Association of COVID-19 Infection With Wearing Glasses in a High-Prevalence Area in Denmark and Sweden

IMPORTANCE: Observational studies have indicated that glasses might protect against contracting COVID-19 through reduced airborne and contact transmission. OBJECTIVE: To investigate the association between wearing one’s own glasses with contracting COVID-19 when adjusting for relevant confounders. DESIGN, SETTING, AND PARTICIPANTS: This cohort study was conducted during the first wave of the COVID-19 pandemic (June to August 2020) in Denmark and Sweden, where personal protective equipment was not recommended for the general population at the time. Employees at Falck, an international rescue corps with different job functions (ambulance, health care, office, and field staff, firefighters, and roadside assistance) participated in the study. EXPOSURES: The main exposure was wearing glasses (also including contact lenses and reading glasses), which was assessed in a questionnaire. Persons wearing glasses were compared with those who did not wear glasses (ie, nonusers). To adjust for potential confounders, information on age, sex, job function, and number of workday contacts were included. MAIN OUTCOMES AND MEASURES: The outcome was COVID-19 infection before (positive polymerase chain reaction test) or during the study period (biweekly voluntary tests with a rapid test). The investigated hypothesis was formulated after collecting the data. RESULTS: A total of 1279 employees in Denmark and 841 in Sweden were included (839 [39.6%] female and 1281 [60.4%] male; 20.5% were aged 60 years). Of these, 829 individuals (64.8%) in Denmark and 619 (73.6%) in Sweden wore glasses. Wearing glasses was inversely associated with COVID-19 infection in the Swedish cohort (odds ratio [OR], 0.61 [95% CI, 0.37-0.99]; P = .047; seroprevalence, 9.3%) but not in the Danish cohort (OR, 1.14 [95% CI, 0.53-2.45]; P = .73; seroprevalence, 2.4%). Adjusting for age, sex, job function, and number of workday contacts in Sweden, wearing glasses no longer was associated with COVID-19 infection (OR, 0.64 [95% CI, 0.37-1.11]; P = .11). When stratifying by job function, a large difference was observed among office staff (OR, 0.20 [95% CI, 0.06-0.70]; P = .01) but not ambulance staff (OR, 0.83 [95% CI, 0.41-1.67]; P = .60) nor health care staff (OR, 0.89 [95% CI, 0.35-2.30]; P = .81). CONCLUSIONS AND RELEVANCE: While wearing one’s glasses was inversely associated with COVID-19 in Sweden in an unadjusted analysis, an association no longer was identified whe