Metabolic Health in older Adulthood – The Retirement Transition

Context – The Influence of Retirement on Health 
An individual’s working life course is usually ended by a work-to-retirement transition (Amick, McLeod, & Bültmann, 2016). Life course transitions, such as retirement, are complex as they involve significant changes in areas like social or psychological status (Amick et al., 2016) and therefore often influence health outcomes. Despite an abundance of research on the influence of retirement on perceived health outcomes, currently no conclusion can be made as findings are mixed (e.g. systematic literature reviews by Schaap, de Wind, Coenen, Proper, & Boot, 2018; van der Heide, van Rijn, Robroek, Burdorf, & Proper, 2013; Xue, Head, McMundd, & Heyn, 2019). Regarding mental health outcomes, results are more consistent and show a beneficial effect of retirement (Mein, Martikainen, Hemingway, Stansfeld, & Marmot, 2003; Oksanen et al., 2011; van der Heide et al., 2013).

Research about the effect of retirement on objectively measured physical health is scarce and inconclusive. Some studies showed a positive association between retiring and physical health, e.g. a reduction in blood pressure and waist circumference within retirees (Xue, Head, & McMunn, 2017), lower cortisol levels (Wong & Shobo, 2016), or lower morbidity levels especially regarding cardiovascular disease (Vallery-Masson, Poitrenaud, Burnat, & Lion, 1981) compared to a working control group. However, other studies found opposite effects like an increase in blood pressure in retirees (Ekerdt, Sparrow, Glynn, & Bossé, 1984; King & Xiang, 2017), an increase in cholesterol levels (Ekerdt et al., 1984), a higher risk for cardiovascular diseases as well as a higher likelihood to classify for the metabolic syndrome (Behncke, 2012) compared to workers. King and Xiang (2017) found no difference in weight, glucose, and cholesterol levels between retirees and a working control group.

These mixed and inconclusive findings might be explained by the different methodologies and research designs used in different studies, e.g. a discrepancy in ages of participants, use of a control group, the contextual setting, definitions and measurement of outcome, follow-up times, response rates or pre-retirement factors like occupational status (Matta et al., 2020). Further, in some studies panel data is used (e.g. Ekerdt et al., 1984; Vallery-Masson et al., 1981) whereas others analyse diary data (Wong & Shobo, 2016) or cross-sectional data (King & Xiang, 2017). More longitudinal research is needed to draw firm conclusions about the effect of the retirement transition on objective physical health. One highly important objective measure of physical health is the metabolic syndrome and its components.

The Metabolic Syndrome 
The metabolic syndrome consists of a cluster of hazardous factors which lead to a 2-fold risk of developing cardiovascular diseases and a 5-fold risk for type 2 diabetes mellitus (Alberti et al., 2009). For a metabolic syndrome classification, at least three of the following five risk factors need to be present: abdominal obesity, elevated blood pressure, reduced high-density lipoprotein cholesterol, elevated fasting glucose, and elevated triglycerides (Alberti et al., 2009). The estimated prevalence of the metabolic syndrome is 25% of the world population (Saklayen, 2018) and the number of affected people is rapidly growing (Cho & Koo, 2018). To the best of our knowledge, only one longitudinal study has investigated the influence of the retirement transition on the metabolic syndrome (Behncke, 2012). Behncke suggests that retirees are more likely to have the metabolic syndrome than comparable individuals who stay in the workforce.

Employment States and the Metabolic Syndrome 
Due to prolonged working lives, there is a need for more research on the influence of employment states on cardiometabolic risk factors among older employees (Soltysik et al., 2019). Employment status is derived from the type of contract a person has (OECD, 2003). Various employment states exist, some examples are standard full-time employment, part-time employment, unemployment, or self-employment. Some information is available regarding differences in metabolic syndrome prevalence among employment states, e.g. the prevalence is higher among non-standard employment types compared to standard full-time employment (Cho & Koo, 2018) and cardiovascular risk is greater among blue-collar workers compared to white-collar workers or unemployed people (Soltysik et al., 2019). However, information about metabolic syndrome incidence among employment states is scarce.

Retirement and the Metabolic Syndrome 
Retiring from paid work could influence the metabolic syndrome because major life-transitions are often accompanied by health status changes (e.g. Wilcox et al., 2003). In a recent systematic review, Xue et al. (2019) argue that retirement might have an effect on the development of cardiovascular disease via changes in health behaviours like low physical activity, smoking, higher alcohol consumption, and poor diet. These health behaviour changes may facilitate the development of the metabolic syndrome or its individual components, i.e. they might have a mediating role on the relation between retirement and the metabolic syndrome. Earlier studies have shown an increase in weight and waist circumference among retirees from active jobs (Nooyens et al., 2005) or a change to a more unhealthy diet after retirement (Hassen et al., 2017). Obesity is an important contributing factor to the metabolic syndrome (Alberti, 2005). Further, retirement was associated with a decreasing blood pressure (Xue, Head, & McMunn 2017) and increasing cholesterol levels (Ekerdt et al, 1984). Consequently, when analysing the effect of retirement on the metabolic syndrome it is important to include a composite measure of the syndrome as well as its individual components, as it is possible that only some of the components are affected.

Another reason for an effect of retiring on the metabolic syndrome is the individual work context that people retire from. Transitioning into retirement might improve metabolic syndrome measures in retirees that experience a relief effect from adverse working conditions. Studies have shown that employees with chronic work-related stress are at higher risk to classify for the metabolic syndrome (Chandola, Brunner, & Marmot, 2006; Watanabe et al., 2018). Consequently, pre-retirement work conditions may have a moderating role on the effect of retirement on the metabolic syndrome.

As stated above, it is currently not clear what the main effect of retiring is on the metabolic syndrome and its individual components. However, whether retiring has a positive or negative effect on health may depend on the individual circumstances that surround retirement. Therefore, it is relevant to include health behaviours as potential mediators, as well as a pre-retirement workplace characteristics as moderators into the analysis of the effect of the retirement transition on the metabolic syndrome.

Project Aims 
The overarching aim of this research project is to investigate objective health outcomes of the retirement transition. The goal of the first paper is to examine the prevalence and incidence of the metabolic syndrome among various employment states in older Lifelines participants, i.e. 40 years and older. Following this overview, in the second paper we will analyse the influence of retirement as a major life course transition on the metabolic syndrome, as well as its individual components. Further, health behaviours will be included in the model as potential mediators on the relation between retirement and the metabolic syndrome. In the third paper we will analyse pre-retirement workplace characteristics as moderators on the effect of retirement on the metabolic syndrome. Currently, there is no population-based longitudinal study that analyses changes of metabolic syndrome measures due to retirement by comparing pre- and post-retirement scores within individuals. Thus, this project will begin to fill this gap in the literature.

Using Lifelines data to answer our research questions is excellent for several reasons. Due to the longitudinal study design of Lifelines, it is possible to compare pre- and post-retirement levels of objectively measured components of the metabolic syndrome within participants as well as to compare these measures with a working control group to control for an effect on health simply due to aging. This is a great advantage over studies on objective cardiometabolic health outcomes that used a cross-sectional design (e.g. King & Xiang, 2017) or did not include within-person scores (Behncke, 2012). Further, the large sample size and extensive data collection (including biomaterials as well as questionnaire data) at various time-points are advantages over other biobanks. This design allows us to not only analyse a change in the metabolic syndrome pre- and post-retirement, but also include measures of health behaviours and working conditions. Lastly, Lifelines data is population-based and broadly representative for the population of the Northern Netherlands when adjusting for some differences in demographics (Klijs et al., 2015). Therefore, the results will largely be generalizable to the broader society.