In the United States, physical inactivity is implicated to cause a high burden of disease and mortality, including 7% of cardiovascular disease, 8% of type-2 diabetes, 12% of common cancers, and 11% of all-cause mortality. In response, several public-health organizations have endorsed the National Physical Activity Plan, a key message of which is that increasing population levels of physical activity will require altering the physical and social environments in which Americans work, play, learn, and travel. Evidence for this assertion is persuasive, but research on built-environment effects on physical activity nonetheless suffers from limitations. The effect of the built environment on cycling, a promising way to increase population levels of physical activity for its ability to also act as transportation, is less studied than that for walking and other forms of recreational physical activity. The use of geographical-positioning-systems (GPS) devices, such as smartphones, makes it possible to objectively characterize cycling and its environmental determinants across space and time. The overall goal of this research is to assess the effect of local bicycle infrastructure on bike ridership and incidence of heart disease, cancer, and mortality in the Atlanta, GA area using objective data from smartphone applications. Atlanta has recently invested in several protected bike lanes and multi-use trails, new infrastructure that makes the average cyclist feel more comfortable. Thus, existing bikers may be more likely to ride a bike and people who never bike might start. As a first aim, this research will assess and quantify whether these new types of infrastructure increase population-level cycling in the area, leveraging existing data generated from smartphone apps. It will employ a difference-in-difference study design to control for potentially differential trends in app use between intervention and control areas. The topic of ?big data? and how it might be leveraged to improve healthcare and genetic research has received considerable attention. Big data may also be useful as part of a multi-sectoral strategy to monitor and improve determinants of population health, but questions of representativeness and validity remain.
The second aim, which is methodological, is to characterize and propose methods for the control of biases that may result from the use of smartphone-generated data, drawing on rigorous training in epidemiologic methods and informing future population-based research relying on this novel data source.
The third aim i s to estimate the burden of cardiovascular disease, cancer, and all-cause mortality potentially averted by this new infrastructure, using methods for estimating the population attributable fraction. This result will provide insight into the public-health impact of this urban-planning intervention.
The importance of regular physical activity to prevent disease and delay death is well recognized. It is less clear how to encourage people to be more active in a society that is designed for them to sit. The proposed research will test whether changing the physical environment in the Southeastern U.S., an area with a high prevalence of physical inactivity and chronic diseases, makes people more active through biking.
