Most humans spend more than 12–16 h per day in the postprandial state. Generally, food intake occurs in a three meal per 24 h fashion with in-between meal snacking (Kant, 2018; Paoli et al., 2019). After meal ingestion, plasma insulin concentrations rise and subsequently increase glucose uptake and -oxidation in peripheral tissues (Thiebaud et al., 1982). Furthermore, insulin activates lipoprotein lipase which promotes the storage of dietary fat in adipose tissue (Lafontan and Langin, 2009). Lowering of plasma insulin concentrations during fasting (e.g., overnight) permits lipolysis, a shift from glucose to fat oxidation resulting in a state of physiological insulin resistance if the fasting is continued (Lafontan and Langin, 2009; Soeters et al., 2012).

Several postprandial signals (e.g., plasma bile acids, fibroblast growth factor 19 (FGF19), lipids) are still increased in the systemic circulation 4–5 h after meal ingestion (Schrezenmeir et al., 1993; Van Nierop et al., 2019; Meessen et al., 2020). From an evolutionary point of view it may be reasoned that the human body is habituated to a lower meal frequency (Meiselman, 2000). Metabolic flexibility (i.e., reciprocal changes in carbohydrate- and fatty acid oxidation) is a characteristic of metabolic health. Continuous/frequent food intake on the other hand may result in reduced metabolic flexibility, obesity and type 2 diabetes mellitus (T2DM) (Smith et al., 2018). These lifestyle-related diseases are associated with modern eating habits and a sedentary lifestyle (Kopp, 2019). Nutritional- and exercise interventions should be the first in prevention- or treatment strategies because of the biological rationale, low cost and non-pharmacological approach (Manson et al., 2004).

The most used strategies of intermittent fasting can be clustered into three groups: alternate day fasting, whole- day fasting and time-restricted feeding (TRF). Alternate day fasting incorporates alternating ad libitum feeding days and fasting days (accompanied with 1 small meal), whilst the whole-day fasting strategy involves 1 or 2 days of complete fasting per week plus and ad libitum feeding on the residual days (Tinsley and La Bounty, 2015). TRF encompasses many different strategies, but is mainly the combination of a limited feeding time-window and “prolonged” fasting (Patterson and Sears, 2017; McAllister et al., 2020). As a result of the wide range of TRF protocols in studies (due to variation in meal timing, duration of fasting time, with or without caloric restriction), the effects of TRF on metabolic parameters and physical performance are scarce and equivocal (Moon et al., 2020; Pellegrini et al., 2020). Some metabolic studies show that TRF may improve whole-body insulin sensitivity whereas other studies reported no effect on glucose metabolism or even reported impaired glucose tolerance (Halberg et al., 2005; Carlson et al., 2007; Soeters et al., 2009; Cienfuegos et al., 2020; Jones et al., 2020). Interestingly TRF (16/8) decreased fat mass without affecting strength in resistance-trained males (Moro et al., 2016).

Therefore, we investigated the physiological effects of eucaloric TRF (22/2, one meal per day in the evening) for 11 days on metabolic health and physical performance in free-living healthy lean individuals compared to normal food intake (three meals per day) in a crossover design.

This randomized crossover nutritional intervention study showed that intended eucaloric meal frequency reduction to one meal per day in the evening, lowered body- and fat mass and increased fatty acid oxidation during exercise, without impairment of aerobic capacity or strength. Furthermore, we found decreased plasma glucose concentrations during the second half of the day.

The loss of body- and fat mass after TRF (with different protocols i.e., 4- and 6-h TRF regimens and 16/8) have been reported previously (Moro et al., 2016; Cienfuegos et al., 2020). This may be the result of enhanced fat utilization throughout the day caused by lower insulin concentrations (Knapik et al., 1988; Tinsley and La Bounty, 2015). This hypothesis is supported by the observation of increased fat oxidation during exercise after the one meal per day period. Although we did not detect significant differences in caloric intake between the intervention periods, it was plausible that there was a caloric deficit of ∼110 kcal/day (1,210 kcal in 11 days) during the one meal per day intervention. Nevertheless, this caloric deficit cannot explain the total loss in body mass (–1.4 kg) and/or fat mass (–0.7 kg) at day 11, since ∼1,210 kcal account for approximately 140 gram fat. Therefore, these findings suggest that a 22 h fast in combination with the preserved daily physical activity and unchanged resting energy expenditure, caused a larger negative energy balance compared to the three meal per day pattern. This is supported by our finding of decreased plasma glucose concentrations throughout the day. However, we did not measure 24 h calorimetry, the golden standard for metabolic flexibility. It could be possible that the participants changed the macronutrient composition of their meals to maintain isocaloric intake during the one meal per day period. Moreover, we cannot exclude the beneficial role of ketone bodies nor the changes in the gut microbiome (Jamshed et al., 2019; Currenti et al., 2021a,b). Furthermore, we did not observe a decrease in lean mass during the one meal period reduction intervention, which may suggest that there was no increased net protein catabolism (e.g., decreased protein intake or use of amino acids for gluconeogenesis). However, it is possible that participants ate more during the 3 meals/day condition—which was undetected by the dietary records due to limitations of this method—to account for a notable portion of the body mass and fat mass difference.

