Van werk naar pensioen: veranderingen in fysieke activiteit, sedentair gedrag en slaap

Physical activity (PA), sedentary behaviour (SB) and sleep are related to multiple health outcomes, including mortality, cardiometabolic health, certain types of cancer, mental health, adiposity, etc. However, about one fourth of the European adult population is not sufficiently physically active, too sedentary and/or does not sleep enough. This places a substantial burden on the health care system and on quality of life, especially taking into account the ageing of the population. Therefore, it is a major society concern to optimise these lifestyle behaviours in the population. The transition from work to retirement might provide an excellent opportunity to improve PA, SB and sleep levels, as increased free time and the necessity to re-invent daily activities can facilitate these changes. Indeed, research has shown that most retirees increase their sleep time and decrease their sedentary time. However, although there is often more time and more opportunity, PA levels seem to decrease when retiring. Unfortunately, these previous studies have mainly used self-reported measures of the behaviours, which are prone to recall bias and social desirability. Also, little is known about predictors of favourable and unfavourable changes in these behaviours during the retirement transition. Therefore, the proximal aim of the present PhD project is to extend the scientific knowledge concerning changes in device-measured PA, SB and sleep during the retirement transition, including the role of socio-economic position (SEP) and social factors as predictors of these changes. The distal aim is to inform the development of future lifestyle interventions targeted at adults undergoing the retirement transition. From the literature study in paper 1 we concluded that the changes in PA and SB seemed to be more favourable for higher SEP retirees, except for changes in PA related to household and caregiving. For the latter PA domain, there seemed to be no difference between the SEP groups. Further, paper 2 describes the results of a longitudinal observational study (n = 95) in which we measured PA, SB and sleep with accelerometers in a convenience sample of adults in Flanders, Belgium. The measurement took place once before retirement and three times after retirement, until one year post-retirement. The data were analysed using compositional data analysis in linear mixed models. Similar to previous findings, we identified an increase in sleep time and a decrease in sedentary time. However, the amount of PA remained stable during the retirement transition in our study, contrasting with previous studies where it decreased. Nonetheless, there appeared to be a trend towards more favourable changes for higher SEP retirees compared to lower SEP retirees. This trend was statistically significant when SEP was indicated by income, but not when education or occupation were used as SEP indicators. In paper 3, based on data from the same sample as in paper 2, we found that the social factors 'group memberships' and 'satisfaction with social relationships' were not predictive for changes in PA, SB and sleep during the retirement transition. However, there were cross-sectional associations between these social factors and the behaviours at pre-retirement and at 12 months post-retirement. Only one previous study has investigated the association between group memberships and PA during the retirement transition. This study did find a longitudinal association. We conclude that social factors might become predictors of PA, SB and/or sleep one year after retirement, but more studies should investigate this before firm conclusions can be drawn. Finally, our qualitative interview study (paper 4) identified changes in barriers and facilitators for engaging in PA during the retirement transition. For instance, under the theme 'changes in time availability and time structure', the facilitator 'time availability' increased, while the barrier 'procrastination of PA' also increased in retirement. Other main themes, inductively identified from the interviews, were 'emotional/mental changes' and 'social changes'. In light of these findings, we proposed some potential implications for future interventions. A limitation of our studies is the use of a convenience sample, which introduces the potential for selection and recruitment bias. As such, caution is warranted when generalizing these findings beyond the studied population. Drawing from our findings, we suggest that future studies should continue to investigate predictors of changes in PA, SB and sleep during the retirement transition. These can be unmodifiable predictors to identify at-risk groups for unfavourable changes, such as SEP, or modifiable predictors, such as social factors. We found that retirees with lower SEP seem to be an important at-risk group for unfavourable changes. Therefore, we also suggest to develop interventions focusing on this target group.