The challenges of quantifying the effects of housing on health using observational data.

Housing is an often overlooked yet fundamental social determinant of health. Like other social epidemiology exposures, housing faces a tension between the promise of modern causal inference methods and the messy reality of complex social processes and reliance on observational data. We use examples from over a decade of research to illustrate some of the key challenges in undertaking causally focused healthy housing research and demonstrate approaches that have been applied to address these challenges. We reflect on the improved understanding these approaches have delivered, and the key gaps and next steps in generating the evidence required to act on housing as a social determinant of health.