Mental wellbeing is increasingly recognized as a major global concern, with widespread claims that we are living through a ‘mental health crisis’ or even an ‘epidemic of depression.’
These narratives attribute worsening mental health to broad societal changes such as increased social media use, economic insecurity, and rising individualism. However, empirical evidence supporting these claims remains inconclusive and often suffers from methodological limitations. Many studies focus on a limited set of countries, conflate diagnostic rates with actual mental health outcomes, or fail to disentangle the effects of age, historical period, and birth cohort.
This dissertation addresses these gaps by applying advanced age-period-cohort (APC) modeling to high-quality data around the world to investigate recent trends in mental health and their structural determinants across diverse national contexts. The dissertation consists of one theoretical background chapter and four empirical chapters organized into two overarching research aims: (1) to map patterns of change in mental health across the life course and between cohorts; and (2) to interpret the social and structural factors driving these changes.
The first empirical chapter critically compares classical and contemporary methods for APC analysis using data from the Korean Welfare Panel Study (KOWEPS). By applying multiple APC techniques to the same dataset, it evaluates whether different approaches converge on similar conclusions about mental health trends. This methodological benchmark clarifies the strengths and limitations of existing APC models and provides a foundation grounding the results of subsequent empirical chapters.
The second empirical chapter turns to the German case, using data from the German Socioeconomic Panel (SOEP) and the German Health Update (GEDA) to investigate whether increases in diagnosed depression reflect broader declines in mental wellbeing. The findings reveal divergent trajectories: while diagnoses have increased, general mental health shows improvements followed by recent declines, while prevalence of distressed mood has modestly improved. These results suggest that rising diagnoses may reflect changes in stigma, medicalization, or treatment-seeking behavior rather than a true deterioration in underlying mental health.
The third empirical chapter presents a global analysis of aggregate mental health trends across cohorts and periods, drawing on longitudinal and cross-sectional datasets from 24 countries around the world (N ≈ 4.2 million observations). It shows that while mental health has declined among recent cohorts in many traditional high-income countries, it has improved in several emerging economies, highlighting the importance of national social, economic, and institutional contexts.
The fourth empirical chapter investigates age patterns in mental wellbeing across countries using the same datasets. Contrary to assumptions of a universal midlife low or U-shape, the results reveal considerable cross-national variation in age gradients. While some countries exhibit midlife peaks in distress, others show flat or even inverted patterns. These findings suggest that life course patterns in mental health are shaped by institutional and cultural contexts, and cannot be assumed to follow a consistent trajectory.
Taken together, this dissertation argues that contemporary changes in mental health are heterogeneous and context-dependent. Claims of a global mental health crisis overlook the complex ways in which structural, cultural, and institutional factors shape mental wellbeing across time and place.
Jos van Leeuwen is a PhD researcher in Social and Political Sciences at the European University Institute. His work examines patterns of mental wellbeing and mental distress in modern societies, with particular attention to how large-scale social change is reflected in life-course trajectories. Drawing on longitudinal and repeated cross-sectional data from more than 20 countries, he applies advanced age–period–cohort and life-course methods to analyse cross-national variation and long-term trends. His research has been published in Population and Development Review and presented at international conferences. He holds Master’s degrees in sociology, clinical psychology, and philosophy. Jos has joined the EUI as a research associate for a new project focusing on the micro-dynamics of life events and wellbeing, and since January 2026 he has started a new job as lecturer in quantitative methods at Vrije Universiteit Amsterdam.
Register