1563-P: Neighborhood Drivability and Diabetes Incidence in Toronto, Canada

Background: Reliance on cars contributes to physical inactivity, and therefore may be a risk factor for type 2 diabetes. We investigated whether living in neighborhoods that are highly conducive to driving is associated with an increased incidence of diabetes. Methods: Working age adults (20-64 yrs) who were living in Toronto, Canada on April 1st 2011, were followed over 6 yrs for incident diabetes using a validated algorithm based on hospital records and physicians’ services claims. For neighborhood drivability, we used a novel index capturing three factors of the built environment: urban sprawl, pedestrian unfriendliness and parking options. Cox regression was used to examine the association between neighborhood drivability quintiles (Q) and diabetes incidence, adjusting for age, sex, income, ethnicity, immigration status and comorbidity, and censoring for death. Results: Among 1,473,994 individuals in our sample (mean age 40.9±12.2, 48.5% male), 77,835 developed diabetes. Overall, there was a direct relationship between drivability and diabetes incidence, however the magnitude of this effect varied by age and income. Among young adults (20-34 yrs), those living in the most drivable neighborhoods (Q5) had a 58% higher incidence of diabetes (adjusted HR: 1.58 (95%CI: 1.47-1.69)) relative to those in the least drivable neighborhoods (Q1), whereas the same comparison in older adults (55-64 yrs) yielded smaller differences (HR: 1.31 (95%CI: 1.26-1.36)). High drivability was most strongly associated with diabetes risk in the middle income neighborhoods with 96% increased risk for young residents (HR:1.96 (95%CI: 1.64-2.33) and a 46% increased risk for older residents (HR: 1.46 (95%CI:1.32-1.62). Associations between drivability and diabetes incidence were significant but of a lesser magnitude in low- and high-income neighborhoods. Conclusion: In our setting, neighborhood drivability is a risk factor for the diabetes incidence among working age adults, especially younger, middle-income populations. Disclosure N. den Braver: None. J. Beulens: None. J. Lakerveld: None. P. Gozdyra: None. F. Wu: None. F. Rutters: None. G.S. Fazli: None. R. Moineddin: None. G. Booth: None.