Hyperglycemia enhances coagulation and reduces neutrophil degranulation, whereas hyperinsulinemia inhibits fibrinolysis during human endotoxemia

Abstract

Type 2 diabetes is associated with altered immune and hemostatic responses. We investigated the selective effects of hyperglycemia and hyperinsulinemia on innate immune, coagulation, and fibrinolytic responses during systemic inflammation. Twenty-four healthy humans were studied for 8 hours during clamp experiments in which either plasma glucose, insulin, both, or none was increased, depending on randomization. Target plasma concentrations were 5 versus 12 mM for glucose, and 100 versus 400 pmol/L for insulin. After 3 hours, 4 ng/kg Escherichia coli endotoxin was injected intravenously to induce a systemic inflammatory and procoagulant response. Endotoxin administration induced cytokine release, activation of neutrophils, endothelium and coagulation, and inhibition of fibrinolysis. Hyperglycemia reduced neutrophil degranulation (plasma elastase levels, P < .001) and exaggerated coagulation (plasma concentrations of thrombin-antithrombin complexes and soluble tissue factor, both P < .001). Hyperinsulinemia attenuated fibrinolytic activity due to elevated plasminogen activator-inhibitor-1 levels (P < .001). Endothelial cell activation markers and cytokine concentrations did not differ between clamps. We conclude that in humans with systemic inflammation induced by intravenous endotoxin administration hyperglycemia impairs neutrophil degranulation and potentiates coagulation, whereas hyperinsulinemia inhibits fibrinolysis. These data suggest that type 2 diabetes patients may be especially vulnerable to prothrombotic events during inflammatory states.

Figure 1

Study design. Twenty-four subjects were studied (4 groups of 6). Somatostatin and glucagon were administered to all subjects using the same doses regardless of clamp type. Glucose was infused at variable rates to achieve plasma concentrations of 5 mM (euglycemic clamps) or 12 mM (hyperglycemic clamps). Insulin was infused at 10 mU·m⁻² BSA·min⁻¹ (lower insulinemic clamps) or 40 mU·m⁻² BSA·min⁻¹ (high insulinemic clamps). In the H_insuE_gluc clamp, insulin infusion was increased to 100 mU·m⁻² BSA·min⁻¹ at T = 2 hours (“Study protocol”). LPS was infused at T = 0 hours. Measurements were performed at indicated time points.

Figure 2

Plasma glucose and insulin concentrations. Mean (± SE) plasma levels of glucose (A) and insulin (B), after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). Glucose was measured every 5 minutes; insulin was measured at time points as described in “Study protocol.”

Figure 3

Neutrophil functions. Mean (± SE) plasma levels of elastase (A) and MPO (B), and whole blood neutrophil numbers (C) after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). *P < .05, **P < .01, and ***P < .001 hyperglycemic versus euglycemic clamps.

Figure 4

Cytokine concentrations. Median (± IQR) plasma levels of TNF-α (A), IL-6 (B), IL-8 (C), and IL-10 (D) after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). P = NS.

Figure 5

Endothelial cell activation. Mean (± SE) plasma levels of soluble E-selectin (A) and von Willebrand factor (B) after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). P = NS.

Figure 6

Coagulation markers. Mean (± SE) plasma levels of TATc (A) and soluble TF (B) after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). *P < .05 and ***P < .001 hyperglycemic versus euglycemic clamps; †P < .05 hyperinsulinemic versus lower insulinemic clamps.

Figure 7

Fibrinolysis markers. Mean (± SE) plasma levels of PA activity (A), PAPc (B), t-PA antigen (C), PAI-1 antigen (D), and PAI-1 activity (E) after LPS administration (4 ng/kg intravenously, T = 0 hours), during one of the following conditions: an L_insuE_gluc clamp (□), an L_insuH_gluc clamp (○), an H_insuE_gluc clamp (■), and an H_insuH_gluc clamp (●). *P < .05, **P < .01, and ***P < .001 hyperinsulinemic versus lower insulinemic clamps; †P < .05 hyperglycemic versus euglycemic clamps.