Abstract
Objectives: To investigate the impact of cannabis use on the infection and survival outcomes of COVID-19.
Study Design: Cross-sectional study based on the UK Biobank (UKB) dataset.
Methods: We identified 13,099 individuals with cannabis smoking history in the UKB COVID-19 Serology Study. The Charlson-Quan Comorbidity Index was estimated using inpatient ICD-10 records. Multivariable logistic regression characterized features associated with COVID-19 infection. Cox models determined the hazard ratios (HR) for COVID-19-related survival.
Results: Cannabis users were more likely to getting COVID-19 (odds ratio: 1.22, P = 0.001) but multivariable analysis showed that cannabis use was a protective factor of COVID-19 infection (adjusted odds ratio: 0.81, P = 0.001). Regular cannabis users, who smoked more than once per month, had a significantly poorer COVID-19-related survival, after adjusting for known risk factors including age, gender, smoking history, and comorbidity (adjusted hazard ratio: 2.81, P = 0.041).
Conclusions: The frequency of cannabis use could be considered as a candidate predictor for mortality risk of COVID-19.