In Reply In large-scale epidemiological studies such as ours,1 measurement of the exposure and outcome is seldom error-free or even with the highest possible precision. Frank and colleagues2 have developed a walkability index using parcel-level information and validated this measure through travel surveys in 2 regions of the United States in 2010. Their walkability index is a composite index including residential density, retail floor area ratio, intersection density, and land use mix, and it was designed to be related to transportation preference.2 Such data were not available for our study, which was conducted in Finland and covered a time period from 2003 to 2013. Similarly, although self-reported smoking is a strong predictor of a range of smoking-related health outcomes,3 some misreporting is still possible and ideally assessment of cotinine concentrations would have complemented self-reports to add precision to the assessment of smoking status. In our study,1 only self-reported data on smoking were available. Furthermore, we could not assess walking ability of the participants or individual differences in perceiving the distance to the nearest tobacco outlet.
Anna Pulakka, Ichiro Kawachi, Mika Kivimäki. Distance From Home to the Nearest Tobacco Outlet May Not Reflect the True Accessibility—Reply. JAMA Intern Med. 2017;177(2):287–288. doi:10.1001/jamainternmed.2016.8596