Iwayama et al.

The Journal of Physiological Sciences (2021) 71:35 The Journal of Physiological

https://doi.org/10.1186/s12576-021-00821-1
Sciences

ORIGINAL PAPER Open Access

Diurnal variations in muscle and liver

glycogen differ depending on the timing
of exercise
Kaito Iwayama1* , Yoko Tanabe2, Fumiya Tanji3, Takahiro Ohnishi4 and Hideyuki Takahashi2

Abstract
It has been suggested that glycogen functions not only in carbohydrate energy storage, but also as molecular sensors
capable of activating lipolysis. This study aimed to compare the variation in liver and muscle glycogen during the day
due to different timing of exercise. Nine healthy young men participated in two trials in which they performed a sin-
gle bout of exercise at 70% of their individual maximal oxygen uptake for 60 min in the post-absorptive (morning) or
post-prandial (afternoon) state. Liver and muscles glycogen levels were measured using carbon magnetic resonance
spectroscopy (13C MRS). Diurnal variations in liver and muscle glycogen compared to baseline levels were significantly
different depending on the timing of exercise. The effect of the timing of exercise on glycogen fluctuation is known
to be related to a variety of metabolic signals, and the results of this study will be useful for future research on energy
metabolism.
Keywords: Liver, Muscle, Glycogen, Post-absorptive exercise, Post-prandial exercise

Background vigorous-intensity aerobic physical activities per week to

The evidence regarding the health benefits of regular maintain a healthy body [4]. However, while these guide-
physical activity is well established, and previous research lines indicate the time, frequency, and intensity of exer-
demonstrates that virtually everyone benefits: men and cise, they do not describe the timing of exercise, such as
women, young children to older adults, women who are post-absorptive or post-prandial states.
pregnant or postpartum, people living with a chronic dis- Recently, the timing of exercise has received much
ease, or people attempting to reduce their risk of disease attention due to its influence on various factors, such as
[27]. Regular exercise can also be expected to achieve food intake [2], circadian rhythm [36, 41], energy sub-
and maintain desired body composition, such as weight strates [39], body composition [3, 10, 24, 33], the adap-
loss and weight gain inhibition, due to increased energy tation of training [6, 38], and glucose metabolism [23,
expenditure [27]. According to the World Health Organi- 37]. While there is still disagreement regarding some of
zation physical activity guidelines, adults are recom- the above, there is general consensus on the energy sub-
mended to undergo 150–300 min of moderate-intensity strates, i.e., exercise performed in the post-absorptive
or 75–150 min of vigorous-intensity physical activity, or state induces higher fat oxidation than exercise per-
an equivalent combination of moderate-intensity and formed in the post-prandial state [39]. It has long been
known that pre-exercise nutritional status affects energy
substrates during exercise [7, 14]. Additionally, post-
*Correspondence: k-iwayama@sta.tenri-u.ac.jp absorptive exercise increases 24-h fat oxidation, even in
1
Faculty of Budo and Sport Studies, Tenri University, 80 Tainoshocho, energy-balanced conditions [15, 31]. The body condi-
Tenri, Nara 632‑0071, Japan tion after an overnight fast is characterized by no energy
Full list of author information is available at the end of the article

