Clinical Chronobiology: A Timely Consideration in Critical Care Medicine

Helen McKenna; Gijsbertus T. J. van der Horst; Irwin Reiss; Daniel Martin


Crit Care. 2018;22(124) 

In This Article

Circadian Dysrhythmia in Critically ill Patients

Critically ill patients are particularly susceptible to circadian disruption (dysrhythmia) due to mistiming, or total loss, of sensory cues disturbing the master regulation within the environment of the intensive care unit, and/or pathology affecting the peripheral clock mechanism at a cellular level. The latter phenomenon is imperfectly understood, but may relate to the inflammatory response, with septic patients seen to have abnormal patterns of melatonin secretion relative to non-septic ventilated patients.[36] The cellular effect outlasts the septic insult, with animal models of sepsis demonstrating rhythm disturbances for weeks after the insult,[37] findings mirrored in human survivors of sepsis even after discharge from the intensive care unit. Rhythmicity can be affected in different ways, leading to a loss of amplitude, asynchronous timing or degradation to a completely chaotic pattern.[38] The clinical consequences of this aspect of critical illness are thought to be manifold but may have been neglected in the past due to the more obvious and immediately life-threatening features of physiological instability. The two most obvious manifestations of circadian dysrhythmia on the intensive care unit are the two complex and interrelated phenomena of sleep disturbance and delirium (Figure 1). Sleep disruption is one of the commonest afflictions reported by survivors of critical illness.[39] It manifests as fragmentation and disruption of sleep architecture,[40] and a displacement of the greater proportion of sleep to the daytime.[41] Disturbed sleep is not only a manifestation of circadian dysthymia, but also a driver of further disarray in peripheral clocks,[42] creating a vicious circle of disruption. Of the many downstream effects of sleep disruption, its impairment of the immune system has the most obvious ramifications for recovery from critical illness.[43] Evidence is also emerging that ICU delirium, a syndrome independently associated with increased mortality and long-term morbidity from critical illness,[44] and affecting up to 80% of critically ill patients,[45] is a clinical manifestation of circadian dysrhythmia (Figure 1). Delirium is associated with a reduction in peak amplitude of a urinary metabolite of melatonin (6-sulfatoxymelatonin).[46] Circadian disruption and sleep disturbance often precedes delirium, this temporal link suggesting a possible causal relationship.[47]

Figure 1.

Venn diagram depicting inextricable relationship between circadian rhythm, sleep deprivation and delirium in critically ill patients

Many of the effects of critical illness circadian dysrhythmia may not be detectable clinically, but the diverse ramifications we see in chronic circadian dysrhythmia, outlined earlier, highlight the degree to which cellular function in different tissues is affected, beyond the overt symptoms of sleep disturbance and cognitive impairment. For example, mitochondrial oxidative phosphorylation itself exhibits circadian oscillation,[48,49] disruption of which may result in mismatching of cellular energetics to demand. In patients with sepsis where cellular bioenergetic failure is already threatened[50] and drugs such as propofol suppress mitochondrial respiration,[51] such a discrepancy may be catastrophic for cell function. We propose that "chronofitness", the correct alignment of peripheral and central clocks, should be a new target in the management of critical illness. The first step will be to identify and address the modifiable risk factors associated with the disruption of circadian rhythms in critically ill patients (Table 2), and implement strategies to prevent such disruption. Here we outline the rationale and evidence for a number of chronobiological strategies to improve outcome in critically ill patients.

The Intensive Care Unit Environment

Inside the intensive care unit, the usual schedule of Zeitgebers is obliterated. In the outside world, the variation between light and darkness ranges from 0.0001 lx on a moonless night to 0.25 lx during a full moon, 1000 lx on the most overcast day and 130,000 lx in brightest sunshine. At night, artificial light leaches into patient areas from many sources – room and corridor lights, the glare from monitors and torches used to measure pupil size (leading to mean night-time illumination of 10 lx in one study[52]) – whilst during the day indoor illumination rarely reaches that of the outside world even on an overcast day (mean daytime illumination 158 lx[52]). Excessive night-time light suppresses the release of melatonin, a key molecule orchestrating circadian rhythmicity in different tissues,[53] and circadian disruption can be seen at a cellular level when cells are exposed to constant light.[54] Animal models of critical illness have demonstrated worse outcomes in the presence of circadian disturbance experimentally induced by constant exposure to light.[55] Imposing a light/dark cycle in a neonatal ICU accelerated body weight gain and shortened time to discharge in preterm infants.[56] "Chronofriendly" ICUs should aim to emulate normal day-time illumination levels with large windows, sufficient artificial lighting and perhaps incorporating blue-enhanced lighting in refurbishments and new builds.[57] A recent review summarised a number of randomised controlled trials in which morning bright-light therapy reduced delirium or improved sleep in acutely ill patients.[58] In a study published after this review, light therapy did not reduce the incidence of delirium in the critically ill and it is possible that it was less effective during early critical illness when patients were heavily sedated.[59] At night, providing sufficient light to perform critical tasks safely must be balanced against the negative effects of interruption of the patient's light/dark cycle. Where possible, night light should be minimised or the patient's eyes protected using eye masks. The use of lights and monitors that emit red rather than blue light at night may blunt inhibition of melatonin secretion.[60] Standards for morning light intensity during the day and night using luxmeters could be initiated.

