Skip to main content

Why meals during resting time cause fat accumulation in mammals? Mathematical modeling of circadian regulation on glucose metabolism


Synchronization of metabolic rhythms regulated by circadian clock and meal timing is essential for maintaining nutrient homeostasis in response to fluctuating food intake in animals. Despite numerous experimental findings on the involvement of circadian regulation of glucose and lipid metabolism, the optimal regulatory strategy for the maintenance of energy homeostasis remains poorly defined. A mathematical framework is useful to assess the circadian regulation of glycogen production/breakdown and de novo lipogenesis/lipolysis by evaluating the contribution of time of the day-dependent activation or the repression of each metabolic process in the maintenance of energy homeostasis. Here, we present a mathematical model that describes the dynamics of glycogen and triglyceride contents, two major forms of energy storage in the body that provide the fuel needed during different phases of food deprivation. By changing peak phases of glycogenesis and fat synthesis, we searched for the optimal phase set that minimizes the risks of two types of possible metabolic dysfunctions: (1) high blood glucose and (2) energy exhaustion. Based on the optimal phase set, we compared the level of fat accumulation between meal timing in the active and resting periods. Our results showed that an increased fat accumulation by food intake in the resting period can be the byproduct of minimizing energy homeostasis risks in the synchronized feeding schedule that animals adopt in nature. Our finding will be useful to schedule an optimal meal timing to prevent metabolic diseases caused by misalignment of biological and social time in modern society.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. Adamovich Y, Rousso-Noori L, Zwighaft Z, Neufeld-Cohen A, Golik M, Kraut-Cohen J, Wang M, Han X, Asher G (2014) Circadian clocks and feeding time regulate the oscillations and levels of hepatic triglycerides. Cell Metab 19(2):319–330.

    Article  Google Scholar 

  2. Albers JW, Pop-Busui R (2014) Diabetic neuropathy: mechanisms, emerging treatments, and subtypes. Curr Neurol Neurosci Rep 14(8):473.

    Article  Google Scholar 

  3. Antonetti DA, Klein R, Gardner TW, Antonetti A (2012) Diabetic retinopathy. N Engl J Med 366(13):1227–1239.

    Article  Google Scholar 

  4. Arble DM, Bass J, Laposky AD, Vitaterna MH, Turek FW (2009) Circadian timing of food intake contributes to weight gain. Obesity 17(11):2100–2102.

    Article  Google Scholar 

  5. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161(1):84–92.

    Article  Google Scholar 

  6. Bass J, Takahashi JS (2010) Circadian integration of metabolism and energetics. Science 330(6009):1349–1354.

    Article  Google Scholar 

  7. Crosby P, Hamnett R, Putker M, Hoyle NP, Reed M, Karam CJ, Maywood ES, Stangherlin A, Chesham JE, Hayter EA, Rosenbrier-Ribeiro L, Newham P, Clevers H, Bechtold DA, O’Neill JS (2019) Insulin/IGF-1 drives PERIOD synthesis to entrain circadian rhythms with feeding time. Cell 177(4):896-909.e20.

    Article  Google Scholar 

  8. Di Lorenzo L, De Pergola G, Zocchetti C, L’Abbate N, Basso A, Pannacciulli N, Cignarelli M, Giorgino R, Soleo L (2003) Effect of shift work on body mass index: results of a study performed in 319 glucose-tolerant men working in a Southern Italian industry. Int J Obes 27(11):1353–1358.

    Article  Google Scholar 

  9. Doi R, Oishi K, Ishida N (2010) CLOCK regulates the circadian rhythms of hepatic glycogen synthesis through transcriptional activation of Gys2. J Biol Chem 285(29):22114–22121.

    Article  Google Scholar 

  10. Fonken LK, Workman JL, Walton JC, Weil ZM, Morris JS, Haim A, Nelson RJ (2010) Light at night increases body mass by shifting the time of food intake. Proc Natl Acad Sci USA 107(43):18664–18669.