Both bile acids and FGF19 concentrations are still increased 4–5 h after food intake (Van Nierop et al., 2019; Meessen et al., 2020). So, from an evolutionary point of view, these prolonged postprandial signals suggest that humans are well prepared for periods with limited food availability. Bile acids are released after food intake and with a function as postprandial signaling molecules within tissues of the enterohepatic cycle by activation of the intestinal and hepatic bile acid receptors such as farnesoid X receptor (FXR). FXR activation results in the production of FGF19 in the ileum, and FGF19 reaches the liver via the portal vein where it exerts unique (e.g., stimulation of fatty acid oxidation) and insulin-like actions (e.g., inhibition of gluconeogenesis, stimulation of glycogen synthesis (Kir et al., 2011; Potthoff et al., 2011). We did not observe changes in fasted- or postprandial bile acid and FGF19 concentrations. This may be explained by the fact that the duration of the overnight fast before the MMT was identical for both interventions, so we may have missed acute effects of TRF on bile acid- and FGF19 levels. Alternatively, skipping two meals may have not changed intestinal transit time or gallbladder kinetics, both important factors that determinate plasma bile acid concentrations (Sips et al., 2018; Meessen et al., 2020).

Increased cholesterol and LDL-C levels are associated with cardiovascular disease (CVD) (Levine et al., 1995). We detected a small increase in LDL-C during the one meal period, whereas the other lipid profiles were unchanged. A previous study reported that one meal per day resulted in increased total cholesterol, HDL-C and LDL-C concentrations, whereas TGs were decreased (Stote et al., 2007). They could not link their findings to dietary fat intake since they controlled food intake. In our study, participants consumed all the calories needed to maintain a stable weight within 2 h. This may have caused a higher hepatic lipid load, which in turn inhibited LDL-C receptor expression and subsequently resulted in elevated plasma LDL-C concentrations (Mustad et al., 1997). Moreover, prolonged fasting in our study could have decreased LDL catabolism in the liver to save as a nutrient reservoir to peripheral tissues (Saävendahl and Underwood, 1999). Importantly, mean plasma glucose concentrations through the day was in the one-daily meal condition, which may an health benefit. This will be particular important if the same will be the case in dysglycaemic and patients with diabetes.

We did not observe impaired physical performance as a result of the reduced meal frequency. Other studies investigating the effects of TRF on physical performance produced conflicting data. Recently, Correia et al. (2020) and Zouhal et al. (2020) reviewed the effects of exercise training and fasting and showed that TRF studies predominantly report no changes or impairment of aerobic, anaerobic and strength performance. Only a few studies showed modest improvements in VO_2max and strength (Tinsley and La Bounty, 2015; Bouguerra et al., 2017). Most of these studies were conducted during the Ramadan, where (unlimited) food intake between dusk and dawn is an important confounder. These kind of endurance performances are likely to be affected by the one meal per day pattern because the active muscle primarily oxidize endogenous carbohydrates (e.g., a rate-limiting step in performance), while fasting depletes these glycogen stores (Soeters et al., 2012; Hawley and Leckey, 2015).

In our study, participants were allowed to ingest their food between 17:00 and 19:00 during the one meal pattern. It is possible that our results would have been different when they were allowed to eat only in the morning, e.g., between 08:00 and 10:00. Early morning TRF increases insulin sensitivity and improves β-cell function and mid-day TRF reduces body weight and fasting glucose and insulin concentrations (Moro et al., 2016; Jones et al., 2020). Conversely, TRF in the afternoon or evening even induced glucose intolerance (Carlson et al., 2007). In the present study, the meal test did not reveal any differences in glucose, insulin, FGF19 or bile acids between the interventions. These data show that the human body can cope with one daily large meal providing all the required energy, without deteriorating metabolic regulation.

A recent review has suggested that lean mass and is generally maintained during TRF including Ramadan period (Keenan et al., 2020) supporting the data in the present study. Importantly, the participants in the present study were not competitive athletes and LBM may be easier to maintain compared to athletes. However, Moro et al. (2020) implemented TRF on elite cyclists and found a small reduction in body fat and maintained LBM. Importantly, Moro showed that power output per kg body weight increased. Brady et al. (2021) also reported reduction in body mass by TRF but did not reduced endurance performance.

The most important limitation of our study is that the study was conducted in free-living healthy individuals. Data on food intake was obtained with food diaries completed by the participants. Nonetheless, our results showed that the participants were most likely compliant to the one meal per day pattern since the glucose sensor showed lower glucose concentrations throughout the day. It is a strong point that we carefully characterized strength, maximal power, and aerobic capacity under controlled conditions and with state-of-the art methodologies.

In summary, we found that one meal per day in the evening lowers body weight and adapts metabolic flexibility during exercise via increased fat oxidation, whereas the physical performance and the postprandial responses to liquid mixed meals were unaffected. Future studies should investigate the effect of TRF in participants with reduced metabolic flexibility (e.g., patients with obesity, metabolic syndrome or type 2 diabetes mellitus). Moreover, the timing of the one meal consumption in early morning may yield larger effects.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethics Committee at Norwegian School of Sport Sciences (Application #56-190618). The patients/participants provided their written informed consent to participate in this study.

EM co-designed the study, executed statistical analyses, and wrote the manuscript. HA co-designed the study, performed clinical experiments, executed statistical analyses, and reviewed the manuscript. TB and AR performed clinical experiments, executed statistical analyses, and reviewed the manuscript. EJ performed clinical experiments and reviewed the manuscript. AK was responsible MD during study and reviewed the manuscript. SO and FS performed laboratory analyses and reviewed the manuscript. EK, JR and AK reviewed the manuscript. JJ and MS designed the study, reviewed, and edited the manuscript. All authors contributed to the article and approved the submitted 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.