Integrated Lifecycle Approach to the Development of Health & Other Capabilities

The goal of this collaborative interdisciplinary project is to deepen understanding of the developmental origins of health in order to assess the effectiveness of early life prevention and investment strategies. Four interrelated tasks are proposed. (1) Using experimental longitudinal data with long-term follow-up, the project will analyze, synthesize, and interpret the impacts of early childhood interventions. Existing studies will be enhanced with new data collected for later ages and from administrative sources. Curricula, questionnaires, populations targeted, and qualifications of teachers in the influential interventions will be systematically compared. The sources of differences and similarities in program outcomes will be investigated. Mediation analyses will be conducted to investigate channels of influence, including parenting. (2) This project will use observational longitudinal data to estimate the causal effects of education, parenting practices, and family investments and environments on the health, healthy behaviors, and other socioeconomic outcomes of persons over their life cycles and across generations. The project unifies the adult health and human capital literature with the literature on early life development by developing dynamic models of the acquisition of health. It extends current frameworks in economics that focus on investment in children as passive recipients to study parent-child, parent-teacher-child interactions in multiple sibling families, with an active role for children. (3) Using experimental longitudinal data for rhesus monkeys, this project will analyze the effects of early adversity on later life health and on genetic expression of early life adversity as it affects immune systems. (4) The project will develop and apply methodologies to account for small sizes of the influential experimental early childhood intervention studies and account for (a) noncompliance with randomization protocols; (b) attrition; (c) nonresponse; (d) multiple hypotheses (adjusting significance levels for the number of hypotheses tested); and (e) the longitudinal nature of the data.