    Article  Google Scholar 

  11. Gómez-Ulla F (2009) Diabetic retinopathy. Curr 5(1):1–2.

    Article  Google Scholar 

  12. Greenwell BJ, Trott AJ, Beytebiere JR, Pao S, Bosley A, Beach E, Finegan P, Hernandez C, Menet JS (2019) Rhythmic food intake drives rhythmic gene expression more potently than the hepatic circadian clock in mice. Cell Rep 27(3):649-657.e5.

    Article  Google Scholar 

  13. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, Leblanc M, Chaix A, Joens M, Fitzpatrick JAJ, Ellisman MH, Panda S (2012) Time-restricted feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell Metab 15(6):848–860.

    Article  Google Scholar 

  14. Inoki K, Kim J, Guan KL (2012) AMPK and mTOR in cellular energy homeostasis and drug targets. Annu Rev Pharmacol Toxicol 52(1):381–400.

    Article  Google Scholar 

  15. Ishikawa K, Shimazu T (1976) Daily rhythms of glycogen synthetase and phosphorylase activities in rat liver: influences of food and light. Life Sci 19(12):1873–1878

    Article  Google Scholar 

  16. Ishikawa K, Shimazu T (1980) Circadian rhythm of liver glycogen metabolism in rats: effects of hypothalamic lesions. Am J Physiol 238(1):E21–E25

    Article  Google Scholar 

  17. Kim H, Zheng Z, Walker PD, Kapatos G, Zhang K (2017) CREBH maintains circadian glucose homeostasis by regulating hepatic glycogenolysis and gluconeogenesis. Mol Cell Biol.

    Article  Google Scholar 

  18. Kumar Jha P, Challet E, Kalsbeek A (2015) Circadian rhythms in glucose and lipid metabolism in nocturnal and diurnal mammals. Mol Cell Endocrinol 418(1):74–88.

    Article  Google Scholar 

  19. Kuroda H, Tahara Y, Saito K, Ohnishi N, Kubo Y, Seo Y, Otsuka M, Fuse Y, Ohura Y, Hirao A, Shibata S (2012) Meal frequency patterns determine the phase of mouse peripheral circadian clocks. Sci Rep 2:711.

    Article  Google Scholar 

  20. la Fleur SE (2003) Daily rhythms in glucose metabolism: suprachiasmatic nucleus output to peripheral tissue. J Neuroendocrinol 15(3):315–322.

    Article  Google Scholar 

  21. Liu S, Brown JD, Stanya KJ, Homan E, Leidl M, Inouye K, Bhargava P, Gangl MR, Dai L, Hatano B, Hotamisligil GS, Saghatelian A, Plutzky J, Lee CH (2013) A diurnal serum lipid integrates hepatic lipogenesis and peripheral fatty acid use. Nature 502(7472):550–554.

    Article  Google Scholar 

  22. Masaki T, Chiba S, Yasuda T, Noguchi H, Kakuma T, Watanabe T, Sakata T, Yoshimatsu H (2004) Involvement of the hypothalamic histamine H1 receptor in the regulation of feeding rhythm and obesity. Diabetes 53(9):2250–2260.

    Article  Google Scholar 

  23. Miller BH, McDearmon EL, Panda S, Hayes KR, Zhang J, Andrews JL, Antoch MP, Walker JR, Esser KA, Hogenesch JB, Takahashi JS (2007) Circadian and CLOCK-controlled regulation of mouse transcriptome and cell proliferation. Proc Natl Acad Sci USA 104(9):3342–3347.

    Article  Google Scholar 

  24. Oike H (2017) Modulation of circadian clocks by nutrients and food factors. Biosci Biotechnol Biochem 81(5):863–870.

    Article  Google Scholar 

  25. Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, Schultz PG, Kay SA, Takahashi JS, Hogenesch JB (2002) Coordinated transcription of key pathways in mice by the circadian clock. Cell 109(3):307–320.

    Article  Google Scholar 

  26. Roesler WJ, Khandelwal RL (1985) Diurnal variations in the activities of the glycogen metabolizing enzymes in mouse liver. Int J Biochem 17(1):81–85

    Article  Google Scholar 

  27. Rosen ED, Spiegelman BM (2006) Adipocytes as regulators of energy balance and glucose homeostasis. Nature 444(7121):847–853.

