Endocrine Abstracts

Background: The potential impact of SARS-CoV-2 on glucose metabolism has been discussed since the first days of the COVID-19 pandemic, and now has been confirmed by research results and clinical practice data about new-onset carbohydrate metabolism disorders after COVID-19. Determining the predictors of their development will help to determine the best treatment tactics.

Aim: To determine predictors of carbohydrate metabolism disorders within 1 year after COVID-19.

Materials and methods: We included consecutive patients hospitalized with viral pneumonia caused by COVID-19. All underwent a clinical and laboratory examination, including assessment of carbohydrate metabolism indicators (blood glucose on admission, fasting plasma glucose, HbA1c, proinsulin, insulin, C-peptide, HOMA-IR). The prospective part of the study included patients without previous history of diabetes mellitus, who had HbA1c > 6.0% during the acute period of the COVID-19. In this group, patients underwent an oral glucose tolerance test (OGTT) 6-8 weeks and 1 year after discharge. In patients, which developed prediabetes and diabetes after COVID-19, we performed a comparative analysis of clinical and laboratory parameters.

Results: We included 155 hospitalized patients with acute phase of COVID-19: median age 59 years [47;72], male/female ratio (%) - 49,4/50,6, BMI – 28,9 kg/m2 [25,4;32,9]. Among these patients, 55 had elevated HbA1c values without previous DM history: HbA1c 6,1-6,4% - 36 patients, HbA1c≥6,5% - 27 patients. During the follow-up 1 year after COVID-19, 33 patients were lost to follow-up, 12 patients developed carbohydrate metabolism disorders, and 10 were euglycemic. We found differences in baseline proinsulin levels between groups with and without carbohydrate metabolism disorders on follow-up: 0,58 mIU/l [0,26; 0,94] and 1,16 mIU/l [0,98; 2,47], respectively (P<0,001). AUC for baseline proinsulin levels estimated 0,906 (95% CI, 0.784-1.000, P<0.001), with cut-off 1,10 mIU/l to distinguish patients with/without carbohydrate metabolism disorders at follow-up. The model had acceptable NPV, 93,8% (83,5%- 99,6%), however, PPV was low, 54,4% (14,1%- 77,2%).

Conclusion: Baseline proinsulin levels with a cut-off at 1,10 mIU/l could be potential predictor of long-term carbohydrate metabolism disorders in patients without previous DM history after COVID-19. However, further studies with extended groups and longer follow-up are needed to confirm this finding.