|dc.description.abstract||Background: The World Health Organization recommends strategies to improve urban design, public transportation,
and recreation facilities to facilitate physical activity for non-communicable disease prevention for an increasingly
urbanized global population. Most evidence supporting environmental associations with physical activity comes from
single countries or regions with limited variation in urban form. This paper documents variation in comparable built
environment features across countries from diverse regions.
Methods: The International Physical Activity and the Environment Network (IPEN) study of adults aimed to measure
the full range of variation in the built environment using geographic information systems (GIS) across 12 countries
on 5 continents. Investigators in Australia, Belgium, Brazil, Colombia, the Czech Republic, Denmark, China, Mexico,
New Zealand, Spain, the United Kingdom, and the United States followed a common research protocol to develop
internationally comparable measures. Using detailed instructions, GIS-based measures included features such as
walkability (i.e., residential density, street connectivity, mix of land uses), and access to public transit, parks, and private
recreation facilities around each participant’s residential address using 1-km and 500-m street network buffers.
Results: Eleven of 12 countries and 15 cities had objective GIS data on built environment features. We observed a
38-fold difference in median residential densities, a 5-fold difference in median intersection densities and an 18-fold
difference in median park densities. Hong Kong had the highest and North Shore, New Zealand had the lowest
median walkability index values, representing a difference of 9 standard deviations in GIS-measured walkability.
Conclusions: Results show that comparable measures can be created across a range of cultural settings revealing
profound global differences in urban form relevant to physical activity. These measures allow cities to be ranked more
precisely than previously possible. The highly variable measures of urban form will be used to explain individuals’
physical activity, sedentary behaviors, body mass index, and other health outcomes on an international basis. Present
measures provide the ability to estimate dose–response relationships from projected changes to the built environment
that would otherwise be impossible.||spa