Although it had significant strengths, this study has some limitations. Importantly, it was cross sectional and cannot therefore establish
causation. Further there is likely to be residual confounding because people who like to walk may choose to live in areas with more public
transport and we did not have information on these preferences.
Although the PTAI measure was a more rigorous public transport accessibility measure than other studies have used, it is still a
relatively limited measure of accessibility. With no information on individual’s travel destinations or behaviours, and inconsistent
information on wider public transport system characteristics across the 42 areas, this study could not adopt more sophisticated regional
public transport accessibility measures such as those advocated above. The reduction of the exposure to a binary measure may have
resulted in a loss of information, although preliminary analysis suggested that dichotomising PTAI at the median was appropriate. Furthermore,
the distance decay component of the exposure measure (i.e. that a shorter walk time to public transport is assigned a higher