Optimum nutrition is the foundation of health and recovery from disease, encompassing the quality and quantity of macro- and micronutrients(Reference Echouffo-Tcheugui and Ahima1). Recently, the timing of nutrient intake about our circadian rhythms – approximately 24-h cycles in sleep-wake patterns – has emerged as a critical component of optimum nutrition. Circadian clocks, present in nearly every organ and brain region of mammals, regulate daily rhythms in behaviour, physiology and metabolism that are crucial for nutrition(Reference Bass2). These rhythms dictate daily patterns in hunger and satiety, influencing food-seeking and ingestive behaviours(Reference Scheer, Morris and Shea3).
In addition to regulating hunger, circadian rhythms play a significant role in various stages of digestion and nutrient absorption. These include daily fluctuations in saliva production, gastric acid secretion, intestinal motility and nutrient absorption(Reference Konturek, Brzozowski and Konturek4). Typically, periods of feeding align with the active or awake phase, while fasting corresponds with the rest or sleep phase. Metabolic processes related to nutrient metabolism, such as pancreatic and gut functions and the metabolism of carbohydrates, lipids and proteins in the liver, muscles and adipose tissue, also follow these circadian rhythms(Reference Bass2,Reference Bass and Takahashi5) .
The coordination of these metabolic rhythms suggests that restricting nutrient intake to specific time windows to better align with the circadian rhythms can optimise the benefits of altering food quality and quantity or mitigate the adverse effects of poor nutrition. Animal studies have demonstrated that time-restricted feeding (TRF) without altering food quantity can reduce the negative impacts of an unhealthy diet and enhance the positive effects of a healthy diet. This review paper explores the scientific foundation of TRF and reviews key findings from animal studies on its effects on various organ functions and related health outcomes.
Circadian rhythm
The concept of TRF is based on circadian rhythm. Circadian rhythms are regulated by cell-autonomous circadian timekeeping mechanism composed of a group of transcription regulators. Basic helix-loop-helix and Per-Arnt-Singleminded (PAS) domain-containing transcription factors CLOCK (or NPAS2)/BMAL1 heterodimer bind to cis-acting E-box in the promoters of their own repressors cryptochrome (Cry1, Cry2) and period (Per-1, -2, -3) and of nuclear hormone receptors Rev-erbs (Rev-erb-α, -β) and retinoic acid receptor-related orphan receptors (RORs) (Ror-α, -β, -γ). RORs and Rev-erbs act as activators and repressors, respectively, in the promoter of Bmal1 to drive its rhythmic expression(Reference Mohawk, Green and Takahashi6). Regulated transcription and post-transcriptional events involving a growing list of nuclear and cytoplasmic proteins generate a ∼24 h rhythm in the mRNA and protein level of most of these 13 different transcriptional regulators that are sustained even in the absence of any external timing cues(Reference Takahashi7). In addition to regulating each other’s expression, these transcription factors also drive the rhythmic expression of several target genes through their respective promoter binding sites. Altogether, a transcriptional basis of circadian rhythm enables a core set of regulators to temporally couple their activity with the synchronous rhythmic expression of hundreds or thousands of genes with peak expression at distinct times (phase) of the day(Reference Zhang, Lahens and Ballance8,Reference Mure, Le and Benegiamo9) .
The clock components are also regulated by the metabolic status of the cell. The stability and function of these clock components are regulated directly or indirectly by co-factors or ligands such as S-adenosyl methionine, alpha-ketoglutarate, NAD+/NADH, NADP+/NADPH, acetyl Co-A, O-GlcNAc and haem, whose steady-state levels are influenced by cellular metabolism(Reference Bass and Takahashi5,Reference Bass and Lazar10) . Additionally, nutrient-sensing pathways, including AMPK, mTOR and FoxO, also modulate the levels or function of several clock components, either directly or indirectly(Reference Panda11). This integration of cellular metabolism and circadian rhythm likely evolved to anticipate and respond to the predictable 24-h rhythm in nutrient availability in nature, programming our genome for a consistent window of time for nutrient ingestion.
Key research leading to time-restricted feeding
The concept of TRF emerged from extensive research in behaviour, metabolism, genetics and genomics, linking circadian rhythms to the timing of nutrition. It is well-established that almost all animals, from experimental models like Drosophila and rodents to humans, exhibit a daily rhythm in food intake that aligns with the active period of the 24-h day-night cycle. This rhythm persists even in constant darkness, without any external timing cues. Early observations highlighted that the timing of nutrition could distinctly affect cellular function, particularly in the hypothalamic brain region – the suprachiasmatic nucleus (SCN). For instance, when rats were fed 14C-labelled deoxy-glucose during their usual night eating time, the glucose tracer was barely detected in the SCN. However, the same tracer is given during the daytime strongly accumulated in the SCN, a pattern that continued even when lights were turned off during the day (subjective day)(Reference Schwartz and Gainer12). This experiment indicated a possible circadian rhythm in nutrient demand or uptake in some tissues.
Similarly, in humans, eating identical meals at breakfast, lunch or dinner results in different metabolic responses. Post-prandial glucose levels rise modestly in the morning but remain elevated for a significantly longer time after dinner(Reference Carroll and Nestel13). Even when healthy adults receive a constant glucose infusion over 24 h, blood glucose levels rise at night and fall around dawn(Reference Van Cauter, Desir and Decoster14). These findings suggest that the body’s metabolic responses are closely tied to the circadian rhythm, further supporting the concept that the timing of food intake can influence health outcomes.
