Job Insecurity and Proxy Suicide Risk in UK Adults: A Longitudinal Analysis of Social Determinants

IntroductionJob insecurity is a known risk factor for suicide. The extent to which perceived job insecurity impacts suicide risk is unclear. This study examined whether perceived job insecurity is associated with suicide risk, operationalised via a proxy score derived from variables empirically linked to suicide risk, and whether this association is confounded by broader social determinants. MethodData were drawn from Waves 10, 12, and 14 of the UK Household Longitudinal Study, comprising 33,364 observations from 20,566 participants between 18 and 66 years old. A proxy for suicide risk was constructed using the following five variables: subjective wellbeing (GHQ-12), loneliness, general health, life satisfaction, and subjective financial situation. Job insecurity was measured via self-reported likelihood of job loss. Generalised Estimating Equations were used to model associations, adjusting for sex, ethnicity, marital status, education, and income. ResultsUnadjusted models showed a significant association between perceived job insecurity and suicide risk. However, in a multivariable model controlling for social determinants including low income, being divorced or separated, and lower education, this association was no longer significant. On the contrary, being male, married or cohabiting, or holding a university degree were associated with lower odds of suicide risk. ConclusionAlthough perceived job insecurity initially appeared to be associated with suicide risk, its effect is largely explained by social determinants. These findings highlight the importance of integrating social determinants into suicide prevention strategies. Targeting financial hardship, relationship instability, and educational disadvantage may be more effective than addressing job insecurity alone. The use of a validated proxy score offers a feasible and scalable approach for assessing suicide risk in large-scale datasets lacking direct measures.