Our findings demonstrate that residents living in less walkable areas, particularly recent immigrants in low-income neighborhoods, have an accelerated risk of developing diabetes compared with those living in more walkable areas. Although diabetes can be prevented through physical activity, healthy eating, and weight loss (25
), the environment in which one lives may pose barriers to achieving these measures that are difficult to overcome. Lessons learned from successful antismoking campaigns in the past suggest that population-level strategies can play a key role in promoting healthy behaviors (26
). Moreover, public policies that target whole populations may prevent as many cases of diabetes as those targeting high-risk individuals alone (27
). Our findings support the concept that neighborhood design potentially could influence the health of urban populations through the opportunities it provides for residents to undertake healthy behaviors. This may have practical implications for community-level interventions aimed at diabetes prevention.
Like many cities in the U.S. and Europe, recent immigrants in our setting are most often of South Asian, East Asian, or African descent, groups that are particularly susceptible to the development of diabetes (7
). Recent immigrants are often underemployed relative to their occupational roles in their home country and therefore experience low socioeconomic status relative to the host country population, leading to their residence in less desirable areas of major urban centers. Although the prevalence of chronic conditions is often lower among recent immigrants, a phenomenon known as the “healthy migrant” effect (30
), this was not the case for diabetes among recent arrivals in our setting, especially for those living in low-income and low walkability areas. The growing trend of rapid urbanization in developing countries, together with rich diets and sedentary lifestyles in those settings, may have resulted in less healthy migrant populations to Canada in recent years, especially with regard to their risk of obesity-related conditions such as diabetes (5
). The role of poverty in accentuating the risk of diabetes in recent immigrants is likely multifactorial. Lower-income populations may rely more on local, low-cost opportunities for physical activity within their neighborhood than their wealthier counterparts because of limited financial resources. Moreover, for complex diseases, such as diabetes, the cumulative effect of multiple environmental insults (e.g., poverty, lack of opportunities for physical activity, and greater exposure to inexpensive, unhealthy foods) may dramatically heighten the risk of disease in genetically susceptible individuals (18
Our findings are supported by other studies demonstrating an association between neighborhood features that discourage physical activity and risk factors for diabetes such as physical inactivity, insulin resistance, and obesity (9
). However, virtually all of these were cross-sectional in design. An exception is an analysis by Berry et al. (34
), which noted several neighborhood characteristics that predicted long-term weight gain, including area poverty and higher levels of perceived traffic. Relatively few studies have examined the relationship between neighborhood features and discrete health outcomes such as diabetes. In the Multiethnic Study of Atherosclerosis, participants who reported having better neighborhood resources for physical activity and healthy foods were 38% less likely to develop diabetes over a 5-year period (35
). Our study adds to this literature by suggesting that neighborhood walkability is also a significant risk factor for the development of diabetes among young and middle-aged adults living in urban settings and that recent immigrants are particularly susceptible to its influence.
In many cities around the world, the walkability of a neighborhood is tied to the era in which it was first developed. In Toronto, as in other North American cities, the older neighborhoods, built predominantly before World War II, have characteristics that enhance their walkability, including higher residential densities, shorter block lengths, and the coexistence of residential and commercial areas in the same neighborhood. The latter provides residents with a choice between walking and other modes of travel to reach routine destinations. During the period after World War II, this style of urban design was largely replaced by a shift toward sprawling developments and the separation of land uses through legislated changes in zoning. This led to tremendous growth of purely residential, automobile-oriented suburban communities. Suburban living dramatically reduces the opportunities residents have to engage in transportation-related physical activity, such as walking or bicycling to work or school or to run errands (9
). The automobile-oriented planning practices of the 1960s and 1970s have been contested by the “New Urbanism” movement, which promotes a return to pre–World War II styles of urban design. Many cities are now setting limits on further suburban sprawl, instead favoring high-density development in major employment and retail areas and along major transportation corridors. In this way, urban landscapes can be modified over time.
Our research approach had both limitations and strengths. One limitation of this work is that it included a single municipality; therefore, our findings may not apply to other settings. However, historical trends in urban development that occurred in Toronto are typical of those that occurred in other cities in North America and elsewhere in the world. Our setting is also one of the most multicultural cities globally, making it an ideal setting for this type of research. Although our analysis was based on individuals, socioeconomic status could be measured at only the area level. That measure was based on very small residential units, which have been shown to be a good proxy for individual income (37
). Our data did not include country or region of origin or timing of immigration, and for that reason we used a proxy measure for recent immigration: recent registration for Ontario health care. Previous research showed that most individuals (>80%) captured using this method are new immigrants; however, the remainder include both immigrants and nonimmigrants who migrate between provinces (21
). This misclassification would tend to reduce the differences between the recent immigrant and long-term resident groups studied. The recent immigrant group is likely to be heterogeneous with respect to factors influencing their rate of assimilation, including their country of origin, prior “Westernization,” and the presence of family members in Canada, which we were not able to account for in our analysis. In addition, ethnoracial composition was not available in our data, but most immigrants to Ontario are from populations that carry a high genetic predisposition for developing diabetes (7
). Furthermore, recent immigrants are well distributed across Toronto, with few ethnic enclaves (38
); thus, clustering of ethnic populations should not account for our findings.
Other area-level factors could have contributed to our findings. Differences in the retail food environment potentially could influence rates of obesity and insulin resistance (32
); however, a previous study found higher concentrations of convenience stores and fast food outlets in the more walkable neighborhoods of Toronto (38
). Last, unmeasured confounders could have contributed to the association between neighborhood walkability and diabetes incidence; therefore, we cannot prove definitively that this relationship was directly causal. We hypothesize that living in a less walkable neighborhood increases one’s risk of diabetes through its effects on physical activity and obesity, based on the vast literature supporting the association between neighborhood environments and these entities, including research conducted in Toronto using our index and similar measures elsewhere (9
). However, information about BMI, physical activity levels, and travel behaviors was lacking in our datasets.
Collectively, this line of research supports the notion that where one live has a profound influence on one’s health. This is a fundamental shift in the paradigm in which we view the etiology of chronic diseases such as diabetes—from a purely biomedical one to one that incorporates the larger world in which we live. Moreover, it opens up other avenues for the prevention of obesity-related diseases that include changes in the way communities are designed (40
). As the fast pace of urbanization continues, particularly in developing countries, lessons learned from industrialized nations will be increasingly valuable. Further research is needed to establish the consistency of these findings in other settings, to identify the world regions and ethnoracial groups most susceptible to these effects, and, most importantly, to more fully understand what impact interventions targeting the built environment might have on levels of obesity and diabetes.