Elucidating the role of unemployment in complex social inequalities in mental health: An intersectional mediation analysis of the cross-sectional Spanish National Health Surveys.
Moreno-Llamas A, Bacigalupe A, et al. • Social science & medicine (1982) • 2026
Complex mental health inequalities seem to be underpinned by the dual processes of higher risks and worse consequences of unemployment, with unemployment completely mediating mental health inequalities in strata of men with manual occupations but only partially in strata of women.
Key Findings
Results
Women, migrants, and those in manual occupations had poorer mental health compared to non-manual native men in intersectional analysis.
Study used 2012 and 2017 cross-sectional Spanish National Health Surveys with n=22,383 adults aged 18-64 years
Response rates were 89.6% (2012) and 74.0% (2017)
Mental health was measured using the GHQ-12 and classified into good and poor
Eight intersectional strata were created by cross-classifying gender (men/women), social class (manual/non-manual), and migration status (born in Spain/migrant)
Non-manual native men served as the reference group for total effect comparisons
Results
Unemployment completely mediated mental health inequalities in strata of men with manual occupations but only partially mediated inequalities in strata of women.
Intersectional mediation analysis was used to decompose total effects into direct and indirect (mediated) components
Unemployment was classified as employed or unemployed
Complete mediation was found for manual occupation men, meaning the mental health disadvantage was fully explained by their higher unemployment
Partial mediation for women's strata indicates that factors beyond unemployment also contribute to their mental health inequalities
Results
Mental health inequality was explained by both higher unemployment prevalence (pure indirect effect) and greater vulnerability to unemployment (mediated interaction effect) in certain intersectional strata.
Native men and women in manual occupations showed mental health inequality explained by both higher unemployment prevalence and greater vulnerability to unemployment
Immigrant women in non-manual occupations also showed mental health inequality explained by both pathways
The 'pure indirect effect' reflects higher exposure to unemployment (higher prevalence)
The 'mediated interaction effect' reflects worse mental health consequences when unemployed (greater vulnerability)
These dual processes were compared against the reference group of non-manual native men
Methods
The study applied an intersectional mediation analysis framework to simultaneously examine the intertwining of gender, social class, and migration status as determinants of mental health inequalities.
Previous research had not considered the intertwinement of multiple social inequalities in this context
Eight intersectional strata were formed from binary categorizations of three social axes: gender, social class, and migration status
The approach allowed decomposition of total mental health inequalities into components attributable to unemployment as a mediator
Both the 2012 and 2017 survey waves were analyzed, covering a period of economic crisis and its aftermath in Spain
Conclusions
Women, migrants, and those in manual occupations face barriers in the labor market that increase their vulnerability to unemployment, which in turn contributes to mental health inequalities.
The study situated findings within the context of documented labor market disadvantages for these groups
Manual occupation workers were identified as more likely to be unemployed
Health policies are recommended to target strata of manual occupations given their higher unemployment likelihood
The findings highlight that both higher risks of unemployment and worse consequences of unemployment underpin complex mental health inequalities
Moreno-Llamas A, Bacigalupe A, Martín U, San Sebastián M, Gustafsson P. (2026). Elucidating the role of unemployment in complex social inequalities in mental health: An intersectional mediation analysis of the cross-sectional Spanish National Health Surveys.. Social science & medicine (1982). https://doi.org/10.1016/j.socscimed.2026.119041