Socioeconomic factors associated with suicidal behaviors in South Korea: systematic review on the current state of evidence

Background The economic and human costs of suicide to individuals, families, communities, and society make suicide a major public health problem around the world. Suicide rates in South Korea are among the highest in the world. This paper is the first systematic review investigating socioeconomic risk factors for suicidal behaviors (suicidal ideation, attempted suicides, and completed suicides) in South Korea. Methods We performed a systematic review in Medline and Web of Science. Empirical studies and peer-reviewed articles on the association between individual socioeconomic factors and suicidal behaviors have been included. A total of 53 studies were included in a descriptive synthesis. Results Overall, 35 studies focused on the association between individual socioeconomic factors and suicidal ideation, 16 were related to suicide attempts, while 10 addressed completed suicides. Low income, unemployment, and financial difficulties were identified as risk factors for all suicidal behaviors. Working in precarious conditions, long working hours, self-employment, changes in employment status, shift work/night-time work, and occupational stress were associated with an increased risk for suicidal ideation. Low educational attainment appears to increase the risk for suicide attempts and completed suicide, but the significance of educational attainment on the reporting of suicidal ideation could not be verified. The primary studies were unable to ascertain whether the place of residence impacts on suicidal behaviors. Conclusions The results highlight the relevance of socioeconomic factors for suicidal behaviors in South Korea. Governmental social spending must be increased and redirected more efficiently so that the economically most vulnerable groups are financially protected and income inequality does not widen. Furthermore, comprehensive prevention strategies at the community level are needed. Future research needs to focus on identifying vulnerable groups for whom the effects of low socioeconomic status may have particularly serious consequences with regard to suicidal behaviors. Supplementary Information The online version contains supplementary material available at 10.1186/s12889-022-12498-1.


Authors
Measure of SI Findings/Strength of association Examined factor: Employment/Occupation (Ahn et al. 2019) Single question: "Have you ever thought of killing yourself during the past year?" -"Unskilled manual workers" highest OR for SI in both genders a gender specific differences: higher OR for female "clerks" and "Service and sales workers" a (Chin et al. 2011) Single  • Females: NS -changes in job status(all): temporary job OR 1.46 (CI=1.12-1.90) a3 ,unemployed OR 1.58 (CI=1.21 -2.06) a3 , daily job NS, self employed NS -gender specific differences in change in job status: change from fulltime job to temporary job, daily job or unemployed only significant in males, for females NS (Kim and Yoon 2018) Single question about whether the respondent contemplated suicide in the past 12 months -Young adults (18 -35 years): out of labor force vs. waged employment OR 2.99 OR (CI=1.08-8.31), unemployment NS, selfemployment NS -Middle-aged adults (36-55 years): NS in all categories (Kim et al. 2017) Single question whether they had seriously -low job security = highest HR for SI Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001 -no job significantly associated with SI, analyzed via x 2 -test -Having no job affects AS through depression and SI.

