Abstract
Sepsis is defined as any life-threatening organ dysfunction caused by a dysregulated host response to infection. It remains an important cause of critical illness and has considerable short- and long-term morbidity and mortality. In the last decades, preclinical and clinical research has revealed a biphasic pattern in the (neuro-)endocrine responses to sepsis as to other forms of critical illness, contributing to development of severe metabolic alterations. Immediately after the critical illness-inducing insult, fasting- and stress-induced neuroendocrine and cellular responses evoke a catabolic state in order to provide energy substrates for vital tissues, and to concomitantly activate cellular repair pathways while energy-consuming anabolism is postponed. Large randomized controlled trials have shown that providing early full feeding in this acute phase induced harm and reversed some of the neuro-endocrine alterations, which suggested that the acute fasting- and stress-induced responses to critical illness are likely interlinked and benefical. However, it remains unclear whether, in the context of accepting virtual fasting in the acute phase of illness, metabolic alterations such as hyperglycemia are harmful or beneficial. When patients enter a prolonged phase of critical illness, a central suppression of most neuroendocrine axes follows. Prolonged fasting and central neuroendocrine suppression may no longer be beneficial. Although pilot studies have suggested benefit of fasting-mimicking diets and interventions that reactivate the central neuroendocrine suppression selectively in the prolonged phase of illness, further study is needed to investigate patient-oriented outcomes in larger randomized trials.