© The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 2 of 8

intake since the supper of the previous night, moreover, day (day 2), two meals (breakfast at 09:00 and lunch at
overnight fasting is the only way to induce a post-absorp- 13:00) were provided, and participants ran at 70% of their
tive state in people with traditional eating habits of three ­VO2max for 60 min, beginning at 07:00 (morning) or 16:00
meals daily. During overnight fasting, substrate oxidation (afternoon) on a treadmill. Participants were instructed
predominantly shifts from glycogenolysis to fatty acid to remain awake and to maintain a sedentary position,
oxidation [30], and hormonal changes activate several except when performing prescribed exercise sessions.
transcription factors and regulate gene expression [11].
A possible physiological background for these aspects is
that glycogen variability is involved in whole-body energy Participants
metabolism. Glycogen in the liver and muscles is a stor- Nine healthy young men were recruited for the present
age source of energy and its quantity affects whole-body study after providing written informed consent. Partici-
energy metabolism. Specifically, decreased liver glyco- pants’ mean physical characteristics were as follows: age,
gen levels stimulate lipolysis in adipose tissue through 23.4 ± 1.3 years; height, 170.1 ± 1.5 cm; body weight,
a central nervous system-mediated mechanism [20], 59.3 ± 1.0 kg; and body fat, 10.1 ± 0.7%, maximal oxygen
and decreased muscle glycogen levels trigger sequential uptake ­( VO2max) was 60.7 ± 1.4 ml/kg/min. At the time
events, including the dissociation of the AMP-activated of the study, no participant had any medical condition
protein kinase (AMPK–glycogen interaction, enhanced or was taking medications or smoking. This study was
activity and altered intracellular localization of AMPK; approved by the Ethics Committee of the Japan Institute
and the upregulated expression of genes associated with of Sports Sciences (IRS-2017-048).
fat oxidation, such as carnitine palmitoyltransferase,
fatty acid translocase, and hormone-sensitive lipase [26].
In this way, in the liver and muscle glycogen is not only Pre‑study evaluation
a source of energy, but also a regulator of whole-body To determine the workload corresponding to 70% of the
energy metabolism. Nevertheless, none of the previ- individual ­VO2max, all participants performed a graded
ous studies has reported fluctuations in liver and muscle exercise test including submaximal and maximal tests
glycogen levels due to post-absorptive or post-prandial using a treadmill (Ohtake Root Kogyo, Japan) at least
exercise. 1 week before the main experiment. The initial running
This study aimed to clarify the effect of exercise in the speed was set at 9.0 km/h and was then increased by
morning or afternoon, i.e., post-absorptive or post-pran- 1.2 km/h every 3 min at each subsequent stage up to a
dial, on liver and muscle glycogen fluctuations during lactate concentration exceeding 4.0 mmol/L. Each stage
the day. Liver and muscle glycogen levels were measured was separated by a 1-min rest, and the rest was extended
over two trials, including a 60-min exercise session in the to 3 min when the lactate concentration exceeded
morning before breakfast or in the afternoon using car- 4.0 mmol/L. Then, the following running speed was set
bon magnetic resonance spectroscopy (13C MRS), which to the stage before the lactate concentration exceeded
is non-invasive. 4.0 mmol/L and was increased by 0.6 km/h every 1 min
until the participants reached voluntary exhaustion. A
Methods capillary blood sample was taken from fingertips to ana-
This study entailed a randomized repeated measures lyze the lactate concentration (Lactate Pro 2, Arkray Co.,
design including two trials with exercise sessions per- Ltd., Kyoto, Japan) immediately before the test, after each
formed in the morning or afternoon; the two trials were running stage, and after 1, 3, and 5 min of exercise to
separated for at least 1 week. exhaustion.
Each trial consisted of two consecutive days. On the Oxygen uptake was considered to be maximal when at
first day (day 1) of each experiment, participants were least two of the following four criteria were met: (1) VO2
permitted to engage in normal activities, such as walk- reached a plateau; (2) heart rate exceeded 90% of the age-
ing up and down stairs, until arriving at the laboratory. predicted maximal value; (3) respiratory exchange ratio
Thereafter, they were instructed to refrain from per- increased above 1.10, and (4) ratings of perceived exer-
forming strenuous exercise and consuming beverages, tion at the end of the test ≥ 19. The highest V
­ O2 for two
such as energy, caffeinated, or alcoholic drinks, except consecutive 15 s during the test was taken as the V­ O2max.
for experimental meals. Participants arrived at the labo- Respiratory gas was continuously collected using the
ratory before 17:30. Their body composition was sub- breath-by-breath method (AE310S, Minato Medical Sci-
sequently measured using the bioimpedance method ence Co., Osaka, Japan). Regression analysis revealed that
(InBody 770; InBody, Tokyo, Japan). They consumed the relative oxygen uptake and treadmill running velocity
supper at 18:30 and went to bed at 23:00. On the second corresponded to 70% of the individual ­VO2max.
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 3 of 8

