The circadian system, which is driven by both an internal circadian clock and external environmental signals, controls the timing of feeding, and sleep-wake cycles (Manoogian and Panda, 2017; Northeast et al., 2020). Food scarcity is a major challenge in terms of energy conservation, with prolonged arousal increasing food-seeking behaviors, which indicates reciprocal control between sleep and feeding regulation processes (Arrigoni et al., 2019). Obese individuals and those with metabolic imbalances and diabetes mellitus experience significant disturbances of sleep-wake cycles and sleep architectures (Wu et al., 2014; Koren et al., 2016; Imam et al., 2020). Likewise, sleep deprivation increases the risk of metabolic changes (Knutson et al., 2007; Davies et al., 2014).

Aging is a complex and unavoidable process, which is accompanied by a series of comorbid conditions, including sleep loss, fatigue, metabolic syndrome, and diabetes (Akerstedt et al., 2018; Smith et al., 2020). With increasing age, sleep becomes progressively fragmented and is characterized by impaired consolidation of non-rapid eye movement (NREM) sleep (Klerman and Dijk, 2008; Mander et al., 2017). Age affects homeostatic processes and the number of hours of sleep needed, resulting in a diminished ability to maintain consolidated sleep (Klerman and Dijk, 2008; Mander et al., 2017). Older adults who sleep less show a trend toward an increased body mass index and waist circumference, as well as an increased risk of obesity (Gildner et al., 2014; Grandner et al., 2015; Mamalaki et al., 2019). These results indicate that metabolic symptoms are associated with poor sleep quality during the process of aging.

In contrast to ad libitum feeding, where an animal has unrestricted access to food, restricted feeding (RF) is a dietary regime involving fasting for specific periods, which restricts energy intake and appears to reduce body weight (Manoogian and Panda, 2017). Previous research suggested that RF might help to improve sleep quality and quantity and lead to an extended life span (Hou et al., 2019; Oike et al., 2020). In human subjects, 10-h time-restricted eating (TRE) reduced body weight and promoted sleep in obese individuals (Gill and Panda, 2015; Kesztyus et al., 2020). It also improved the cardiometabolic health of patients with metabolic diseases, including controlling blood pressure and decreasing atherogenic lipid cholesterol and glycated hemoglobin (Wilkinson et al., 2020), as well as that of diabetic mice (Hou et al., 2021). Body weight loss via RF may help to lower adiposity, insulin resistance, plasma cholesterol, and plasma triglycerides (Chaix et al., 2014). Based on the aforementioned results from human and animal studies, RF seems to be an effective lifestyle, non-pharmacological intervention to prevent metabolic symptoms and improve health-related quality of life. However, whether RF has beneficial effects on metabolic health and sleep in older adults remains unclear.

The aim of the present study was to investigate the effect of RF on metabolic parameters and sleep-wake architectures. In the current study, aged mice were assigned to an RF protocol for 2 weeks. During this period, their body weights, blood biochemistry profiles, including blood glucose, cholesterol levels, amylopsin, lipase, alanine aminotransferase and alkaline phosphatase activity, total protein, albumin, and globulin levels, and sleep-wake cycles were monitored. The results revealed that RF improved metabolic symptoms and sleep-wake profiles in aging mice, thereby indicating that a non-pharmacological intervention can help to prevent obesity and obesity-associated disorders in the adult population.

In the current study, we found that RF improved metabolic symptoms by reducing body weights and blood glucose and cholesterol levels. Furthermore, RF led to anticipation of food through increased wakefulness, with reduced sleep. At the beginning of RF, sleep decreased in the RF group in the dark period. After stable food entrainment was achieved around day 7 in the RF group, the amount of time spent in wakefulness in the 2 h prior to food availability (18:00–20:00) increased, thereby increasing the mean duration of awake episodes. There was no significant difference in the amount of wakefulness, NREM sleep, and REM sleep during the dark period in the RF group as compared with that in the control group. During the RF regime, dopamine levels in the midbrain increased in the RF group, suggesting that dopamine transmission in the midbrain might contribute to increased wakefulness during food anticipation.

