From midlife to early old age: health trajectories associated with retirement

Abstract

Background: Previous studies report contradictory findings regarding health effects of retirement. This study examines longitudinally the associations of retirement with mental health and physical functioning.

Methods: The participants were 7584 civil servants from the Whitehall II cohort study aged 39-64 years at baseline and 54-76 years at the last follow-up. Self-reported mental health and physical functioning were assessed using the Short Form Medical Outcomes Survey questionnaire, and the scales were scored as T-scores (mean [SD] = 50 [10]). Retirement status and health were assessed with 6 repeated measurements over a 15-year period.

Results: The associations between retirement and health were dependent on age at retirement, reason for retirement, and length of time spent in retirement. Compared with continued employment, statutory retirement at age 60 and early voluntary retirement, respectively, were associated with 2.2 (95% confidence interval = 1.7 to 2.8) and 2.2 (1.7 to 2.7) points higher mental health and with 1.0 (0.6 to 1.5) and 1.1 (0.8 to 1.4) points higher physical functioning. Retirement due to ill health was associated with poorer mental health (-0.7 points [-1.62 to 0.2]) and physical functioning (-4.5 points [-5.1 to -3.9]). Within-subject analyses suggested a causal interpretation for statutory and voluntary retirement, but health selection for retirement due to ill health.

Conclusions: Longitudinal analyses of repeat data suggest that health status improves after statutory and voluntarily retirement, although the improvement seems to attenuate over time. By contrast, the association between retirement due to ill health and subsequent poor health seems to reflect selection rather than causation.

Figure 1

Average scores (and 95% confidence intervals) of mental health and physical functioning by age.

Figure 2

Trajectories of mental health and physical functioning as a function of retirement status and age. Mental health and physical functioning are scored as T-scores (mean [SD] = 50 [10]). The vertical lines are 95% confidence intervals. See eTable 1 (http://links.lww.com) for the percentages of observations belonging to the various retirement status groups by age group. See eTable 2 (http://links.lww.com) for statistical details of the multilevel model.

Figure 3

Associations between length of retirement (number of follow-up phases being retired) and differences in mental health and physical functioning. The values are within-participants regression coefficients from separate multilevel models for each reason for retirement and indicate the within-participant difference in health associated with retirement compared to not being retired. Mental health and physical functioning are scored as T-scores (mean [SD] = 50 [10]). The vertical lines are 95% confidence intervals. See eTable 3 (http://links.lww.com) for statistical details.