Table 3. Effect of forest disturbance (deforestation, roads, fire, and selective logging) on malaria incidence at the municipality level in the five timber production states (Roraima, Pará, Rondônia, Acre, and northern Mato Grosso).
| Model 1§ | Model 2 | Model 3 | Model 4 | ||||||||||
| Variable* | SD† | RR‡ | 95% CI | p-value | RR‡ | 95% CI | p-value | RR‡ | 95% CI | p-value | RR‡ | 95% CI | p-value |
| % deforestation (1997–2000) | 4.2 | 0.95 | (0.73, 1.24) | 0.71 | – | – | – | – | – | – | – | – | – |
| % deforestation (2000–2001) | 1.9 | 1.01 | (0.79, 1.27) | 0.96 | – | – | – | – | – | – | – | – | – |
| % deforestation (2001–2002) | 1.2 | 0.91 | (0.74, 1.12) | 0.36 | – | – | – | – | – | – | – | – | – |
| % deforestation (2002–2003) | 1.3 | 1.05 | (0.81, 1.35) | 0.72 | 1.07 | (0.85, 1.35) | 0.57 | 1.15 | (0.86, 1.54) | 0.33 | 1.09 | (0.83, 1.43) | 0.52 |
| % deforested land (2003) | 0.7 | 1.15 | (0.97, 1.37) | 0.12 | 1.21 | (1.01, 1.44) | 0.04 | 1.24 | (1.03, 1.48) | 0.02 | 1.22 | (0.98, 1.51) | 0.07 |
| Unpaved road density(m/km2) (2001) | 18.5 | – | – | – | 1.42 | (1.13, 1.79) | 0.003 | 1.40 | (1.10, 1.76) | 0.01 | 1.26 | (1.00, 1.58) | 0.05 |
| Paved road density (m/km2) (2001) | 7.9 | – | – | – | 1.20 | (0.97, 1.48) | 0.09 | 1.17 | (0.94, 1.46) | 0.15 | 1.21 | (0.96, 1.52) | 0.11 |
| % affected by old fire(1996–1999) | 17.9 | – | – | – | – | – | – | 0.86 | (0.69, 1.07) | 0.18 | 0.96 | (0.77, 1.21) | 0.75 |
| % affected by recent fire(2000–2002) | 12.2 | – | – | – | – | – | – | 1.15 | (0.91, 1.47) | 0.25 | 1.15 | (0.92, 1.46) | 0.22 |
| 0–7% remaining forestslogging¶ | – | – | – | – | – | – | – | – | – | – | 1.72 | (1.18, 2.51) | 0.005 |
| >7–43% remainingforests logged¶,# | – | – | – | – | – | – | – | – | – | – | 0.39 | (0.23, 0.67) | 0.001 |
Variable (year of acquisition).
Residual standard deviation is the unit of change for all forest disturbance risk factors.
Risk Ratio.
Models are adjusted for several sociodemographic and environmental risk factors at the municipality level including: percent of population who migrated in the previous 2 years, male to female ratio, average number of people per household, percent rural population, percent of households living under minimum wage, average transportation costs to the nearest capitol, percent GDP growth from 2000 to 2005, and land cover in 2003 including percent of municipality that was water, remaining forest, and savanna.
Logging occurred between 1999–2002.
# 43% of a county’s remaining forest was the maximum amount of selective logging in a Brazilian Amazon county between 1999–2001.