Table 1.
Summary table reporting reviewed results on the association between COVID-19 cases/deaths and air pollution (PM2.5, PM10 and NO2).
| References | Period | Area of Study | Aim | Data analysis | PM2.5 | PM10 | NO2 |
|---|---|---|---|---|---|---|---|
| Zhu et al. (2020) | From Jan 23rd to Feb 29th | 120 cities of China | Temporal association between daily confirmed cases and air pollution (PM2.5, PM10 and NO2) | Generalized Additive Model (GAM) | A 10-μg/m3 PM2.5 increase (lag0–14) was associated with a 2.24% increase of daily confirmed new cases | A 10-μg/m3 PM10 increase (lag0–14) was associated with a 1.76% increase of daily confirmed new cases | A 10-μg/m3 NO2 increase (lag0–14) was associated with a 6.94% increase in daily confirmed new cases |
| Jiang et al. (2020) | From Jan 25th to Feb 29th | Wuhan, XiaoGan and HuangGang (China) | Temporal association between daily confirmed cases and air pollution (PM2.5, PM10 and NO2) | Multivariate Poisson regression | Wuhan (RR = 1.036, CI:1.032–1.039); XiaoGan (RR = 1.059, CI = 1.046–1.072); HuangGang (RR = 1.144, CI = 1.12–1.169) | Wuhan (RR = 0.964, CI: 0.961–0.967); XiaoGan (RR = 0.961, CI = 0.950–0.972); HuangGang (RR = 0.915, CI = 0.896–0.934) | Wuhan (RR = 1.056, CI: 1.053–1.059); XiaoGan (RR = 1.115, CI = 1.095–1.136); HuangGang (no association found) |
| Li et al. (2020) | From Jan 26th to Feb 29th in 2020 | Wuhan and XiaoGan | Temporal association between daily confirmed cases and air pollution PM2.5, PM10 and NO2) | Simple linear regression | Wuhan (R2 = 0.174, p < 0.05); XiaoGan (R2 = 0.23, p < 0.01). | Wuhan (R2 = 0.105; p > 0.05); XiaoGan (R2 = 0.158, p < 0.05). | Wuhan (R2 = 0.329, p < 0.001); XiaoGan (R2 = 0.158, p < 0.05). |
| Yao et al. (2020) | Data up to March 22nd | 49 cities of China | Spatial association between fatality rate and air pollution (PM2.5 and PM10) | Multiple linear regression | χ2 = 15.25, p = 0.004; A 10 μg/m3 increase in PM2.5 was associated with a 0.24% (0.01%–0.48%) increase in fatality rate | χ2 = 13.53, p = 0.009; A 10 μg/m3 increase in PM10 was associated with a 0.26% (0.00%–0.51%) increase in fatality rate | / |
| Ogen (2020) | Data up to the end of Feb | 66 administrative regions in Italy, Spain, France and Germany | Spatial association between deaths counts and air pollution (NO2) | Descriptive analysis: percentage of deaths in three NO2 μmol/m2concentration range (0–50; 50–100; 100–300) | / | / | 83% of fatality cases are associated with NO2 > 100 μmol/m2 |
| Zoran et al. (2020a) | From Jan 1st to Apr 30th | Milan (Italy) | Temporal association between total cases, daily confirmed cases and total deaths and air pollution (PM2.5 and PM10) | Pearson coefficient correlation | R = −0.39; R = 0.25; R = −0.53 | R = −0.30; R = 0.35; R = −0.49 | / |
| Zoran et al. (2020b) | From Jan 1st to Apr 30th | Milan (Italy) | Temporal association between total cases, daily confirmed cases and total deaths and air pollution (NO2) | Pearson coefficient correlation | / | / | R = −0.55; R = −0.35; R = −0.58 |
| Bontempi (2020b) | From Feb 10th to March 12th | 7 provinces of Lombardy, Italy; 6 provinces of Piedmont, Italy; | Spatial description of PM10 exceedances versus COVID-19 cases | Descriptive analysis: Number of days of PM10 exceeding 50 μg/m3 and COVID-19 incidence | / | Lombardy: PM10 exceeding between 0 and 8, COVID-19 incidence % between 0,03 and 0,49. Piedmont: PM10 exceeding between 3 and 12, COVID-19 incidence % between 0,01 and 0,03. | / |
| Coccia (2020b) | Data up to April 7th | 55 Italian Provinces | Spatial association between confirmed cases and air pollution (PM10) | Hierarchical multiple regression model | / | COVID-19 in North Italy has a high association with air pollution of cities measured with days exceeding the limits set for PM10 | / |
| Fattorini and Regoli (2020) | Data up to April 27th | 71 Italian provinces | Spatial association between total confirmed cases and air pollution (PM2.5, PM10 and NO2) | Pearson regression coefficient analysis | R2 = 0.340; p < 0.01 | R2 = 0.267; p < 0.01 | R2 = 0.247; p < 0.01 |
| Frontera et al. (2020) | Data up to 31st March | Italian regions | Spatial association between total confirmed cases and air pollution (PM2.5) | Pearson regression coefficient analysis | R2 = 0.64; p < 0.01 | / | / |
| Frontera et al. (2020) | Data up to 31st March | Italian regions | Spatial association between deaths and air pollution (PM2.5) | Pearson regression coefficient analysis | R2 = 0.53; p < 0.05 | / | / |
| Wu et al. (2020) | Data up to April 04th | 3000 counties in the U.S.A. | Prediction of risk of COVID-19 deaths in the long-term average exposure to fine particulate matter (PM2.5) | Zero-inflated negative binomia models | Long-term exposure increase of 1 μg/m3 in PM2.5 is associated with a 15% increase in the COVID-19 death rate. | / | / |
| Adhikari and Yin (2020) | From March 1st to Apr 20th | Queens county, New York (U.S.A) | Temporal association between daily confirmed cases and total deaths and air pollution (PM2.5) | Negative binomial regression model | Estimate on cases values = −0.4029 (CI%: 0.6478–0.6896); Estimate on deaths value = −0.1151 (CI%: 0.7966–0.9971) | / | / |
| Bashir et al. (2020) | From March 4th to April 24th | California | Association between confirmed cases and air pollution (PM2.5, PM10 and NO2) | Spearman and Kendall correlation tests | Kendall r (−0.359); Spearman r (−0.453) | Kendall r (−0.287); Spearman r (−0.375) | Kendall r (−0.514); Spearman r (−0.736) |
| Bashir et al. (2020) | From March 4th to April 24th | California | Association between deaths and air pollution (PM2.5, PM10 and NO2) | Spearman and Kendall correlation tests | Kendall r (−0.339); Spearman r (−0.429) | Kendall r (−0.267); Spearman r (−0.350) | Kendall r (−0.485); Spearman r (−0.731) |
| Vasquez-Apestegui et al. (2020) | Data up to June 12th | 24 districts of Lima, Perù | Spatial association between total confirmed cases and air pollution (PM2.5) | Multivariate regression model | Crude coefficient = 0.083, p < 0.05 | / | / |
| Vasquez-Apestegui et al. (2020) | Data up to June 12th | 24 districts of Lima, Perù | Spatial association between deaths and air pollution (PM2.5) | Multivariate regression model | Crude coefficient = 0.0016, p < 0.01 | / | / |
| Vasquez-Apestegui et al. (2020) | Data up to June 12th | 24 districts of Lima, Perù | Spatial association between case fatality rate and air pollution (PM2.5) | Multivariate regression model | Crude coefficient = −0.014, p > 0.05 | / | / |