BackgroundSARS-CoV-2 caused the COVID-19 pandemic in 2020. The virus is likely to show seasonal dynamics in European climates as other respiratory viruses and coronaviruses do. Analysing the association with meteorological factors might be helpful to anticipate how cases will develop with changing seasons.
MethodsRoutinely measured ambient daily mean temperature, absolute humidity, and relative humidity were the explanatory variables of this analysis. Test-positive COVID-19 cases represented the outcome variable. The analysis included 54 English cities. A two-stage meta-regression was conducted. At the first stage, we used a quasi-Poisson generalized linear model including distributed lag non-linear elements. Thereby, we investigate the explanatory variables' non-linear effects as well as the non-linear effects across lags.
ResultsThis study found a non-linear association of COVID-19 cases with temperature. At 11.9°C there was 1.62-times (95%-CI: 1.44; 1.81) the risk of cases compared to the temperature-level with the smallest risk (21.8°C). Absolute humidity exhibited a 1.61-times (95%-CI: 1.41; 1.83) elevated risk at 6.6 g/m3 compared to the centering at 15.1 g/m3. When adjusting for temperature RH shows a 1.41-fold increase in risk of COVID-19 incidence (95%-CI: 1.09; 1.81) at 60.7% in respect to 87.6%.
ConclusionThe analysis suggests that in England meteorological variables likely influence COVID-19 case development. These results reinforce the importance of non-pharmaceutical interventions (e.g., social distancing and mask use) during all seasons, especially with cold and dry weather conditions.