson, & Keeley, 1998). Church, mosque or other important religious attendance, i.e. how often someone attends religious meetings, is one of the most widely used questions to investigate the level of religious involvement (Koenig, 2005; Moreira-Almeida et al., 2006). There are several studies available to support the positive effect of church attendance on suicidality. According to Duberstain et al (2004) suicides were less likely to engage in

4

religious practice or community activities and they had lower levels of social interaction. Several studies have revealed religious commitment expressed in church attendance being closely inversely associated with suicidal ideation and approval of suicide (da Silva, de Oliveira, Botega, Marin-Leon, Barros, & Dalgalarrondo, 2006; Siegrist, 1996; Stack & Lester, 1991). At the ecological level Lester (1987) has found the proportion of church attendance being negatively related to suicide rates, but Neeleman (1998) argued that compared to rates of religious belief, levels of attendance were only weakly associated with lower suicide rates, which was interesting as macro-level associations between religion and suicide tend to be attributed to the effects of social cohesion.

Subjective religiosity Another question widely used to investigate the level of religious involvement is subjective religiosity, the importance of religion in someone’s life (Moreira-Almeida et al., 2006). In post-modern societies, personal beliefs are at least as relevant as integration in religious institutions for the explanation of individual and group behavior (Neeleman, 1998; Stack, 1983). As stated by Walker and Bishop (2005), higher internalized, intrinsic religiosity was associated with less depression and thus, less suicidal thoughts. Further, intrinsic religiosity was found to be a more robust variable than both extrinsic religiosity and social support.

Moral objection and condemnation Suicide is a condemned act in most of the known religions in the world, especially in the three monotheistic religions: Judaism, Christianity and Islam. However, the strength of this condemnation has varied over time and within the religions themselves. Even though conservative church members (Catholic and Orthodox) have been the most

5

outspoken against suicide (Kelleher et al., 1998), little work has documented differences between Catholics and Protestants in their attitudes toward suicide (Stack, 1992). According to Pescosolido and Georgianna (1989), within Christianity, only two groups (Catholics and Reformed Churches) claimed to have an official policy towards suicide. All others claimed no official statement other than the sixth commandment (“You shall not murder”). Both Hindus and Buddhists believe in rebirth and in karma, which generates rebirth and influences both character and the fate of human beings. This facilitates the thinking that one can put an end to one’s life and not regard it as the final step. However, Buddhism is primarily against suicide as it is thought to be one of the greatest sins that could be committed. According to the prevailing belief among Buddhists, anyone who commit suicide will be reborn at least a lower level of life, i.e. as an animal (Bolz, 2002). Although in general the Hindu religion has strongly opposed destruction of a body that had not lived its allocated time, there are situations in which religion is traditionally rather tolerant of suicide. If a person is unable to reach more perfection on earth, either because of a serious handicap or because the environment makes life impossible, he or she is considered socially dead already and departure from life on earth will not alter the next life in any way (Tousignant, Seshadri, & Raj, 1998). Islam is much firmer about the sinfulness of suicide than Hinduism and Buddhism. The same may be true for Islam versus some Christian sects (Lester, 2006). The Islamic doctrine regarding suicide is well known: persons taking their own life will be denied entry to heaven. Islam unequivocally forbids the taking of one’s life. Suicide is considered a sin and subsequently a crime (Khan & Reza, 2000). In some Islamic countries confusion has erupted over the difference between completed suicide and attempted suicide. Although many Islamic countries have punitive laws against attempted suicide, there are no strong principles against it and no restrictions for

6

obtaining the required medical services. Similarly, because of religious and social reasons, suicide is considered a shameful act in the family and one that must be concealed (Khan, 1998). Several researchers have revealed association between attitude to suicidal behavior – moral objection opposed to tolerance and acceptance – and religiousness. Religious persons tend to be less accepting towards suicide (Dervic et al., 2004; Eskin, 2004; Neeleman, 1998). Other researchers claim that for example traditional Catholic culture may be merely protecting a person from expressing the more intense suicidal feelings, but not from suicide itself (Kirby, 2001; Pritchard & Baldwin, 2000).