Excessive night-time noise leads to sleep deprivation.[61] Noise reduction can be achieved by staff behaviour modification and the use of ear plugs in patients; the latter showing promising signs of being able to reduce ICU delirium.[61] Measurement of sound levels on ICU may help staff to understand which aspects of patient care create the most noise and facilitate the development of local guidelines to help minimise environmental noise pollution.[62] Other disturbances such as automated non-invasive blood pressure measurement, physical examination, turning and washing should be rationalised and cohorted at night, which as part of a bundle of interventions has been shown to improve sleep in the critically ill.[63]


The consumption of food is probably the most powerful Zeitgeber for peripheral clocks. Most mammals do not eat at night and enter a period of natural fasting, during which time they switch from using primarily glucose as a fuel source to ketone bodies.[64] Yet it remains common practice to feed critically ill adults continuously, whether via the enteral or parenteral route. Time-restricted feeding is a well-described intervention that has a profound effect on the circadian rhythm and may be a powerful tool in the prevention of metabolic disorders.[65,66] Much of the benefit seen in animal models exploring this line of work appears to relate to the maintenance or entrainment of the circadian rhythm.[67] Numerous animal model studies have demonstrated that changes to normal feeding schedules have significant metabolic consequences,[68] including reduction in the amplitude and an alteration of phase in hepatic metabolite levels.[69] Daytime feeding in nocturnally active mice shifted the liver clock by 180° and resulted in significantly decreased survival rates in a sepsis model, in comparison to mice fed at night.[70] Eating out of phase with the SCN rhythm destroys the normal relationship between this central timekeeper and the peripheral clocks, and the resulting conflict manifests as gastrointestinal and metabolic disease.[71] This has been proposed as a mechanism for the increased incidence of obesity and metabolic disorders in shift workers.[72] A further interesting twist is the bi-directional circadian relationship between the gut microbiome and its host,[73] with the circadian rhythm of one influencing the other. Whilst studies have been undertaken to look at outcome differences between bolus versus continuous feeding in subsets of critically ill patients,[74,75] they did not consider circadian factors in their design. A recent and eloquent personal perspective on this topic by Paul Marik[76] highlighted that providing continuous protein or amino acid supplementation limits skeletal muscle protein synthesis whilst intermittent feeding promotes it. His article concludes by stating: "Continuous tube feeding is unphysiological and likely harmful and should be abandoned".[76]

Physical Activity

Loss of daytime physical activity may be absolute in severe critical illness, as the result of sedation, paralysis, muscle weakness or critical illness polyneuropathy or myopathy. Posture on the ICU is predominantly supine, rather than upright, for much of the day. There is a chronobiological argument for incorporating early basic patient mobilisation, even passively, through physiotherapy, at the times relevant to that patient. Some have suggested that early mobilisation may contribute to circadian health, and early mobilisation has been suggested as being an essential component of any strategy to reduce delirium in the critically ill.[77] Individual chronotyping (through questioning relatives) may be especially valuable here; perhaps increases in compliance with valuable physiotherapy could be achieved in awake patients, by targeting their morningness or eveningness.


Sedatives can be a necessary evil in critical care medicine, but their side-effect profile may extend beyond our current awareness.[78] Administration of sedatives to the critically ill worsens sleep patterns and delirium, and drives circadian dysrhythmia,[79,80] with desynchronisation of the rhythmical secretion of melatonin demonstrated in sedated, mechanically ventilated patients.[81] Benzodiazepines appear to be the greatest offenders in this group.[81,82] When possible, minimising the use of sedative drugs is likely to be beneficial for critically patients,[78] and will require the optimisation of analgesia. In particular, it is advisable to avoid the knee-jerk reaction to treat sleep disturbance with sedative medications, which do not provide the active biological functions of sleep, and exacerbate sleep disturbance and delirium in the critically ill.