Examined factor: Income (Hong et al. 2011)
Dichotomous question whether the respondent had ever attempted suicide in the past 12 months -pro rich inequalities in suicide attempts (the poorer, the more suicide attempts) -Income-related inequalities increased strongly between 2005 -2007 Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001 Gender-and age-specific analysis -suicide rate generally high among individuals in low-income quintiles, people in medical aid group most likely to commit suicide -self-employed at higher risks of committing suicide than wage earners in the same income class -effect of income on suicide is pronounced for males -Effect of income on suicide only marginally influenced by age Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001 Gender-specific analysis -individuals with lower levels of educational attainment exhibited higher rates of suicide -differences in suicide mortality between educational groups grew larger across consecutive years -inequalities were more salient in men than in women (Lim et al. 2015) Gender-and age specific analysis -RR's of college and high school were particularly large, difference between high school and middle school graduate or less was attenuated -increase in inequalities over time (=bigger differences in suicide deaths between educational groups over time) -college education linked to a smaller likelihood respectively risk of committing suicide -mortality rates for suicide in younger women with the lowest education level increased dramatically from 1995 to 2010 Examined factor: Socioeconomic Place of residency (Chan et al. 2015) Gender-and age specific analysis -inverse relationship between the degree of urbanicity and regional suicide rates, more pronounced among women -Increases over time in the suicide rates among youth and working-age adults were greater in large urban centers and in rural regions. Gender-specific analysis Age-standardized suicide rates (100,000) between 1995 -2005 highest in rural areas followed by urban and metropolitan areas -consistently positive association between district deprivation and suicides over time Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001 Evaluation of lifetime prevalence of risk factors for suicidal ideation and suicide attempts -high prevalence of lifetime suicidal behavior in Jeollabuk-do Province is higher than in Korea as a whole -most significant risk factors were found to be social support, family disharmony, anger, depression and low self-esteem (Chan et al. 2015) Exploration of geospatial trends of suicide in South Korea from 1992 to 2012 -Suicide rates increased during the study period -Inverse relationship between the degree of urbanicity and regional suicide rates.
-Suicide in South Korea as a dynamic process, regional variation linked to migration and urbanization (Chin et al. 2011) Investigation of suicidal ideation and its associated factors by sex in adults -Less SI among those who were married, had higher educational attainment, were nonsmokers, were under low stress and had good self-rated health -Effects of stress and depression on SI is higher for men, effects of sleep and subjective body image is higher for women (Cho et al. 2007) Examination of possible socioeconomic mortality differences in adolescents -Clear socioeconomic gradient for all-cause mortality in boys, less evident pattern in girls -transport accidents were the most common cause of death among adolescents -marginally significant socioeconomic difference in suicide mortality for 10-19-year-olds (Choi et al. 2019) Identification of the impact of poverty on suicide risk in older adults -suicide and poverty rates for older adults is highest in South Korea among OECD-Countries -Elderly people living in poverty had a higher risk of suicide than those classed as having a high income -Hazard ratio for suicide risk by income class differed significantly by sex and age group; males and young-old adults aged 60-74 in the poverty group with significantly higher suicide rates compared with females and adults aged 75 or older (Fukai et al. 2020) Determination multidimensional (physical, mental, social, spiritual) health status in relation to the presence of depressive symptoms and suicidal ideation in the general population -Physical, mental, social, and spiritual self-rated health statuses were all found to be associated with an individual's predisposition to depression and suicidal ideation -Spiritual health status affected depression only in men and mental health status affected depression only in women; social health status in men and mental health status in women were associated with suicidal ideation  -year period (1998-2007) -persistent pro-rich inequality in depression, suicidal ideation and suicide attempts over the past decade = People with higher incomes were less likely to be affected -inequalities doubled over the time-period of 10 years.
-need for expanded social protection policies for the less privileged in the population (Jeong and Chun 2019) Identification of individual and regional characteristics that influence suicidal ideation among the elderly population -age, frequency of communication with friends, religious activity, social activity, leisure activity, trust in neighbors, subjective stress level, depressive symptoms, and subjective health status were significantly associated with suicidal ideation -the lower the regional deprivation level, the higher the suicidal ideation odds ratio -most significant effects found in rural areas. (Jo et al. 2015) Investigation of the association between in-school students' parttime work and 1-year suicide attempts -middle school students' part-time work may increase suicide attempts -no association between part-time work and suicide attempts among high school students -Of 58,900 included participants, 9,265 experienced SI -Participants with food insecurity, low household income, and those who lived alone showed a significant increase in SI and were more likely to experience it (Kang et al. 2015) Determination of the prevalence, pattern, and predictors of SI, and suicide attempts of adolescents in the past 12 months -The prevalence rates for SI and suicide attempts were 19.1% and 4.9%, respectively, among adolescents -Being female, having poor SES and poor perceived academic performance, subjective feelings of depression, cigarette smoking, alcohol use, perceived general medical health, and experience of any sexual intercourse increased the risk for SI and suicide attempts -Suicide attempt rates among Korean female adolescents were highest at age 13 years (Kang et al. 2014) Assess the prevalence, incidence, and persistence of SI of elderly adults (≥65y) and to investigate the psychosocial factors associated with these -Prevalence of SI at baseline, incidence at follow-up, and persistence were 11.5%, 9.6%, and 36.5% -SI at baseline is independently associated with no current employment, lower monthly income, stressful life events, more severe pain, presence of disability, depressive symptoms, and smoking -Incidence of SI is independently predicted by baseline status of unmarried, deficit in social support, severe pain, presence of depressive symptoms, and smoking Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001

