The results of the systematic review show that unemployment and low income are major individual risk factors for all suicidal behaviors. Working in precarious conditions (job insecurity, part-time work), long working hours, changes in employment status, shift work and night-time work and occupational stress were associated with an increased risk for suicidal ideation. The included studies provide evidence that the strength of the association between a particular risk factor and suicidal behavior is influenced by age and gender. Thus, the influence of income was shown to be particularly strong for men, and low educational attainment was a risk factor for suicide attempts and completed suicides especially for women and younger adults. Occupational stressors differed between the sexes as well, low job control for men and interpersonal conflict for women were associated with suicidal ideation. Furthermore, low SES appears to increase the risk for suicidal behaviors. Included studies were unable to ascertain whether and to what extent place of residence influenced the three investigated suicidal behaviors. This might be due to the design review, which only considered individual socioeconomic factors, as this allowed only a small number of studies to be eligible. Some studies identified that socioeconomic variables were indirectly associated with suicidal behavior, as they affect the mental and physical health.
The reasons for unemployment and low income contributing to suicidal behaviors are manifold and interdependent. Occupation serves to provide social and financial resources [72]. Unemployed individuals are at risk of facing cessations in social inclusion in everyday life, which consequently can lead to a loss of self-esteem and decreased levels of social standing [73]. Low income reduces access to essential goods, causes chronic stress and unhealthy behaviors closely related to suicide, such as excessive drinking and smoking [27, 74, 75].
In some studies included in this systematic review, income was significantly associated with suicidal behavior in the working-age population only, respectively the influence of income decreased with increasing age [37, 38, 58]. This may be due to the life events and responsibilities associated in the working-age, such as financial support of children and paying parents’ medical bills [76]. However, since most studies consider low income to be a general risk factor, it can be assumed that although age influences the strength of the effect, low income is a weighty predictor of suicidal behavior among all ages. Other studies also concluded that the effects of low income on suicidal behavior are stronger in men, which might be explained by the traditional ideas of Korean society [31, 40, 53, 66].
One reason for the strong association between working in agriculture, forestry and fishery (AFF) or military services and completed suicides might be the convenient access to lethal suicidal methods in this occupational fields (e.g. firearms and pesticides) [47, 77]. Another finding of the analyses between occupational groups and suicidal behavior is that blue collar worker respectively manual workers were more likely to report suicidal ideation compared to white- and pink-collar worker [20, 49, 57]. Of particular note are unskilled manual workers, who represented the largest risk group. This might be due to the more physically demanding work and the precarious working conditions as well as the higher likelihood of part-time work. All of these are factors investigated in this systematic review that were found to significantly increase the risk for suicidal ideation [46, 56, 71].
However, all included studies regarding occupational group and suicidal ideation are cross-sectional. Therefore, causality cannot be established. In addition, the economic crisis in 2008 has led to major changes related to occupation [47, 78]. For that reason, for all studies dealing with employment and occupation, the different data collection periods must be considered. However, self-employment has been observed as a risk factor for suicidal behavior, which may be due to exploitative working circumstances (such as long and irregular working hours), lower health insurance coverage and an exclusion from state welfare programs.
Occupational stressors associated with suicidal intention differed by gender: Low job control and lack of reward for men and interpersonal conflict for women remained significant after adjusting for confounders [45, 69]. These results might be explained by the influence of Confucianism, which shaped the traditional gender roles in South Korea. Accordingly, men serve as the primary wage earners and hold control over their work. Job insecurity could thus negatively affect men’s self-image, and, thereby, suicidal thoughts may arise. Further studies also indicate that Korean men’s job satisfaction is more dependent on (extrinsic) reward [79]. Women tend to be psychologically more susceptible to interpersonal stress, the cause of which they often see in themselves [80]. Job insecurity was found to be particularly a risk factor for head of households, as they are more responsible for supporting their families [41].
Interestingly, compared to attempted suicide and completed suicides, the level of education is hardly decisive for the development of suicidal ideation. One explanation might be that people with higher levels of education have better coping strategies [81]. However, most of the included studies lack an adjustment for determinants that may affect the educational attainment, i.e. support and availability of parents or gender differences in terms of educational opportunities [82, 83].