Religious affiliations and denominations Religion in general, regardless of type, has been found inversely associated with suicide risk (Faria et al., 2006). It has been suggested by Dervic et al (2004) that the assessment of the presence or absence of religious affiliation may be more useful than the evaluation of an association between specific religious denomination (i.e. Catholic versus Protestant) and suicidal behavior. However, a comparison of suicide rates according to the prevalent religious denomination in countries brings to light a most remarkable difference between countries of Islam and countries of any other prevailing religion. In Muslim countries, where committing suicide is most strictly forbidden, the total suicide rate is close to zero. In Hindu and Christian countries, the total suicide rate is around 10 per 100,000 (Hindu 9.0; Christian 11.2). In Buddhist countries, the total suicide rate is distinctly higher at 17.9 per 100,000 population. At 25.6, the total suicide rate is markedly highest in secular countries, which in this analysis included countries where religious observances had been prohibited for a long period of time (Bertolote & Fleischmann, 2002).

7

Population level data introduce the risk of “ecologic fallacy” and limit their usefulness for determining individual level risk factors for suicide (Clarke, Bannon, & Denihan, 2003; Hilton, Fellingham, & Lyon, 2002). Although population-based findings do not take personal levels of religiosity into consideration, they might indicate the importance of the religious context, i.e. the prevalence of a religion in a country, in relation to suicide deaths, as a major cultural factor in the determination of suicide (Bertolote & Fleischmann, 2002).

Conceptualization of current study Koenig et al (2005; 2001) have defined religion as an organized system of beliefs, practices, rituals, and symbols designed to facilitate closeness to the sacred or transcendent. At the same time, religion is a wide concept and comprises different dimensions: religious belief, religious affiliation and denomination, organizational religiosity, non-organizational religiosity, subjective religiosity, religious commitment/motivation, religious “quest”, religious experience, religious well-being, religious coping, religious knowledge, and religious consequence. In the current study two dimensions of religion are in focus: religious affiliation/denomination and subjective religiosity. Koenig et al (2005; 2001) have explained the content of these dimensions as follows. Religious affiliation refers to identification with a particular religious group, but is not equivalent to membership in a religious group or adherence to the beliefs or practices of that group. Affiliation and denomination are often used interchangeably, although they are not exactly the same. Denomination usually refers to the specific group within Protestant Christianity with which the person is affiliated. Subjective religiosity taps that internal sense of God’s and religion’s importance in the individual’s life. How religious does the person

8

consider himself to be? This is an entirely subjective dimension that relies upon selfreport. The aim of the current study was to investigate the hypothesis whether religiosity assessed in two dimensions – religious denomination and subjective religiosity – is a protective factor against attempted suicide from a cross-cultural perspective.

Material and methods General description In 2000 the World Health Organization (WHO) launched the worldwide intervention study on suicidal behavior SUPRE-MISS (Suicide Prevention – Multisite Intervention Study on Suicidal Behaviors) with the main objective to reduce the mortality and morbidity associated with suicidal behavior. The methodology of SUPRE-MISS was elaborated by WHO expert group and adapted to local conditions of each participating site (WHO, 2002). SUPRE-MISS had three components: (1) an intervention study of suicide attempters seen in emergency settings in defined catchment areas; (2) a community survey in the same catchment areas, served also as the control group for suicide attempters; and (3) a qualitative community description of the basic socio-cultural characteristics of the target communities. All three components of the study were conducted in seven culturally diverse places around the world: Brazil (Campinas), Estonia (Tallinn), India (Chennai), Islamic Republic of Iran (Karaj), South Africa (Durban), Sri Lanka (Colombo), Viet Nam (Hanoi). The research protocol was approved by the relevant ethics committee in each site. The detailed description of the study and the characteristics of the suicide attempters as well as the community survey were presented previously elsewhere (Bertolote et al., 2005; Fleischmann, Bertolote, De Leo, Botega, Phillips, Sisask et al., 2005). 9

