|
Abstract
|
BACKGROUND: Wide and busy roads are barriers to pedestrians, possibly affecting physical activity, independent mobility, social interaction, and physical and mental health. This 'barrier effect' is a pressing problem in many African cities, where, despite the growth in car traffic, walking is still the dominant mode, especially for the poor, women, and older people. This study quantifies the barrier effect of roads for the first time in an African city (Praia, capital of Cabo Verde), mapping the effect at the city level, and analysing its distribution across areas and social groups. The indicators account for land use (i.e., what is on the other side of the road).
Methods: We estimated, for each building, indicators of the barrier effect for trips to houses of other people and to food shops. To overcome lack of data, we manually compiled and mapped many of the variables used (road characteristics, food shops, informal settlements, census data). Barrier effects were defined as the proportions of the areas of other buildings and of food shops within 600m that are located across main roads. We analysed the distribution of the effect using descriptive statistics, cumulative frequencies, bivariate associations, and regression models.
Results: Strong barrier effects were found across the city, in some areas reaching 70% (i.e., roads curtail 70% of the walking accessibility potential of residents). The effect tends to be higher in older informal settlements than in formally planned areas and newer informal settlements, and it disproportionately affects individuals aged 65+. The effect is higher for households with high, medium, or low socio-economic status than for those with very high and very low status. The indicators were robust to changes in the assumptions, including type of roads included, maximum walking distance, and attractiveness of buildings and food shops. The indicators were also able to capture differences between effects in areas near roads and with many potential destinations across the road and areas near similar roads but with no destination across the road. Simpler indicators (e.g., considering only distance to roads) were less variable (as they do not consider land use).
Conclusions: The barrier effect of roads is generally high and affects some areas with vulnerable populations. Mitigation measures (e.g., good crossing facilities) are needed. The indicators used performed well and are a good option for the assessment of barrier effects in data-poor contexts.
Language: en |