Intermittent feeding and circadian rhythm in critical illness

Although the survival rates in critically ill patients are increasing worldwide, longer-term outcomes after admission in the Intensive Care Unit (ICU) are often poor, with up to 80% of patients suffering from long-term complications including impairments in sleep, physical function, and cognitive and psychological health [1]. Circadian rhythms, i.e., 24-h cycles, are central to physiological, psychological, and behavioural processes. Disruptions in circadian rhythms are associated with complications such as immune system disruption, delirium, long-term cardiovascular consequences, neurodegenerative diseases, type 2 diabetes mellitus, and increased mortality [2▪,3▪]. With the ICU environment being so drastically different from daily life with ongoing clinical and environmental changes, these disruptors likely contribute to impairments in circadian rhythms. Supporting circadian health in critically ill patients may help improve metabolism and reduce psychological health impairment and delirium during the post-ICU recovery phase. Therefore, it is essential to understand how critical illness affects circadian rhythms in order to develop intervention strategies and chronotherapy to minimise disruption of patients’ circadian rhythms in the ICU.

Nutritional support forms an essential part of standard clinical care in critically ill patients, thereby improving clinical outcomes. Although current nutritional guidelines [4,5] specify recommendations on the quantity and quality of the provided energy, macro- and micronutrients, strategies towards the timing and mode of feeding have been largely understudied. When food is consumed affects various physiological functions, including the sleep/wake cycle, core body temperature, (skeletal muscle) insulin sensitivity, whole-body metabolic health, and mental alertness. This has been referred to ‘chrononutrition’, i.e. synchronisation of eating with the body's entrained circadian rhythms, which has led to an enormous scientific and public interest in time-restricted eating diets; a dietary strategy that alters meal timing and incorporates more extended daily periods of fasting into the diet, without restricting the total energy intake. Time-restricted eating has been shown to reduce risk factors for type 2 diabetes mellitus and cardiovascular disease [2▪,6▪].

In the ICU, continuous and intermittent feeding (intermittent, bolus, or cyclic) are the most common enteral nutrition administration strategies. Continuous feeding is standard practice, as the slow release of nutrients into the stomach is thought to enhance feeding tolerance, reduce the risk of regurgitation, and lower respiratory complications, as well as being convenient. In contrast, intermittent feeding is more physiological as it mimics eating patterns in everyday life, thereby maintaining regular gastrointestinal hormone secretion and digestion, and it gives patients more mobility. Studies in animals and healthy humans [7–11] have suggested that intermittent feeding results in improved insulin sensitivity, increased muscle protein synthesis, activation of fasting-induced autophagy and ketogenesis, and the preservation of circadian rhythms in contrast to continuous feeding. However, in critically ill patients, only a handful of studies have directly compared the effect of intermittent versus continuous feeding on clinical outcomes, and these have been discussed in earlier reviews [12–17]. This review aims to provide an overview of studies published in the last 18 months on the effect of timing of nutritional support on metabolic outcomes in critically ill patients, with a specific interest in circadian alignment during and post-ICU admission.

Multiple physiological processes in peripheral tissues such as gastrointestinal function, muscle, and other vital organs are all under circadian regulation. The master regulator is in the hypothalamus's suprachiasmatic nucleus, primarily entrained by the light/dark cycle. At the molecular level, the circadian clock is based on the transcriptional/translational feedback loop of proteins such as Cryptochrome (CRY), Period (PER), Brain and muscle Arnt-like protein (BMAL), and Circadian Locomotor Output Cycles Kaput (CLOCK) that take ∼24 h to complete. However, nutrient signalling molecules directly regulate clock genes; activation of insulin-mTOR pathways increases the stability and translation of PER proteins [18,19,20^▪▪], whereas fasting activates AMP-activated protein kinase (AMPK) and nicotinamide phosphoribosyltransferase (NAMPT) pathways reducing the stability and transcription of CRY and PER [21–23]. In this way, changes in insulin and cAMP due to mistimed meals will influence hundreds of downstream ‘clock-controlled’ genes in peripheral tissues responsible for metabolic processes, including gastrointestinal function, glycaemic control, and muscle metabolism.