After discovering circadian clock components and recognising their presence in all tissues, researchers became interested in understanding the molecular mechanisms by which circadian clocks regulate behaviour, physiology and metabolism. Concurrently, evidence accumulated showing disruptions in metabolic homeostasis in animals with loss-of-function mutations in clock components(Reference Rudic, McNamara and Curtis15,Reference Turek, Joshu and Kohsaka16) . Genomics studies cataloguing the transcriptional outputs of circadian clocks in different organs of mice fed a standard diet ad libitum revealed circadian rhythms in chromatin modifications and mRNA levels of numerous genes encoding metabolic regulators and nutrient-metabolising enzymes(Reference Panda, Antoch and Miller17–Reference Vollmers, Schmitz and Nathanson20). Circadian proteomics and metabolomics studies further supported these findings, highlighting the molecular basis of circadian regulation of metabolism(Reference Robles, Humphrey and Mann21–Reference Brown, Gaspar, Sassone-Corsi and Christen24).
Three additional sets of studies contributed to the development of the TRF concept. Mechanistic investigations into the circadian regulation of metabolism revealed that several nutrient-sensing pathways are reciprocally regulated by the circadian clock. Notably, the discovery that the fasting-sensing AMPK pathway(Reference Lamia, Sachdeva and DiTacchio25) and the feeding-sensing mTOR pathway(Reference Lipton, Yuan and Boyle26,Reference Crosby, Hamnett and Putker27) can influence the stability or function of clock components suggested that changes in nutrition timing could alter the timing or phase of the clock. This explained earlier observations in nocturnal mice, where restricting food access to the daytime shifted the phase of the liver’s circadian clock(Reference Damiola, Le Minh and Preitner28,Reference Stokkan, Yamazaki and Tei29) . Such ‘food resetting’ of peripheral clocks, along with ‘light-resetting’ of the SCN clock(Reference Hatori and Panda30), underscored how feeding patterns within the 24-h light-dark cycle affect the circadian system and metabolism.
Research has shown that animals exposed to constant light or chronic jet lag, due to random changes in the light-dark cycle, experience dampened circadian rhythms, disrupted activity-rest patterns and behavioural disorders(Reference Hatori and Panda30). This led to the hypothesis that frequent, round-the-clock food intake or random daily meal timing could weaken peripheral circadian clocks and their outputs, leading to compromised metabolism. Conversely, consolidating food intake into a consistent period could sustain robust circadian clocks and improve metabolic health.
When mice have ad libitum access to a standard diet, they typically consume over 80 % of their daily energy intake during the night(Reference Kohsaka, Laposky and Ramsey31) and exhibit circadian rhythms in the expression of 10–15 % of protein-coding genes(Reference Zhang, Lahens and Ballance8,Reference Panda, Antoch and Miller17) . However, when access to the standard diet is restricted to an 8-h window during the day, the total daily energy intake does not change, but the liver’s circadian clock realigns to daytime feeding. This consolidated feeding results in a substantial increase in the number of liver transcripts showing circadian rhythm. In contrast, mice fasted for 24 h show a dampened rhythm in liver circadian clock components and a dramatic reduction in the number of liver circadian transcripts(Reference Vollmers, Gill and DiTacchio32).
Parallel studies on diet-induced obesity (DIO) models further reveal the impact of feeding behaviour on liver circadian rhythm. Mice fed an obesogenic diet ad libitum shift their feeding behaviour to consume a relatively higher percentage of their diet during the day, reducing their night-time food intake(Reference Kohsaka, Laposky and Ramsey31). This alteration dampens the daily rhythm of food intake and subsequently dampens the liver’s circadian rhythm. These findings highlight how changes in feeding behaviour, driven by diet quality, can affect circadian rhythms in the liver.
Altogether, these complementary studies examining the effects of ad libitum feeding of a standard diet or a high-fat diet on liver circadian rhythms and the impact of feeding consolidation led to the idea that TRF could modulate the effects of nutrition quality and quantity. This suggests that not only what and how much we eat but when we eat can also significantly influence metabolic health and circadian regulation.
Time-restricted feeding and nutrition quality
As mentioned earlier, mice fed an obesogenic diet ad libitum exhibit a dampened circadian rhythm and are also known to develop metabolic diseases. To test the relative contribution of obesogenic diet and the disrupted eating pattern to obesity and metabolic diseases in this DIO model, we subjected mice to an isoenergetic obesogenic diet either ad libitum or within an 8 h TRF during the night(Reference Hatori, Vollmers and Zarrinpar33). These TRF rodents that eat the same calories as the ad libitum DIO cohort from the same food source exhibit improved molecular rhythms in circadian clock components and are largely protected from high-fat diet-induced obesity and related metabolic illnesses(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34) . It is noteworthy that all these mouse cohorts, irrespective of diet type and eating patterns (ad libitum or TRF), were housed under identical light:dark cycle, which presumably had the same entraining effects on the SCN clock. Even mice fed an obesogenic diet during daytime (day TRF) largely protected mice from obesity and metabolic illnesses(Reference Sherman, Genzer and Cohen35). Since the global rise in obesity and related co-morbidities is known to arise from obesogenic diet, the protective as well as therapeutic effects of TRF on body weight gain, adiposity, glucose intolerance, dyslipidaemia and steatohepatitis have raised a substantial interest in the translational potential of TRF in combating obesity and metabolic diseases.
