Mealtime: A circadian disruptor and determinant of energy balance?

Circadian rhythms play a critical role in the physiological processes involved in energy metabolism and energy balance (EB). A large array of metabolic processes, including the expression of many energy‐regulating endocrine hormones, display temporal rhythms that are driven by both the circadian clock and food intake. Mealtime has been shown to be a compelling zeitgeber in peripheral tissue rhythms. Inconsistent signalling to the periphery, because of mismatched input from the central clock vs time of eating, results in circadian disruption in which central and/or peripheral rhythms are asynchronously time shifted or their amplitudes reduced. A growing body of evidence supports the negative health effects of circadian disruption, with strong evidence in murine models that mealtime‐induced circadian disruption results in various metabolic consequences, including energy imbalance and weight gain. Increased weight gain has been reported to occur even without differences in energy intake, indicating an effect of circadian disruption on energy expenditure. However, the translation of these findings to humans is not well established because the ability to undertake rigorously controlled dietary studies that explore the chronic effects on energy regulation is challenging. Establishing the neuroendocrine changes in response to both acute and chronic variations in mealtime, along with observations in populations with routinely abnormal mealtimes, may provide greater insight into underlying mechanisms that influence long‐term weight management under different meal patterns. Human studies should explore mechanisms through relevant biomarkers; for example, cortisol, leptin, ghrelin and other energy‐regulating neuroendocrine factors. Mistiming between aggregate hormonal signals, or between hormones with their receptors, may cause reduced signalling intensity and hormonal resistance. Understanding how mealtimes may impact on the coordination of endocrine factors is essential for untangling the complex regulation of EB. Here a review is provided on current evidence of the impacts of mealtime on energy metabolism and the underlying neuroendocrine mechanisms, with a specific focus on human research.

signalling to the periphery, because of mismatched input from the central clock vs time of eating, results in circadian disruption in which central and/or peripheral rhythms are asynchronously time shifted or their amplitudes reduced. A growing body of evidence supports the negative health effects of circadian disruption, with strong evidence in murine models that mealtime-induced circadian disruption results in various metabolic consequences, including energy imbalance and weight gain.
Increased weight gain has been reported to occur even without differences in energy intake, indicating an effect of circadian disruption on energy expenditure. However, the translation of these findings to humans is not well established because the ability to undertake rigorously controlled dietary studies that explore the chronic effects on energy regulation is challenging. Establishing the neuroendocrine changes in response to both acute and chronic variations in mealtime, along with observations in populations with routinely abnormal mealtimes, may provide greater insight into underlying mechanisms that influence long-term weight management under different meal patterns. Human studies should explore mechanisms through relevant biomarkers; for example, cortisol, leptin, ghrelin and other energy-regulating neuroendocrine factors. Mistiming between aggregate hormonal signals, or between hormones with their receptors, may cause reduced signalling intensity and hormonal resistance.
Understanding how mealtimes may impact on the coordination of endocrine factors is essential for untangling the complex regulation of EB. Here a review is provided on current evidence of the impacts of mealtime on energy metabolism and the underlying neuroendocrine mechanisms, with a specific focus on human research.

K E Y W O R D S
chrononutrition, circadian disruption, circadian rhythms, energy balance, energy expenditure

| INTRODUC TI ON
The influence of circadian rhythms on energy balance (EB) has become a topic of increasing interest with a new-found pursuit to identify whether meal timing or energy distribution across the day can impact on weight management and metabolic health. Life on earth is characterised by continuous rhythms arising from evolutionary adaptations to earth's natural 24-hour light/dark cycle. As humans, we have evolved an active light phase primarily designed for energy replenishment, reproduction and activity, and an inactive dark phase in which to sleep, recover and regenerate. To achieve these daily cycles, input is required to inform the body of the time of day, and outputs are required to relay this information between central and peripheral tissues. 1,2 New research in murine and human models highlights the importance of circadian rhythms with respect to regulating energy metabolism, and the metabolic health consequences that may occur from disruption of these rhythms. How time of eating results in changes in clock genes and then impacts on metabolic health is currently not well defined. Specifically, the effect of meal timing on energy expenditure (EE) and EB remains controversial. Figure 1 illustrates the complex regulation of EB, with temporal input from the central clock (in the brain), activity and feeding, as required to synchronise the temporal excretion of neuroendocrine hormones, which in turn regulate energy intake (EI) and EE. Here, we review current evidence regarding the influence of circadian rhythm disruption, and specifically differences in mealtime, on EE with a focus on underlying endocrine mechanisms involved in energy regulation. Because the term circadian disruption is not clearly defined, we address the effects of potent desynchronisation protocols as circadian disruptors, as well as the more subtle effects of altering mealtime, where the extent of circadian disruption may be less obvious. The first part of the review addresses the effects of circadian disruption on whole body EE. The second part addresses the underlying endocrine changes that may contribute to chronic alterations in energy regulation.

