This study is looking at part of the stress response in those who are critically ill after cardiac surgery. Most patients will recover quickly after cardiac surgery and will be discharged home within a week. Some will not and remain on CICU sedated, ventilated and needing powerful drugs to maintain their blood pressure. The cause of this is usually the ‘inflammatory response’ – a set of reactions which when localised (such as a twisted ankle) promote rapid healing of the injured part. When the injury is larger (such as major surgery, burns or infection), this set of reactions can occur in all tissues of the body leading to inflammation of the whole body with subsequent multiple organ dysfunction and in some cases death. One of the hormones that protects against an uncontrolled inflammatory response is the steroid hormone cortisol. Doctors often give synthetic cortisol type drugs to patients on the ICU in an attempt to reduce this inflammatory response. However, if used in too large a dose, they can have significant side effects such as immunosuppression, poor wound healing and a diabetic state. There is still disagreement over whether steroids should be used and if so, how much.

People produce their own cortisol in a diurnal rhythm (it is high when you first wake in the morning and lowest in the late afternoon – this is why people feel sleepy at around 4pm). This rhythm is actually produced by discreet pulses (an ultradian rhythm). However, no one has ever looked at what happens to these pulses during critical illness. We think that both the size and frequency of these pulses is important for correct hormone functioning. It has been shown that these pulses change dramatically when a patient is chronically unwell and early results from current work we are doing shows large derangements in those having routine cardiac surgery. This study is designed to see what happens to these pulses when people are critically ill. Once we know what is ‘normal’ during critical illness, we can begin to give steroids to patients in a more tailored way.


The vast majority of patients who have cardiac surgery recover rapidly and are discharged home within a week. Some do not and become critically ill. They spend prolonged periods on CICU sedated and ventilated, require inotropes and vasopressors to maintain their cardiac function and blood pressure and suffer acute kidney injury 1 requiring haemofiltration. A number of aetiologies contribute to this state including major surgery itself, massive blood transfusion, infection and reperfusion. Whatever the cause, the final common pathway is that of systemic inflammation. Activation of leukocytes leads to release of cytokines and eicosanoids prompting a cascade of events that results in vasodilation, loss of endothelial barrier function and oedema, myocardial depression and perturbation of microvascular flow – particularly in the lungs and kidneys, where it causes renal and pulmonary dysfunction and failure.

One of the key hormones that protects against systemic inflammation is the steroid hormone cortisol – part of the hypothalamic-pituitary-adrenal (HPA) axis. Humans with a pre-existing poorly functioning HPA axis succumb rapidly to even minor illness, but steroid replacement protects against these challenges. Thus, activation of the HPA axis is vital in illness, although prolonged activation can become maladaptive. Indeed, treatment with synthetic glucocorticoids results in a plethora of unwanted side effects including susceptibility to infection, poor wound healing and blood glucose derangements.

It has long been known that cortisol is secreted in a circadian fashion. However, it has only recently been discovered that underlying this is an ultradian rhythm in which secretion of both ACTH and cortisol occurs in a pulsatile manner2 3-5(Fig 1). It is actually the changes in cortisol pulse amplitude, and to a lesser extent frequency, that makes up the circadian rhythm. These pulses are critical for signal transduction 6 – trough levels between the pulses allow the glucocorticoid receptor (GR) to regenerate and maintain activity. Disruption of normal pulses appears to be an important factor for glucocorticoid mediated side effects in conditions such as obstructive sleep apnoea7 and in some of our ongoing work we are beginning to show that pulse characteristics change immediately after routine cardiac surgery, although we have little information about what happens when patients are critically ill.

We know that tissue levels of the free (and biologically active) glucocorticoid correlate very closely with the changes in plasma levels8 and that changes in pulse characteristics have dramatic effects on the level of glucocorticoids within tissues. With increased pulse frequency, mean plasma cortisol levels may not change greatly, but since at the peak of each pulse the concentration of cortisol usually exceeds the capacity for cortisol binding globulin there will be a disproportionately great effect on tissue cortisol levels. It is clear therefore, that both in health and disease it is not just the amount of cortisol that is secreted, but the way in which it is secreted that is important for physiological function.

Our group has shown that ultradian pulsatiltity continues throughout the peri-operative period of cardiac surgery.

Adrenal dysfunction in critical illness is thought to be both common and important, although there remains significant disagreement over its diagnosis and management. Partly, this is because the thresholds for ‘normality’ and ‘insufficiency’ are unknown. This is not surprising given the fact that there have been no studies on the endogenous dynamics of cortisol secretion in the critically ill. Previous work has been limited to four-hourly cortisol measurements and short synacthen tests, both of which are totally inadequate for the assessment of ultradian rhythms1


This study is designed to describe changes in the ultradian rhythms of cortisol that occur when patients are critically ill.

Return to trial summary