Short-term TRF experiments lasting up to six months have demonstrated beneficial effects in mice fed a standard diet. TRF either sustained or increased muscle mass(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36–Reference Williams, Crown and Lyons38) , which is associated with better metabolic health. Additionally, long-term studies on mice fed a standard diet, where some form of TRF was inadvertently imposed, showed an increase in lifespan compared to mice fed an isoenergetic standard diet ad libitum (Reference Mitchell, Bernier and Mattison39). These studies of TRF on mice fed a standard diet revealed that TRF can further enhance the health benefits of a healthy diet.
Time-restricted feeding and nutrition quantity
A moderate reduction in calorie intake, known as calorie restriction (CR), is well-documented to confer several health benefits, including increased lifespan. TRF can enhance the lifespan-extending effects of CR. Standard CR protocols in rodents inadvertently impose time restrictions as well(Reference Acosta-Rodriguez, de Groot and Rijo-Ferreira40). Typically, the control group has ad libitum access to a standard diet, while the CR group receives a single daily meal with 10–30 % fewer calories than the control group’s daily intake. CR mice usually consume their food ration within less than 8 h, thus experiencing both CR and TR(Reference Acosta-Rodriguez, de Groot and Rijo-Ferreira40).
Controlled studies where CR mice are fed small meals distributed throughout the 24 h have shown a median lifespan increase of 10 % compared to ad libitum-fed (ALF) controls. However, TRF of a low-calorie diet during the night-time extends lifespan by up to 35 %(Reference Acosta-Rodriguez, Rijo-Ferreira and Izumo41,Reference Deota and Panda42) . These lifespan studies and the previously described TRF studies on mice fed a high-fat diet were conducted on the same strain of mice in independent laboratories. Altogether, these studies support the idea that TRF interacts with both the quality and quantity of nutrition. It can mitigate the adverse effects of a disease-promoting diet, enhance the benefits of a standard diet and amplify the lifespan-extending effects of CR.
Similar effects of TRF on nutrition quality and quantity have also been observed in Drosophila (Reference Gill, Le and Melkani43–Reference Guo, Livelo and Melkani48). These findings in both flies and rodents provide robust experimental systems to identify the molecular mechanisms underlying TRF’s impact on organ and organismal health, as well as to test related hypotheses.
Pleiotropic effects of time-restricted feeding
Systematic studies evaluating the effects of TRF on various disease models in laboratory animals have shown that TRF has pleiotropic effects across multiple organ systems. These include improvements in hypertension, glucose intolerance, cardiac function decline, metabolic dysfunction-associated steatotic liver disease/metabolic dysfunction-associated steatohepatitis, dysbiosis, dyslipidaemia, Huntington’s disease and Alzheimer’s disease(Reference Zarrinpar, Chaix and Panda49–Reference Zhang, Wang and Lin52). Most of these studies focused on the TRF of an obesogenic diet due to its obvious public health relevance as a potential tool to combat obesity and related metabolic diseases.
However, it is important to note that obesity does not always correlate with metabolic diseases, and vice versa. Animal studies have shown that TRF can reduce the risk of metabolic diseases even without significant changes in body weight or adiposity(Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Deota and Bhardwaj53) . The health benefits of TRF span various disease-relevant organ systems. The remainder of this article will concentrate on a few key metabolic organs, with a specific emphasis on the liver, to further explore the health benefits of TRF.
Time-restricted feeding and liver function
Human metabolic health is closely tied to the metabolism of glucose, fats (including free fatty acids and TAG) and cholesterol, with the liver playing a crucial role in managing all three. The following discussions are based primarily on time-series analyses of tissue or plasma samples along 24 h. Unlike the single time-point analyses typical of most metabolic studies, time-series analyses reveal molecular changes during both the fasted and fed states. Consequently, most references to changes in gene, protein and metabolite expression or function pertain to either the fasted or fed state.
Time-restricted feeding and the control of blood glucose
Both hormonal regulation of glucose (glucagon, insulin and incretins) and cell-autonomous regulation of glucose metabolism are modulated by both the circadian clock and nutrient-sensing mechanisms. Time-series analyses of gene expression, protein abundance, post-translational modifications and metabolite levels in rodents that undergo TRF have shed light on potential mechanisms by which TRF maintains glucose homeostasis.
TRF reduces insulin resistance through mechanisms that are not yet fully understood and are likely to involve multiple organs. Both fasting and post-prandial serum insulin levels are relatively high in high-fat and ALF mice, but TRF reverses this trend(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36,Reference Zhang, Wang and Lin52–Reference Zarrinpar, Chaix and Yooseph54) . Insulin activates mTOR, which, through a signalling cascade, phosphorylates ribosomal protein S6. This mTOR-initiated phosphorylation of S6 enhances the translation of mRNAs involved in ribosome biogenesis and protein translation and thereby boosts overall translation of proteins necessary for cell growth and division(Reference Zoncu, Efeyan and Sabatini55). In mice, which naturally feed at night, phospho-S6 levels are typically higher during the night(Reference Lipton, Yuan and Boyle26). However, mice fed a high-fat diet ad libitum exhibit an inverted pS6 rhythm, with higher levels during the day. TRF corrects this abnormal temporal activation of pS6, aligning the higher levels with the fed state and indicating normal post-prandial insulin signalling. During the fed state, improved insulin signalling and higher glucokinase (Gck) levels under TRF can direct glucose towards glycogen synthesis. Consequently, TRF mice accumulate increased levels of liver glycogen(Reference Hatori, Vollmers and Zarrinpar33). These coordinated changes in gene expression and metabolite levels suggest that TRF reprogrammes glucose metabolism away from gluconeogenesis and towards anabolic pathways.
