Neighbourhood correlates of average population walking: using aggregated, anonymised mobile phone data to identify where people walk.

Understanding and monitoring socio-spatial patterns of population walking mobility can inform urban planning and geographically targeted health promotion strategies aimed at increasing population levels of physical activity. In this study we use aggregated, anonymous mobile phone mobility data to examine the association between neighbourhood physical and social characteristics and residents' weekly walking behaviour across 313 neighbourhoods in a large metropolitan region of Queensland, Australia. We find that residents in neighbourhoods that are highly fragmented by streets with speed limits above 50 kmph, residents in neighbourhoods with high retail density and those living is economically disadvantaged neighbourhoods walk fewer kilometres and minutes on average per week than their counterparts. These findings can inform urban planning policy on the minimum specifications required in newly developing neighbourhoods and provide targets for retro-fitting features into existing neighbourhoods.