The incidence of many chronic diseases, including sleep disorders, obesity, cancer, and neurodegenerative disorders, has increased in the aged population (Lopez-Otin et al., 2016). Pharmacological approaches aimed at combating sleep disorders and associated morbidities and improving the quality of life in this population have proved ineffective. Thus, there is increased interest in identifying non-pharmacological strategies. One such approach is intermittent fasting (IF), which involves fasting for extended periods, with intervening periods of normal food intake, on a recurring basis (Mattson et al., 2017). TRE is a subcategory of IF. In TRE, food intake is restricted to a time window of 8 h or less per day. Unlike IF, the focus in TRE is on how many hours per day food consumption is permitted rather than on how long a person must fast (Chaix et al., 2014). The available evidence suggests that the IF protocol produces equivalent weight loss when compared to the TRE regime (Rynders et al., 2019). TRE is very helpful for reducing weight gain linked to late night snacking. In the present study, we chose an RF regime of 2 h during the active phase of mice to test the effects of RF on metabolic health and sleep-wake cycles. At the end of the 14-day period, the mean body weight of the mice in the RF group had decreased 30.2% as compared with the baseline value. Previous studies showed that most body weight loss in RF regimes is due to a decrease in white adipose tissue, especially visceral fat stores (Barzilai et al., 1998; Finkel, 2015; Akbari et al., 2019). In contrast to food restriction, which is thought to slow down the aging process, obesity is believed to exacerbate this process. According to some research, adverse effects of obesity on the function of mitochondria and nutrient-signaling pathways might result in an imbalance in insulin and blood glucose (Bentley et al., 2018). The mechanistic target of rapamycin (mTOR) controls cell growth and metabolism and energy level sensing. Previous studies suggested that mTOR activation promotes cell growth and proliferation by inducing protein synthesis and lipogenesis and increasing glucose metabolism (Jacinto et al., 2004; Hagiwara et al., 2012). In contrast, inhibition of mTOR reduced mRNA translation and protein synthesis (Saxton and Sabatini, 2017) and increased the lifespan (Harrison et al., 2009; Liu and Sabatini, 2020). Food restriction activated adenosine monophosphate-activated protein kinase, which then inhibited and suppressed mTOR signaling to slow aging and increase resistance to pathologies, including metabolic diseases (Burkewitz et al., 2014; Sunjaya and Sunjaya, 2021). These findings are consistent with RF-mediated body weight loss in the aged mice in the present study.

We also found that RF improved cardiometabolic symptoms by lowering blood glucose levels and cholesterol levels in aged mice. Previous studies showed that time RF maintained glucose homeostasis (Mitchell et al., 2015) and lowered fasting blood glucose levels in diabetic mice (Wei et al., 2019). In rodent liver, an RF regime upregulated the metabolic rate, which increased fatty acid β oxidation and reduced hepatic glucose production (Hatori et al., 2012). RF also regulated the metabolic rate by inducing gluconeogenesis and balancing glucose levels. In addition, RF improved the expression of a rate-limiting enzyme, cholesterol 7α-hydroxylase, which promotes bile acid production from cholesterol (Froy and Miskin, 2010; Longo and Panda, 2016; Froy, 2018). We found no significant difference in the serum concentrations of total protein, albumin, and globulin or alkaline phosphatase and alanine aminotransferase activity in the RF group versus the control group. These results indicated that RF can improve cardiometabolic health, without deleterious effects on liver function.