Glucagon is a key pancreatic hormone that regulates hepatic gluconeogenesis. It acts through its receptor in the liver to trigger cAMP Response Element-Binding Protein (CREB) phosphorylation, which subsequently activates downstream target genes to promote gluconeogenesis, particularly during the fasted state(Reference Mayr and Montminy56). In ALF obese mice, however, CREB remains phosphorylated during both the fasted and fed states for unknown reasons. This persistent CREB phosphorylation during the fed state supports excessive gluconeogenesis, likely contributing to elevated blood glucose levels. Under TRF, liver pCREB levels are elevated during the fasted phase and decline during the fed state(Reference Vollmers, Gill and DiTacchio32,Reference Hatori, Vollmers and Zarrinpar33) . This desirable rhythm in CREB phosphorylation under TRF likely arises from at least two mechanisms. During the fed state, pancreatic alpha cells may reduce glucagon expression or release, and increased CRY expression in the liver can further suppress glucagon signalling downstream of its receptor(Reference Zhang, Liu and Dentin57). The reduced pCREB levels during the fed state correlate with decreased expression of target genes such as pyruvate carboxylase (Pcx), phosphoenolpyruvate carboxykinase 1 (Pck1) and glucose-6-phosphatase (G6pc), which mediate key steps in gluconeogenesis. This redirection reduces gluconeogenesis and shifts some pyruvate to the TCA cycle, leading to elevated levels of TCA cycle intermediates like malate, fumarate and citrate in the TRF liver(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Lin and Le36) .
Reduced expression of G6pc decreases the flux of glucose-6-phosphate (G6P) being dephosphorylated to glucose and released from the liver. Instead, increased G6P in the TRF liver can act as a substrate-level activator of G6P dehydrogenase (G6pdx), whose expression is also elevated in TRF livers for unknown reasons(Reference Hatori, Vollmers and Zarrinpar33). G6pdx mediates the first committing step of the pentose phosphate pathway (PPP), likely diverting some G6P towards the PPP. The PPP is a major source of NADPH, which is used to reduce glutathione. Consequently, levels of reduced glutathione and several intermediate products of the PPP are elevated in TRF livers(Reference Hatori, Vollmers and Zarrinpar33).
The pentose sugars from the PPP and intermediates of the TCA cycle are substrates for nucleotide biosynthesis. The rate-limiting steps in these pathways are mediated by Thymidine kinase 1 (Tk1) and Uridine monophosphate synthase (Umps), whose expressions are circadian and activated by the clock component BMAL1(Reference Fustin, Doi and Yamada58). TRF increases the expression of Bmal1, as well as its targets Tk1 and Umps. Consequently, higher levels of several nucleotides are observed in the liver of TRF mice(Reference Hatori, Vollmers and Zarrinpar33).
Nutrient sensing pathways such as AMPK, CREB, mTOR and Sirtuins directly or indirectly affect the expression or function of clock components(Reference Bass and Takahashi5,Reference Ramsey, Marcheva and Kohsaka59) . There are parallels between daily rhythms in these pathways and circadian clock components. Ad libitum access to an obesogenic diet dampens feeding rhythms and disrupts the normal expression or function of nutrient-sensing pathways, leading to reduced peak levels of many clock component mRNAs or proteins while having little effect on their trough levels. Conversely, TRF increases the peak expression levels of both activators and repressors of core clock components, including Cry1, Rev-erbα, Per2 and the activator Bmal1 (Reference Hatori, Vollmers and Zarrinpar33).
TRF enforces a relatively longer daily fasting period, which activates pathways implicated in mediating the benefits of caloric restriction on glucose homeostasis. These include increased expression and functional states of AMPK, FoxO, ATF, FGF21, Sirtuins and a modest increase in ketone bodies (ketogenesis)(Reference Vollmers, Gill and DiTacchio32–Reference Chaix, Zarrinpar and Miu34,Reference Hua, Li and Yang60) . In dietary patterns similar to TRF, daily activation of autophagy mediators or markers is also documented(Reference Martinez-Lopez, Tarabra and Toledo61). Under TRF, these mediators of fasting or feeding are likely active only for a few hours towards the end of the daily fasting period or after feeding begins.