| CIRC AD IAN RHY THMS AND THEIR ROLE IN ME TABOLIS M
The suprachiasmatic nucleus (SCN) is located in the hypothalamus within the brain and is the primary regulator of circadian rhythms.
The SCN receives photic input from the retina, relaying temporal information to the brain and peripheral tissues. The SCN maintains a self-sustaining 24-hour rhythm with output to peripheral tissues sent via neural (autonomic), hormonal (hypothalamo-pituitary) and Specifically, many of the functions involved in energy metabolism and regulation of EB are under strong circadian control and have been reviewed in detail previously. 5,6 In addition to photic inputs, other factors, such as food (both quality, quantity and timing) and physical activity, can act as entrainment cues (zeitgebers). Although the timing of activity can induce phase-shifts in the master clock, 7,8 feeding time primarily influences peripheral clock timing with little to no effect on the SCN. [9][10][11] Therefore, in an interactive loop, circadian rhythms can drive EI and regulate energy metabolism, yet energy intake and activity can also influence the timing of clock genes and their local tissue activity ( Figure 1).
Regular circadian rhythms help to maintain normal body functions and enable anticipation of events required for survival, including regulation of the timing of sleep, activity, digestive processes and metabolism (both storage and breakdown of fuel sources). 1,2 Both central and peripheral rhythms are evident in many key metabolic processes, including regulation of EB, from the most basic cellular level though to whole body energy metabolism. For example, circadian rhythms can influence genes and gene products involved in rate-limiting steps of cellular metabolism. A good example is the supply of NAD +, which exhibits a daily rhythm as a result of circadian oscillations in nicotinamide phosphoribosyltransferase (NAMPT), which controls a rate-limiting step in the salvage of NAD+. 12,13 Various endocrine signals involved in the regulation of energy metabolism display circadian oscillations 6 and circadian variations in overall whole body EE and macronutrient balance have been observed as a result of oscillations in preferential nutrient uptake and the use of macronutrients at specific times of the day. 5,6 Circadian variation in resting metabolic rate (RMR) has been observed in a number of human studies, with RMR or CO 2 production peaking around 5.00 to 6.00 pm, and with a trough at approximately 5.00 am. 14, 15 The thermic effect of food (TEF) has also been reported to be greater in the morning compared to the evening. [16][17][18][19] The reasons for this are not clearly understood, although it was suggested to be the result of many of factors, including insulin resistance in the evening, 20-22 reduced nutrient uptake in the evening resulting in lower energy intensive processes such as hepatic and muscle glycogen synthesis, which both display diurnal variation and peak during the active phase, [23][24][25] as well as lower rates of futile substrate cycling and/or reduced rates of protein turnover. 16 Following the consumption of identical meals, glucose, insulin and free fatty acid levels are reportedly elevated in the evening compared to the morning, indicating a lower uptake and storage of nutrients. 19 This may in part explain the lower evening TEF measured in some studies. Furthermore, circadian clock gene regulation in peripheral tissues is likely to be responsible for reducing nutrient absorption in the evening, allowing for more readily available fuel to prepare for the onset of fasting.
Nutrient oxidation also appears to be under circadian control, with higher carbohydrate oxidation in the morning and greater fat oxidation in the evening. 15 These findings are evidence of the robust circadian regulation involved in processes of energy metabolism.

