LAUSR

April 9, 2019 1859 Nay Aung, MBBS Steffen E. Petersen, MD, DPhil, MPH In Response: We thank Vieira et al for their interest in and comments on our recent article1 describing the association between ambient air pollution and cardiac imaging phenotypes. Vieira and colleagues argue that despite several strengths afforded by a large sample size, our study may lack generalizability because the study sample is unlikely to be representative of the population as a whole. It is well recognized that the UK Biobank cohort is not representative of general population because of a “healthy volunteer” selection bias,2 which was also emphasized in our original article. However, we would like to highlight that valid and generalizable assessment of exposure-disease relationships can be conducted in the presence of a sufficiently large number of individuals with different levels of exposures investigated with high internal validity.3 With regard to the comment on the inclusion of individuals with cardiovascular risk factors, additionally restricting the analysis to individuals without risk factors would have halved the sample size. All subsequent investigations would be underpowered to uncover the subtle phenotypic changes (and small effect sizes) observed in this study. Hence, we elected to account for the potential confounding influences of cardiovascular risk factors by statistically adjusting for them in our regression models. Lastly, it is beyond the scope of our study to evaluate the underlying mechanistic pathways promoting cardiac remodeling. We reported that systemic hypertension does not appear to have an important influence on the relationship between particulate matter with an aerodynamic diameter <2.5 μm and cardiac structural alteration and postulated that oxidative stress may be responsible for the changes observed in our findings. As Vieira et al noted, the development of cardiovascular disease may well be mediated by particulate matter with an aerodynamic diameter <2.5 μm-associated endothelial dysfunction, which itself may be driven by an increased burden of reactive oxygen species in endothelial cells.4