The studies included in the systematic review were insufficient to conclude whether and to what extent place of residence affects suicidal ideation. On the one hand, this could be due to the relatively low amount of studies and study subjects. On the other hand, studies suggest that not the place or residency, but rather community factors such as availability of social welfare facilities and social cohesion within the neighborhood are more relevant when determining predictors for suicidal ideation [84, 85]. However, regional deprivation levels and specific features of neighborhood environment (air pollution, noise, availability of resources) have not been assessed in the primary studies.
Future research should further focus on contextual factors within high-risk spatial clusters as well as on identifying at-risk groups in order to derive targeted prevention strategies for them, as protection policies for vulnerable populations are much needed to prevent suicidality. Since all socioeconomic variables were studies separately, future studies should examine the interdependence between socioeconomic variables to better assess the effect size of a risk factor. In investigating the relationship between socioeconomic variables and suicidal behavior, research should focus on identifying vulnerable groups in order to derive more concrete recommendations for action by policymakers. Gender- and age-specific differences in the effect of socioeconomic factors on suicidal behavior suggest that further studies are needed that address these differences more specifically, with particular attention to the tension between collectivist Confucianism and the phenomenon of a hitherto unknown burgeoning individualism. Overall, the results highlight the need for meaningful suicide prevention in South Korea, which must have two goals: 1) Developing effective strategies to minimize risk factors and support vulnerable groups, and 2) breaking the stigma of suicidal ideation and mental illness as a sign of failure and weakness.
In addition to political and economic policies and changes to access to health resources, communities and municipalities play a critical role in suicide prevention. Effective support within communities can help protect vulnerable individuals from suicide by building and improving social relationships and coping skills. Assistance and counseling centers must be created that reflect and address the specific characteristics of victims of suicidal behavior in a region (e.g. high old-age poverty). Gatekeeper, most of whom are the first point of contact in the health care system, need to be trained to recognize linguistic, behavioral, and situational warning signs and learned to intervene in areas at risk for suicide. With the introduction of the “Standardized Suicide Prevention Program for Gatekeeper Intervention in Korea“, the government has taken an important first step [86]. However, effective community suicide prevention goes beyond the involvement of gatekeepers and includes the collaboration of schools, workplaces, leisure programs and places of worship to meet local needs and resources with regard to prevention strategies. Comprehensive suicide prevention efforts at the community level can help reduce what may be the biggest problem: the stigma surrounding suicidal behavior and mental illness.
Limitations
Several factors need to be considered when interpreting the result; first of all, related to the primary studies themselves. Foremost, the studies are very heterogeneous. All studies focusing either on suicidal ideation or suicide attempts are based on self-reports. Both selection bias and social desirability may lead to an underestimation. Most studies assessed suicidal ideation only by one dichotomous question (yes/no) rather than by a scale consisting of multiple items, covering severity, duration, and variability of suicidal ideation. Furthermore, the risk factors and suicidality have bidirectional effects: For example, a suicide attempt could also decrease economic productivity, resulting in unemployment and poverty. Therefore, the possibility of reverse causality has to be taken into consideration. Although the adjustment of essential confounders is crucial in this area, many studies were secondary analyses of surveys. For that reason, some important variables for these secondary analyses might be missing. Due to the lack of available data, the majority of studies did not control for mental health variables and the outcomes of studies are difficult to compare with those of studies that did. In addition, most studies only collected information about the participants’ current status of health and did not gather information on participants’ history of mental or physical illness.
In addition, limitations occur due to choices made for the systematic review. The literature search is based on two databases and restricted to articles in English language. Thus, studies published in Korean language and grey literature have not been considered. Since there was no restriction on the year of publication, the studies may come to different conclusions, which should be understood in the light of economic and social transformations over time. In this regard, it is worth mentioning the two economic crises at the end of the 1990s and in 2008/2009, which may have changed the impact of the risk factors. We did not assess risk of bias of primary studies. The influence of confounders has not been measured uniformly, and it is unsure whether flaws in the design, conduct or analysis of a study may lead to biases.