Data collection Suicide attempters were identified between 2002 and 2004 by medical staff in one or more emergency care settings within a catchment area with a population of at least 250,000 in each of participating sites. Once medically stable, the suicide attempters were asked to fill in a consent form and were thereafter interviewed by clinically experienced and specially trained psychiatrists, medical doctors, psychologists or psychiatric nurses. Within the community survey in the same catchment areas at least 500 randomly selected community members from general population were interviewed. These community members served also as controls for suicide attempters. The survey was conducted between 2002 and 2004. All participants provided informed consent. The interviewers were specially trained nurses, psychologists, medical students, medical doctors, family health workers, and public health professionals. In total 2819 suicide attempters and 5484 controls were interviewed. Detailed numbers of research subjects by participating sites are given in Table 1. --Table 1 approximately here ---

Instruments The questionnaire for suicide attempters, based on the European parasuicide study interview schedule (EPSIS) of the WHO/EURO multicentre study on suicidal behavior, and a common survey instrument for community survey were developed, translated and pilot-tested in each site (Kerkhof, Bernasco, Bille-Brahe, Platt, & Schmidtke, 1999; WHO, 2002).

10

The following religion related questions were asked from both suicide attempters and controls: 1. What is your religious denomination? 2. Do you consider yourself to be a religious person? To assess suicidal behavior during the life-time, the controls were asked the following questions: 1. Have you ever seriously thought about committing suicide? 2. Have you ever made a plan for committing suicide? 3. Have you ever attempted suicide? If the answer was “yes” to any of these questions, a person was classified as suicidal and excluded from logistic regression analysis (Table 1).

Statistical analysis Statistical analysis was performed with SPSS program (version 14.0). Differences between suicide attempters and control group were evaluated by chi-square test. Binary logistic regression analysis was performed to calculate gender and age adjusted odds ratios (OR) with 95 % confidence interval (95% CI). The level of statistical significance was set at α = 0.05.

Results Religious denomination The results of the study revealed large diversity of religious denominations across participating sites. Predominant religions were Catholicism and Protestantism in Brazil, Protestantism and Orthodox besides a great amount of persons without religious denomination in Estonia, Hinduism in India, Islam (Shi'ite) in Islamic Republic of Iran, various denominations without any of them prevailing in South Africa, and Buddhism 11

in Sri Lanka. In Viet Nam most of the people reported no religious denomination (Table 2). Differences between suicide attempters and control group in the pattern of distribution of religious denominations were significant at p < 0.001 level in Estonia, India, Sri Lanka, Brazil and South Africa. The differences were non-significant in Islamic Republic of Iran (p = 0.197), where in both groups the main religious denomination was Islam, and in Viet Nam (p = 0.859), where in both groups the majority of persons had no religious denomination. --Table 2 approximately here --Males and females had similar pattern of distribution as total results with two exceptions only – suicide attempters and control group were not significantly different in females in India (p = 0.067) and in males in Brazil (p = 0.852).

Effect of religious denomination and subjective religiosity In India, Sri Lanka and Islamic Republic of Iran all controls and/or suicide attempters had some kind of religious denomination, therefore the odds ratio was not calculable. In Estonia religious denomination had protective effect and in South Africa risk effect. In Brazil and Viet Nam the role of religious denomination was statistically non-significant (Table 3). --Table 3 approximately here --Considering him/herself as a religious person (subjective religiosity) was protective factor against suicide attempt in Brazil, Estonia, Islamic Republic of Iran and Sri Lanka.

12

In South Africa it turned out to be a risk factor. In India and Viet Nam the odds ratio of subjective religiosity was statistically non-significant (Table 4). --Table 4 approximately here ---

Discussion The aim of the current study was to find out whether different dimensions of religiosity (religious denomination and subjective religiosity) protect against attempted suicides. To the best of our knowledge this is the first individual level study conducted at the same time in culturally different sites, which enables to investigate the effect of religiosity on suicide attempts from a cross-cultural perspective.