| ME ALTIME A S A ZEITG EB ER
Meal timing is a potent zeitgeber in peripheral clocks. This is based not only on the time of eating, but also nutritional cues. Macronutrients F I G U R E 1 Circadian influences in the regulation of energy balance. Neuroendocrine hormones responsible for regulation of energy balance are excreted from multiple peripheral organs at specific times of the day in response to temporal input from neural, hormonal (predominantly melatonin) and behavioural signals from the suprachiasmatic nucleus (SCN). Hormonal signals relay messages within the periphery and to the energy regulatory centres in the arcuate nucleus (ARC) in the hypothalamus where temporal expression of receptors allows for effective signalling. Within the ARC, activation of the neuropeptide Y (NPY)/agouti-related protein (AgRP) and proopiomelanocortin (POMC)/cocaine-and amphetamine-related transcript (CART) neurones express their respective neuropeptides to increase or decrease appetite and reduce or increase energy metabolism and energy expenditure respectively. Mealtime and activity can also influence the timing and amplitude of peripheral clock genes and hence timing of hormone secretion. Differential input from energy intake (EI)/energy expenditure (EE) and the SCN may result in circadian desynchrony with either misaligned or attenuated circadian rhythms. Chronic misalignment may result in energy imbalance through dysregulation of peripheral energy metabolism and dysregulated signalling to the brain in relation to EE and EI  HFD. 37 The translation of such findings to humans is currently limited and likely to be a feature of future research. However studies reporting no changes [53][54][55][56] or decreases in body weight 57,58 are often overlooked and, indeed, there are large variations in energy intake, expenditure and balance between studies in response to desynchronisation protocols and specific outcomes may be species specific. Despite this, most studies do show a level of metabolic disturbance including reduced EE and impaired glucose tolerance. 53,54,57 In human studies, the effects of circadian desynchrony on EE and EB are also unclear. Imposing desynchrony through long or short days had no effect on total daily EE. 59 however, when broken down by gender, females actually showed a significant increase in fasting and postprandial EE during misalignment which was not seen in males. 65 However, in these studies, shift, may differentially influence those at risk of weight gain. 69,70 However, regardless of maintaining EB, metabolic health implications may occur in response to abnormal meal timing and circadian misalignment. Many of these studies show differences in weight gain despite similar calorie intake, indicating that mealtime must alter EB through altered EE. Although the effects of these extreme shifts in the time of feeding found in rodent studies may be extrapolatable to shift workers, there is a need to understand the more subtle effects of shifting meals, such as from earlier to later in the day. 81 Mealtime studies in rodent models have begun to address these more subtle effects of time of eating and redistribution of energy intake, across the normal wake-phase, on body weight and EB. One study reported that skipping dinner resulted in significantly lower body weight compared to skipping the first meal of the day or having three meals per day. 82 Significantly greater weight gain has been observed in mice that were subjected to a 4-6-hour delay in the onset of wake-phase feeding compared to mice allowed to eat ad lib. 83, 84 Yoshida et al 84 confirmed that this weight gain came from greater energy intake compared to mice allowed to eat ad lib. (ad lib. mice consumed 65% of their calories in the first 6 hours of their wake-phase). In a second protocol comparing delayed vs ad libitum feeding, the mice were subjected to energy restriction and matched energy intake. Despite identical intakes, the ad lib. feeding mice lost a greater amount of body weight compared to those in the delayed feeding group. This indicates that later meal timing can contribute to weight gain through reducing EE when energy intake is controlled, as well as through increasing energy intake where food is provided ad lib. Thus, studies in murine models tend to support the idea that eating earlier in the active phase can improve body weight regulation through both energy intake and EE.

