Table 2. Associations of environmental and health outcomes with urban landscape profiles adjusted for social environment and country.
| Urban Landscape Profiles | Scattered Pixels | Proximate Stones | Proximate Inkblots | Contiguous Large Inkblots |
|---|---|---|---|---|
| Environmental outcomes | Coef (95% CI) | Coef (95% CI) | Coef (95% CI) | |
| Lack of greenness (1-NDVI) a | 0 (Reference) | −24.16 (−49.27, −3.27) * | −15.39 (−43.23, 7.57) | −26.55 (−59.48, −0.00) * |
| PM2.5 (μg/m3)a | 0 (Reference) | 17.46 (5.35, 30.97) * | 21.58 (6.72, 38.50)* | 40.91 (17.45, 69.05)* |
| NO2 (ppb) a | 0 (Reference) | 23.54 (−3.07, 57.32) | 34.53 (2.83, 75.99) * | 363.94 (187.01, 649.94) * |
| Carbon footprint (CO2 emissions/hab) | 0 (Reference) | −0.08 (−0.18, 0.03) | −0.12 (−0.25, 0.00) | 0.13 (−0.12, 0.39) |
| Health outcomes (age adjusted) | ||||
| NCDs mortality rate (per 100,000 hab) | 0 (Reference) | 9.08 (−18.6, 36.76) | 3.53 (−23.29, 30.34) | −4.30 (−39.68, 31.08) |
| Non-intentional injuries mortality rate (per 100,000 hab) | 0 (Reference) | −3.95 (−9.21,1.3) | −4.43 (−9.92,1.07) | −12.15 (−18.41, −5.9)* |
| Hypertension prevalence (%) | 0 (Reference) | −0.04 (−0.52, 0.44) | 0.26 (−0.31, 0.84) | 0.21 (−0.68,1.09) |
| Diabetes prevalence (%) | 0 (Reference) | 0.20 (−0.12,0.52) | 0.22 (−0.14, 0.57) | 0.58 (0.11, 1.05) * |
| Obesity prevalence (%) | 0 (Reference) | −0.99 (−2.61, 0.64) | −1.38 (−3.2, 0.43) | −2.03 (−3.99, −0.07) * |
*
p < 0.05.
a
Log-transformed variables. The results are reported as % of increase of the outcomes comparing each profile to the reference (scattered pixels). To do this, we applied the following formula: (exp(coef)-1) × 100. NDVI, Normalized Difference Vegetation Index; PM2.5, Particulate Matter that have a diameter of less than 2.5 μm; NO2, Nitrogen dioxide; CO2, Carbon dioxide; NCDs, non-communicable diseases. Linear regression models adjusted for social environment index and country.