P347A prospective study of rotating night shift work and incident depression in the nurses' health study 2

Evidence linking shift work to mental health, and more specifically to depression and depressive symptoms, has been mixed, and largely based on cross-sectional studies, with limited information on potential confounding factors. We, therefore, plan to assess the association between rotating night shift work and incident depression (assessed as self-reported physician-diagnosis and anti-depressive medication intake) over 10 years of follow-up (2001-2011) in the Nurses' Health Study 2 cohort (N = 33,992), a prospective cohort study with detailed information on a number of important confounders. We hypothesise that longer duration of rotating nightshift work is associated with an increased risk of incident depression. We will also examine whether this association is modified by chronotype - a proxy for individual phase of entrainment. It has been suggested that later chronotypes, while their baseline risk of developing depression might be higher than that of morning chronotypes, are less impacted by shift work, including rotating night shift work. We will use Cox proportional hazard models to estimate age- and multivariable-adjusted hazard ratios (HRs) and 95% CIs of developing depression across exposure categories of lifetime history of rotating night shift work (none (ref),