There is a growing understanding that there are social gradients in health, teenage birth rates and causes of morbidity and mortality. However, the more ultimate causes and the more precise patterns that underlie this variation is yet largely unknown (Nettle 2010). The overall project aims to better understand individual and area level effects on reproductive decision-making, health and a range of “risky” behaviours related to mortality in Northern Ireland.
Individuals who grow up in deprived areas tend to speed up reproduction and favour high risk behaviours related to (ill)health which sometimes lead to premature death. To date, studies attempting to answer questions about the effect of area characteristics on reproduction, health behaviours and mortality have often been restricted by cross-sectional data, crude death rates (i.e. on country or county level) and lacked individual-level analysis. We address these methodological shortcomings by making use of NILS unique features, including its longitudinal aspect and large sample size. We will construct multilevel event history analyses for individuals included in the 2001 census, with focus on two outcomes: (a) mortality from high risk/self-inflicted factors (as compared to deaths from extrinsic factors); and (b) reproductive events of men and women from 2001. Ultimately, we are interested in examining how individual and area level factors might explain observed variation in these behaviours and importantly how effects of individual factors might be mediated by area level characteristics. Part one uses NILS.