In animal studies, RF during both light and dark periods can lead to food entrainment by increasing locomotor activity (Liu et al., 2012; Caba et al., 2021). In previous studies, when the RF time occurred during the light period in adult animals, the time spent in both NREM and REM sleep decreased, with increased wakefulness (Mieda et al., 2004; Szentirmai et al., 2010; Northeast et al., 2019). In adult mice, when the RF time occurred during the light period, the time spent in sleep-wake stages changed, with increased wakefulness for 2–3 h preceding food availability. However, RF during the light period resulted in a reversal in the light-dark pattern of sleep-wake stages (Mieda et al., 2004). Other studies showed that RF for 4–6 h in the dark period improved sleep-wake cycles in diabetic (Hou et al., 2019) and Huntington’s murine models (Wang et al., 2018). Although these later studies defined sleep in adult animals by immobility, without EEG/EMG recordings, their findings suggested that RF during the active phase has a beneficial effect on circadian rhythms. In the present study, food was provided at 20:00 (lights off) and removed 2 h later, as described previously (Liu et al., 2012). At baseline conditions, when the animals in both the RF and control groups had access to chow and water ad libitum, there were no significant differences in the sleep-wake rhythms of the two groups. After the start of the RF protocol, both NREM and REM sleep decreased in the dark period but not in the light period in the RF group as compared with that in control group. Stable food entrainment was achieved in the RF group around 7 days after the commencement of the RF protocol. At this point, wakefulness, characterized by an increased mean duration of wakefulness and a decreased number of wakefulness episodes, dramatically increased during the 2-h period before food availability. We found no significant difference in the sleep-wake cycles in the RF and control groups during the 10-h light (08:00–18:00) and dark periods (20:00–8:00). During RF days 7–14, the sleep-wake profiles in the RF group during each 24-h period were similar, without any significant difference in the mean duration and number of wakefulness, NERM sleep, and REM sleep episodes or any significant difference in the total amount of wakefulness. In contrast, previous studies showed that caloric restriction and fasting increased the amount of wakefulness in rats, with longer term food deprivation (>7 days) causing sleep to become fragmented (Borbely, 1977; Szentirmai et al., 2006; Castro-Faundez et al., 2016). The use of aged mice and a different RF schedule in our study may account for the discord between our findings and those in the literature.

In addition to RF affecting sleep-wake vigilance, RF also altered sleep and wakefulness quality, as quantified by the EEG power density for wakefulness, NREM sleep, and REM sleep. During the 2 h before food availability, the EEG power density for wakefulness under the RF protocol decreased in the frequency of 3.75–4.25 Hz, which was associated with an increased EEG power density in NREM sleep in the frequency bands of 1.75–2.25 Hz and 3.5–4.5 Hz. These observations indicated that the mice in the RF group showed a low inclination for arousal during RF and that both food anticipation and circadian rhythms might regulate sleep-wake cycles. In our study, we detected no significant difference in the EEG power density during the 10-h light period (08:00–18:00) in the RF group compared to that in the control group. Our findings are in accordance with those of studies on wild-type mice and diabetic mice, in which the authors reported that homeostatic regulation of sleep and wakefulness remained intact during an RF protocol (Hou et al., 2019; Northeast et al., 2019).

Previous studies showed RF increased c-Fos protein expression in the dorsomedial nucleus, lateral hypothalamus (LH), perifornical area, and tuberomammillar nucleus, which indicated that FAA may be regulated by networks distributed throughout the brain or that more essential nuclei controlling FAA remain to be unidentified (Angeles-Castellanos et al., 2004). The dopaminergic system, including dopamine and its receptors, is known to be involved in regulating sleep-wake cycles (Qu et al., 2008; Luo et al., 2018) and reward-motivating behavior (Berridge et al., 2010; de Lartigue and McDougle, 2019). As shown in previous studies, genetic depletion of tyrosine hydroxylase or concurrent ablation of dopamine D1 and D2 receptors in mice induces mortality a few weeks after birth. The feeding behavior and physiological function of the gastrointestinal system were altered in these mice, indicating a strong relationship between the dopamine system and gastrointestinal function (Zhou and Palmiter, 1995; Kobayashi et al., 2004; Palmiter, 2007). Previous research also demonstrated that dopamine is crucial for maintaining wakefulness. For example, knockout (KO) of dopamine D2 receptors significantly decreased wakefulness, with an increase in both NREM and REM sleep in mice (Qu et al., 2010). Brain neuronal calcium signaling by dopaminergic neurons in the brain was shown to be higher during wakefulness and lower during sleep, and calcium signals increased significantly during transitions from NREM sleep to wakefulness (Eban-Rothschild et al., 2016; Dong et al., 2019), which indicated that dopaminergic neurons in the midbrain are essential for initiating and maintaining wakefulness (Eban-Rothschild et al., 2016; Oishi et al., 2017a; Yang et al., 2018; Ciric et al., 2019). Studies also showed that dopaminergic midbrain signals govern neuron activity in the striatum and nucleus accumbens to initiate and maintain wakefulness for food seeking (Eban-Rothschild et al., 2016; Luo et al., 2018). The aforementioned findings support the results of our study, in which c-Fos protein expression in the midbrain and striatum and dopamine concentration in the midbrain increased prior to food availability in the RF group.