    Article  Google Scholar 

  28. Salgado-Delgado R, Angeles-Castellanos M, Saderi N, Buijs RM, Escobar C (2010) Food intake during the normal activity phase prevents obesity and circadian desynchrony in a rat model of night work. Endocrinology 151(3):1019–1029.

    Article  Google Scholar 

  29. Saponaro C, Gaggini M, Carli F, Gastaldelli A (2015) Subtle balance between lipolysis and lipogenesis: a critical point in metabolic homeostasis. Nutrients 7(11):9453–9474.

    Article  Google Scholar 

  30. Shan Z, Li Y, Zong G, Guo Y, Li J, Manson JE, Hu FB, Willett WC, Schernhammer ES, Bhupathiraju SN (2018) Rotating night shift work and adherence to unhealthy lifestyle in predicting risk of type 2 diabetes: results from two large US cohorts of female nurses. BMJ 363:k4641.

    Article  Google Scholar 

  31. Shi Z, Riley M, Taylor A, Noakes M (2017) Meal-specific food patterns and the incidence of hyperglycemia in a Chinese adult population. Br J Nutr 118(1):53–59.

    Article  Google Scholar 

  32. Shimizu H, Hanzawa F, Kim D, Sun S, Laurent T, Umeki M, Ikeda S, Mochizuki S, Oda H (2018) Delayed first active-phase meal, a breakfastskipping model, led to increased body weight and shifted the circadian oscillation of the hepatic clock and lipid metabolism-related genes in rats fed a high-fat diet. PLoS ONE 13(10):1–17.

    Article  Google Scholar 

  33. Stokkan KA, Yamazaki S, Tei H, Sakaki Y, Menaker M (2001) Entrainment of the circadian clock in the liver by feeding. Science 291(5503):490–493.

    Article  Google Scholar 

  34. Stucchi P, Gil-Ortega M, Merino B, Guzmán-Ruiz R, Cano V, Valladolid-Acebes I, Somoza B, Le Gonidec S, Argente J, Valet P, Chowen JA, Fernández-Alfonso M, Ruiz-Gayo M (2012) Circadian feeding drive of metabolic activity in adipose tissue and not hyperphagia triggers overweight in mice: Is there a role of the pentose-phosphate pathway? Endocrinology 153(2):690–699.

    Article  Google Scholar 

  35. Tahara Y, Shibata S (2018) Entrainment of the mouse circadian clock: Effects of stress, exercise, and nutrition. Free Radic Biol Med 119:129–138.

    Article  Google Scholar 

  36. Tahara Y, Aoyama S, Shibata S (2017) The mammalian circadian clock and its entrainment by stress and exercise. J Physiol Sci 67(1):1–10.

    Article  Google Scholar 

  37. Tsai LL, Tsai YC, Hwang K, Huang YW, Tzeng JE (2005) Repeated light-dark shifts speed up body weight gain in male F344 rats. Am J Physiol Endocrinol Metab 289(2):E212–E217.

    Article  Google Scholar 

  38. Ueda HR, Chen W, Adachi A, Wakamatsu H, Hayashi S, Takasugi T, Nagano M, Nakahama KI, Suzuki Y, Sugano S, Iino M, Shigeyoshi Y, Hashimoto S (2002) A transcription factor response element for gene expression during the circadian night. Nature 418(6897):534–539.

    Article  Google Scholar 

  39. Vetter C, Dashti HS, Lane JM, Anderson SG, Schernhammer ES, Rutter MK, Saxena R, Scheer FAJL (2018) Night shift work, genetic risk, and type 2 diabetes in the UK Biobank. Diabetes Care 41(4):762–769.

    Article  Google Scholar 

  40. Virtanen P, Gommers R, Oliphant TE et al (2020) SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods 17:261–272.

    Article  Google Scholar 

  41. Wang XS, Armstrong MEG, Cairns BJ, Key TJ, Travis RC (2011) Shift work and chronic disease: epidemiological evidence. Occup Med 61(2):78–89.

