Comparison of the association between COVID-19 and new onset diabetes in a regional and national dataset

Abstract

Aims: This study examined new onset diabetes following COVID-19 infection in large datasets. Methods: A retrospective cohort study design was used in three distinct databases: data from a large academic medical center (University of Kentucky Healthcare, UKHC), claims data from privately insured patients in the Merative MarketScan (MarketScan) database, and a subset of the MarketScan database with lab tests. Control groups included patients from the prepandemic timeframe (prepandemic cohort), as well as COVID-era patients without documented COVID infection (COVID − cohort). Results: Incident diabetes in COVID infected patients was higher in all datasets. In the UKHC and MarketScan datasets, the adjusted Hazard ratios (HR) were 3.46 and 2.13 in UKHC and MarketScan. Incident diabetes was elevated even in patients who were treated in the outpatient setting and much higher in patients who were treated in the inpatient setting, especially those who required inpatient interventions. In MarketScan, the HR for developing type 1 diabetes was 1.61. Conclusions: Persons infected with COVID-19 during the initial 18 months of the pandemic demonstrated a higher incidence of new onset diabetes, which was evident in all populations. Diabetes incidence was especially high in an academic center population with higher health disparities, and in patients with a more severe infection with the virus.