Effect of SARS-CoV-2 on glycemic control in post-COVID-19 diabetic patients

Abstract

Aim: To study the clinicodemographic profile, outcomes, and post-COVID change in glycemic control among treated COVID-19-infected patients with poorly controlled or well-controlled diabetes mellitus (DM).

Methods: Adult COVID-19-infected patients who tested positive with rapid antigen test or RT-PCR admitted were included in this prospective observational study. Patients were divided into well-controlled and poorly controlled diabetes group based on HbA1c values at admission. Telephonic follow-up and HbA1c estimation was done after three months. Clinical and laboratory investigations performed were compared between both groups. Hazard ratios (HRs) for mortality risk in both well-controlled and poorly controlled COVID-19 patients with DM was done by Cox proportional hazard models.

Results: Out of 260 patients, 140 (53.84%) and 120 (46.15%) were poorly and well-controlled diabetics respectively. One hundred sixty-three patients (62.69%) were male, and the mean age was 52.67 ± 15.69 years. Severity, duration of hospital stay, steroid duration, insulin requirement and mean HbA1C, both at admission and after three months, were significantly higher in poorly controlled group than the well-controlled group (P < 0.005). With increase in age, the HR for all-cause mortality increased by 1.15 times (95% CI, 1.05-1.25; P = 0.0025) in well-controlled than poorly controlled group, whereas with increase in FBS at admission, the HR for all-cause mortality increased by 1.03 times in poorly controlled than well-controlled group (95% CI, 1.01-1.06; P = 0.0044).

Conclusion: Our results show that well-controlled blood glucose levels or improved glycemic control are associated with a better outcome in patients with COVID-19 and pre-existing type 2 diabetes mellitus.

Keywords: COVID-19; glycemic control; post covid diabetes.

Figure 1

Mean plot of HbA1C at different time points between well-controlled and poorly controlled diabetic groups

Figure 2

Hazard ratios for outcomes in well-controlled and poorly controlled diabetes mellitus cohorts under Cox proportion model. *indicates statistical significance. Note: Cox proportional hazard model is done for each variable, and the significant variables are included in the multivariate Cox proportional hazard model. Hazard ratios are shown on a log scale. Error bars represent the limits of the 95% confidence interval for the hazard ratio