    Article  Google Scholar 

  42. Wehrens SMT, Christou S, Isherwood C, Middleton B, Gibbs MA, Archer SN, Skene DJ, Johnston JD (2017) Meal timing regulates the human circadian system. Curr Biol 27(12):1768-1775.e3.

    Article  Google Scholar 

  43. Zani F, Breasson L, Becattini B, Vukolic A, Montani JP, Albrecht U, Provenzani A, Ripperger JA, Solinas G (2013) PER2 promotes glucose storage to liver glycogen during feeding and acute fasting by inducing Gys2 PTG and GL expression. Mol Metab 2(3):292–305.

    Article  Google Scholar 

Download references


This work was done by the Japan Society for the Promotion of Science Research Fellowship (DC1) and Grant-in-Aid for Japan Society for the Promotion of Science Fellows to A.H (18J20316). We are grateful to Camila Caldana, Yusuke Nakane, and Shigenobu Shibata for their helpful advice from the point of view of experimental research. We also thank Shingo Gibo, Hiroshi Ito, Gen Kurosawa, and Motohide Seki for their useful comments on mathematical modeling.

Author information



Corresponding author

Correspondence to Akane Hara.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


Appendix A: Dependence of costs on the threshold of hyperglycemia and energy depletion

In this Appendix, we discuss the dependency of energy depletion costs on the values of thresholds \(x_{G}^{U}\) and EL. In Sects. 3.2 and 3.3, we calculated the risks of energy exhaustion (CE) and hyperglycemia (CG) setting \(x_{G}^{U} = 2.0\) and EL = 4.0 in Eq. (9a) and (9b), based on the two regulations. First, we set thresholds that did not result in CE = 0 or CG = 0 in multiple conditions of (φF, φL) so as to choose one condition that achieves the lowest risk. Second, we did not consider the cases where the energy levels were below the threshold throughout the period. We considered that if a chronic lack of energy occurs, the animal cannot survive and the amount of food intake L(t) will increase. Varying \(x_{G}^{U}\) and EL, these regulations are violated under certain conditions as follows.

Threshold of energy exhaustion

We calculated the risks of energy exhaustion by varying the threshold to EL = 2.0, 4.0, 8.0 (Fig. 5a). The result of EL = 4.0 was the same as the result shown in Sect. 3.1.

Fig. 5

Threshold dependences of risks of energy exhaustion and excess glucose levels. We calculated two possible risks, denoted by Eq. (9b) and (9c) for 24 × 24 = 576 combinations of (φF, φL). The light-colored cells represent a higher level of risk, as shown in the scales. The condition achieved the lowest risk in the white frames. The white and black bars along the axis represent the resting and active times, respectively. (a) Risk of energy exhaustion. (i) EL = 1.0. The risks become zero in the multiple conditions represented by the black area. (ii) EL = 2.0. The lowest risk is achieved at (φF, φL) = (9, 23). (iii) EL = 4.0. The lowest risk is achieved at (φF, φL) = (1, 0). (b) The risk of hyperglycemia. (i) \(x_{G}^{U} = 1.0\). The lowest risk is achieved at (φF, φL) = (19, 3). (ii) \(x_{G}^{U} = 2.0\). The lowest risk is achieved at (φF, φL) = (19, 0). (iii) \(x_{G}^{U} = 4.0\). The risks become zero in the black area

When we set EL = 2.0, the threshold was too low to evaluate the best conditions. This is because some conditions of (φF, φL) resulted in CE = 0 (black cells in Fig. 5b) and we could not find the optimum (φF, φL).

In contrast, when we set EL = 8.0, the threshold was too high because the peaks of energy levels were below the threshold EL = 8.0 in some (φF, φL). Furthermore, energy exhaustion occurred throughout the day, even in the feeding phases. Under such conditions, the lowest risk was achieved (φF, φL) = (0, 21) in the condition where φF was during the active period, which is differs from EL = 4.0. Therefore, the threshold of energy exhaustion has an effect on the result of searching for the condition of lowest risks.

