Risk of New-Onset Type 2 Diabetes among Vaccinated Adults after Omicron or Delta Variant SARS-CoV-2 Infection

Abstract

Methods This cohort study used Singapore s national COVID-19 registry to construct cohorts with a positive or negative SARS-CoV-2 test result between September 1, 2021, and December 31, 2022. Vaccination status and new-onset T2D were determined using the National Immunization Registry5 and the National Diabetes Database (eMethods 1 in Supplement 1). Covariates included demographics (age, sex, ethnicity, socioeconomic status), vaccination status at test (index) date (T0), and comorbidities. Individuals were excluded for death within 30 days of T0, missing sociodemographic data, SARSCoV-2 infection within 300 days of T0, or preexisting T2D (eFigure in Supplement 1). The primary outcome of interest was new-onset T2D 31 to 300 days after T0 among individuals with vs without SARS-CoV-2 infection. This studywas considered national public health research under the Singapore Infectious Diseases Act; thus, patient consent and institutional review board approval were not required. Reporting followed the STROBE guideline. Baseline characteristics and between-group standardized mean differences (SMDs) were computed; overlap weighting was used to adjust for differences. Risk of new-onset T2D was estimated using competing-risks regression with death as a competing risk, with overlap weights applied; subgroup analyses included ethnicity, COVID-19 vaccination, and infection severity. Sensitivity analyses were conducted with inverse probability weights or the doubly robust approach and the negative-outcome control (solid-organ malignant neoplasm).We constructed a historical cohort for influenza hospitalizations separately; risk of new-onset T2D after hospitalization for COVID-19 vs influenza was contrasted using competing-risks regression (eMethods 2 in Supplement 1). Analyses were conducted using R, version 4.3.1 (R Project for Statistical Computing). 95%CIs excluding 1 indicated significance. Results We compared individuals with vs without SARS-CoV-2 infection during Delta (82 212 vs 531 855) and Omicron (972 610 vs 1 039 276) predominance. SMDs were below 0.05 after weighting (Table 1). No overall increased risk of new-onset T2D was observed during Delta (hazard ratio [HR], 0.99 [95% CI, 0.92-1.06]) or Omicron (HR, 1.00 [95%CI, 0.97-1.03]) predominance; sensitivity analyses did not notably alter these estimates (Table 2). In subgroup analyses, increased postinfectious risk of T2D was observed during Delta predominance among Indian individuals but not among Chinese or Malay individuals. During Omicron predominance, elevated postinfectious risk of T2D (HR, 1.50 [95%CI, 1.06-2.11]) was observed among unvaccinated or partially vaccinated individuals but not among fully vaccinated or boosted individuals. T2D riskwas not elevated among patients with mild SARS-CoV-2.(Table Presented).infection, but those hospitalized with COVID-19 had increased postinfectious risk. Elevated postinfectious risk persisted in hospitalized patients with COVID-19 during Omicron predominance, even after excluding steroid-treated individuals. Compared with historical influenza hospitalizations, patients with COVID-19 had 45.0% higher postacute risk of T2D during Delta predominance vs 56.0% higher risk during Omicron predominance. For the negative-outcome control (malignant neoplasm), between-group risk did not differ notably.infection, but those hospitalized with COVID-19 had increased postinfectious risk. Elevated postinfectious risk persisted in hospitalized patients with COVID-19 during Omicron predominance, even after excluding steroid-treated individuals. Compared with historical influenza hospitalizations, patients with COVID-19 had 45.0% higher postacute risk of T2D during Delta predominance vs 56.0% higher risk during Omicron predominance. For the negative-outcome control (malignant neoplasm), between-group risk did not differ notably. Discussion No overall increased risk of new-onset T2D after SARS-CoV-2 infection was observed in our cohort. Earlier pre-Omicron studies reported increased (27%-46%) diabetes risk after SARS-CoV-2 infection.1-3 Although increased risk was observed following Omicron infection in Hong Kong adults4 and no notable decrease was observed among boosted Korean individuals (vs primary vaccination),6 58.4% of Hong Kong individuals had not completed primary vaccination and 66% of Korean individuals were boosted4,6; 89.4% of our population was boosted. Interethnic differences in insulin resistance pathophysiology may influence diabetogenic risk after infection. Study limitations include results that may not be fully generalizable to populations of different ethnicities or lower vaccination uptake and absence of physical measurements. We observed no overall increased risk of new-onset T2D after mild SARS-CoV-2 infection among a highly vaccinated and boosted multiethnic Asian cohort during Omicron predominance; risk was elevated among Indian individuals during Delta predominance, unvaccinated individuals, and hospitalized patients with COVID-19. Risk of T2D after hospitalization was higher for the COVID-19 cohort vs the historical influenza cohort. Future research is required to evaluate the risk of new-onset T2D after COVID-19 vs other viral infections.