New-onset diabetes following SARS-CoV-2 infection in the Omicron era: a matched cohort study

Background Earlier studies have suggested that SARS-CoV-2 infection increases the risk of subsequently developing diabetes. However, reports are more limited and inconsistent for infection with the Omicron variant and how COVID-19 vaccination may alter the risks. Methods A population-based, matched cohort study was conducted by using linked registry data. Individuals aged ≥16 years diagnosed with COVID-19 between 15 December 2021 and 31 December 2022 were matched to those without COVID-19 by age, sex, and COVID-19 vaccine recency. Cause-specific Cox models were used to estimate the association between COVID-19 and the initiation of diabetes treatment, adjusting for relevant demographic, healthcare-utilization, and health-related factors. Negative control outcomes were assessed. Results Among 5 736 501 matched pairs with and without COVID-19, followed for a median of 200 days, 45 816 initiated diabetes treatment. Compared with those without COVID-19, individuals with COVID-19 had a 14% higher risk of subsequently initiating diabetes treatment [adjusted hazard ratio (aHR) 1.14; 95% confidence interval (CI) 1.12; 1.17]. The risk was higher for those who had received up to two COVID-19 vaccine doses compared with those who had received a booster within the last 90 days (aHR 1.22; 95% CI 1.18, 1.25 vs aHR 1.08; 95% CI 1.04, 1.12). Similar associations were observed between COVID-19 and negative control outcomes while the incidence patterns differed. Conclusion While we observed a small increased risk of diabetes following SARS-CoV-2 infection during the Omicron-dominant period, the true causal effect could be small or even null given the potential unmeasured confounding and detection bias. Future research on post-acute COVID-19 outcomes should consider including negative control outcomes to better detect potential biases.