Environmental cues such as light/dark phase, temperature changes, and physical activity can synchronise the circadian clock with the external environment, with food consumption being the most potent entrainer for peripheral clocks. In the gut, nutrient uptake, gastric motility, gastric acid and gastrointestinal hormones production, nutrient absorption, and the gut microbiome are under circadian regulation [24^▪▪,25^▪▪]. Glucose metabolism is also under circadian control [26]; the hepatic clocks regulate glucose production, whereas the pancreatic clocks regulate insulin secretion according to time of day with much less secretory capacity at night. In contrast, the muscle clock regulates glucose uptake through reduced glucose transporter translocation at night versus day [27▪]. Moreover, circadian disruption can acutely impact glycaemic control through impairments in beta-cell function and peripheral insulin sensitivity [26].

Diurnal rhythms of central phase markers such as body temperature, blood pressure, heart rate, and sleep patterns in critically ill patients are highly disturbed during ICU admission [28–30] and continue to be disrupted for weeks after discharge [29,31^▪▪,32]. The pathophysiological response to critical illness might primarily drive this disruption in circadian rhythms, whereas non-physiological clinical factors, such as mechanical ventilation, medications, and sedation, may further contribute. Moreover, critical illness comes with pain, fatigue, stress, and cognitive dysfunctions such as delirium, which might further exacerbate circadian disruption [3▪]. ICU patients are exposed to frequent patient care interactions, noise, persistent light, and often to continuous enteral feeding, which are potentially modifiable factors that could mitigate circadian disruption. Therefore, further understanding of the extent of circadian disruptors and circadian health in ICU patients is needed.

A handful of studies have assessed the rhythmic expression of clock genes in critically ill patients [28,33–35], and the studies published in the last 18 months are summarised in Table 1. Studies that have quantified circadian rhythms of clock genes in critically ill patients early in ICU admission show no rhythmic expression of crucial clock genes compared to healthy controls [28,33]. However, these studies vary in patient population and are limited to neurology patients and patients with and without sepsis (type of patients not specified). Following admission, the time to inclusion is critical when circadian health is assessed, as circadian rhythm disturbance occurs rapidly [34]. In addition, baseline circadian health (i.e., at home or on the ward) might be variable and is an important confounding factor. For the assessment of circadian disruption (i.e. rhythmicity), the frequency of blood samples is critical for modelling analysis and varied in the published studies from 2- to 6 hourly in 24 h, with recommendations being 2-hourly with at least 4-hourly [36] so as not to underpower the analysis.

The relationship between circadian rhythm disruption and clinical and environmental factors in critical illness has been largely unexplored. Maas et al. published two additional papers (using the original larger dataset of n=112 critically ill patients [33]) to associate changes in clock genes with melatonin levels, light/dark phase, nutritional intake, and physical activity levels [37^▪▪,38^▪▪]. No associations were found between clock gene amplitudes and illness severity (SOFA scores), encephalopathy (Glasgow Coma Score), rest-activity rhythmicity (daily pattern of activity and rest), and melatonin levels. Low day-time light intensity levels, frequent nursing care, and night-time noise were highly prevalent. Nutritional intake (only available in n = 43/112 patients) was inadequate, with 39% receiving some bolus feeding (enteral or oral); however, no detailed nutritional intake data was reported. Physical activity levels in critically ill patients were drastically lower compared with ambulatory and bedrest healthy controls. No relation between feeding regime and/or physical activity levels and clock gene expression was made, so it remains yet uncertain how disruptors such as light, noise, nutrition, and/or physical acitivity affects circadian rhythm in ICU patients.

The optimal feeding mode for critically ill patients has become an ongoing debate in critical care nutrition. As reviewed previously, most continuous versus intermittent feeding [12–17] studies in critical illness are aimed to improve nutritional intake targets. To date, the few studies that have been conducted have included relatively small patient cohorts and have failed to show clear clinical benefit; well-controlled RCTs comparing metabolic effects of altered meal timing in critically ill patients remain scarce. Supporting evidence to understand the possible effect of intermittent feeding on glucose, gastrointestinal, and muscle metabolism in critically ill patients is discussed in the following sections.

