UV, ozone, and COVID-19 transmission in Ontario, Canada using generalised linear models

To, Teresa; Zhang, Kimball; Maguire, Bryan; Terebessy, Emilie; Fong, Ivy; Parikh, Supriya; Zhu, Jingqin; Su, Yushan

BACKGROUND: Quantifying the impact of environmental factors on COVID-19 transmission is crucial in preventing more cases. Ultraviolet (UV) radiation and ozone (O(3)) have reported antimicrobial properties but few studies have examined associations with community infectivity of COVID-19. Research suggests UV light can be preventative while the effect of O(3) is contested. We sought to determine the relationship between UV, O(3), and COVID-19 incidence in Ontario, Canada. METHODS: In our time series analyses, we calculated daily incidence rates and reproductive number (R(t)) from 34,975 cases between January and June 2020 across 34 Ontario Public Health Units. We used generalised linear models, adjusting for potential confounders, to calculate point estimates (PE) and 95% confidence intervals (CI) for UV and O(3). Analyses were further stratified by age groups and outbreaks at institutions versus community. RESULTS: We found that 1-week averaged UV was significantly associated with a 13% decrease (95% CI: 0.80–0.96) in overall COVID-19 R(t), per unit increase. A negative association with UV was also significant among community outbreaks (PE: 0.88, 95% CI: 0.81–0.96) but not institutional outbreaks (PE: 0.94, 95% CI: 0.85–1.03). A positive association of O(3) with COVID-19 incidence is strongly suggested among institutional outbreak cases (PE: 1.06, 95% CI: 1.00–1.13). CONCLUSION: Our study found evidence to support the hypothesis that higher UV reduced transmission of COVID-19 and some evidence that ground-level O(3) positively influenced COVID-19 transmission. Setting of infection should be strongly considered as a factor in future research. UV and O(3) may explain some of COVID-19's seasonal behaviour.