Background and Purpose
Background: Walkability is a combination of built environment characteristics that may encourage or inhibit commuting and leisure time physical activity (PA). Several studies have supported its use and found a positive association between walkability and adult’s PA in high-income countries. However, to this date such associations have not been analyzed in regions with low and middle- income countries, such as Latin America (LA).

Objectives
The aim of this study was to examine the association between walkability and walking for commuting and leisure in adults from Curitiba, Brazil.

Methods
The Spaces for PA in Adults Study (ESPACOS Project) was conducted in Curitiba, Brazil. Data were collected in 2010 in thirty-two census tracts selected according to GIS-based walkability and income. Participants were 697 adults (52.0% women) randomly sampled and age between 18 and 65 years. The International Physical Activity Questionnaire was used to capture self-reported walking (leisure and transportation). The main effects of walkability and income and their interaction were the focus of these analyses. The dependent variable was walking at least 150 min/week so logistic regressions with random intercept adjusting for census tract clustering were performed and interaction terms between quadrants of walkability and income were included to test the moderator effect of quadrants income on the walkability-PA association.

Results
Walking for commuting at least 150 min/week was 21.1% in low-walkable areas and ranged from 33.5 to 35.0% in those areas classified as being high walkable. Walking for leisure for at least 150 min/week was achieved by roughly 12% of participants and did not vary across quadrants of walkability and income. After adjusting for all individual confounders (e.g.), quadrants of walkability versus income showed an independent association with walking for transport (prevalence odds ratio [pOR] =2.10; 95% confidence interval [CI]=1.31-3.37; p=0.002). No association was found between walkability and walking for leisure (pOR=1.01; 95%CI=0.63-1.62; p=0.971). No interaction between walkability and income was found for transport (pOR=0.91; 95%CI=0.36 -2.32; p=0.840) or leisure walking (pOR=0.86; 95%CI=0.34-2.19; p=0.753).

Conclusions
This was the first study that examined the association between walkability and walking outcomes in a developing country with identical protocol, design and instruments employed high- income countries. The main findings emerging from this study showed that adults living in highly walkable areas are more likely to achieve recommended levels of PA through walking as mode of transport. Additionally, no significant interaction between neighborhood walkability and income was found, indicating that walkability is strongly associated with walking for commuting regardless of the participants’ neighborhood income. Finally, walkability was not associated with walking for leisure. These results extend the evidence by demonstrating that built environment is an important PA correlate in developing countries. This study confirms findings from high-income countries and showed that walkability is positively associated with PA of adults living in Latin America.