Religious denomination Sites included in WHO SUPRE-MISS study differed substantially in the religiositysecularity spectrum. Some sites were very religious (India, Islamic Republic of Iran, Sri Lanka) and some were very or rather secular (Viet Nam, Estonia). Brazil and South Africa were in between. The prevailing religious denominations differed also to a large extent. It seems that large diversity of religious denominations within a site like in South Africa, where the category “other” was particularly over-represented compared to other sites, is not favorable in terms of having protective effect against attempted suicide. South Africa was the only site where religious denomination turned out to be risk factor. South Africa has been described as “The Rainbow Nation” because of its cultural diversity. There are a variety of ethnic groups and a greater variety of cultures within each of the groups. While cultural diversity is seen as a national asset, the interaction of cultures results in 13

the blurring of cultural norms and boundaries at the individual, family and cultural group levels (Wassenaar, van der Veen, & Pillay, 1998). In Brazil the religious denomination Catholicism was more frequent among the control group than among suicide attempters and Protestantism was more frequent among suicide attempters than among the control group. While religious denomination had no effect on suicide attempt in Brazil, it could be assumed that the risk effect of Protestantism could neutralize the protective effect of Catholicism. However, this is only a speculation, as for analysis in current study all denominations of Christianity were added together and differentiating the effect of denominations within Christianity was not the issue of interest. At the same time it is known from a study by Botega et al (Botega, Barros, Oliveira, Dalgalarrondo, & Marin-Leon, 2005), that in Brazil the lifetime prevalence of suicidal ideation among Protestants was even lower than among Catholics. In Estonia having religious denomination provided protective effect against suicide attempt. Usually Estonia acknowledges its protestant heritage, but in the SUPRE-MISS study dissimilarly the Orthodox prevails over Protestantism. This is probably due to the peculiarity of the site selected: in Tallinn (the capital city of Estonia) the proportion of Slavic inhabitants is bigger than in Estonia as a whole (according the data for year 2002 derived from Estonian Statistical Office 37% versus 26%) and Slavic people are mainly Orthodox.

Subjective religiosity Subjective religiosity was protective factor in four sites out of seven (Brazil, Estonia, Islamic Republic of Iran, Sri Lanka). It can be concluded that subjective religiosity plays even more important role considering the protective effect against attempted suicide than does religious denomination. In two sites (India and Viet Nam) the effect

14

was statistically non-significant. It is known from previous research, that in India the subjective religiosity does not protect against attempted suicide, it protects against completed suicide (Vijayakumar, 2003). The results from Viet Nam can be attributed to its secularity, which may influence the overall way of thinking and mentality. In South Africa the subjective religiosity turned out to be a risk factor, which is an exceptional result again as it was also true for the effect of religious denomination. As mentioned above, it can be explained by the cultural diversity and in the blurring of cultural norms within the site.

Limitations Within SUPRE-MISS study the information about religious denomination and subjective religiosity was collected from investigated subjects by asking direct questions. Even with clinically experienced and specially trained interviewers the possibility remains, that the information could be incomplete due to respondents’ memory bias and unwillingness to report honestly sensitive issues like religiosity. Religiosity has even more different faces, which were not assessed with SUPRE-MISS instrument. Other dimensions of religiosity might also play an important role in some cultures and religions. Although, the strength of the current study was the common study design applied in all culturally different participating sites. At the same time a question may arise whether respondents with considerably different cultural background perceive similarly the questions asked about religion, even if the instruments were adapted to local conditions and pilot-tested before main study begun.

