Table 3.
Multivariable logistic regression results for census tract COVID-19 positivity (≥ 10 % vs < 10%) based on SES factors, community health factors, and environmental exposures: 397 census tracts, Seattle, King County, WA, July 12, 2020
| SES model | Community health model | Environmental exposures model | |||
|---|---|---|---|---|---|
| Adjusted OR (95% CI) |
Adjusted OR (95% CI) |
Adjusted OR (95% CI) |
|||
| People of color (%) | 1.01 (0.98, 1.03) | Cardiovascular mortality rate | 1.27 (0.70, 2.29) | PM2.5 | 3.71 (2.16, 6.38)** |
| < 185 % federal poverty level (%) | 1.05 (1.02, 1.08)** | YPLL | 1.01 (0.97, 1.04) | Ozone | 1.06 (0.90, 1.24) |
| No high school diploma (%) | 1.05 (1.00, 1.10)* | BMI | 8.23 (4.00, 16.9)** | NOx | 0.94 (0.89, 0.98)* |
| Unemployed (%) | 0.96 (0.85, 1.09) | Insurance (%) | 1.02 (0.97, 1.07) | Population near heavy traffic (%) | 1.02 (1.00, 1.04) |
| Overcrowded housing (%) | 1.25 (1.10, 1.42)** | Toxic releases from facilities | 1.00 (1.00, 1.00) | ||
| Overcrowded housing (%)+ | 1.42 (1.25, 1.60) | ||||
*p <0.05; **p < 0.01
+The overcrowded housing indicator was included in the environmental exposures model to verify whether the correlation between air pollutants and COVID-19 infection was driven by high population density, as has been hypothesized elsewhere [37]