“…However, the ORs associated with a 10 g/m 3 increase in NO x were close to unity, indicating a lack of association regardless of statistical power. Previous studies on background levels of nitrous oxides and stroke have produced both negative [7,9,17,[19][20][21]24] and positive results [4-6, 13, 14, 16, 22, 23] , and have all used shared monitoring station data, whereas we used individually modeled data. Background concentrations of NO x have been shown to be an indicator for ultrafine particulate air pollution in an urban setting close to our study area [31] .…”