Early life determinants of socioeconomic and health inequalities : life course studies within and across generations

Sammanfattning: The role of the early life social and maternal environments in the production of health inequalities was widely documented in the past few decades but the underlying mechanisms were often subject to suboptimal analytic practices. The developmental origins of health and disease (DOHaD) paradigm came with a promise to better document the early life programming of adult diseases, with an emphasis on the postnatal modifiers of prenatal influences. However, DOHaD as a transgenerational phenomenon (i.e., across more than 2 generations) is rarely acknowledged in human observational studies, despite growing evidence from animal experiments. Moreover, evidence is currently lacking as to whether health inequalities attributable to early life disadvantages have become narrower or wider over historical time in response to major shifts in social structures. This PhD thesis builds on life course epidemiology to provide improved insights and generate new knowledge on the early life origins of health inequalities, their persistence over time, and transmissions across generations. To this end, the project focuses on a range of social and biological disadvantages as predictors of attained socioeconomic position (SEP) and adult health outcomes within the life course and across generations, using register-based and survey data from Sweden. In Study I, I used the inverse odds weighted approach to causal mediation analysis and demonstrated that two adult socioeconomic indicators – education and occupation – jointly mediated about one third of the association between parental SEP and cardiovascular mortality in adulthood. The health behavioral risk factors – smoking, alcohol drinking, poor diet, physical inactivity – and body mass index additionally mediated around one tenth of the association. In Study II, SEP was measured across four time points over the life course: at birth, around age 10, in mid adulthood, and late adulthood. The longitudinal trajectories of SEP were constructed using latent class analysis, with the aim to investigate the associations between the latent SEP trajectories and mortality. Compared to lifetime low SEP, upward mobility trajectories showed reduced hazard ratios of all-cause mortality and mortality from a variety of causes including cardiovascular diseases. Study III aimed to investigate the early life determinants of adult SEP, assessed as an index of education and occupation combined, within and across generations. The association between parental SEP and participants’ SEP became weaker in the offspring generation when compared to their parents, but the association between small-for-gestational age and adult SEP did not exhibit any significant change. Further, lower SEP of grandparents, unmarried status of grandmothers, and paternal preterm birth were associated with lower SEP attained by the offspring. In Study IV, I mainly extended the analysis of Study III to the incidence of ischemic heart disease (IHD) and found that the associations of parental SEP and birthweight-for-gestational age with adult IHD did not change between generations. Additionally, low SEP of grandparents, younger and older childbearing ages of grandmothers, and paternal preterm birth increased the risks of IHD in the adult offspring. To conclude, the thesis highlights the importance of some early life social and biological disadvantages in shaping future socioeconomic and health outcomes not only over the life course but also over historical time and across generations, whilst adding to the scant literature on transgenerational developmental programming of health in humans. Future life course research can benefit from shedding lights into the mechanisms through which health inequalities originating in early life persist over time and are transmitted across multiple generations.