Seasonal variation in SARS-CoV-2 transmission in The Netherlands, 2020-2022: statistical evidence for an inverse association with solar radiation and temperature.

Abstract

In temperate regions, respiratory viruses such as SARS-CoV-2 are better transmitted in winter than in summer. Understanding how the weather is associated with SARS-CoV-2 transmissibility can enhance projections of COVID-19 incidence and improve estimation of the effectiveness of control measures. During the pandemic, transmissibility was tracked by the reproduction number <i>R</i>_<i>t</i>. This study aims to determine whether information about the daily temperature, absolute humidity and solar radiation improves predictions of <i>R</i>_t in The Netherlands from 2020 to 2022, and to quantify the relationship between <i>R</i>_<i>t</i> and daily weather data. We conducted a regression analysis, accounting for immunity from vaccination and previous infection, higher transmissibility of new variants and changes in contact behaviour due to control measures. Results show a linear association between log<i>R</i>_<i>t</i> and daily solar radiation and temperature, indicating a ratio of <i>R</i>_<i>t</i> in winter versus summer of 1.7 (95% CI, 1.4; 2.1). The possibility that this association arises from unrelated seasonal patterns was dismissed, as weather data from earlier years provided poorer fits with only small effect sizes. This suggests a causal relationship between solar radiation and temperature with SARS-CoV-2 transmissibility, enhancing confidence in using this relationship for short-term predictions and other epidemiological analyses.