Retirement, mortality and self -reported health: An application of life course epidemiology

Research literature examining retirement as a risk factor for mortality is both sparse and heterogeneous. Retirement classifications and definitions are poorly developed and lack standardization. A systematic review of the literature used a modified Cochrane and ‘best evidence synthesis’ approach to answer the question Does the research literature support the view that type of retirement is a risk factor for mortality?” A literature search identified 988 studies related to retirement and mortality. A total of 976 studies were excluded based on content and quality which left 12 studies for review, data extraction and evidence synthesis. By combining all types of retirement together, there is a sufficient level of evidence for retirement as a risk factor for mortality. However, when stratified by retirement type (on-time, early and health-related), the results are much different: (1) a mixed level of evidence for health-related retirement as a risk factor for mortality; (2) a mixed level of evidence against early retirement as a risk factor for mortality; and (3) a mixed level of evidence neither for nor against on-time retirement as a risk factor for mortality. ^ The belief that retirement leads to early death may be an artifact of individuals retiring due to poor health and subsequently dying as a result of their poor health status and not actually dying due to the retirement transition. Similarly, an individual that is healthy and retires early may lead a longer life—not because they retired, but because they are healthy. ^ The current peer-reviewed literature provides very few high quality studies on the effect of retirement on mortality. There is a critical need for more research in this area due to the large number of current retired individuals and the rapid growth of retired individuals in the very near future. Especially important is our understanding of the role of health on retirement and retirement on health. ^ The relationship between the type of retirement and health is not well documented. Moen’s life course approach to retirement and health may was used to help clarify the relationship between retirement and health by accounting for the circumstances surrounding the retirement transition. ^ We identified heads of households and spouses, ages 18 to 60, from the Panel Study of Income Dynamics (PSID) that reported working or transitioned from work to initial retirement or work to initial permanent disability between 1984 and 2005 (n = 18,757). Pre-retirement self-reported health (excellent, good and poor) was the exposure of interest. Retirement status (working = no, retired = yes) was the primary outcome of interest. Additionally, retirement was stratified into three retirement types: (1) early (≤ age 59 without disability); (2) on-time (≥ age 60 without disability); and (3) health-related (permanent disability at any age). Retirees that reported ‘good’ health pre-retirement were 72% more likely to be retired than those retirees that reported ‘excellent’ health. Retirees that reported poor’ health pre-retirement were 4.7 times as likely to be retired when compared to those retirees that reported ‘excellent’ health. Among retirees, early retirees were 26% less likely to be in ‘good’ and 44% less likely to be in ‘poor’ health when compared to on-time and health-related retirees that reported ‘excellent’ pre-retirement health. On-time retirees were 25% more likely to be in ‘good’ health and less likely to be in ‘poor’ health when compared to early and health-related retirees that reported ‘excellent’ pre-retirement health. Health-related retirees were 51% more likely to report ‘good’ health and 4.1 times as likely to report ‘poor’ health when compared to early and on-time retirees that reported ‘excellent’ pre-retirement health. ^ Health status appears to drive the retirement transition. The belief that retirement leads to early death may be an artifact of individuals retiring due to poor health and subsequently dying as a result of their poor health status and not actually dying due to the retirement transition. Similarly, an individual that is healthy and retires early may lead a longer life—not because they retired, but because they are healthy. ^ The role of health on retirement and the impact of retirement on health is complex and cannot be assessed with simple exposure-disease models. Moen’s life course approach was used to facilitate a more complete explanation of the relationship between retirement and health. ^ We identified heads of households and spouses, ages 18 to 60, from the Panel Study of Income Dynamics (PSID) that transitioned from work to initial retirement or work to initial permanent disability between 1985 and 2003 ( n = 2,208). Type of retirement was the exposure of interest and was categorized as: early (≤ age 59 without disability); on-time (≥ age 60 without disability); and health-related (permanent disability at any age). Post-retirement health status (excellent/good and poor) was the primary outcome of interest. ^ Retirees in ‘poor’ post-retirement health were about 2.7 times as likely to be health-related retirees as early retirees (OR = 2.65, 95% CI = 1.89, 3.72). Additionally, retirees in ‘poor’ post-retirement health were only 16% more likely to be on-time retirees than early retirees (OR = 1.16, 95% CI = 0.84, 1.60). ^ Effectively managing occupational health and safety as well as chronic health conditions that are affected by work and non-work factors is necessary to limit the effect of negative health trajectories on an individual’s health in later life.