Experimental meals Tokyo, Japan). Blood glucose levels were determined by

Experimental meals were designed to achieve energy- blood samples taken from participants’ fingertips (Medis-
balanced conditions, assuming a resting metabolic rate of afe FIT, Terumo, Tokyo, Japan) [13].
24.0 kcal/kg/day according to estimated energy require-
ments for Japanese individuals [1]. The physical activity Experimental exercise and physical activity
factor was assumed to be 1.75 (2471 ± 39 kcal/day) on Participants ran for 60 min at 70% of their ­VO2max on a
day 1 and 1.85 on day 2 (1732 ± 33 kcal for breakfast and treadmill (Ohtake Root Kogyo, Japan).
lunch) based on a previous study using a room-size meta- During the experimental 60-min running session, res-
bolic chamber [17]. Participants consumed 6.2 ± 0.1 g/kg piratory gas was continuously collected using the breath-
of body weight of carbohydrates (370 ± 7 g/day on day 1 by-breath method (AE310S, Minato Medical Science Co.,
and 4.4 ± 0.0 g/kg of body weight (260 ± 5 g for breakfast Osaka, Japan). Immediately after completing the 60-min
and lunch on day 2. Experimental meals comprised 15% exercise, lactate concentration was measured by blood
protein, 25% fat, and 60% carbohydrate; these values were samples taken from participants’ fingertips.
expressed as percentages of the total energy intake. Non-exercise-related activity was estimated using a
hip-worn accelerometer device (ActiGraph GT9X Link,
Glycogen measurements ActiGraph, Pensacola FL), which was validated using a
Muscle and liver glycogen contents were measured using three-axial accelerometer and data filtering technology
13
C MRS at 20:00 on day 1, and at 06:00, 08:00, 12:00, [28]. Participants wore accelerometers upon wakening
15:00, and 17:00 on day 2. Muscle glycogen was assessed on day 1 until the final glycogen measurement on day 2,
using the calf, as it is one of the major muscles that con- except during experimental exercise.
tract during running [5]. 13C MRS was performed using
a clinical 3-Tesla superconducting magnetic resonance Statistical analyses
system (Magnetom Verio and Magnetom Skyra, Siemens, Data in the main text and figures are presented as
Germany) as previously described [18, 35]. Briefly, the means ± standard errors. To determine the sample size,
13
C MRS muscle spectra were collected in 15-min blocks the effect size (d) from the present data was expected to
at 200-ms intervals, resulting in 4500 acquisitions using be 0.7 for the comparison of changes in muscle and liver
a 13C–1H double-tuned surface coil, 10 cm in diameter glycogen fluctuation during the day. The results of the
(Takashima Seisakusho, Tokyo, Japan) that was set on power analysis using G*Power version 3.1.9.2 (Univer-
the calf muscle. Muscle glycogen levels were quantified sity of Dusseldorf, Germany) indicated a minimum sam-
by comparing them with an external standard solution ple size of eight to ensure a power of 0.8 and an α level
(120 mM glycogen from oysters and 50 mM KCl). Simi- of P < 0.05. Considering the possibility of dropouts, we
larly, the 13C MRS liver spectra were collected in 16-min recruited nine participants. Mean condition pair values
blocks at 160-ms intervals, resulting in 6000 acquisitions were compared using a paired t test. Muscle and liver
using the same coil as that for muscle measurement; the glycogen levels, respiratory gas analysis during exercise
coil was set on the right side of the trunk by the liver. sessions, and blood parameters between both trials were
Liver glycogen levels were quantified by comparing them compared using a two-way repeated-measures analysis
with an external standard solution (200 mM glycogen of variance to identify the main effect of trial and time.
from oysters and 150 mM KCl). When the first meas- The interaction between trial and time was determined
urement was taken in both the muscle and liver, the coil using Bonferroni’s correction with post hoc pairwise
position was marked on the skin with a felt-tipped pen comparisons. Changes from baseline in muscle and liver
to ensure that 13C MRS data were collected at the same glycogen between both trials were compared using a two-
position during subsequent measurements. way repeated-measures analysis of variance to identify
the main effect of trial and time. The interaction between
trial and time was determined using Bonferroni’s (trial)
Blood sampling and analyses
and Dunnett’s (time) correction with post hoc pairwise
Blood samples were collected from the antecubital vein
comparisons. Statistical significance was set at P < 0.05.
in commercially available vacuum-sealed serum collec-
All statistical analyses were performed using SPSS statis-
tion tubes (Nipro, Osaka, Japan) at the same time as gly-
tical software version 24 (IBM Japan, Tokyo, Japan).
cogen measurements. Serum samples were obtained by
centrifugation at 3000 rpm for 10 min at 4 °C and were
stored at − 80 °C until analysis. From the obtained sam- Results
ples, insulin, and triglyceride levels were measured in All participants completed both trials and there were no
an independent laboratory (LSI Medience Corporation, significant differences in body mass, body fat, or fat-free
mass between trials.
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 4 of 8

During the experimental exercises in the two trials, time (P < 0.01) and time and trial interactions (P < 0.01)
participants ran on a treadmill at a speed equivalent were observed for change from baseline in muscle glyco-
to 70% of their ­VO2max for 60 min (12.3 ± 0.4 km/h). gen, although the main effect of trial (P = 0.11) was not
This intensity was lower than the lactate threshold statistically significant. Similarly, significant main effects
of each individual calculated in the pre-study evalu- of trial (P < 0.05), time (P < 0.01) and time and trial inter-
ation (14.8 ± 0.6 km/h). No significant differences actions (P < 0.01) were observed for change from baseline
were observed in the average heart rate (morning, in liver glycogen.
153 ± 4 bpm; and afternoon, 150 ± 3 bpm; P = 0.12), post- Significant main effects of trial (P < 0.05) and time
exercise blood lactate (morning, 1.7 ± 0.4 mM; and after- (P < 0.01) were observed for blood glucose levels,
noon, 1.5 ± 0.3 mM; P = 0.16), and the rating of perceived although the main effect of trial and time interaction
exertion (morning, 11.3 ± 0.9; and afternoon, 11.1 ± 0.7; was not statistically significant (Fig. 2a). Significant
P = 0.65) between trials. The results of respiratory analy- main effects of time were observed on serum insulin lev-
sis during exercise are shown Table 1. els (P < 0.01), although the main effect of trial and time
Table 2 shows the time course of glycogen levels in interaction was not statistically significant (Fig. 2b). Sig-
the muscle and liver. Significant main effects of trial nificant main effects of time (P < 0.01) and trial and time
(P < 0.05) and time (P < 0.01) and time and trial interac- interactions (P < 0.05) were observed on serum triglycer-
tions (P < 0.01) were observed for muscle glycogen lev- ide levels. At 15:30, serum triglyceride levels were signifi-
els. Significant main effects of time (P < 0.01) and trial cantly higher in the afternoon than in the morning trial
and time interactions (P < 0.01) and a significant trial (P < 0.05, Fig. 2c).
tendency (P = 0.06) were observed when analyzing liver Participants maintained an inactive and sedentary life-
glycogen levels. Figure 1 shows the change from baseline style during both experimental days, except when they con-
in muscle and liver glycogen. Significant main effects of ducted exercises. No significant differences were observed

Table 1 Results of respiratory analyses

10 min 20 min 30 min 40 min 50 min 60 min Average

O2 uptake (ml/min)
Morning 2494 ± 51 2623 ± 52 2618 ± 51 2620 ± 54 2614 ± 56 2624 ± 55 2599 ± 51*
Afternoon 2470 ± 45 2577 ± 46 2569 ± 50 2566 ± 54 2564 ± 55 2561 ± 54 2551 ± 49
CO2 production (ml/min)
Morning 2315 ± 70 2439 ± 65 2423 ± 65 2414 ± 72 2414 ± 73 2419 ± 71 2404 ± 68**
Afternoon 2408 ± 66 2541 ± 67 2531 ± 67 2521 ± 73 2508 ± 75 2506 ± 77 2502 ± 69
Ventilation (L/min)
Morning 66.8 ± 3.2 72.0 ± 3.2 73.6 ± 3.5 74.5 ± 4.0 75.1 ± 4.2 75.4 ± 4.2 72.9 ± 3.6
Afternoon 68.6 ± 3.1 74.6 ± 3.2 75.3 ± 3.2 75.1 ± 3.2 75.3 ± 3.5 75.8 ± 3.5 74.1 ± 3.2
Respiratory exchange ratio
Morning 0.93 ± 0.02 0.93 ± 0.01 0.92 ± 0.01 0.92 ± 0.01 0.92 ± 0.01 0.92 ± 0.01 0.92 ± 0.01**
Afternoon 0.97 ± 0.01 0.99 ± 0.01 0.98 ± 0.01 0.98 ± 0.01 0.98 ± 0.01 0.98 ± 0.01 0.98 ± 0.01
Values are shown as mean ± standard errors. *P < 0.05, **P < 0.01 vs afternoon trial

Table 2 Results of muscle and liver glycogen measurements

Day 1 Day 2
20:00 6:00 8:00 12:00 15:00 17:00

Muscle glycogen (mM)

Morning 86 ± 7 88 ± 7 59 ± 7** 65 ± 6** 73 ± 5** 80 ± 6
Afternoon 90 ± 6 93 ± 6 87 ± 6 90 ± 7 97 ± 7 72 ± 7
Liver glycogen (mM)
Morning 245 ± 20 185 ± 14 130 ± 16** 169 ± 18* 202 ± 18** 213 ± 20
Afternoon 239 ± 15 193 ± 17 187 ± 14 206 ± 14 266 ± 21 207 ± 20
Values are shown as mean ± standard errors. *P < 0.05, **P < 0.01 vs afternoon trial
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 5 of 8

40

Supper
a

Exercise / Rest
Exercise / Rest
Sleep

Breakfast

Lunch
20
Muscle glycogen (mM)

0
*
*,‡

**,‡
-20 *,‡
‡

-40
18:00 22:00 2:00 6:00 10:00 14:00 18:00
Time (h)
150
Supper

Exercise / Rest

Exercise / Rest
b
Sleep

Breakfast

Lunch
100
Liver glycogen (mM)

50

0 †
†
‡
‡
-50
†
**, ‡
‡
-100 *, ‡

**, ‡
-150
18:00 22:00 2:00 6:00 10:00 14:00 18:00
Time (h)
Fig. 1 Change from baseline (at 20:00 on day 1) in muscle (a) and liver (b) glycogen in the morning (closed circles) and afternoon trials (open
circles). Values are shown as means ± SEs. *P < 0.05 vs. afternoon trial. **P < 0.01 vs. afternoon trial. †P < 0.05 vs. baseline. ‡P < 0.05 vs. baseline

between the morning and afternoon trials in non-exercise in the morning was associated with relatively lower liver
physical activity (day 1, 325 ± 27, 317 ± 24 counts/min, and muscle glycogen during the day compared to exercise
P = 0.81; and day 2; 139 ± 9, 138 ± 5 counts/min, P = 0.86) in the afternoon.
and step counts (day 1, 6746 ± 735, 6687 ± 651, P = 0.94; Compared with baseline values (i.e., at 20:00 on day
day 2, 1733 ± 239, 1639 ± 176, P = 0.28). 1), liver glycogen levels decreased after overnight fast-
ing in the morning (-23%) and afternoon (-21%) trials.
Discussion A key function of liver glycogen is to maintain blood
The purpose of this study was to determine the effect of glucose levels by releasing glucose into the bloodstream
the timing of exercise on fluctuations in liver and mus- [12]; therefore, its degradation after overnight fasting
cle glycogen levels. Our main findings were that exercise is reasonable [18]. The morning trial induced further
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 6 of 8

carbohydrates at breakfast and lunch to replenish the

150
a glycogen degradation by the overnight fast. Conse-
quently, it was found that liver glycogen fluctuations
during the day differ depending on exercise timing. A
shortage in liver glycogen directly facilitates lipolysis
Blood glucose (mg/dL)

100
in white adipose tissue by activating a liver–brain–adi-
pose neurocircuitry, this demonstrates the presence of
a “glycogen depletion signal” [20, 40]. The fat utilization
50 signal may continue to increase fat oxidation after exer-
Exercise/Rest

Exercise/Rest
cise in the morning as decreased liver glycogen levels
Breakfast

remain severely depressed after breakfast [15]. In addi-

Lunch

0
tion to fat metabolism, plasma glucose clearance rates
6:00 8:00 10:00 12:00 14:00 16:00 18:00 are higher in the morning compared with the evening in
Time (h) response to similar glucose intake [34]. Since increased
30 liver glycogen synthesis leads to improved glucose tol-
b
Lunch
Exercise/Rest

Exercise/Rest
Breakfast

erance [29], a significant transient decrease in liver

glycogen by exercise may be beneficial in improving
glucose tolerance. It has been suggested that optimiz-
20 ing the timing of exercise is beneficial for improving
Insulin (µU/ml)

glucose homeostasis [9] and diurnal variation in liver

glycogen with different timing of exercise would pro-
10
vide useful insights for carbohydrate/fat metabolism.
Muscle glycogen levels did not decrease after overnight
fasting in the morning (+ 2%) and afternoon (+ 3%) tri-
als compared with each baseline value. Diurnal muscle
0 glycogen variation has been observed to be markedly
6:00 8:00 10:00 12:00 14:00 16:00 18:00
small on sedentary days [18, 35] because muscle glyco-
Time (h) gen does not contribute to the maintenance of blood
200 glucose. Thus, there were no significant differences in
c
Breakfast

Lunch
Exercise/Rest

Exercise/Rest

pre-exercise muscle glycogen levels compared to base-

line levels in each study, but the diurnal variation dif-
150
Triglyceride (mg/dL)

fered significantly depending on the timing of exercise.

The respiratory exchange ratio during the experimen-
100 tal exercise was significantly lower in the morning trial
* than in the afternoon trial (Table 1). Nonetheless, the net
change in muscle glycogen due to experimental exercise
50 did not differ between trials. Previous research indicates
that net muscle glycogen breakdown is similar between
post-absorptive and post-prandial exercise, but there is
0
6:00 8:00 10:00 12:00 14:00 16:00 18:00
a significant increase fat oxidation and decrease in the
Time (h) intramyocellular triglyceride in type I muscle after post-
Fig. 2 Changes in blood glucose (a), insulin (b), and triglyceride absorptive compared to post-prandial exercise [8]. It has
(c) levels in the morning (closed circles) and afternoon trials (open been suggested that intramyocellular triglyceride content
circles). Values are shown as means ± SEs. *P < 0.05 vs. afternoon trial is a marker or mediator of muscle insulin resistance [22].
Furthermore, muscle glycogen is not only a carbohy-
drate energy reserve, but also a molecular sensor capable
low liver glycogen levels (-46%, at 08:00 on Day 2) by of activating signaling pathways in response to exercise,
experimental exercise, and liver glycogen remained sig- including the nuclear translocation of AMPK and upreg-
nificantly decreased relative to baseline, even at the last ulation of genes responsible for fat oxidation [26]. Addi-
measurement. However, there was no significant differ- tionally, exercise-induced muscle glycogen depletion
ence in liver glycogen levels before experimental exer- can increase subsequent insulin sensitivity and prevent
cise compared to baseline in the afternoon trial. This glucose from being diverted to de novo lipogenesis [21].
was due to the fact the participants consumed enough When considering carbohydrate/fat metabolism and
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 7 of 8

glucose tolerance, it is potentially beneficial to know the may also be applied to consider the optimal exercise tim-
timing of exercise and muscle glycogen fluctuations. ing for patients with diabetes.
In summary, the diurnal fluctuations in liver and mus-
Acknowledgements
cles glycogen varied depending on the timing of exercise. Not applicable.
Previous studies suggest that the timing of exercise may
affect energy metabolism over 24 h by altering the pat- Authors’ contributions
KI and FT conceived and designed research. KI analyzed data, interpreted
tern of glycogen fluctuations [15, 16, 31]. Furthermore, results of experiments and prepared figures; KI, YT, and TO performed experi-
acute [8, 32] and chronic [24, 38] post-absorptive exercise ments; KI and HT drafted manuscript, edited and revised manuscript. All
have been observed to induce different metabolic adap- authors approved final version of manuscript.

tations than post-prandial exercise, and have been asso- Funding

ciated with diurnal variation of glycogen storage. Future This study was supported by the Japan Society for the Promotion of Science
studies are needed to evaluate the effect of the timing of (Grant-in-Aid for Research Activity start-up 16H07476).
exercise on glycogen fluctuations in the liver and muscles Availability of data and materials
over a longer period. The datasets used and/or analyzed during the current study are available from
One limitation of the present study is that it did not the corresponding author on reasonable request.
measure liver and muscle glycogen levels after exercise in
the afternoon trial, that is, the effect of supper and sleep Declarations
on post-exercise glycogen fluctuation. Thus, the effect of Ethics approval and consent to participate
morning or afternoon exercise on the circadian rhythm This study was approved by the Ethics Committee of the Japan Institute of
of glycogen variability remains unclear. Although the Sports Sciences (IRS-2017-048). Written informed consent was obtained from
all participants.
nocturnal energy expenditure and substrate did not dif-
fer depending on the timing of exercise [15, 17], it would Consent for publication
be interesting to investigate liver and muscle glycogen Written informed consent for publication was obtained from all participants.
levels the following morning after supper and sleep to Competing interests
examine the relationship between glycogen fluctuations The authors declare that they have no competing interests.
and linked physiological indicators. Another limita-
Author details
tion is that the post-exercise carbohydrate intake timing 1
Faculty of Budo and Sport Studies, Tenri University, 80 Tainoshocho, Tenri,
might not have been optimal for glycogen resynthesis. Nara 632‑0071, Japan. 2 Faculty of Health and Sport Sciences, University
Consuming carbohydrates immediately after exercise is of Tsukuba, Ibaraki, Japan. 3 Sport Medical Science Research Institute, Tokai Uni-
versity, Kanagawa, Japan. 4 Medical Center, Japan Institute of Sport Sciences,
better for glycogen resynthesis than consuming carbohy- Tokyo, Japan.
drates 2 h after exercise [19]. For the purpose of glyco-
gen measurement, participants in this study consumed Received: 26 May 2021 Accepted: 5 November 2021
breakfast 90 min after exercise rather than immediately
in the morning trial. However, differences in intake tim-
ing do not affect glycogen synthesis when evaluated over
References
8 h or 24 h [25]. Therefore, we believe that the effect of 1. Anon (2015) Dietary reference intakes for Japanese (2015). Ministry of
dietary intake timing in this study on eventual glycogen Health Labour and Welfare of Japan, Tokyo
storage is negligible. Finally, the intensity and duration of 2. Bachman JJ, Deitrick RW, Hillman AR (2016) Exercising in the fasted
state reduced 24-hour energy intake in active male adults. J Nutr Metab
the experimental exercise may be too intense for the gen- 2016:1984198. https://​doi.​org/​10.​1155/​2016/​19841​98
eral population. The energy expenditure during exercise, 3. Blankenship JM, Rosenberg RC, Rynders CA, Melanson EL, Catenacci VA,
which is feasible for many people, may be lower than in Creasy SA (2021) Examining the role of exercise timing in weight man-
agement: a review. Int J Sports Med 42:967–978
the experimental exercise of the present study, and con- 4. Bull FC, Al-Ansari SS, Biddle S et al (2020) World Health Organization 2020
sequently, the variability in glycogen may be smaller. guidelines on physical activity and sedentary behavior. Br J Sports Med
54:1451–1462. https://​doi.​org/​10.​1136/​bjspo​rts-​2020-​102955
5. Cappellini G, Ivanenko YP, Poppele RE, Lacquaniti F (2006) Motor Patterns
in human walking and running. J Neurophysiol 95:3426–3437
Conclusions 6. Chtourou H, Souissi N (2012) The effect of training at a specific time of
Exercise performed in the morning before breakfast was day: a review. J Strength Cond Res 26:1984–2005
7. Coyle EF, Coggan AR, Hemmert MK, Lowe RC, Walters TJ (1985) Substrate
associated with a relative decrease in liver and muscle usage during prolonged exercise following a preexercise meal. J Appl
glycogen during the day compared to exercise performed Physiol 59:429–433. https://​doi.​org/​10.​1152/​jappl.​1985.​59.2.​429
in the afternoon. The effect of the timing of exercise on 8. De Bock K, Richter EA, Russell AP, Eijnde BO, Derave W, Ramaekers M,
Koninckx E, Leger B, Verhaeghe J, Hespel P (2005) Exercise in the fasted
glycogen fluctuation, which has been shown to be related state facilitates fibre type-specific intramyocellular lipid breakdown
to a variety of metabolic signals, will be useful in future and stimulates glycogen resynthesis in humans. J Physiol 564:649–660.
studies of energy metabolism. The results of this study https://​doi.​org/​10.​1113/​jphys​iol.​2005.​083170
Iwayama et al. The Journal of Physiological Sciences (2021) 71:35 Page 8 of 8

9. Gabriel BM, Zierath JR (2019) Circadian rhythms and exercise - re-setting Endocrinol Metab 302:E1343–E1351. https://​doi.​org/​10.​1152/​ajpen​do.​
the clock in metabolic disease. Nat Rev Endocrinol 15:197–206 00004.​2012
10. Gillen JB, Percival ME, Ludzki A, Tarnopolsky MA, Gibala MJ (2013) Interval 27. Piercy KL, Troiano RP, Ballard RM, Carlson SA, Fulton JE, Galuska DA,
training in the fed or fasted state improves body composition and George SM, Olson RD (2018) The physical activity guidelines for Ameri-
muscle oxidative capacity in overweight women. Obesity 21:2249–2255. cans. JAMA 320:2020–2028. https://​doi.​org/​10.​1001/​jama.​2018.​14854
https://​doi.​org/​10.​1002/​oby.​20379 28. Robusto KM, Trost SG (2012) Comparison of three generations of
11. Goldstein I, Hager GL (2015) Transcriptional and chromatin regulation ActiGraph activity monitors in children and adolescents. J Sport Sci
during fasting—the genomic era. Trends Endocrinol Metab 26:699–710. 30:1429–1435. https://​doi.​org/​10.​1080/​02640​414.​2012.​710761
https://​doi.​org/​10.​1016/j.​tem.​2015.​09.​005 29. Ros S, Zafra D, Valles-Ortega J, Garcia-Rocha M, Forrow S, Dominguez J,
12. Gonzalez JT, Fuchs CJ, Betts JA, van Loon LJC (2016) Liver glycogen Calbo J, Guinovart JJ (2010) Hepatic overexpression of a constitutively
metabolism during and after prolonged endurance-type exercise. Am J active form of liver glycogen synthase improves glucose homeostasis. J
Physiol 311:E543–E553. https://​doi.​org/​10.​1152/​ajpen​do.​00232.​2016 Biol Chem 285:37170–37177
13. Hara H, Mihara M, Narushima M, Todokoro T, Yamamoto T, Koshima I 30. Rui L (2014) Energy metabolism in the liver. Compr Physiol 4:177–197.
(2012) Blood glucose measurement for flap monitoring to salvage flaps https://​doi.​org/​10.​1002/​cphy.​c1300​24
from venous thrombosis. J Plast Reconstr Aesthetic Surg 65:616–619. 31. Shimada K, Yamamoto Y, Iwayama K, Nakamura K, Yamaguchi S, Hibi M,
https://​doi.​org/​10.​1016/j.​bjps.​2011.​11.​026 Nabekura Y, Tokuyama K (2013) Effects of post-absorptive and postpran-
14. Horowitz JF, Mora-Rodriguez R, Byerley LO, Coyle EF (1997) Lipolytic dial exercise on 24 h fat oxidation. Metabolism 62:793–800. https://​doi.​
suppression following carbohydrate ingestion limits fat oxidation during org/​10.​1016/j.​metab​ol.​2012.​12.​008
exercise. Am J Physiol 273:E768–E775. https://​doi.​org/​10.​1152/​ajpen​do.​ 32. Stocks B, Dent JR, Ogden HB, Zemp M, Philp A (2019) Postexercise skeletal
1997.​273.4.​E768 muscle signaling responses to moderate- to high-intensity steady-state
15. Iwayama K, Kawabuchi R, Park I, Kurihara R, Kobayashi M, Hibi M, Oishi S, exercise in the fed or fasted state. Am J Physiol 316:E230–E238. https://​
Yasunaga K, Ogata H, Nabekura Y, Tokuyama K (2015) Transient energy doi.​org/​10.​1152/​ajpen​do.​00311.​2018
deficit induced by exercise increases 24-h fat oxidation in young trained 33. Schoenfeld BJ, Aragon AA, Wilborn CD, Krieger JW, Sonmez GT (2014)
men. J Appl Physiol 118:80–85. https://​doi.​org/​10.​1152/​jappl​physi​ol.​ Body composition changes associated with fasted versus non-fasted
00697.​2014 aerobic exercise. J Int Soc Sports Nutr 11:54. https://​doi.​org/​10.​1186/​
16. Iwayama K, Kurihara R, Nabekura Y, Kawabuchi R, Park I, Kobayashi M, s12970-​014-​0054-7
Ogata H, Kayaba M, Satoh M, Tokuyama K (2015) Exercise increases 34. Sonnier T, Rood J, Gimble JM, Peterson CM (2014) Glycemic control is
24-h fat oxidation only when it is performed before breakfast. EBioMed impaired in the evening in prediabetes through multiple diurnal rhythms.
2:2003–2009. https://​doi.​org/​10.​1016/j.​ebiom.​2015.​10.​029 J Diabetes Complications 28:836–843
17. Iwayama K, Ogawa A, Tanaka Y, Yajima K, Park I, Ando A, Ogata H, Kayaba 35. Takahashi H, Kamei A, Osawa T, Kawahara T, Takizawa O, Maruyama K