Author Aim Main results
-Persistent SI is independently predicted by stressful life events and depressive symptoms at baseline (Kang et al. 2017) Determination of the relationship between shift work and mental health, particularly insomnia, depression, and SI among electronics production workers -low education and unemployed young adults had significantly higher rates of attempts -lowest income level associated with significantly higher rates of attempts in only young adult women -Association between SI and suicide attempts was attenuated among those with the lowest and highest income, enhanced among other groups (Kim and Yoon 2018) Analysis of the extent to which socioeconomic attainments (education, household income, and employment status) are associated with risk of SI among adults -Young adults with 2-year colleges and higher, had a lower likelihood of SI -Middle-aged respondents with 4-year colleges and higher had an increased likelihood of SI -Individuals who were not employed had an increased risk of SI -The likelihood of SI, decreased with household income among middle-aged adults (Kim et al. 2006) Investigation of the effects of social class on suicide risk -proportions and OR'S for suicide are higher in young people and for divorced subjects -OR's are higher for residents of rural areas, and people in lower social classes Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001  (Kim et al. 2018) Investigation the relationship between occupational stress and SI, focusing on gender differences among employees -All job stresses contributed to increased risk for SI in men and all job stresses except inadequate job control and organizational system were risk factors for SI in women -After adjustment for depression, only inadequate job control and lack of reward remained significant for SI in men, and only interpersonal conflict remained significant in women (Kim et al. 2019b) To examine how employment transitions between permanent employment, precarious employment, and unemployment are associated with SI among working employees and whether individuals who transition between these three states are more vulnerable than those who remain -Compared to the "permanent to permanent" group, individuals in the "permanent to precarious" (OR 1.74), "permanent to unemployment" (OR 1.97), "precarious to precarious" (OR 1.86), and the "precarious to unemployment" (OR 1.43) groups had higher odds of SI -magnitude of such odds was generally higher than that of individuals at annual unemployment or precarious states (Kim et al. 2019c) Analysis of suicide and suicide by pesticide rates among workers from 2003-2017 and their associations with occupational groups, and the influence of three major economic indices on these factors -Age, gender, and time period effects, as well as the effects of macroeconomic indicators, differ by occupational group for suicide and suicide by pesticide → Workers in agriculture, forestry, and fishing showed higher suicide rates than the general population -Suicides from pesticides largely declined in all occupational groups after 2012, when the pesticide ban policy was implemented -real gross domestic product had a positive relationship with suicides overall only in the last five-year period examined in this study; the unemployment rate had a positive relationship across the board (Ko et al. 2014) Identification of the relationship of SEP and suicide attempts using both the subjective and objective measures of SEP in the Korean adolescents -higher risks of suicide attempts were found in the Korean adolescents with the low level of both the subjective and objective SEP in the middle school and with the low subjective SEP and high objective SEP in the high school -subjective SEP is far more important in the high school students than objective SEP Comparison of suicidal behaviors between the employed and standardly employed workers -Persons self-employed, compared with standard workers, were more likely to report SI in both small and middle to large businesses -likelihood for suicide attempt was only significant in persons self-employed in small businesses (Min et al. 2015) Comparison of SI and suicide attempts in precarious workers and their non-precarious counterparts -Employees with precarious work were more likely to exhibit SI and suicide attempts than employees with non-precarious work -Odds were largely attenuated when adjusting for income, education, and depressive feelings → indicates a strong association between suicidal risk and socioeconomic indicators and/or depression (Moon and Park 2012) Investigation of the influence of socio-demographic status on SI among middle-aged adults -SI most prevalent among middle-aged women who engaged in manual labor -Absence of a spouse is significantly associated with suicidal ideation in middle-aged men (Park and Lee 2015) Investigation of the age-specific contributions of sociodemographic factors, health status, and health behaviors to SI -factors associated with SI differ by age -Perceived bad health, stress, and depression had a significant influence on SI in all age groups; depression was strongest predictor among all ages -Disease and marital status did not have a significant effect on SI among those 45-64 years old, education had no effect among those 65 years or older Abbreviations: NS = not significant. a = multivariate/multiple regression (adjusted). b = confirmed in univariate analysis, no multivariate/multiple analyses (unadjusted). c = only confirmed in univariate analysis, NS in multivariate/multiple analysis; 1 = p < 0.05, 2 = p < 0.01, 3 = p < 0.001

Author Aim
Main results (Park et al. 2016a) Investigation of the risk factors of SI and their PAF (population attributable fraction, measure to quantify the attributable burden of exposure on a population) of the elderly population -Advanced age and economic status (more than $25,700 per year) were significantly associated with decreased SI risk -Factors that were associated with increased risk were poor social support, sleep problems, chronic illness, poor subjective health, functional impairment, and depression → depression as biggest risk factor for SI (Park and Lee 2016) Examination of the factors that affect suicide attempts of adolescents in multicultural families -risk factors: Female gender, residence in large cities (compared with in rural areas), living with relatives/alone/with friends/in a dormitory or living in a facility (compared with living with family), high and low socio-economic status (compared with a middle-level), high and low academic performance (compared with a middle level), severe perceived stress (compared with non-severe stress), conflicts with a teacher (compared with conflicts with parent), foreign-father/-parent families (compared with foreign-mother family) ( -factors of gender, economic status, the overall amount of behavior problems, the tendency to internalizing and externalizing problems, somatic problems, thought problems, delinquent behavior, and aggressive behavior were independent predictors of adolescent SI and selfharm behavior (Sohn et al. 2014) Investigation of the difference in medical care expenditure between those who committed suicide from non-illness-related causes and their age-and gender-matched controls -medical care expenditures increased in only the last 3 months prior to suicide in the adolescent group -significant associations with being a suicide completer and having a rural residence, low socioeconomic status, and high medical care expenditure -After stratification, a significant positive association with medical care expenditures and being a suicide completer was found in the adolescent and young adult groups, no significant results were found in the elderly groups for both men and women