Small area and individual level predictors of physical activity in urban communities: a multi-level study in Stoke on Trent, England.

Reducing population physical inactivity has been declared a global public health priority. We report a detailed multi-level analysis of small area indices and individual factors as correlates of physical activity in deprived urban areas. Multi-level regression analysis was used to investigate environmental and individual correlates of physical activity. Nine individual factors were retained in the overall model, two related to individual intentions or beliefs, three to access to shops, work or fast food outlets and two to weather; age and gender being the other two. Four area level indices related to: traffic, road casualties, criminal damage and access to green space were important in explaining variation in physical activity.