Retirement timing, living arrangements and mental health: a multi-method investigation using Survey of Health, Ageing and Retirement in Europe.

Abstract

Aims: The relationship between retirement and mental health remains unclear, with mixed evidence likely driven by methodological challenges and heterogeneity across population subgroups. This study explores the relationship between retirement, time since retirement and mental health, focusing on differences between those living alone and cohabiting.

Methods: We used data from 10,698 participants (48,815 observations) in the Survey of Health, Ageing and Retirement in Europe, waves 1-8, spanning 20 countries and 16 years. Participants were included if they were employed at baseline and retired during follow-up. We applied a triangulated approach using (1) multilevel linear spline models to examine trajectories and (2) fixed-effects instrumental variable analyses (IV) using statutory retirement age as an instrument to establish causal effects. The outcome was depressive symptoms measured using the EURO-D scale (higher scores indicate more depressive symptoms).

Results: Multilevel spline models showed minimal change in depressive symptoms before retirement. At retirement, a small decline in EURO-D scores was indicated (-0.07 points per year [95% CI: -0.13 to -0.0003]), followed by an increase starting two years post-retirement (0.09 per year, 95% CI: 0.02 to 0.16). IV analyses showed that retirement was associated with fewer depressive symptoms (-0.26, 95% CI: -0.45 to -0.07). Individuals living alone reported higher depressive symptoms, but cohabitation status did not moderate the relationship between retirement and mental health.

Conclusions: Retirement is associated with a short-term reduction in depressive symptoms, but the effect diminishes over time. While individuals living alone consistently report higher depressive symptoms, cohabitation status does not moderate the retirement-mental health relationship.