15

References Bertolote, J. M., & Fleischmann, A. (2002). A global perspective in the epidemiology of suicide. Suicidologi, 7(2), 6-7. Bertolote, J. M., Fleischmann, A., De Leo, D., Bolhari, J., Botega, N., De Silva, D., et al. (2005). Suicide attempts, plans, and ideation in culturally diverse sites: the WHO SUPRE-MISS community survey. Psychological Medicine, 35(10), 14571465. Bolz, W. (2002). Psychological analysis of the Sri Lankan conflict culture with special reference to the high suicide rate. Crisis, 23(4), 167-170. Botega, N. J., Barros, M. B., Oliveira, H. B., Dalgalarrondo, P., & Marin-Leon, L. (2005). Suicidal behavior in the community: prevalence and factors associated with suicidal ideation. Revista Brasileira de Psiquiatria, 27(1), 45-53. Clarke, C. S., Bannon, F. J., & Denihan, A. (2003). Suicide and religiosity - Masaryk's theory revisited. Social Psychiatry and Psychiatric Epidemiology, 38(9), 502506. da Silva, V. F., de Oliveira, H. B., Botega, N. J., Marin-Leon, L., Barros, M. B., & Dalgalarrondo, P. (2006). [Factors associated with suicidal ideation in the community: a case-control study]. Cadernos de Saúde Pública, 22(9), 18351843. Dervic, K., Oquendo, M. A., Grunebaum, M. F., Ellis, S., Burke, A. K., & Mann, J. J. (2004). Religious affiliation and suicide attempt. The American Journal of Psychiatry, 161(12), 2303-2308. Duberstein, P. R., Conwell, Y., Conner, K. R., Eberly, S., Evinger, J. S., & Caine, E. D. (2004). Poor social integration and suicide: fact or artifact? A case-control study. Psychological Medicine, 34(7), 1331-1337.

16

Durkheim, E. (1897/2002). Suicide: A Study in Sociology. London and New York: Routledge Classics. Eskin, M. (2004). The effects of religious versus secular education on suicide ideation and suicidal attitudes in adolescents in Turkey. Social Psychiatry and Psychiatric Epidemiology, 39(7), 536-542. Faria, N. M., Victora, C. G., Meneghel, S. N., de Carvalho, L. A., & Falk, J. W. (2006). Suicide rates in the State of Rio Grande do Sul, Brazil: association with socioeconomic, cultural, and agricultural factors. Cadernos de Saúde Pública, 22(12), 2611-2621. Fleischmann, A., Bertolote, J. M., De Leo, D., Botega, N., Phillips, M., Sisask, M., et al. (2005). Characteristics of attempted suicides seen in emergency-care settings of general hospitals in eight low- and middle-income countries. Psychological Medicine, 35(10), 1467-1474. Greening, L., & Stoppelbein, L. (2002). Religiosity, attributional style, and social support as psychosocial buffers for African American and white adolescents' perceived risk for suicide. Suicide & Life-Threatening Behavior, 32(4), 404-417. Hilton, S. C., Fellingham, G. W., & Lyon, J. L. (2002). Suicide rates and religious commitment in young adult males in Utah. American Journal of Epidemiology, 155(5), 413-419. Kelleher, M. J., Chambers, D., Corcoran, P., Williamson, E., & Keeley, H. S. (1998). Religious sanctions and rates of suicide worldwide. Crisis, 19(2), 78-86. Kerkhof, A., Bernasco, W., Bille-Brahe, U., Platt, S., & Schmidtke, A. (1999). European Parasuicide Study Interview Schedule (EPSIS). In U. Bille-Brahe (Ed.), Facts and Figures: WHO/EURO. Copenhagen: WHO Regional Office for Europe. Khan, M. M. (1998). Suicide and attempted suicide in Pakistan. Crisis, 19(4), 172-176.

