Glucocorticoid treatment increases cholesterol availability during critical illness: effect on adrenal and muscle function

Abstract

Hypocholesterolemia hallmarks critical illness though the underlying pathophysiology is incompletely understood. As low circulating cholesterol levels could partly be due to an increased conversion to cortisol/corticosterone, we hypothesized that glucocorticoid treatment, via reduced de novo adrenal cortisol/corticosterone synthesis, might improve cholesterol availability and as such affect adrenal gland and skeletal muscle function. In a matched set of prolonged critically ill patients (n = 324) included in the EPaNIC RCT, a secondary analysis was performed to assess the association between glucocorticoid treatment and plasma cholesterol from ICU admission to day five. Next, in a mouse model of cecal ligation and puncture-induced sepsis, septic mice were randomized to receive either hydrocortisone (1.2 mg/day) (n = 17) or placebo (n = 15) for 5 days, as compared with healthy mice (n = 18). Plasma corticosterone, cholesterol, and adrenocortical and myofiber cholesterol were quantified. Adrenal structure and steroidogenic capacity were evaluated. Muscle force and markers of atrophy, fibrosis and regeneration were quantified. In a consecutive mouse study with identical design (n = 24), whole body composition was assessed by EchoMRI to investigate impact on lean mass, fat mass, total and free water. In human patients, glucocorticoid treatment was associated with higher plasma HDL- and LDL-cholesterol from respectively ICU day two and day three, up to day five (P < 0.05). Plasma corticosterone was no longer elevated in hydrocortisone-treated septic mice compared to placebo, whereas the sepsis-induced reduction in plasma HDL- and LDL-cholesterol and in adrenocortical cholesterol was attenuated (P < 0.05), but without improving the adrenocortical ACTH-induced CORT response and with increased adrenocortical inflammation and apoptosis (P < 0.05). Total body mass was further decreased in hydrocortisone-treated septic mice (P < 0.01) compared to placebo, with no additional effect on muscle mass, force or myofiber size. The sepsis-induced rise in markers of muscle atrophy and fibrosis was unaffected by hydrocortisone treatment, whereas markers of muscle regeneration were suppressed compared to placebo (P < 0.05). An increased loss of lean body mass and total and free water was observed in hydrocortisone-treated septic mice compared to placebo (P < 0.05). Glucocorticoid treatment partially attenuated critical illness-induced hypocholesterolemia, but at a cost of impaired adrenal function, suppressed muscle regeneration and exacerbated loss of body mass.