Hyperglycemia in sepsis is a risk factor for development of type II diabetes

Abstract

Background

Hyperglycemia is frequent in sepsis, even in patients without diabetes or impaired glucose metabolism. It is a consequence of inflammatory response and stress, so its occurrence is related to severity of illness. However, not all severely ill develop hyperglycemia and some do even in mild disease. We hypothesized the existence of latent disturbance of glucose metabolism that contributes to development of hyperglycemia and that those patients might have increased risk for diabetes.

Methods

Patients admitted with sepsis and no history of impaired glucose metabolism were included and divided in the hyperglycemia group (glucose ≥7.8 mmol/L) and normoglycemia group. Severity of sepsis was assessed. Surviving patients without diabetes at discharge were followed-up for 5 years to investigate risk for development of diabetes.

Results

Hyperglycemia was related to severity of sepsis. Follow-up was finished for 55 patients with hyperglycemia, of which 8 (15.7%) developed diabetes, and 118 patients with normoglycemia, of which 5 (4.2%) developed diabetes (P = .002). Relative risk for developing type 2 diabetes was 4.29 (95% CI, 1.35-13.64).

Conclusion

Patients with hyperglycemia in sepsis who are not diagnosed with diabetes before or during the hospitalization should be considered a population at increased risk for developing diabetes.

Introduction

Hyperglycemia is common during acute infection, particularly in patients with sepsis [1], as is in other critical illnesses and is often referred to as critical illness hyperglycemia (CIH). It may arise in patients with previously diagnosed diabetes or impaired glucose metabolism, or as a first manifestation of previously undiagnosed diabetes. However, it frequently occurs in patients who have normal glucose metabolism before and after the septic episode.

Increase of blood glucose during infection, as in other acute insults to the organism, is a part of adaptive response, mediated by inflammation and neuroendocrine mechanisms. Activation of hypothalamic-pituitary-adrenal axis in stress leads to release of cortisol, but stress response also increases secretion of other anti-insulin hormones: catecholamines, glucagon and growth hormone [2], [3]. Proinflammatory cytokines, predominantly interleukin-1 and tumor necrosis factor α, cause hyperglycemia and peripheral insulin resistance by promoting the same hormones, but also by altering insulin receptor signaling [4], [5], [6], [7], [8]. In muscle and fat cells, inflammation-mediated insulin resistance decreases glucose uptake, whereas in hepatocytes, it causes ongoing gluconeogenesis regardless of hyperglycemia and increased insulin release. In addition, despite hyperglycemia and peripheral insulin resistance insulin concentrations in sepsis may be normal or even decreased, due to suppression of pancreatic β cells caused by proinflammatory cytokines and stimulation of α receptors by catecholamines [9], [10]. All these mechanisms are physiologic responses to infection and are probably present in all patients with sepsis, but evident hyperglycemia is not present in all of them. It is most obviously associated with severity disease, and has been associated with unfavorable outcome in sepsis [1].

Hyperglycemia itself was shown to be harmful during critical illness. Tight glucose control during sepsis has been shown to be associated benefit in survival [11], [12], [13], [14] and is now a part of the Surviving Sepsis Campaign Guidelines [15].

The association between severity of illness, hyperglycemia and outcome is clear, but all patients with severe infections, even with organ failure, do not develop hyperglycemia and some will have increased blood glucose even in milder disease.

Our hypothesis therefore was that hyperglycemia in non-diabetic patients with sepsis is not only a marker of disease severity but is also a marker of latent disturbance in glucose metabolism and that it could be associated with increased risk of developing type 2 diabetes or impaired glucose metabolism in the period following recovery from sepsis.