Time-restricted feeding and hepatic lipid metabolism
The complementary functions of nutrient-sensing pathways and circadian clock components are believed to improve hepatic lipid homeostasis in the TRF liver. The increased level of AMP in TRF liver can activate AMP kinase, which phosphorylates acetyl-CoA carboxylase (ACC)(Reference Vavvas, Apazidis and Saha62). ACC catalyses the carboxylation of acetyl-CoA to malonyl-CoA, the first and rate-limiting step in fatty acid synthesis(Reference Winder63). Phosphorylated ACC is enzymatically inactive, suggesting that TRF may reduce de novo lipogenesis. Additionally, the increased peak expression level of the circadian clock repressor Rev-erb-α acts as a repressor of several genes implicated in lipid synthesis, such as fatty acid synthase (Fasn)(Reference Cho, Zhao and Hatori64). Reduced Fasn mRNA levels and increased relative phospho-ACC levels synergistically reduce fatty acid synthesis, leading to lower levels of several long-chain free fatty acids, including myristate and palmitate(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34) . Another clock repressor, Per2, may indirectly contribute to reduced lipogenesis in TRF liver by inhibiting the lipogenic transcriptional regulator PPARγ. PPARγ levels, typically elevated in the liver of high-fat ALF mice, return to standard diet levels in TRF mice(Reference Chaix, Zarrinpar and Miu34). This reduction in PPARγ expression, along with increased mRNA levels of its inhibitor Per2, correlates with reduced expression of PPARγ target genes, including stearoyl-CoA desaturase1 (Scd1), unsaturated fatty acid elongase Elovl5 and fatty acid transporter Cd36. The expression of other Elovl genes is also reduced in the TRF liver. This reduced expression of key enzymes in fatty acid desaturation and elongation leads to more than a 50 % decline in several unsaturated fatty acids, including oleate (18:1n9), palmitoleate (16:1n7), vaccenate (18:1n7) and eicosenoate (20:1n9 or 11)(Reference Hatori, Vollmers and Zarrinpar33).
TRF also promotes increased lipolysis and β-oxidation. The increased expression of PPAR alpha (Ppara), hepatic lipases (Lipc and Lipe), β-oxidation enzymes such as enoyl-CoA delta isomerase 2 (Eci2) and enoyl-CoA hydratase and 3-hydroxyacyl CoA dehydrogenase (Ehhadh) and reduced expression of the lipid droplet-associated lipolysis inhibitor gene (cell death-inducing DFFA-like effector C) Cidec in the TRF liver support more lipolysis and smaller lipid droplets in hepatocytes. Further lipid oxidation via the β-oxidation pathway is negatively regulated by fatty acid synthesis. Malonyl-CoA, a product of ACC activity in the first step of fatty acid synthesis, allosterically inhibits mitochondrial carnitine palmitoyltransferase 1 (CPT1), which is essential for the transit of long-chain fatty acids and acylcarnitine esters into the mitochondria for β-oxidation. In the ALF liver, elevated levels of malonylcarnitine imply increased inhibition of CPT1 and reduced β-oxidation caused by impaired entry of fatty acids into the mitochondria. In the TRF liver, reduced ACC activity decreases malonylcarnitine levels, promoting fatty acid oxidation and leading to increased levels of β-hydroxybutyrate, an end product of β-oxidation(Reference Hatori, Vollmers and Zarrinpar33). Additionally, Fabp1 (fatty acid binding protein 1), which binds to and clears potentially toxic unesterified long-chain fatty acids from the cytoplasm, exhibits a moderate increase in expression in the liver of TRF mice.
A beneficial byproduct of the overall reduction in the free fatty acid pool in the TRF liver is reduced inflammation. Free fatty acids are substrates for the increased production of pro-inflammatory long-chain ω-6 fatty acids, such as dihomo-linoleate (20:2ω6) and arachidonate (20:4ω6). The ALF liver is prone to more oxidative stress due to relatively less reduced glutathione, a major cellular antioxidant. Oxidation of arachidonate and linoleic acid in the livers of ALF mice further increases levels of pro-inflammatory eicosanoids: 15-Hydroxyeicosatetraenoic acid (HETE), 5-HETE and 13-hydroxyoctadecadienoic acid (HODE)(Reference Hatori, Vollmers and Zarrinpar33). In contrast, reduced hepatic free fatty acids, along with a glutathione-enriched cellular environment in the livers of TRF mice, attenuate levels of pro-inflammatory lipids and oxidative damage to hepatocytes. This reduction in hepatic stress may result in lower plasma alanine aminotransferase (ALT) levels, a biomarker of fatty liver disease, in TRF mice. TRF liver histology also suggests a healthier liver, with significantly less hepatic steatosis compared to ALF mice(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36,Reference Chaix, Deota and Bhardwaj53) . Volume analyses of serial block-face scanning electron microscope images of the liver samples reveal increased volumes of mitochondria and endoplasmic reticulum (ER) relative to the livers of ALF mice(Reference Hatori, Vollmers and Zarrinpar33).
Time-restricted feeding and cholesterol metabolism in the liver
The mammalian liver synthesises up to 20 % of the body’s total cholesterol pool and also regulates cholesterol breakdown for the production of bile acids and other sterols. Both the diurnal rhythms in food intake and the clock component Rev-Erb influence cholesterol biosynthesis and its breakdown into bile acids through the transcriptional regulation of several genes(Reference Vollmers, Gill and DiTacchio32,Reference Zhang, Zhao and Lu65) . The mRNA expression of two key enzymes involved in the de novo cholesterol biosynthetic pathway, squalene epoxidase (Sqle) and 7-dehydrocholesterol reductase (Dhcr7), typically shows daily rhythms in the liver. However, in TRF liver, their expression was higher, indicating a potential increase in de novo cholesterol biosynthesis. Despite this, TRF mice typically exhibit reduced cholesterol levels compared to mice fed a high-fat diet ad libitum. This suggests that the potential increase in de novo cholesterol synthesis in TRF mice is counteracted by increased cholesterol breakdown. The mRNA expression of two cytochrome P450 monooxygenase enzymes at the committing steps of the classical and acidic pathways of bile acid synthesis, Cyp7a1 and Cyp7b1, respectively, serves as an indicator of cholesterol catabolic activity. Relative to ALF livers, the peak expression of Cyp7a1 was elevated in TRF groups, and the expression of Cyp7b1 was elevated throughout 24 h in the TRF livers(Reference Hatori, Vollmers and Zarrinpar33).
