The term built environment (BE) refers to the surroundings of our homes, schools, or workplaces that can include buildings, streets, stores, parks, and open spaces. The BE has an influence on our health and weight that can act through complex mechanisms, from the quality of housing to access to healthy food and to safe places to exercise and play. In recent studies, shorter distances from home to supermarkets and grocery stores were linked to lower body weight. Recently developed metrics of the neighborhood BE have included distances from home to fast foods and convenience stores as well as to parks and trails. However, the causal links between selected BE features and body weight outcomes are still waiting to be established.
To study the causal impact of BE on health it is important to have data on diets and physical activity: intermediate behavioral variables in the causal BE-obesity pathway. The need to include them has been stressed before.
To address the complexity of this question, Dr. Adam Drewnowski and colleagues (Public Health Sciences Division) combined data on the neighborhood BE along with information on diet, physical activity, and weight from the Seattle Obesity Study (SOS II), a longitudinal cohort of 440 adult residents of King Co, WA. Unlike many past studies in this area, the SOS II allowed the investigators to examine the BE-obesity pathway in a longitudinal manner.
This combination of data allowed the investigators to answer the following questions: 1) What was the relation at baseline between neighborhood BE within a certain distance of home and diet and physical activity? 2) Was there a relation at baseline between behavioral variables and obesity prevalence, adjusting for confounders? 3) Lastly, was there a relation between behavioral variables at baseline and 12-month weight loss or weight gain? The answers from this study were recently published in BMC Public Health.
The investigators first obtained geographical information of participant’s home addresses. They then defined neighborhood BE measures as counts and densities of food sources and physical activity locations within 800 meters of home. This data was made possible by “the development of GIS (geographic information systems) techniques by our colleagues at the Urban Form Lab allowed us to track distances from home to grocery stores, fast foods, and parks,” notes Dr. Drewnowski. In addition, tax parcel property values were obtained. For intermediate behavioral variables, diet quality was defined by using Healthy Eating Index (HEI 2010) scores, constructed via data from food frequency questionnaires, and physical activity was obtained by self-report. Health outcomes included weights and heights, which were measured at baseline and at the following 12 months. Multivariable regressions examined the associations among BE measures at baseline, health behaviors (HEI-2010 and physical activity) at baseline, and health outcomes both cross-sectionally and longitudinally.
None of the neighborhood BE metrics were associated with either diet quality, or with meeting physical activity guidelines. However, higher property values predicted better diets and more physical activity. In addition, better diets and more physical activity were associated with lower obesity prevalence at baseline and 12 months, but did not predict weight change.
Dr. Drewnowski gives his two cents, “Asked to identify the best predictor of health, I would say - your zip code and the value of your home. But how do zip codes "work"? The conventional answer is that wealthier areas have better access to healthier foods, more supermarkets, farmers markets and fewer fast foods. Wealthier areas may have easier access to physical activity locations. We say that the observed links between neighborhood features and health need to involve diet and physical activity in some way. If so, then shorter distances to food sources ought to be linked to higher diet quality. However, our study showed that none of the conventional neighborhood BE metrics were associated either with diet quality, or with meeting physical activity guidelines.”
“Areas with high prevalence of obesity have been called ‘obesogenic’- yet many studies did not track the diet or physical activity pathways,” Dr. Drewnowski elaborates. “We say that the distance to the nearest supermarket does not predict diet quality - what truly mattered were education and income. Only higher property values did predict better diets and more physical activity.”
Taking a step back and looking at the big picture, Dr. Drewnowski explains, “Our studies get at the heart of social inequities in health. We view obesity not as a genetic disease but as a manifestation of social inequalities. Our studies have managed to predict obesity levels in Seattle King County by census block.”
Funding for this study was provided by the National Institutes of Health.
Drewnowski A, Aggarwal A, Tang W, Hurvitz PM, Scully J, Stewart O, Moudon AV. 2016. Obesity, diet quality, physical activity, and the built environment the need for behavioral pathways. BMC Public Health. 16(1): 1153. PMCID: PMC5105275.
Basic Sciences Division
Human Biology Division
Maggie Burhans, Ph.D.
Public Health Sciences Division
Vaccine and Infectious Disease Division
Clinical Research Division
Julian Simon, Ph.D.
Clinical Research Division
and Human Biology Division
Arnold Digital Library