Uncertain Jobs, Unequal Odds: South Korea’s Pandemic Findings on Suicide Attempts

The finding is that job loss and deteriorating employment conditions were both associated with higher odds of reported suicide attempts, with a stronger association for job loss.

Losing a job can unsettle much more than a household budget, disrupting daily routines, social connections and a person's sense of security. During the COVID-19 pandemic, some people lost their employment completely, and others faced less favourable work arrangements, raising questions about how these different experiences were connected to mental health.

The research paper, "Association Between Changes in the Employment Environment and Suicide Attempts During the COVID-19 Pandemic", published in Swiss Archives of Neurology, Psychiatry and Psychotherapy, examines those differences using South Korean survey data. Its central finding is that job loss and deteriorating employment conditions were both associated with higher odds of reported suicide attempts, with a stronger association for job loss.

Researchers analysed 172,970 adults who were economically active before January 2020, separating employment experiences into four categories. The findings offer a closer look at employment disruption during the pandemic, but the survey cannot establish whether a job change happened before an attempt or whether employment changes caused the outcome.

Four employment experiences captured in a national survey

The analysis used the 2021 Korea Community Health Survey, drawing on interviews conducted between August and October 2021 with adults aged 19 and older. From an original survey population of 229,242, researchers selected respondents with valid employment-change and suicide-attempt responses, then excluded 1,008 people with missing information on other characteristics.

Participants described employment changes compared with the period before January 2020. The categories covered no change, job loss through dismissal or business closure, deteriorated conditions such as a move to a less favourable workplace or part-time work, and improved conditions. People who had not been economically active before the pandemic were excluded.

The final sample included 127,738 people reporting no change, 7,150 reporting job loss, 34,362 reporting deteriorated conditions and 3,720 reporting improvements. Across these groups, 520 respondents reported an attempt during the preceding year, representing a survey-weighted prevalence of 0.32%. The outcome measured reported attempts, rather than thoughts about suicide or deaths.

Job loss showed the strongest employment-related association

The survey-weighted proportion reporting an attempt was 1.21% among people who had lost their jobs, compared with 0.43% among those reporting worsening conditions, 0.23% among those reporting no change and 0.22% among those reporting improvements. These percentages describe the observed groups before accounting for differences in their other characteristics.

Researchers used survey weights and adjusted their comparisons for age, sex, marital status, education, region, household income, smoking, alcohol consumption, self-rated health, physical activity, hypertension and diabetes. Job loss remained associated with approximately 2.55 times the odds of a reported attempt compared with unchanged employment, and deteriorated conditions were associated with around 1.44 times the odds.

A direct comparison between the two adverse employment categories found that job loss was associated with approximately 1.78 times the odds seen with deteriorated conditions. The findings support examining these experiences separately, since combining everyone affected by employment disruption into one category could obscure differences between losing work and remaining in less favourable employment.

Improved conditions had an adjusted odds ratio of 0.68 compared with no change, but the estimate was too uncertain to establish a protective association. Only 10 respondents in that category reported an attempt, limiting the precision of the result; better employment circumstances cannot be described as preventing attempts on the basis of these data.

Age and income shaped the patterns, with limits on interpretation

Adults aged 19–29 had around 10.58 times the odds of reporting an attempt compared with adults aged 70 or older, women had around 2.17 times the odds compared with men, and people in the lowest household-income group had around 2.90 times the odds compared with the highest-income group.

Other characteristics associated with higher adjusted odds included smoking, poorer self-rated health, lower educational attainment and being unmarried, widowed or divorced. People rating their health as poor had around 8.45 times the odds compared with those rating it as good. These findings describe associations within the analysed population and do not establish that any single characteristic caused an attempt.

Separate tests examined whether the relationship between employment change and reported attempts varied across sex, age and income groups. The patterns differed by age and income, but there was no clear evidence of a difference by sex. Women's higher adjusted odds in the main analysis therefore do not demonstrate a stronger response to employment disruption.

The age comparison grouped people aged 64 or younger against those aged 65 or older, and the income comparison combined the two lower categories and the two higher categories. The reported tests do not establish that younger, older, poorer or wealthier people were categorically more susceptible to a particular employment change.

Why the findings call for closer follow-up

Financial strain, uncertainty, disrupted routines and reduced social connection explain the employment-related patterns. The researchers did not directly test these pathways, and the employment categories did not capture the timing or severity of changes, individual wage losses or detailed working conditions. Depression, stress and self-esteem were not tested as explanations for the associations.

The timing problem is essential because employment changes were measured relative to before January 2020, and attempts covered the year preceding interviews in 2021. An attempt could have happened before or after an employment change, and pre-existing mental illness, occupation or household shocks could have influenced both experiences.

Self-reported answers may be affected by imperfect recall or reluctance to disclose an attempt, and the survey did not clinically verify intent or capture fatal outcomes. Excluding respondents with missing information could also have affected the results. Findings from this selected South Korean population in 2021 cannot automatically be extended to other countries, later periods or people outside the employment-defined sample.