Neighbourhood environments and risk of incident atrial fibrillation: limitations and future directions

We have read the paper by Garg et al. ¹ with great interest. The authors found no significant associations of the social and built environments with a lower incidence of atrial fibrillation in a sample of American adults. In our view, there are at least two key limitations to take into account when interpreting the findings.

First, as the authors correctly mentioned, the use of a one-mile buffer around each participant’s home may not truly represent the built environment to which participants were exposed. For example, a study conducted in the United States found that a large amount of participants’ objectively measured physical activity occurred outside of the common but arbitrarily defined residential buffers (i.e. 0.5, 1 and 5 miles). ² Testing a number of different sizes of buffers or using activity spaces extracted through global positioning system points may help in capturing the built environment that is relevant to cardiovascular health. We admit, however, that identifying the appropriate geographical scales for exposure is also an ongoing challenge in the broader field of place and health.

Second, there are multiple pathways between the built environment and cardiovascular health. For instance, walkable neighbourhoods are likely to have more commercial destinations that promote walking. However, some of these destinations provide not only healthy but also unhealthy food. In addition, recent evidence suggests that the beneficial effects of walkable neighbourhoods on cardiovascular health may be offset by the adverse effects of air pollution in such areas. ³ Air pollution is increasingly recognised as a risk factor for atrial fibrillation. ⁴ In particular, long-term exposure to air pollution has been found to be associated with a higher risk of atrial fibrillation. ⁵ The complex relationships between different built environment exposures need to be disentangled to determine accurately the association between such exposure and cardiovascular health. For example, it is possible to isolate the effect of healthy food stores by adjusting for the availability of unhealthy food (if data are available). Findings from models with multiple built environment exposures can be theoretically important. However, the challenge will be how to translate the findings into urban design practice as it may be difficult to modify the co-presence of these environmental factors in the actual environment (e.g. healthy and unhealthy food in stores, air pollution concentrated in walkable environments).

We acknowledge that as the role of urban design attributes on cardiovascular health is a new interdisciplinary area of research, only a handful of studies exist on this topic. Notably, most of these studies have the same limitations. Meanwhile, the authors should be congratulated for their contribution and extensive work to shed light on this important topic.