17

Khan, M. M., & Reza, H. (2000). The pattern of suicide in Pakistan. Crisis, 21(1), 3135. Kirby, M. (2001). Suicide rates among the elderly in different religious cultures. International Journal of Geriatric Psychiatry, 16(9), 920. Koenig, H. G. (2005). Faith And Mental Health: Religious Resources for Healing. West Conshohocken: Templeton Foundation Press. Koenig, H. G., McCullough, M. E., & Larson, D. B. (2001). Handbook of Religion and Health. New York: Oxford University Press. Lester, D. (1987). Religiosity and personal violence: a regional analysis of suicide and homicide rates. The Journal of Social Psychology, 127(6), 685-686. Lester, D. (2006). Suicide and Islam. Archives of Suicide Research, 10(1), 77-97. Moreira-Almeida, A., Neto, F. L., & Koenig, H. G. (2006). Religiousness and mental health: a review. Revista Brasileira de Psiquiatria, 28(3), 242-250. Neeleman, J. (1998). Regional suicide rates in the Netherlands: does religion still play a role? International Journal of Epidemiology, 27(3), 466-472. Neeleman, J., de Graaf, R., & Vollebergh, W. (2004). The suicidal process; prospective comparison between early and later stages. Journal of Affective Disorders, 82(1), 43-52. Nisbet, P. A., Duberstein, P. R., Conwell, Y., & Seidlitz, L. (2000). The effect of participation in religious activities on suicide versus natural death in adults 50 and older. The Journal of Nervous and Mental Disease, 188(8), 543-546. Pescosolido, B. A., & Georgianna, S. (1989). Durkheim, suicide, and religion: toward a network theory of suicide. American Sociological Review, 54(1), 33-48. Pritchard, C., & Baldwin, D. (2000). Effects of age and gender on elderly suicide rates in Catholic and Orthodox countries: an inadvertent neglect? International Journal of Geriatric Psychiatry, 15(10), 904-910.

18

Siegrist, M. (1996). Church attendance, denomination, and suicide ideology. The Journal of Social Psychology, 136(5), 559-566. Stack, S. (1983). The effect of religious commitment on suicide: a cross-national analysis. Journal of Health and Social Behavior, 24(4), 362-374. Stack, S. (1992). Marriage, family, religion, and suicide. In R. Maris, A. Berman, J. Maltsberger & R. Yufit (Eds.), Assessment and Prediction of Suicide (pp. 540552). New York: Guildford. Stack, S. (2000). Suicide: a 15-year review of the sociological literature. Part II: modernization and social integration perspectives. Suicide & Life-Threatening Behavior, 30(2), 163-176. Stack, S., & Lester, D. (1991). The effect of religion on suicide ideation. Social Psychiatry and Psychiatric Epidemiology, 26(4), 168-170. Tousignant, M., Seshadri, S., & Raj, A. (1998). Gender and suicide in India: a multiperspective approach. Suicide & Life-Threatening Behavior, 28(1), 50-61. Vijayakumar, L. (2003). Psychosocial risk factors for suicide in India. In L. Vijayakumar (Ed.), Suicide Prevention: Meeting the Challenge Together pp. 149-162). India: Orient Longman. Vijayakumar, L., John, S., Pirkis, J., & Whiteford, H. (2005). Suicide in developing countries (2): risk factors. Crisis, 26(3), 112-119. Walker, R. L., & Bishop, S. (2005). Examining a model of the relation between religiosity and suicidal ideation in a sample of African American and White college students. Suicide & Life-Threatening Behavior, 35(6), 630-639. Wassenaar, D. R., van der Veen, M. B., & Pillay, A. L. (1998). Women in cultural transition: suicidal behavior in South African Indian women. Suicide & LifeThreatening Behavior, 28(1), 82-93.

19

WHO (2002). Multisite Intervention Study on Suicidal Behaviours SUPRE-MISS: Protocol of SUPRE-MISS. Geneva: WHO.