The gene expression of cholesterol anabolic and catabolic enzymes is also regulated by the sterol regulatory element-binding transcription factors Srebf1 and Srebf2. Accordingly, hepatic sterol regulatory element-binding protein (SREBP) levels were elevated in TRF mice compared to ALF mice. Specifically, the active cleaved form of SREBP was rhythmic, with higher expression during the dark/feeding phase in the liver of TRF mice, while the total active form was reduced in ALF livers(Reference Chaix, Zarrinpar and Miu34). Together, these changes in gene and protein expression lead to a net increase in the breakdown of cholesterol, resulting in a concomitant reduction in cholesterol levels and an increase in various bile acids in the TRF livers(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36) .
Time-restricted feeding and adipose tissue physiology
Rodents on TRF consistently show a reduction in adipose tissue mass while preserving lean mass(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36) , a phenomenon also observed in some humans undergoing time-restricted eating (TRE). The difference in body weight between ad libitum and TRF groups fed an obesogenic diet for 18 weeks can be as high as 28 %, primarily due to differences in adiposity. Concomitant with this reduction in adipose tissue mass, there is a decrease in serum leptin and an increase in adiponectin(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34) . Although most gene expression and biochemical studies focus on the liver, little is known about the changes in gene expression and metabolites in adipose tissues. Nonetheless, the phenotype of TRF mice suggests that an overall increase in fatty acid oxidation may contribute to their reduced adiposity.
Both white adipose tissue (WAT) and brown adipose tissue (BAT) in mice fed an obesogenic diet ad libitum show hypertrophy and whitening of the brown fat(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36,Reference Chaix, Deota and Bhardwaj53) . Conversely, the TRF WAT has smaller adipocytes, and the BAT exhibits more ‘browning,’ as observed under H & E staining of tissue sections. The expression of master regulators of lipid synthesis and lipid oxidation in WAT and BAT changes under TRF. For instance, while the lipogenic transcription factor PPARγ shows a reduction in expression in the BAT, its expression increases in the WAT of TRF mice(Reference Deota, Lin and Chaix66). Along with a reduction in PPARγ expression in the liver of TRF mice, this change might reflect a shift in lipid storage from the liver and BAT in ALF mice to storage in the canonical lipid storage organ, WAT, under TRF.
There is also a paradoxical increase in PGC1α expression in the WAT of TRF mice(Reference Deota, Lin and Chaix66). Since PGC1α generally promotes fatty acid oxidation, the parallel increase in both PPARγ and PGC1α might indicate that both lipogenic and lipid degradation programmes in the WAT help maintain lipid homeostasis. However, these gene expression changes, without corresponding measurements of protein levels or lipid flux, make these observations highly preliminary.
The relatively higher bile acids in TRF mice(Reference Hatori, Vollmers and Zarrinpar33,Reference Zarrinpar, Chaix and Yooseph54) have led to speculation that some of the bile acids in circulation might bind to the bile acid receptor TGR5 in BAT. TGR5 is known to activate uncoupling protein (Ucp1) expression in BAT and increase thermogenesis(Reference Watanabe, Houten and Mataki67). It is also possible that TRF mice exhibit a relatively lower core body temperature during the fasting period, which can also activate Ucp1 expression. While the mechanism remains unclear, the brown fat from TRF mice does exhibit elevated levels of Ucp1 expression, which may account for some lipid oxidation as thermogenic loss(Reference Hatori, Vollmers and Zarrinpar33). Such predicted role of adipose tissue thermogenesis has been verified by genetic studies(Reference Hepler, Weidemann and Waldeck68).
In addition to an imbalance in lipid synthesis and oxidation in adipose tissue contributing to obesity, the activation of tissue-resident macrophages is also known to increase tissue inflammation and contribute to obesity-related comorbidities. Histological examination of WAT in multiple TRF experiments has shown reduced levels of crown-like structures in WAT, which are largely infiltrating macrophages. In parallel, the WAT in TRF mice also shows a significant reduction in many pro-inflammatory cytokines, including TNFα, IL6 and CXCL2(Reference Hatori, Vollmers and Zarrinpar33).
Time-restricted feeding and cardiac health
One major adverse outcome of chronic circadian disruption, such as shift work, is an increased risk of CVD(Reference Chellappa, Vujovic and Williams69). CVD remain the leading cause of death worldwide(Reference Lozano, Naghavi and Foreman70). Consequently, there is a growing interest in the circadian regulation of heart health. Time-series gene expression studies in various components of the cardiovascular system, including the atrium, ventricle, aorta and endothelial cells, have identified hundreds to thousands of genes exhibiting circadian rhythms(Reference Rudic, McNamara and Reilly71,Reference Bray, Shaw and Moore72) . The functional significance of these rhythms for heart health is demonstrated in various genetic perturbation models(Reference Paschos and FitzGerald73). For example, Bmal1 knockout mice show increased susceptibility to atherosclerosis, partly due to systemic metabolic disruption(Reference Pan, Bradfield and Hussain74). Additionally, constitutive overexpression of the dominant negative circadian clock Δ19 mutant (CCM) in mouse cardiomyocytes disrupts normal circadian gene expression and cardiac function, leading to increased fatty acid oxidation, lactate release, elevated fractional shortening, bradycardia and a longer R–R interval(Reference Bray, Shaw and Moore72). Many cardiac functions in CCM mice parallel those found in aging animals or humans, and aging is a significant risk factor for heart diseases.