Threshold of hyperglycemia

We calculated the risks of hyperglycemia by varying the threshold to \(x_{G}^{U} = 1.0, 2.0, 4.0\) (Fig. 5b). The result of \(x_{G}^{U} = 2.0\) is shown in Sect. 3.2 [the lowest risk was achieved in (φF, φL) = (18, 1)].

When we set \(x_{G}^{U} = 1.0\), the condition of lowest risk was achieved at (φF, φL) = (19, 4). When we set \(x_{G}^{U} = 4.0\), many conditions of (φF, φL) achieved CG = 0 (Fig. 5b) and the threshold was too low to find the best condition for the lowest CG. The conclusion did not depend on \(x_{G}^{U}\) much compared to EL. This is because the glucose levels declined to near zero during fasting phases, and the lack of glucose occurred in all of the conditions during the resting periods. Then, chronic hyperglycemia did not occur where \(x_{G}^{U} = 1.0, 2.0, 4.0\), unlike in the case of varying EL.

Appendix B: Dependence of costs on relative energy production efficiency

In this Appendix, we discuss the dependency of energy depletion costs on the relative energy production efficiency, denoted by f in Eq. (8b). A larger f represents a higher fat efficiency as an energy resource compared to the efficiency of glucose. From one molecule of glucose, about 30 ATPs were obtained as a result of aerobic respiration. On the other hand, from one molecule of palmitic acid, 106 ATPs were produced through β-oxidation. If we consider that one molecule of triglyceride, consists of three palmitic acids, 318 ATPs can be obtained from the triglyceride. Thus, the standard relative ATP production efficiency is set \(f = \frac{318}{{30}} = 10.6\). We calculated the costs for energy depletion to vary f = 5.0, 10.6, and 15.0 (Fig. 6). In Sect. 3.1, we set f = 10.6 as a standard value (Fig. 6b)). We examined the smaller value f = 5.0 (Fig. 6a) and the larger value f = 15.0 (Fig. 6c).

Fig. 6

Relative energy production efficiency depends on the risk of energy depletion. We calculated the risks of energy depletion, denoted by Eq. (8a) for 24 × 24 = 576 combinations of (φF, φL) changing the relative energy production efficiency (f). The light-colored cells represent a higher level of risk, as shown in the scales. The condition achieved the lowest risk in the white frames. The white and black bars along the axis represent the resting and active times, respectively. (a) f = 5.0. The lowest risk is achieved at (φF, φL) = (23, 23). (b) f = 10.6. The lowest risk is achieved at (φF, φL) = (11, 22). (c) f = 15.0. The lowest risk is achieved at (φF, φL) = (9, 23)

The optimal timing of glycogenesis (φL) did not depend on f and the risks of energy depletion were minimum around φL = 23 (Fig. 6a–c). To maintain the glycogen level during the resting periods, the peaks at the end of the resting periods were optimal, as explained in Sect. 3.2.

The optimal conditions for fat production (φF) were varied f. If f = 5.0, the value of risk did not depend on φF so much (Fig. 6a), and increasing f, the optimal conditions became dependent on φF (Fig. 6b, c), since fat had a greater impact on the energy level, and the production phase became more important with increasing f. The φF minimum risk was achieved at the end of the active periods when f = 5 (Fig. 6a), but the optimal φF was shifted to the end of the resting period when f = 10.6 and f = 15.0 (Fig. 6a, b).

As discussed in Sect. 3.1, fat can compensate for the lack of glucose at the end of the resting periods when the efficiency of fat as an energy resource is large enough (e.g. f = 10.6 and f = 15.0). Thus, producing fat at the end of the resting period can be an optimal strategy. However, when f is small (f = 5.0), the fat is unable to overcome the lack of glucose and energy depletion tends to occur when either fat or glucose is at a low level. Therefore, synchronized production of both fat and glycogen was a better strategy than shifted production timing.

