SARS-CoV-2 infection and COVID-19 vaccination and the risk for new-onset type 1 diabetes: a register-based population study in Sweden

Aims/hypothesis Reports from several countries suggested increased incidence of type 1 diabetes during the COVID-19 pandemic, but causality has remained unclear. We investigated incident type 1 diabetes related to SARS-CoV-2 infection and vaccination in Swedish children and adults and assessed whether vaccination modified infection-related risk. Methods We assembled a register-based cohort of all residents aged <80 years on 1 January 2020 and births during follow-up (1 January 2020–31 December 2023). Risk windows after infection (and after each vaccine dose) were 0–30 days, 31–180 days, 181–365 days, and 1–2 years. Incident type 1 diabetes was defined by the earliest ICD-10 diagnosis E10 in the National Diabetes Register or in the National Patient Register. Cox regression with calendar time as the timescale was used with time-varying exposures; analyses were stratified by age (children <18 years; adults 18–79 years), with age-appropriate covariate adjustment. Sensitivity analyses for children were restricted to ages 12–17 years. Results The cohort included 2,650,492 children (3813 incident type 1 diabetes) and 6,870,328 adults (4453 incident type 1 diabetes). SARS-CoV-2 infection was associated with increased type 1 diabetes risk within 2 years in children (hazard ratio [HR] 1.22; 95% confidence interval 1.10, 1.36) and adults (1.10; 1.00, 1.20), driven largely by the 0–30-day window (5.41; 4.34, 6.74 in children, 3.33; 2.69, 4.12 in adults). Vaccination did not modify infection-associated risk (interaction p >0.5). When vaccination was investigated as the exposure, children showed lower HRs (0.77; 0.67, 0.88) within 2 years, but this association was not observed in ages 12–17 (1.00; 0.80, 1.26); adults showed a small excess risk within 0–30 days after dose 1 (1.32; 1.07, 1.62), but not in later windows or doses. Conclusions/interpretation Short-term elevations in incident type 1 diabetes diagnoses were observed in the first 30 days after infection and after first vaccination, and did not persist, supporting detection/diagnostic acceleration rather than a sustained causal effect of SARS-CoV-2 infection or vaccination on type 1 diabetes.