Large Differences in Publicly Visible Health Behaviours across Two Neighbourhoods of the Same City

Background: There are socioeconomic disparities in the likelihood of adopting unhealthy behaviours, and success at giving them up. This may be in part because people living in deprived areas are exposed to greater rates of unhealthy behaviour amongst those living around them. Conventional self-report surveys do not capture these differences in exposure, and more ethological methods are required in order to do so. Methodology/Principal Findings: We performed 12 hours of direct behavioural observation in the streets of two neighbourhoods of the same city which were similar in most regards, except that one was much more socioeconomically deprived than the other. There were large differences in the publicly visible health behaviours observed. In the deprived neighbourhood, we observed 266 more adults smoking (rate ratio 3.44), 53 more adults drinking alcohol (rate ratio not calculable), and 38 fewer adults running (rate ratio 0.23), than in the affluent neighbourhood. We used data from the Health Survey for England to calculate the differences we ought to expect to have seen given the individual-level socioeconomic characteristics of the residents. The observed disparities between the two neighbourhoods were considerably greater than this null model predicted. There were also different patterns of smoking in proximity to children in the two neighbourhoods. Conclusions/Significance: The differences in observed smoking, drinking alcohol, and physical activity between these two neighbourhoods of the same city are strikingly large, and for smoking and running, their magnitude suggests substantial area effects above and beyond the compositional differences between the neighbourhoods. Because of these differences, individuals residing in deprived areas are exposed to substantially more smoking and public drinking, and less physical activity, as they go about their daily lives, than their affluent peers. This may have important implications for the initiation and maintenance of health behaviours, and the persistence of health inequalities.