Although much has been learned about risk factors contributing to suicide, a fundamental understanding of the suicide process remains unknown, and prevention efforts have been less than fully successful (IOM, 2002). Most previous studies of suicide have been restricted to one domain of possible risk factors, and comprehensive studies on both social and personal (psychological and psychiatric) risk factors are rare. Most of those studies are generally from medical perspectives and are exploratory in nature and therefore lack theoretical generalization. Mann and colleagues (1999) developed and tested a stress-diathesis theory of suicide, but this is only a clinical model based on and for psychiatric patients. Heeringen's (2003) psychobiological model of suicidal behavior, which focuses on the process of the state-trait interaction, seems more generalizable but, again, is neurobiological in nature. Through overcoming such deficiencies, this paper attempts to conceptualize a basic paradigm that incorporates the available theories, hypotheses, and findings explaining suicide in the world today.

In the United States, the factors most consistently associated with suicide, affecting over 90% of all people who die by suicide, are mental illness, substance use disorders (SUD) and alcohol use disorders (AUD). However, the fact that only a small proportion of persons with these disorders actually die by suicide and only approximately half of the Chinese suicides have been diagnosed with any mental problems (Phillips et al., 2002) raises important questions that challenge the psychiatric model of suicide (NIMH, 2003). On the other hand, sociological studies of suicide tend to overestimate the importance of social variables by failing to measure psychiatric disorders with such a rigorous instrument as the SCID (the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders) (Spitzer, Williams, Gibbon, & First, 1988).

Durkheim's (1951) classical theory of social integration and regulation explaining egoistic, altruistic, anomic, and fatalistic suicide is in theoretical and practical conflict with the psychopathological theories prevalent in today's world (Durkheim, 1951; Heeringen, 2003; Mann, Waternaux, Haas, & Malone, 1999). Suicide rates and patterns vary greatly among different geographic areas and different racial and ethnic groups around the world, and the radical differences in demography of suicide reflect cultural influences. Society and culture play enormous roles in dictating how people respond to and view mental health and suicide, and cultural variables have a far-ranging impact on suicide. Therefore, suicide prevention requires an understanding of how suicide varies with social and cultural forces and how it relates to individual, group and contextual experiences (IOM, 2002).

Table 1 presents the case-control differences in the means and percentages of the major variables by gender. The suicide group and the control group differed significantly on all factors but the marital status. The age ranged from 15 to 34 years with a mean of 26.9 years (SD=6.3) for the suicide sample and 25.9 years (SD=6.2) for the controls. Combined male and female, suicides were older than living controls. However, the difference was not significant for females in the study. The number of years of formal education ranged from 0 to 16 with a mean of 7.4 (SD=2.8) for the suicides and from 2 to 18 with a mean of 9.2 (SD=2.4) for the living controls. The number of education years was lower for the suicides than for the controls for both men and women. There was no marital status (never married vs. ever-married) difference between the suicides and controls. Among those ever married, the groups of divorced and widowed were so small that statistical comparisons were not feasible. About 28.8 percent (n=113) of the suicides believed in any religion, but only 16.8 percent (n=70) of the living controls believed in any religion. The case-control difference in religion was larger for men than for women. While 47.7 percent (n=187) of the suicides were diagnosed with at least one mental disorder, including alcohol and substance abuse, only 2.6 percent (n=11) of the living controls were with any kind of mental disease, including alcohol and substance abuse. Table 1 also shows that suicides scored significantly higher than controls on the Confucian traditional gender value, failed aspiration, economic deprivation, and lack of coping skills. The case-control differences in all four strain measures were significant for both men and women.

The risk factors for suicide were estimated by multiple logistic regressions with alpha = .05 to identify statistically significant results because the analyses were used to examine a priori hypotheses pertaining to the risk factors that were measured. Table 2 presents the results separately for men and women.

Marital status was not entered in the logistic multiple regression model, as it did not distinguish the case-control status for either men or women. In the regression model, the strongest predictor of suicide was psychiatric diagnosis. For rural young men and women in China, the diagnosed mental disorder is highly correlated with suicide risk. The second strongest predictor of suicide was coping skills. For both rural young men and rural young women, the higher the level of coping skills, the lower the risk of suicide. More years of education reduced suicide risk for men but not for women, and age was not related to suicide risk in the regression analyses. The tendency that suicides were more religious than controls evidenced in the bivariate analysis failed to reach a significant level in the multiple regression analyses.

The score on Confucianism was not associated with suicide risk in the multiple regression analyses. Failed aspiration was related to suicide risk for men but not necessarily for women. As expected, both economic deprivation and lack of coping skills were significantly associated with suicide risk for both men and women. Economic deprivation was directly associated with suicide, while coping skills were inversely associated with suicide.

To assess the association between strain and suicide and the role of mental illness in the association, we performed mediation analyses with MPlus (Muthen & Muthen, 2006) using a logistic Structural Equation Model (SEM) with a binary endogenous variable (1=suicide; 0=living control) and a mediator (1=any mental illness; 0=no mental illness). We assumed multi-normality of the error terms.

First, a path diagram was drawn to clarify the relationship. Strain as a latent construct was formed from the four strain measures: value strain, aspiration strain, deprivation strain, and coping strain using confirmatory factor analyses (CFA) in MPlus, fixing the factor variance to one. Gender and age were controlled for in the analyses.