In both normal and db/db mice, TRF enhances cardiac function and reinstates daily fluctuations in heart rate and blood pressure by suppressing sympathetic activity and decreasing cortisol levels during the light/inactive phase in mice(Reference Tsai, Villegas-Montoya and Boland75,Reference Hou, Su and Duncan76) . Additionally, TRF diminishes the size of atherosclerotic lesions and macrophage infiltration in plaques, as well as reduces plasma cholesterol in the APOE*3-Leiden.CETP mouse model of dietary hyperlipidaemia(Reference In Het Panhuis, Schonke and Modder77). Furthermore, TRF prevents as well as attenuates the progression of atherosclerotic plaque buildup in (LDL-cholesterol receptor) Ldlr -/- mice, but not in (Apo E) ApoE -/- mice(Reference Chaix, Lin and Ramms78). Since mutations in the Ldlr gene are the predominant cause for familial hypercholesterolaemia and TRF lowers LDL-cholesterol and VLDL-cholesterol in Ldlr -/- mice with parallel reduction in atherosclerotic plaque burden(Reference Chaix, Lin and Ramms78), this finding may have translational significance for familial hypercholesterolaemia in humans.
The effect of TRF on age-related cardiac decline has been extensively studied in Drosophila. The Drosophila model is well-established for exploring the genetic basis of cardiac function deterioration due to aging or an energy-dense diet(Reference Ocorr, Perrin and Lim79–Reference Melkani, Trujillo and Ramos82). The Drosophila heart, cylindrical in shape with a conical structure at the anterior end, pumps haemolymph through rhythmic diametric expansion (diastole) and contraction (systole) of the heart tube. By combining a semi-intact in vivo preparation with high-speed imaging, cardiac rhythms can be imaged for a sufficient period to quantitatively assess various cardiac function parameters. TRF of young Drosophila for 12 h every day did not reduce their daily caloric intake, yet the flies did not gain weight up to 7 weeks of age – a substantial portion of their lifespan(Reference Gill, Le and Melkani43). ALF flies showed characteristic features of cardiac aging, marked by increased diastolic interval, systolic interval, heart period, arrhythmia and reductions in systolic diameter, diastolic diameter and radial contractility. In contrast, TRF flies exhibited attenuation of these age-dependent cardiac function deteriorations in both male and female flies, independent of strain background(Reference Gill, Le and Melkani43).
Gene expression profiling of head, body and heart tissues from flies revealed that the gene expression signature of TRF is distinct from that of caloric restriction. In the Drosophila heart, TRF led to the downregulation of a cluster of genes encoding mitochondrial electron transport chains, which may indirectly reduce reactive oxygen species and protect the heart. TRF also led to a coordinated increase in the expression of various subunits of an ATP-dependent chaperonin complex (called T-complex polyppetide (TCP) or Chaperonin containing TCP-1 (CCT) complex), known to facilitate the proper folding of cytoskeletal proteins. While the CCT complex expression increased in the heart, the expression of genes encoding various cytoskeletal monomers or subunits reduced by as much as 20 %. Increased chaperonin expression coupled with reduced expression of cytoskeleton genes likely reduced the overall fraction of unfolded or misfolded cytoskeletal proteins and improved cardiac function. Genetic perturbation of CCT subunits or mitochondrial electron transport chain (ETC) components supported the hypothesis that these gene expression changes contributed to improved cardiac function in flies(Reference Gill, Le and Melkani43). While these experiments are yet to be verified in vertebrates, genome-wide association studies have identified a point mutation in CCT7 associated with ventricular arrhythmia(Reference Erdmann, Stark and Esslinger83). Overall, the Drosophila TRF studies have suggested that changes in mitochondrial function and improvement in proteostasis may underlie some of the benefits of TRF. Interestingly, improved proteostasis is a hallmark of all mice subjected to TRF.
Another outcome from the Drosophila TRF experiments was the observation that TRF flies showed improved sleep relative to the ALF flies. Sleep improvement was observed in group-housed flies, while no difference in individual sleep was found in older ALF or TRF flies when individually housed(Reference Gill, Le and Melkani43). Together, these observations suggest that sleep improvement under group-housed conditions is not due to a change in overall sleep pressure (which remained similar in individually housed flies), but rather to an increase in arousal threshold. Self-reported sleep improvement is also found in human volunteers undergoing TRF. As sleep improvement benefits cardiac health, TRF might contribute to improved heart health through both direct tissue-autonomous mechanisms and indirect system-wide effects by improving sleep.
Time-restricted feeding and brain health
Most studies on rodents and humans have primarily focused on body weight, body composition and metabolic health. However, preliminary results from animal experiments suggest that TRF may also benefit brain health, specifically by protecting against or delaying the onset of neurodegenerative diseases. Gene expression and phenotypic assessments of TRF and ad libitum animals in both flies and mice have revealed several molecular signatures supporting this notion(Reference Hatori, Vollmers and Zarrinpar33,Reference Chaix, Zarrinpar and Miu34,Reference Chaix, Lin and Le36,Reference Gill, Le and Melkani43,Reference Deota, Lin and Chaix66) : (a) TRF reduces mitochondrial electron transport chain activity, potentially lowering reactive oxygen species levels. (b) TRF improves the structural integrity of the ER and mitochondria. (c) TRF enhances the integrative stress response by slightly reducing mTOR activation and increasing the unfolded protein response. (d) TRF improves motor coordination in rodents. (e) TRF leads to sleep improvement in flies.
