Tight and stable glucose control is associated with better prognosis in patients hospitalized for Covid-19 and pneumonia

Aims To investigate possible associations of glucose patterns with outcomes of Corona Virus Disease 19 (COVID-19) using continuous glucose monitoring (CGM) in 43 patients hospitalized for COVID-19 mild-to-moderate pneumonia, regardless of diabetes. Methods Prospective observational study conducted during two pandemic waves in 2020-2021. Glucose sensor metrics of 7-day recording were obtained from blinded CGM. Respiratory function was evaluated as arterial partial pressure of oxygen (PaO2) to fraction of inspired oxygen (FiO(2)) ratio (PaO2:FiO(2)). Results PaO2:FiO(2) ratio was positively correlated with time in tight range (TITR) 70-140 (r = 0.49, p < 0.001) and time in range (TIR) 70-180 (r = 0.32, p < 0.05), and negatively correlated with average glucose (r =- 0.31, p < 0.05), coefficient of glucose variation (CV) (r =- 0.47, p < 0.01) and time above range (TAR) > 140 (r =- 0.49, p < 0.001). No relations were observed with HbA1c. Multivariate regression analysis showed that normal respiratory function at time of CGM removal correlated positively with TITR 70-140 mg/dL (p < 0.01), negatively with CV and TAR > 140 mg/dL (both p < 0.05) and not with TIR 70-180 and average glucose. Conclusions Lower glucose variability and optimal glucose control, expressed as CV and TITR, are CGM metrics predictive of a better prognosis in COVID-19 patients with pneumonia.