Appendix C: Circadian oscillation is necessary to maintain robust peak phases of substances against timings of food intakes

The equilibria of the model (Eqs. 1, 5, and 6) can be derived by calculating \(x_{G}^{*} ,x_{F}^{*}\), and \(x_{L}^{*}\) that satisfy \(\dot{x}_{G} = \dot{x}_{F} = \dot{x}_{L} = 0\), as follows:

$$ \begin{array}{*{20}c} {x_{G}^{*} = \frac{aL\left( t \right)}{{\overline{{\gamma_{F} }} + d_{G} }},\quad x_{F}^{*} = \frac{{\overline{{\gamma_{F} }} }}{{d_{F} }}\frac{aL\left( t \right)}{{\overline{{\gamma_{F} }} + d_{G} }},\quad x_{L}^{*} = \frac{{\overline{{\gamma_{L} }} }}{{\overline{{\gamma_{\beta } }} }}\frac{aL\left( t \right)}{{\overline{{\gamma_{F} }} + d_{G} }},} \\ \end{array} $$

where circadian oscillation is not considered, \(\overline{{\gamma_{F} }} ,\overline{{\gamma_{L} }}\), and \(\overline{{\gamma_{\beta } }}\) can be assumed as constant values. The equilibria change in response to L(t), represented in Eq. (2a) and (2b), dG and dF are represented in Eq. (3a) and (3b), respectively. There are two equilibria corresponding to the feeding (L(t) = 1) and fasting (L(t) = 0) periods as follows:

$$ \begin{array}{*{20}c} {x_{G}^{*} = x_{F}^{*} = x_{L}^{*} = 0,\quad where\,L\left( t \right) = 0\quad \left( {i.e.\, fasting\,period} \right),} \\ \end{array} $$


$$ \begin{array}{*{20}c} {x_{G}^{*} = \frac{a}{{\overline{{\gamma_{F} }} + d_{G} }}, \quad x_{F}^{*} = \frac{{\overline{{\gamma_{F} }} }}{{d_{F} }}\frac{a}{{\overline{{\gamma_{F} }} + d_{G} }}, \quad x_{L}^{*} = \frac{{\overline{{\gamma_{L} }} }}{{\overline{{\gamma_{\beta } }} }}\frac{a}{{\overline{{\gamma_{F} }} + d_{G} }},\quad where\,L\left( t \right) = 1 \,\left( {i.e. \,feeding\,period} \right). } \\ \end{array} $$

Without the circadian oscillation of γF, γL, and γβ, the peak time of xG, xF, and xL in the two schedules were shifted by 12 h (lower row in Fig. 7), and the timings of the food intakes were directly reflected in the peak times of glucose, fat, and glycogen. In contrast, the peak phases did not dramatically shift depending on the two feeding schedules when the circadian oscillation was considered, where φF = 6, φL = 18 (upper row in Fig. 7). Considering the circadian regulation of the production rates, the peak times of glucose, fat, and glycogen were robust to changes of the timing of food intake.

Fig. 7

Daily patterns of variables with and without circadian oscillation of γL, γF, and γβ. Upper row: dynamics of glucose, fat, and glycogen with circadian oscillation, where φF = 6 and φL = 18. Lower row: dynamics without circadian oscillation, where the production rate of fat, glycogen, and glycogenolysis are fixed at the mean values of the maximum and minimum levels of the production rates (γF = 0.11, γL = 1.025, and γβ = 0.25; the maximum and minimum levels are shown in Table 1). The white and black bars along the axis represent the resting and active periods, respectively. The bold and dotted lines show the dynamics of food intake in the active period and in the resting period, respectively

Circadian regulation on the production rates also contributed to reduction of the risks of energy depletion and hyperglycemia. With the production rates fixed at the mean values of the maximum and minimum levels (γF = 0.11, γL = 1.025, and γβ = 0.25), the risks of energy depletion and high glucose level were 289.90 and 369.60, respectively, which were larger than the minimum risks under the optimal phase conditions: 35.33 at (φF, φL) = (11, 22) and 255.29 at (φF, φL) = (18, 1).

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hara, A., Satake, A. Why meals during resting time cause fat accumulation in mammals? Mathematical modeling of circadian regulation on glucose metabolism. J. Math. Biol. 83, 26 (2021).

Download citation


  • Circadian rhythm
  • Chrononutrition
  • Nutrient metabolism
  • Obesity

Mathematics Subject Classification

  • 92-10
  • 92B25