If these results from non-neuronal tissues extend to the nervous system, we could expect a reduction in disease severity in animal models of neurodegenerative diseases under TRF. Two recent studies have shown that mouse models of Huntington’s disease subjected to TRF exhibit improvements in motor coordination, sleep and autonomic nervous system function(Reference Wang, Loh and Whittaker84,Reference Whittaker, Loh and Wang85) . Additionally, gene expression markers of the disease in the striatum were normalised in TRF mice. TRF has also been shown to prevent the deterioration of memory loss in a mouse model of Alzheimer’s disease(Reference Whittaker, Akhmetova and Carlin51). Moreover, TRF improved spatial working memory and long-term synaptic potentiation magnitude in obese mice without weight loss(Reference Davis, Paul and Yates86). These early indications in animal models raise the hope that TRF in humans may delay the onset or reduce the severity of neurodegenerative diseases.
Perspective on the future of time-restricted feeding/time-restricted eating research and translation
The demonstrated success of TRF in combating obesity, metabolic diseases, cancer and dementia in animals is raising substantial hope for its long-term potential in improving public health and reducing healthcare costs. This rising interest in TRF is expected to lead to more laboratory research to fill many knowledge gaps. Currently, most laboratory studies are done on young male rodents, so extending these studies to middle-aged or older animals and including both sexes will offer better insight into the effects of TRF on age-related diseases in both sexes.
In humans, TRE can inadvertently lead to modest CR in some, but not in all participants(Reference Manoogian, Chow and Taub87,Reference Ezpeleta, Cienfuegos and Lin88) . This has raised the question of whether the TRE benefits in humans, specifically on weight loss, arise from CR or time restriction. Few human studies done so far have shown that time restriction often accompanies caloric reduction as well. Sometimes the TRE participants can reduce energy intake by as much as 20 %. Even this approach of caloric reduction without an overt attempt at calorie counting is remarkable because the most extensive and long-term caloric reduction study (CALERIE-2) achieved only 12 % caloric reduction against the targeted reduction of 25 %(Reference Kraus, Bhapkar and Huffman89). As mentioned earlier, in rodents, TRE can accentuate the effects of CR on lifespan, yet there is no existing biomarker that can differentiate between the effects of CR alone v. CR+TR in mice. Until such biomarkers are developed in animals, it is still premature to conclude that the effects of TRE in humans is entirely due to CR. Moreover, implementing CR in humans at a population level has been difficult as it involves extensive literacy on nutrition and calorie counting. Hence, studies focused on the time and effort spent on CR v. TR in humans will be important to quantify the return on investment from a CR-centric v. TR-centric approach.
There have been very few human studies where the primary outcome is not weight loss. One of such outcomes is the reduction of blood glucose among people with pre-diabetes. The current standard of care for combating pre-diabetes is the guidelines based on the outcomes of the Diabetes Prevention Program. In this study, a reduction in energy intake coupled with increased physical activity led to a 0·1 % reduction in HbA1C, which translated to a 58 % reduction in transition from pre-diabetes to type 2 diabetes in the subsequent 2 years(Reference Knowler, Barrett-Connor and Fowler90). As the incidence of pre-diabetes has increased globally, large-scale studies on the effectiveness of TRE in reducing blood glucose among pre-diabetes will be worth exploring. Similarly, the effectiveness of TRE as an adjunct therapy in addition to pharmacotherapy for the management of type 2 diabetes or other metabolic diseases including hypertension and dyslipidaemia is also worth exploring. In summary, animal studies have built a sound basis for translational research on TRE for the prevention and management of chronic metabolic diseases.
Conclusion
The domain of science addressing how timing of nutrition can influence health outcomes is closely intertwined with the field of circadian rhythm, and both fields are relatively young. Only in the year 2000, elegant experiments in mice demonstrated that timing of food intake can influence the circadian clock in the liver without much effect on the central circadian oscillator in the brain(Reference Damiola, Le Minh and Preitner28). Subsequent studies on the interaction of food timing and liver clock as well as the genomic studies demonstrating the gene expression and function of several key metabolic processes show circadian expression ultimately led to the concept of TRF, first demonstrated in mice(Reference Hatori, Vollmers and Zarrinpar33). Nevertheless, within a few years, the field has rapidly expanded and, at least in animals, has demonstrated several health benefits involving multiple organ systems. In parallel, the rich field of nutrition science examining the effects of nutrition quality and quantity on health outcomes has also made great advances. Altogether these studies have produced a wealth of gene expression and gene function data which can be compared with the known molecular basis of human diseases to judicious implementation of optimum quality, quantity and timing of nutrition for disease prevention, management and recovery.
Acknowledgements
NA.
Author contributions
S.P. wrote the first draft of the manuscript with input from S.D., H.L. and G.M. All authors edited and refined the draft.
Financial support
This work was supported by National Institutes of Health (NIH) grants AG065992 to SP and GCM. Additionally, the research in SP lab is supported by the NIH (CA258221), Clara Wu and Joe Tsai Foundation and Wu-Tsai Human Performance Alliance, Tiny Foundation.
Competing interests
S.P. is the author of the books The Circadian Code and The Circadian Diabetes Code.
Open access