Environmental entrainers such as sleep/wake phase, food intake, and physical activity can reset or re-align circadian clocks in peripheral tissues [2▪]. Eating during the inactive phase in animals completely inverts the expression of core clock genes in muscle, adipose tissue, and liver [48]. To date, this has been poorly investigated in humans. Only two studies have recently investigated the effects of limiting meal timing on circadian clocks through repeated tissue sampling in health. Lundell et al.[49^▪▪] showed that time-restricted eating did not alter clock genes in muscle, but changes in the circadian regulation of metabolites, including amino acids, were observed. Zhao et al.[27▪] took four repeated adipose tissue biopsies over 24 h and observed that time-restricted eating restored 3 out of 12 clock genes and rhythm to 450 genes in adipose tissue that were arrhythmic at baseline (personal communication, manuscript accepted, in preprint). Two other studies have reported that time-restricted eating induced changes in clock genes at different time points: time-restricted eating between 8 am and 2 pm decreased PER1 at 8 pm and increased CRY1/2 and RORα at 8 am and 8 pm [50]. Increased amplitude in BMAL1, CRY1, PER2, and RORα was also reported in white blood cells of patients with type 2 diabetes who ate three meals in 12 h versus six meals in 15 h [51].

No studies have been conducted to assess the impact of time of nutritional intake on circadian rhythms in critical illness. However, the time of day of meal ingestion affects the postprandial glucose response and shifting meal intake to earlier in the day improves glucose tolerance throughout the day in healthy adults and those with overweight or obesity [52,53]. In contrast to continuous feeding, intermittent feeding might reduce glucose intolerance and insulin resistance, which is relevant in critically ill patients, as up to 75% of patients show stress-induced hyperglycaemia [54]. The current clinical management of elevated glucose concentrations in the ICU is exogenous insulin administration with continuous enteral nutrition. However, intensive exogenous insulin therapy has been associated with negative consequences, including hypoglycaemic events, increased insulin administration, and increased mortality [55]. Moreover, continuous enteral nutrition to manage glucose levels is supported by limited evidence. Insulin (and IGF-1) has recently been recognised as a circadian entrainer and, as such, can serve as a primary signal of feeding time to cellular clocks throughout the body [19]. Therefore, intermittent feeding might be an effective strategy for preserving or re-aligning circadian rhythms, besides optimising glycaemic control. Moreover, incorporating overnight fasting periods has been suggested to be effective for metabolism, as research in healthy individuals has shown that fasting periods activate ketogenesis and autophagy [50]. A recent pilot study [56^▪▪] tested the feasibility of a period of fasting (alternating 12 h feeding with 12 h fasting) in 70 prolonged critically ill patients, showing that a 12 h nutrient interruption can initiate a metabolic fasting response by increased serum bilirubin and plasma beta-hydroxybutyrate and decreasing insulin requirements and serum IGF-I. Although the effect of intermittent or cyclic feeding on metabolism and clinical outcomes in critically ill patients needs further investigation, intermittent feeding regimes including overnight fasting periods might be relevant to preserve or even reset circadian misalignment. Other potential modifiable clinical and environmental disruptors for circadian rhythms in ICU patients, including light, temperature, physical activity, noise, sleeping medication, and nursing and medical interventions, are summarised in Fig. 1[57▪,58].

Studies in the last 18 months have further shown that intermittent feeding can increase nutritional intake. The suggested effect on improved glycaemic control, impaired gastric intolerance, and muscle mass maintenance compared to continuous feeding is minimal and based upon low-quality evidence. Ramifications on circadian misalignment in gastrointestinal, glucose, and muscle metabolism highlight the degree to which different tissues are affected, with studies in health showing that time-restricted eating can induce changes in peripheral clock genes. Chronotherapy, i.e., aligning meal timing, physical activity, and/or medication, are potential strategies to preserve or reset peripheral circadian rhythms; however, it is not known how these strategies can affect circadian rhythms in ICU patients. Interventional strategies to preserve circadian health could include the use of eye masks and earplugs, intermittent or cyclic day-time feeding, daily mobilisation, light-, and/or melatonin therapy during and post-ICU admission. However, the effect on metabolism and clinical outcomes is as yet unknown.

None.

Papers of particular interest, published within the annual period of review, have been highlighted as: