Clinical evidence exists that patients with diabetes are at higher risk for Coronavirus disease 2019 (COVID-19). We investigated the physiological origins of this clinical observation linking diabetes with severity and adverse outcome of COVID-19.
Publication mining was applied to reveal common physiological contexts in which diabetes and COVID-19 have been investigated simultaneously. Overall, we have acquired 1,121,078 publications from PubMed in the time span between 01 and 01-2000 and 17-04-2020, and extracted knowledge graphs interconnecting the topics related to diabetes and COVID-19.
The Data Mining revealed three pathophysiological pathways linking diabetes and COVID-19. The first pathway indicates a higher risk for COVID-19 because of an upregulation of Angiotensin-converting enzyme 2. The other two important physiological links between diabetes and COVID-19 are liver dysfunction and chronic systemic inflammation. A deep network analysis has suggested clinical biomarkers predicting the higher risk: Hypertension, elevated serum Alanine aminotransferase, high Interleukin-6, and low Lymphocytes count.
The revealed biomarkers can be applied directly in clinical practice. For newly infected patients, the medical history needs to be checked for evidence of a long-term, chronic dysregulation of these biomarkers. In particular, patients with diabetes, but also those with prediabetic state, deserve special attention.
Three pathophysiological pathways linking diabetes and Coronavirus disease 2019 (COVID-19) were revealed by Data Mining.
Diabetics with upregulated Angiotensin-converting enzyme 2 could be of higher risk for COVID-19.
Liver dysfunction with chronically increased serum Alanine aminotransferase indicates adverse outcome of COVID-19.
Clinical biomarkers for chronic inflammation, in particular Interleukin-6, predict the severity of COVID-19.