20

Table 1 Number of suicide attempters (SA) and control group (CG) included in the study SA Brazil Estonia India Islamic Republic of Iran South Africa Sri Lanka Viet Nam TOTAL

162 332 680 632 570 300 143 2819

Non-suicidal 420 433 486 433 371 632 2079 4854

CG Suicidala 96 67 14 71 129 52 201 630

Total 516 500 500 504 500 684 2280 5484

a

Suicidal - persons reporting during their life-time suicidal behaviour (attempts, plans, thoughts)

21

Table 2 Religious denomination, suicide attempters (SA) in comparison with control group (CG)

Brazil

CG SA

Estonia

CG SA

India

CG SA

Islamic Republic of Iran

CG SA

South Africa

CG SA

Sri Lanka

CG SA

Viet Nam

CG SA

N % N % N % N % N % N % N % N % N % N % N % N % N % N %

None 42 8.1 23 14.9 243 48.8 223 67.2 0 0.0 1 0.1 0 0.0 0 0.0 83 16.7 24 4.2 1 0.2 0 0.0 2074 91.1 134 93.1

Protestantism 100 19.4 47 30.5 58 11.6 33 9.9 0 0.0 24 3.5 0 0.0 0 0.0 81 16.3 31 5.5 3 0.5 2 0.7 16 0.7 0 0.0

Catholicism 344 66.7 75 48.7 21 4.2 6 1.8 16 3.2 44 6.5 0 0.0 0 0.0 68 13.7 74 13.1 84 12.7 54 18.0 40 1.8 1 0.7

Jewish 0 0.0 0 0.0 3 0.6 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 1 0.2 0 0.0 0 0.0 1 0.3 13 0.6 0 0.0

Islam 0 0.0 0 0.0 2 0.4 0 0.0 23 4.6 25 3.7 502 99.6 632 100.0 13 2.6 28 5.0 161 24.3 38 12.7 2 0.1 0 0.0

Hinduism 0 0.0 1 0.6 0 0.0 0 0.0 460 92.0 571 84.0 0 0.0 0 0.0 62 12.5 114 20.2 116 17.5 31 10.3 0 0.0 0 0.0

Orthodox 0 0.0 0 0.0 127 25.5 65 19.6 0 0.0 0 0.0 0 0.0 0 0.0 1 0.2 0 0.0 0 0.0 0 0.0 2 0.1 0 0.0

Buddhism 2 0.4 0 0.0 1 0.2 1 0.3 0 0.0 1 0.1 0 0.0 0 0.0 0 0.0 1 0.2 291 43.9 171 57.0 128 5.6 9 6.3

Other 28 5.4 8 5.2 43 8.6 4 1.2 1 0.2 14 2.1 2 0.4 0 0.0 188 37.8 293 51.9 7 1.1 3 1.0 2 0.1 0 0.0

Total 516 100.0 154 100.0 498 100.0 332 100.0 500 100.0 680 100.0 504 100.0 632 100.0 497 100.0 565 100.0 663 100.0 300 100.0 2277 100.0 144 100.0

22

Table 3 Religious denominationa, suicide attempters (SA) in comparison with control group (CG), binary logistic regression analysis adjusted for gender and age OR Brazil Estonia India Islamic Republic of Iran South Africa Sri Lanka Viet Nam

0.73 0.53 not calculable not calculable 6.19 not calculable 1.48

95% CI

p-value

lower 0.41 0.38

higher 1.31 0.73

0.289 < 0.001

3.49

10.95

< 0.001

0.73

2.98

0.277

a

Religious denomination - yes (any) versus none

23

Table 4 Subjective religiositya, suicide attempters (SA) in comparison with control group (CG), binary logistic regression analysis adjusted for gender and age OR Brazil Estonia India Islamic Republic of Iran South Africa Sri Lanka Viet Nam

0.20 0.55 0.71 0.62 2.88 0.29 1.22

95% CI lower 0.12 0.39 0.47 0.45 2.01 0.15 0.78

higher 0.32 0.78 1.08 0.84 4.02 0.57 1.89

p-value < 0.001 0.001 0.109 0.002 < 0.001 < 0.001 0.380

a

Subjective religiosity - considering him/herself as religious person

24