Skip to main content

Socioeconomic Status and Health: Disentangling causal pathways in a life course perspective

Final Report Summary - SESANDHEALTH (Socioeconomic Status and Health: Disentangling causal pathways in a life course perspective)

Health differences according to socioeconomic status (SES) can be attributed to three causal mechanisms: SES affects health (social causation), health affects SES (health selection), and common background factors influence both SES and health (indirect selection). Our project examines these main causal pathways in order to determine their relative contribution of social causation and health selection to health inequalities, but we also study specific mechanisms that link SES and health over the life course, explore different measurements for SES and health, methods, and datasets. We use retrospective survey data from 10 European countries from SHARELIFE (n≈20,000), from England (SHARE, n≈5,000) and Finnish register data that cover an 11 percent sample of the Finnish population. The main causal analysis is done with structural equations models in a cross-lagged panel design, in order to determine the relative importance in terms of explanatory power of social causation and health selection in the life course from childhood to old age.
We find that both SES and health heavily depend on their prior status, but SES more than health. In the transition from childhood to working ages, social causation and health selection are of equally small strength. Turning to the second phase (transition from working ages to old age) causation increases while selection decreases which makes causation the dominant mechanism in older age. The contribution of common background factors remains difficult to assess, because their theoretical definition and empirical measurement still poses major problems. Our results show that both social causation and health selection are responsible for health inequalities, but their relative importance changes over age. Life course modelling can complement causal analysis by revealing interactions between the processes of SES and health and their contribution to health inequality. If both mechanisms contribute, also both can and should be used to tackle health inequalities, with a focus on certain age groups.
With additional preparatory research on methodologies we tried to underpin the above research activities by exploring and comparing methods such as latent growth models, the structured regression approach and specific methods to address the issue of measurement error, and by exploring different ways to measure SES and health. For these analyses the Finnish register data allowed us to study mortality to a much greater extent that it was possible with survey data. We also investigated specific mechanisms in or between certain life course stages, for example health selection in the transition to retirement and the effect of childhood SES on health in older ages. For the latter we found that the effect of childhood SES is mediated almost entirely by education and adult SES, which suggests that childhood poverty can be compensated when it comes to prevention of health problems and health inequality in old age.
While our project dealt with a complex but general question using general measures of SES and health and covering the whole life course, it produced first general findings on the relative importance of social causation and health selection, and offers a theoretical and methodological framework. This can be used by future research on more specific causal mechanisms between aspects of SES and health using more specialized methods and data to contribute to better knowledge on the lifelong interaction between SES and health and the most important intervention points (age groups, mechanisms, policies) to reduce health inequality.
By the end of the four-year project we published seven articles in major peer-reviewed international journals and published or got accepted six other publications such as book chapters. Five more papers are under review at top journals in the field. We presented our findings with 30 presentations, keynote lectures and organized sessions on conferences in 10 different countries.