Abstract
AimsDysglycemia, including the three domains hyperglycemia, hypoglycemia, and increased glycemic variability (GV), is associated with high mortality among critically ill patients. However, this association differs by diabetes status, and reports in this regard are limited. This study aimed to evaluate the associations between the three dysglycemia domains and mortality in critically ill patients by diabetes status and determined the contributing factors for dysglycemia.MethodsThis retrospective study included 958 critically ill patients (admitted to the ICU) with or without DM. Dysglycemia was defined as abnormality of any of the three dimensions. We evaluated the effects of the three domains of glucose control on mortality using binary logistic regression and then adjusted for confounders. The associations between dysglycemia and other variables were investigated using cumulative logistic regression analysis.ResultGV independently and similarly affected mortality in both groups after adjustment for confounders (DM: odds ratio [OR], 1.05; 95% confidence interval [CI]: 1.03-1.08; p <0.001; non-DM: OR, 1.07; 95% CI, 1.03-1.11; p = 0.002). Hypoglycemia was strongly associated with ICU mortality among patients without DM (3.12; 1.76-5.53; p <0.001) and less so among those with DM (1.18; 0.49-2.83; p = 0.72). Hyperglycemia was non-significantly associated with mortality in both groups. However, the effects of dysglycemia seemed cumulative. The factors contributing to dysglycemia included disease severity, insulin treatment, glucocorticoid use, serum albumin level, total parenteral nutrition, duration of diabetes, elevated procalcitonin level, and need for mechanical ventilation and renal replacement therapy.ConclusionThe association between the three dimensions of dysglycemia and mortality varied by diabetes status. Dysglycemia in critical patients is associated with excess mortality; however, glucose management in patients should be specific to the patient’s need considering the diabetes status and broader dimensions. The identified factors for dysglycemia could be used for risk assessment in glucose management requirement in critically ill patients, which may improve clinical outcomes.
Highlights
Most studies have reported on the high incidence of dysglycemia and its independent association with mortality among critically ill patients [1,2,3,4,5,6]
Some recent studies have reported that the association between hypoglycemia and mortality is stronger in patients with diabetes mellitus (DM) [7], which is inconsistent with the findings of other studies
We investigated the association between 3 domains of glucose control and mortality in acutely ill patients depending on their diabetes status
Summary
Most studies have reported on the high incidence of dysglycemia (including the 3 dimensions hyperglycemia, hypoglycemia, and increased glycemic variability [GV]) and its independent association with mortality among critically ill patients [1,2,3,4,5,6]. Acta Diabetologica pre-existing diabetes present a “blunted effect” to increased GV and hyperglycemia This has been attributed to the likely higher tolerance to acute glucose fluctuation in patients with DM [13,14,15]. Some recent studies have reported that the association between hypoglycemia and mortality is stronger in patients with DM [7], which is inconsistent with the findings of other studies Most of these studies did not consider comorbidities, disease severity, inflammation level, insulin therapy and other in ICU treatments (including glucocorticoid use and nutrition therapy), which known as confounding factors when estimating the effects of glucose metrics [16]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
