Design: Cross-sectional analysis of data from residences with varying accessibility and greenness.

 

Setting: Seattle, Washington.

 

Subjects.: Stratified random sample of residents, stratified by accessibility and greenness. Response rate: 17.5%, 529 respondents.

 

Measures: Accessibility and greenness were measured objectively by Geographic Information Systems (GIS) Network Analysis and normalized difference vegetation index (NDVI), respectively. Self-reported destinations, natural features, walking trips, BMI, and importance of destinations were measured through a postal survey.

 

Results: Objective accessibility were related to walking trips per month (r2 = .110, p <.0001), as was subjective greenness (r2 = .051, p <.0001), although objective measures of actual greenness were not. In areas with high accessibility, BMI was lower in areas that had high NDVI, or more greenness (r2 = .129428, model p <.0001; t-test of interaction p = .0257). Low NDVI areas were associated with overestimation of the number of destinations within walking distance (F1, 499 = 11.009, p = .001).

 

Conclusions: Objective and subjective measurements of accessibility and greenness led to an understanding of variation among walking trips and BMI in different neighborhoods.