Physical activity and mental health in middle-aged and older adults: A population-based perspective

Research output: Types of ThesisDoctoral ThesisInternal

61 Downloads (Pure)


It is well-known that physical activity is beneficial for health. Besides the benefits for physical health, more physical activity has also been related to better mental health, e.g., less feelings of depression or anxiety. As physical activity is just one part of the ‘24-hour movement behaviors’, i.e., physical activity, sedentary behavior and sleep, studies that accurately take into account the 24-hour time window of these behaviors are warranted. Within this thesis, I aimed to gain insight into the association between physical activity and mental health among middle-aged and older adults from the general population. Additionally, I explored some of the potential mechanisms underlying these associations.

In Chapter 2, I discussed physical activity, mental health and their associations. In Chapter 2.1, determinants and trajectories of physical activity during the Covid-19 pandemic were investigated. We determined that over a 6-week period, from April 2020 onwards, at least 59% of participants of the population-based Rotterdam Study did not meet the World Health Organization (WHO) guidelines on physical activity. We observed five distinct trajectories of physical activity over this 6-week period at the beginning of the pandemic, of which four were steady over time and one was increasing in levels of activity. Analyses on determinants of these trajectories showed that participants in the ‘steadily low’ trajectories were more often older and lower educated, and less often retired. Also, they were reporting poorer physical health, more depressive symptoms, a less healthy diet, more smoking and lower alcohol use. In Chapter 2.2, we investigated mental health symptoms among middle-aged and older adults from the general population. Using unsupervised hierarchical clustering analyses, we identified three clusters of psychiatric symptoms across clinical diagnoses, which represented ‘Mixed’ symptoms, ‘Depressed affect and nervousness’, and ‘Troubled sleep and interpersonal problems’. We also identified four groups of participants within our population that potentially have similarities in their pathophysiology even though their clinical diagnoses might be different or even lacking. Chapter 2.3 reported on the associations between 24-hour movement behaviors and mental health, i.e., depressive and anxiety symptoms. Using compositional isotemporal substitution analyses, we found that more moderate-to-vigorous physical activity was associated with less depressive symptoms when replacing sedentary behavior or sleep. For anxiety symptoms, we did not find an association with any of the substitutions of 24-hour movement behaviors. The same type of analyses were applied to sleep quality in Chapter 2.4. We found that more sedentary behavior or light physical activity was associated with a poorer sleep quality when replacing sleep, but not when replacing any other activities. We did not find an association between physical activity and sleep quality. Altogether, the results of this chapter emphasize the need for novel and overarching approaches to study 24-hour movement behaviors, e.g., compositional data approaches, and mental health, e.g., transdiagnostic approaches.

I explored some of the associations with potential underlying mechanisms in Chapter 3. We investigated the potential bidirectional association between physical activity and brain structure in Chapter 3.1. Using a cross-lagged panel model approach, we concluded that physical activity at baseline was not associated with a healthier brain structure over time. On the contrary, we did identify an association of baseline brain structure with physical activity over time, i.e., participants with a healthier brain structure at baseline remained more physically active over time. In Chapter 3.2, we investigated the role of polygenic risk for multiple psychiatric phenotypes in depression risk. We concluded that a higher polygenic risk score for depression was associated with both higher cross-sectionally measured depressive symptoms as well as longitudinally measured depressive events. These associations were stronger for more severe phenotypes of depression, e.g., major depressive disorder. In Chapter 3.3, we investigated the longitudinal association of markers of innate and adaptive immunity with depression risk in the general population. We found that markers of innate immunity, i.e., granulocytes and platelets, were related to more depressive symptoms in cross-sectional analyses. Consistently, the relative measures of markers of innate and adaptive immunity, i.e., granulocyte-to-lymphocyte ratio, platelet-to-lymphocyte ratio and systemic inflammation index, were associated with more depressive symptoms. However, we did not identify a longitudinal association of innate and adaptive immunity with incident depressive events. Based on these studies, genetic factors are confirmed to be at play in the development of depression and the potential protective effect of physical activity should be studied among participants across different levels of genetic susceptibility. Based on our studies, brain health and the immune system could not be identified as relevant mechanisms on the pathway from physical activity to mental health, however, studies with longer follow-up time and more repeated measures are needed.

Lastly, Chapter 4 provided a general discussion on the main findings of this thesis, methodological considerations, implications for public health and recommendations for future research.
Original languageEnglish
Awarding Institution
  • Erasmus University Rotterdam
  • Ikram, Arfan, Supervisor
  • Voortman, Trudy, Supervisor
  • Luik, Annemarie, Co-supervisor
Award date15 Mar 2023
Place of PublicationRotterdam
Print ISBNs978-94-6419-712-9
Publication statusPublished - 15 Mar 2023


Dive into the research topics of 'Physical activity and mental health in middle-aged and older adults: A population-based perspective'. Together they form a unique fingerprint.

Cite this