When everyone is living in poverty, poverty is not a stressor. With everything else being held equal, a person's miserable living condition is perceived in relation to others, and the person feels frustrated. Relative poverty or deprivation may be objectively measured, but must be subjectively perceived by the individual who is at suicide risk. Agnew (1992) derived much of his strain theory from the justice literature and linked the psychological frustration resulting from the perceived injustice to deviant behaviors.

High aspiration for future can also become frustration and strain if the reality does not allow the individual to move to or to reach the goal the individual aspired. We divided the subjects who had an aspiration into those who realized their wishes and those who did not realize their wishes and confirmed that a failed aspiration is a strain that leads to suicide. For the suicides in the current sample, the top four failed aspirations were (1) making more money, (2) harmonious family, (3) love and marriage, and (4) job security. Most people should have experienced a crisis in life, but few people in crisis situation were so frustrated as to kill themselves.

Only those who do not have a coping ability become extremely frustrated and seek death as a solution to solve the problems inflicted by the crisis. We measured the coping skills with a standardized instrument and confirmed that those individuals who have better coping skills have lower suicide risk.

Mental illness is a mediator in Chinese suicide. Psychological strains increased the level of mental illness, and mental illness then increased the risk of suicide. However, approximately half of the Chinese suicides in this study were not diagnosed with any psychiatric problems. Psychological strain by itself can also be a strong predictor of suicide in Chinese culture.

The effect of religion/religiosity is reversed in the Chinese culture -- it tends to be detrimental for Chinese rural young people in terms of suicidality. Because of beliefs and social integration, sociologists since Durkheim have argued that religion protected people from suicide (Stack, 2000). Previous studies noted suicide rates are lower in religious countries than in the secular ones (Breault, 1986; Stack, 1983) and low religiosity is associated with suicidal ideation, as well as suicidal behaviors (Cook, Pearson, Thompson, Black, & Rabins, 2002), although no studies have established an association between specific religious denomination (i.e., Catholic versus Protestant) and suicidal behavior (Dervic et al., 2004). In China, Buddhism and Taoism, foundations of traditional Chinese culture, are different from Western religions in terms of supernatural being, afterlife, rituals, and organization (Zhang & Xu, 2007). In Chinese religions, there is not a single God to worship, and there is a lack of social support system and coping mechanisms, as the majority of the religious people do not meet regularly. Compared with an institutionalized religion, social interaction and social support are absent in private worships (Pescosolido & Georgiana, 1989). Different from all the mainstream religions in the West, Chinese religions are often associated with superstition as the saying of zongjiao mixin (religious superstition). To some Chinese individuals, being religious is equivalent to being superstitious, and death is a solution to all the problems and the beginning of a new life (incarnation). Therefore, it is possible that those who take religion to the extreme are likely to think about starting a new life by ending this current one quickly.

We should also understand the finding that religiosity is related to Chinese suicide, in line with the strain theory of suicide. As an atheist country, China now has less than 10 percent of its population claiming themselves religious, and this percentage has reflected the rapid growth in the past 30 or so years since China opened its door to the West (Badham, 2008; Yao & Badham, 2008). Church and prayer are still considered to be deviant to the majority of the Chinese (Zhang, et al., 2010). Additionally, those churchgoers in China may also experience the psychological frustration caused by their religious beliefs that conflict with the mainstream culture in the larger society. The value conflict strain increases when the two conflicting values are both internalized in an individual. In sum, three differences between the Western and Chinese religious practices account for the different effects of religion on suicidality between the two cultures: (1) the lower level of institutionalization of religions in China offers less social support to believers than in the West; (2) the Buddhist incarnation may encourage the superstitious believers to gloriously end their own life; (3) the Chinese religious believers are treated as a minority group, both due to its small percentage and political disadvantage. Value strains can be experienced between the main stream values in society and their subcultural values.

Psychological strains are presumed to be the up-stream reason that leads to both mental problem and suicidality. Further studies in various populations and more sophisticated measures should be conducted to improve and refine this strain theory of suicide, as this knowledge may be critical for reducing suicides at all three levels of prevention. For first degree prevention (universal), the government needs to carry out and reinforce the policies that help reduce economic polarizations in society so as to bring down certain individuals' anger and frustration caused by the aspiration strain, as well as the relative deprivation strain. At the behavioral level, efforts to reduce strain include increasing coping skills through early childhood education interventions. For second degree prevention (selective), for support reasons, we need to identify individuals who are outliers socially in a village or community (relatively poorer than their peers), and social welfare systems should work for those people in need. For third degree prevention (indicated), we have to develop some psychological counseling strategies (strain reduction through cognitive therapy) to care those individuals at the suicidal edge, or screen and refer for treatment people who develop mental illness.

One unique contribution of this study to sociological research of suicide is that it employed a rare blend between psychiatric and social predictors of suicide, where the Chinese version of the SCID was administered to measure psychiatric disorders of both suicides and living controls. The study shows that social factors influence suicide risk independent of the dominant risk factor for suicide, psychiatric illness. Further, the strain theory is first tested in Chinese culture. Hopefully, the present paper will stimulate work in other cultures to test the strain theory of suicide and in a variety of fields beyond just suicide studies.