M, Zhang S, Tanji F, Nabekura Y, Yamamoto K, Tokuyama K (2020) Effects (2015) 13C MRS reveals a small diurnal variation in the glycogen content
of exercise before breakfast on plasma free fatty acid profile and 24-h fat of human thigh muscle. NMR Biomed 28:650–655. https://​doi.​org/​10.​
oxidation. Metabol Open 8:100067. https://​doi.​org/​10.​1016/j.​metop.​2020.​ 1002/​nbm.​3298
100067 36. Tanaka Y, Ogata H, Kayaba M, Ando A, Park I, Yajima K, Araki A, Suzuki
18. Iwayama K, Onishi T, Maruyama K, Takahashi H (2020) Diurnal variation C, Osumi H, Zhang S, Isihara A, Takahashi K, Shoda J, Nabekura Y, Satoh
in the glycogen content of the human liver using 13C MRS. NMR Biomed M, Tokuyama K (2020) Effect of a single bout of exercise on clock gene
33:e4289. https://​doi.​org/​10.​1002/​nbm.​4289 expression in human leukocyte. J Appl Physiol 128:847–854
19. Ivy JL, Katz AL, Cutler CL, Sherman WM, Coyle EF (1988) Muscle glycogen 37. Terada T, Wilson BJ, Myette-Cote E, Kuzik N, Bell GJ, McCarggar LJ, Boule
synthesis after exercise: effect of time of carbohydrate ingestion. J Appl NG (2016) Targeting specific interstitial glycemic parameters with high-
Physiol 64:1480–1485. https://​doi.​org/​10.​1152/​jappl.​1988.​64.4.​1480 intensity interval exercise and fasted-state exercise in type 2 diabetes.
20. Izumida Y, Yahagi N, Takeuchi Y et al (2013) Glycogen shortage during Metabolim 65:599–608
fasting triggers liver–brain–adipose neurocircuitry to facilitate fat utiliza- 38. Van Proeyen K, Szlufcik K, Nielens H, Ramaekers M, Hespel P (2011) Ben-
tion. Nat Commun 4:2316. https://​doi.​org/​10.​1038/​ncomm​s3316 eficial metabolic adaptations due to endurance exercise training in the
21. Jensen J, Rustad PI, Kolnes AJ, Lai YC (2011) The role of skeletal muscle fasted state. J Appl Physiol 110:236–245. https://​doi.​org/​10.​1152/​jappl​
glycogen breakdown for regulation of insulin sensitivity by exercise. Front physi​ol.​00907.​2010
Physiol 2:112. https://​doi.​org/​10.​3389/​fphys.​2011.​00112 39. Vieira AF, Costa RR, Macedo RCO, Coconcelli L, Kruel LFM (2016) Effects of
22. Kelley DE, Goodpaster BH, Storlien L (2002) Muscle triglyceride and aerobic exercise performed in fasted v. fed state on fat and carbohydrate
insulin resistance. Annu Rev Nutr 22:325–346 metabolism in adults: a systematic review and meta-analysis. Br J Nutr
23. Mancilla R, Krook A, Schrauwen P, Hesselink MKC (2020) Diurnal regula- 116:1153–1164. https://​doi.​org/​10.​1017/​S0007​11451​60031​60
tion of peripheral glucose metabolism: potential effects of exercise 40. Yahagi N (2017) Hepatic control of energy metabolism via the autonomic
timing. Obesity 28:S38–S45 nervous system. J Atheroscler Thromb 24:14–18. https://​doi.​org/​10.​5551/​
24. Nybo L, Pedersen K, Christensen B, Aagaard P, Brandt N, Kiens B (2009) jat.​RV160​02
Impact of carbohydrate supplementation during endurance training on 41. Yamanaka Y, Hashimoto S, Takasu NN, Tanahashi U, Nishide S, Honma S,
glycogen storage and performance. Acta Physiol 197:117–127. https://​ Honma K (2015) Morning and evening physical exercise differentially
doi.​org/​10.​1111/j.​1748-​1716.​2009.​01996.x regulate the autonomic nervous system during nocturnal sleep in
25. Parkin JA, Carey MF, Martin IK, Stojanovska L, Febbraio MA (1997) Muscle humans. Am J Physiol 309:R1112–R1121
glycogen storage following prolonged exercise: effect of timing of
ingestion of high glycemic index food. Med Sci Sport Exerc 29:220–224.
https://​doi.​org/​10.​1097/​00005​768-​19970​2000-​00009 Publisher’s Note
26. Philp A, Hargreaves M, Baar K (2012) More than a store: regulatory roles Springer Nature remains neutral with regard to jurisdictional claims in pub-
for glycogen in skeletal muscle adaptation to exercise. Am J Physiol lished maps and institutional affiliations.