| Meal timing on EB and EE
In humans, breakfast skipping under EB conditions has regularly been shown to have no effect on RMR or total daily EE compared to consumption of breakfast. [85][86][87] Although EE is lower in the biological morning, this is compensated for by higher EE later in the day and evening, indicating a redistribution of EE across the day. Indeed, respiratory chamber studies of total daily EE suggest that the lower evening TEF seen with breakfast skipping may be more apparent than real because TEF is actually lower but longer, continuing well into the night and thereby causing an apparently higher sleeping metabolic rate. 87,88 The effects of mealtime appear to be more evident in weight loss studies in which energy distribution is manipulated so that the majority of energy intake is consumed in the morning or consumption of lunch is earlier in the day, resulting in significantly greater weight loss. 89  Understanding the circadian nature of energy-regulating endocrine hormones, as well as how they respond to and influence circadian disruption, may provide a means to understand the impacts of desynchrony on metabolic health and EB. Table 1 illustrates the roles and rhythmicity of several key hormones involved in EB regulation.
Below we discuss the impacts of circadian disruption on endocrine regulation of EB.
Forced desynchrony protocols are a potent circadian disruptor that we can draw on to assess the effects of circadian disruption on endocrine hormones. Currently, the effect of circadian disruption on the expression of endocrine signals is inconsistent. One example is cortisol. In some cases, the cortisol pattern has tracked the changes in timing of the new behavioural cycle without any negative changes in the amplitude, profile or overall mean concentration. 59,61 However, one study found flattened rhythms with phase advanced protocols and suppression of mean concentrations with phase delay protocols. 60 92 However, they found no differences in fasting ghrelin, leptin, glucose or insulin between the groups. 182 Interestingly, they also looked specifically at neuronal circuits that regulate energy homeostasis and found significant increases in dopamine transporter in the striatum and serotonin transporter in the thalamus with early feeding and decreases with late feeding. They hypothesise that morning predominant calorie intake may impact on EB through positively reinforcing brain reward circuitries involved in the hedonic aspects related to food. 182 These findings could theoretically reduce overconsumption, improve dietary compliance and improve satiety. If true, these findings could display a more measurable impact on body weight regulation under ad lib.

| The effects of mealtime on endocrine factors
feeding conditions, where they could improve EB through mechanisms regulating EI more so than EE. Further research is required to understand how endocrine differences resulting from breakfast skipping and energy distribution may contribute to differential EB.
Time-restricted feeding has become a novel approach to potentially improve EB. Although, in humans, the negative EB and weight loss currently appear to be related to reduced energy in- • Reducing uptake of nutrients from the blood (partially by reducing insulin sensitivity) • Maintaining adequate concentrations of enzymes required for macronutrient breakdown • Stimulating the breakdown of nutrient reserves (increased lipolysis, proteolysis and glycolysis) • Converting energy to readily available forms (gluconeogenesis) Robust circadian pattern Peak in the early hours of the wake phase and a nadir in the evening/early night Regulated by circadian variation in ACTH release as well as circadian regulation of tissue specific sensitivity to ACTH in the adrenal gland Acute stressors and meals can stimulate ACTH and cortisol release. This response is generally superimposed above the 24-hour endogenous cycle without an overall effect on the phase or amplitude Highly regulated by the sleep/wake cycle. Decreases at the onset of sleep even during the day and higher cortisol at night during nocturnal awakenings Glucocorticoid receptors are located on various tissues indicating a role of cortisol as a messenger to peripheral tissues Receptors identified in the liver, muscle, pancreatic β-cells, white adipose and gut tissue Direct impact of glucocorticoids or dexamethasone, (a potent synthetic glucocorticoid) have been shown in synchronising, shifting or reducing the amplitude of local clock genes in the liver, skeletal muscle and white adipose tissue Insulin [123][124][125][126][127][128][129][130][131][132] Primary role is to regulate blood glucose levels. Peripherally, insulin is highly anabolic: • Attaches to receptor sites on target cells to enable the uptake of glucose from the blood stream • Enhances synthetic processes (fatty acid and triacylglycerol synthesis, protein synthesis, glycogenesis) • Suppressed catabolic processes (lipolysis, proteolysis, gluconeogenesis and glycogenolysis) Centrally, insulin acts as a deterrent to weight gain. Readily crosses the blood-brain barrier and binds to receptors in the ARC in the hypothalamus to repress food intake via: • Decreasing the expression of the orexigenic neuropeptides NPY and AgRP • Increasing the expression of anorexigenic neuropeptides POMC and CART Primarily released in response to glucose intake However, its secretion and receptor sensitivity show diurnal variation.
In healthy individuals, β-cell responsiveness to glucose and whole-body insulin sensitivity are impaired later in the day In obese individuals, rhythms in insulin sensitivity are attenuated, phase delayed or even absent, with inverted or absent rhythms reported in individuals with type 2 diabetes
Increases energy availability through processes in hepatocytes: • Increases glycogenolysis and gluconeogenesis • Decreases glycolysis and glycogenesis.
• Increases fatty acid breakdown in the liver and decreases lipogenesis.
• Promotes amino acid breakdown • Therefore, may reduce hepatic lipid accumulation May stimulate lipolysis in adipocytes in humans however only if insulin levels are low.