In addition to dopamine, several other neurotransmitters regulate food intake and sleep and are therefore thought to regulate FAA (Pendergast and Yamazaki, 2018). Pharmacologically increasing serotonin levels by a selective serotonin reuptake inhibitor, fluvoxamine, suppressed FAA, whereas decreasing serotonin levels by a tryptophan hydroxylase inhibitor enhanced FAA in mice (Rozenblit-Susan et al., 2016). However, FAA was unaffected by time RF in both serotonin transporter and tryptophan hydroxylase-2 KO mice, which indicated that serotonin is not necessary for regulating FAA (Gallardo et al., 2020). Some studies suggested that a neuropeptide knows as orexin or hypocretin may be involved in the regulation of food anticipation behavior. Orexin, which strongly promotes food intake and wakefulness, is produced by lateral and posterior hypothalamic neurons (Cun et al., 2014; Soya and Sakurai, 2020; Concetti and Burdakov, 2021). Administration of orexin into the lateral ventricle stimulated food consumption in rodents (Sakurai et al., 1998), and ablation of orexin neurons or KO of the orexin gene in rodents attenuated food intake and obesity (Hara et al., 2005; Fujiki et al., 2006), suggesting that the orexinergic system contributes to food entrainment. The aforementioned findings are consistent with the results of our study, in which c-Fos expression significantly increased in the lateral hypothalamus in the RF group. In another study, FAA was attenuated in orexin KO mice but not eliminated, and core body temperatures and food consumption in the orexin KO mice and control mice were similar (Kaur et al., 2008). Research also demonstrated that KO of the orexin system results in narcolepsy and sleep fragmentation (Coffey et al., 2021). These results suggest that orexin neurons may sense the metabolic status and modulate arousal to promote food-seeking behavior during RF regimes. The exact role of orexin neurons in the metabolic balance and sleep-wake regulation during RF in aged mice needs to be elucidated in the future.

At present, there is little consistency in RF protocols in animal studies. For example, in many animal studies on RF, food was provided during the inactive period when animals spend more time sleeping. In other studies, food was provided during the active period. Moreover, the feeding window has also varied from 2 to 8 h per day in different animal studies (Northeast et al., 2020; Caba et al., 2021). An inappropriate RF protocol can result in erroneous conclusions (Pendergast and Yamazaki, 2018). A standardized and optimized experimental protocol for RF is needed to elucidate the mechanism of RF in improving metabolic health and sleep-wake cycles.

The present study has some limitations. As sleep-wake cycles vary with age and gender (Dorsey et al., 2020), one limitation is that our study included only aged male mice. Thus, it was not possible to compare sex- and age-related differences in the RF regime. Another limitation of the study was that although the RF regime was associated with benefits in the aging mice, such a regime does not resemble RF in humans because the sleep structure and food consumption behavior of humans and rodents differ greatly. This is a problem common to all RF studies that utilize animal models.

In summary, food restriction in aged mice had a profound impact on body weights, blood glucose and cholesterol levels, and amylopsin and lipase activities. In addition, RF changed the sleep-wake profiles of the mice, with an increase in the mean duration of wakefulness and a decrease in the number of wakefulness episodes. Food restriction also influenced sleep and wakefulness quality during the 2-h period prior to food availability. Our results indicate that RF during aging may prohibit or delay the onset of age-related diseases by improving metabolic health, without severe deleterious effect on sleep. They also suggest that RF could be a novel, non-pharmacological intervention to prevent metabolic disorders in the elderly population.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by the Animal Care and Use Committee of Anhui Medical University.

QX, Y-XX, and M-KZ conceived and designed the study. Y-XX, G-YL, H-QB, S-CW, W-HC, and G-JC carried out the experiments. QX, G-YL, P-PZ, and QJ analyzed the data. C-BG, M-KZ, G-HC, D-FZ, and QX wrote the manuscript. All authors contributed to the study and approved the final version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.