Article Text

Download PDFPDF

Physical activity
The multiple scales of area effects on health: from evidence to implications for public health practice and policy
  1. M. Riva
  1. Department of Geography, Institute of Hazard and Risk Research, Wolfson Research Institute, Durham University, Durham, UK

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


    In area effects on health research, the influence of area characteristics on health is conceptualised and measured in multiple ways. Each conceptualisation, and its empirical application, is therefore likely to provide a unique identification and understanding of the environmental determinants of health and their potential mechanisms of influence, and may convey different information for public health practice and policy. Using as an example the influence of the built environment on physical activity in Montreal, Canada, the objectives of this presentation are to: (1) present results of associations between built environment characteristics and walking measured using different statistical and spatial approaches; and (2) discuss how these findings may inform public health practice and policy to promote active living.


    Data on walking behaviours of 2716 adults were linked to area-level data on population density, land use mix and accessibility of services, and analysed using multilevel modelling and spatial analysis procedures.


    In multilevel models, greater land use mix (OR 2.85, 95% CI 1.44 to 5.64) and accessibility of services (OR 1.05, 95% CI 1.02 to 1.08), but not population density, are significantly associated with more walking. These findings are extended by illustrating the scale dependence of area effects, eg population density influences walking significantly only in small areas defined by a 500 m street-network buffer around individuals’ residence (OR 1.04; 95% CI 1.01 to 1.06) and not in larger areas delimited by 750 m or 1000 m buffers. More local patterns of area effect are further revealed, eg more mixed land use is associated with higher levels of walking in some but not all areas of the city, suggesting that the significance and magnitude of associations between built environment characteristics and walking are not the same across space.


    Application and findings of the three approaches lead to distinct yet complementary conclusions about the environmental determinants of physical activity and the scales at which they may be operating. Findings are discussed in terms of their implications for public health practice and policy, eg acquiring and inventorying data on population health as well as area characteristics for surveillance purposes, developing and implementing health promotion and prevention strategies based on extent data at the local and at different geographical levels, and evaluating policy interventions aimed at changing the local environment. The challenges of targeting actions at the local area level are underlined and examined in light of recent national (UK and Canada) and international recommendations and guidance to create environments conducive to physical activity.