Possible roles in satiety
• May modulate satiety -has been shown to reduce food intake and inhibit hunger, likely through brain -liver axis • Although somewhat counterintuitive, it may be due to cross-reactivity with GLP-1 receptors.
• It may arise from hepatic metabolic changes or from glucagon working directly in the CNSPossible capacity to increase energy expenditure: Infused glucagon shown to increase EE in some human studies however the Effects of endogenous glucagon on EE remain unclear Secreted primarily in response to nutrient availability in the body. Eg hypoglycaemia, prolonged fasting, exercise and protein rich meals.

Role in energy balance
Circadian/diurnal or food related rhythm

Role in regulation of other endocrine signals
Ghrelin [145][146][147][148][149][150][151][152][153][154][155] Primary recognised for its effects in inducing hunger Centrally elicits hunger and food intake via: • Activating GHS-Rs in the hypothalamus which express NPY/AgRP and found that, after 8 weeks of restricting intake to between 1.00 pm to 9.00 pm compared to 8.00 am to 9.00 pm, there was a significant reduction in inflammatory markers and increased adiponectin. Adiponectin has anti-inflammatory functions, interacts with AMPK and stimulates peroxisome proliferator-activated receptor gamma coactivator 1α) protein expression and mitochondrial biogenesis. It can also act in the brain to increase EE and hence may have contributed to differential weight loss in the TRF group.
However, Sutton et al, 98 reported that early TRF for 5 weeks had no effect on inflammatory markers high-sensitivity C-reactive protein and interleukin-6, as well as no changes in cortisol. They also note a significant increase in 8-isoprostane (a marker of oxidative stress) in the control group which was not detected in the early TRF group. Therefore, early TRF may have been preventative   [193][194][195][196][197] This emphasises the need to consider individual responses to meal timing and circadian desynchrony and to focus on methods that will identify those most at risk of negative health consequences. This advocates a precision nutrition approach that current studies and healthcare models do not yet address. Furthermore, the effects of chronodisruption from late feeding may be a secondary effect of poor quantity and quality of sleep. Sleep loss can induce hypercortisolemia, elevated C-reactive protein, increased secretion of pro-inflammatory cytokines, [198][199][200] and reduced circulating levels or leptin and increase ghrelin. 201,202 In addition, the central and peripheral administration of certain neuropeptides (including insulin, CCK, ghrelin and leptin) can impact on the timing and quality of sleep. 203,204 Thus, food at night may increase circulating appetite-related hormones in the evening, subsequently impacting on sleep. This, in turn, could result in increased inflammation and circulating neuroendocrine hormones, which drive appetite and reduce EE.

| CON CLUS IONS
The evidence regarding the capacity of meal timing to cause circadian disruption, altered EB, and subsequent weight gain and metabolic disorders in human studies is inconclusive. Current findings suggest that differential mealtime can alter the excretion of many energy-regulating endocrine hormones through altering their temporal phase and amplitude or by suppression of entire rhythms.
Theoretically, this could contribute to desynchrony between synergistic hormones and their receptors, leading to reduced signalling and apparent hormone resistance. Presently, however, there is minimal evidence that these alterations in endocrine signalling are directly linked to any alteration in EE in human studies. Small reductions in TEF that have been reported with later meals appear to be balanced out by the redistribution of EE and no differences in total daily EE. There is no clear evidence of any effect of mealtime induced circadian disruption on EE in humans. However, the technical assessment of EE in humans is challenging and a lack of sensitive measures, changes in habits under laboratory and research settings, and individual human phenotypes contribute to the challenges in measuring such small effects. Despite this, some studies have reported greater weight loss with earlier eating and TRF protocols.
Further well controlled research is necessary that aims to understand whether meal timing (and the extent of mealtime differences) causes changes in mechanisms regulating EI and/or EE and whether there is further interplay between mealtime, macronutrient intake and specific populations.

ACK N OWLED G EM ENTS
This paper is linked to the UK Clock Club meeting discussion on 14 January 2019 for which we gratefully acknowledge funding support

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflicts of interest.

DATA AVA I L A B I L I T Y
Data sharing is not applicable to this article because no new data were created or analysed in this review.

PEER R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jne.12886.