Aim:Life course epidemiology is the study of how exposures during gestation, childhood, adolescence and adulthood influence later health and wellbeing. Analysis of data from across the life course enables us to examine the dynamic ways in which variables of interest change and interact across a person's life span, determinants of these changes, and how the pattern of change relates to later health. A key focus within life course studies is analysis of mediation, i.e. the chain of intermediate processes that links an exposure to an outcome. Well-executed mediation studies are a crucial next step for the field of life course epidemiology. They represent an opportunity to increase understanding of pathways and mechanisms by studying the causal chains that explain exposure-outcome associations. Mediation analysis can therefore provide aetiological insight and identify potential intervention targets. In recent years, the epidemiological literature has given much more consideration to some of the statistical issues in mediation analysis than was the case previously; epidemiologists are now much more likely to give careful thought to mediator-outcome confounding and the potential resultant collider bias, and methods are now available that allow for exposure-mediator interactions or for situations where a confounder of the mediation-outcome association is caused by the exposure. Counterfactual theory has been used to show that the 'total' effect of an exposure on an outcome can be decomposed into the 'natural' direct and indirect effects, and methods for identifying these effects have been developed. However, several challenges remain. In this proposal, I will address 2 key issues in mediation analysis: 1) mediation analysis with repeated measures data, and 2) causal mediation analysis using Mendelian Randomization (MR), and will apply appropriate methods to address the following 2 questions of public health importance in the field of adiposity and cardiometabolic health:i) Whether the trajectory of growth and development affects cardiometabolic health independently of final attained adiposity.