Reevaluating diabetes and COVID-19 outcomes using national-level data

Purpose: Using a US nationally representative survey of adults, we aimed to evaluate the association between prevalent diabetes and the uptake of COVID-19 testing, rate of positive testing and symptom severity. Methods: Data were sourced from the 2020-2021 National Health Interview Survey. COVID-19 outcomes were defined as: (1) test uptake (2) test positivity (3) diagnosis of COVID-19 and (4) severe disease symptoms with a positive COVID-19 test result. We compared the prevalence of COVID-19 outcomes by diabetes status and examined their associations using multivariate adjusted logistic and ordered logistic regression models. Results: The prevalence of test uptake and test positivity were 50.7 % and 9.4 % in the US population, respectively. 10.3 % were diagnosed with COVID-19 infection by health professionals. There were no statistically significant differences in the outcomes by diabetes status. However, individuals with diabetes were more likely to have severe symptoms. In adjusted regression model, we found no significant associations of diagnosed diabetes with all outcomes. Conclusions: Our findings contrast with prior evidence derived from hospitalized patients. Researchers and policy makers are encouraged to review the properties of data sources and their impact on public health recommendations, particularly in response to future pandemics.