Diminuição da gordura corporal regional após treinamento de intervalo de alta intensidade a longo prazo

Prévia do material em texto

J Phys Fitness Sports Med, 6 (2): 103-110 (2017)
DOI: 10.7600/jpfsm.6.103
JPFSM: Regular Article
Decrease in regional body fat after long-term high-intensity interval training
Koichiro Azuma1*, Yusuke Osawa2, Shogo Tabata1, Fuminori Katsukawa2, Hiroyuki Ishida2,
Yuko Oguma2, Toshihide Kawai3, Hiroshi Itoh3, Shigeo Okuda4, Shuji Oguchi5,
Atsumi Ohta5, Haruhito Kikuchi5, Mitsuru Murata5 and Hideo Matsumoto1
Received: October 26, 2016 / Accepted: January 18, 2017
Abstract High-intensity interval training (HIIT) has recently received much attention as a 
new option for aerobic training. Despite its smaller time requirement, HIIT has been reported 
to have a greater effect than continuous moderate-intensity training on fat loss, especially a 
decrease in truncal adiposity. We therefore examined whether long-term HIIT preferentially 
modulates truncal adiposity rather than peripheral adiposity, especially thigh adiposity, where 
local muscle energy consumption increased profoundly during HIIT. We also examined the as-
sociation between changes in adipose tissue distribution and serum adiponectin level. Twelve 
healthy male participants (28-48 years old) were assigned to a group that performed HIIT using 
only a leg ergometer (L-HIIT, n = 7) or to a group that performed HIIT using both leg and arm 
ergometers (LA-HIIT, n = 5) twice weekly for 16 weeks. The training programs consisted of 
8 to 12 sets of >90% V・O2 peak for 1 min, with 1 min of very light active recovery. Body com-
position analyses as well as aerobic fitness and measurements of serum adiponectin were per-
formed at baseline and after intervention. A linear improvement in aerobic fitness was observed 
along with a decrease in leg fat (5.4 ± 1.7 vs. 5.1 ± 1.7 kg, p < 0.05) near the main working 
muscles during HIIT in the combined (L+LA-HIIT) group. Moreover, there was an association 
of decrease in leg fat or thigh adiposity with improvement in aerobic fitness in the combined 
group (ρ = -0.59, p < 0.05; and ρ = -0.71, p < 0.05, respectively). Visceral adiposity was de-
creased in L-HIIT (115 ± 45 vs. 100 ± 47 cm2, p < 0.05), however no decrease was observed in 
total fat or truncal fat in either group. No change was observed in serum adiponectin concentra-
tion in either group. Changes in serum adiponectin were associated with changes in visceral 
adiposity in the combined group (ρ = -0.72, p < 0.01). Regional rather than whole-body fat loss 
was observed after a 16-week HIIT program.
Keywords : high-intensity interval training (HIIT), regional adiposity, adiponectin, aerobic fitness
Introduction
 Loss of fat mass is known to occur with a relative in-
crease in energy expenditure against energy intake. There-
fore, exercise volume, which is the product of exercise 
intensity, duration, and frequency, is mainly responsible 
for exercise-induced fat loss. High-intensity interval train-
ing (HIIT) has recently received much attention as a new 
option for aerobic training, not only among athletes, but 
also people with obesity and those with time-constraints 
because it can be performed in a safe and time-efficient 
manner1). HIIT is unique in its focus on exercise inten-
sity rather than training volume as the main contribu-
tor to its effects. It has been suggested that, even with 
a significantly smaller exercise volume, high-intensity 
exercise has a more favorable effect on fat loss than does 
continuous moderate-intensity exercise2,3). Few studies, 
however, have examined the effect of long-term HIIT on 
body composition4-7). Trapp et al. showed that HIIT had a 
significant reduction in fat mass and trunk fat, compared 
with energy-matched, longer-duration continuous endur-
ance training among young women. They also found a fat 
loss in legs compared to arms in HIIT only, though only 
DXA scan was used for the analyses of body fat distribu-
tion. The same group has also reported loss in visceral fat 
as well as total and trunk fat among overweight young 
men by HIIT, though thigh adiposity was not examined in 
detail. A greater understanding of the effects of long-term *Correspondence: azumakx@keio.jp
1 Institute for Integrated Sports Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
2 Sports Medicine Research Center, Keio University, 4-1-1 Hiyoshi, Kohoku-ku, Yokohama City, Kanagawa 223-8521, Japan
3 Department of Internal Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
4 Department of Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
5 Department of Laboratory Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
104 JPFSM : Azuma K, et al.
HIIT on body composition, especially body fat distribu-
tion, can facilitate planning effective structured exercise 
programs for individuals with time constraints who desire 
to lose body fat. 
 Adiponectin, an adipocytokine, favorably influences 
the development of atherosclerosis and energy homeo-
stasis8), and has received much attention in relation to the 
metabolic effects of exercise, partly due to the effects of 
adiponectin on the modulation of skeletal muscle energy 
metabolism in animal models9-11). In fact, studies have 
revealed that long-term exercise that improves fitness 
levels, increases insulin sensitivity, and reduces body fat, 
is also associated with increased resting adiponectin lev-
els12-14). However, it is unclear whether the exercise itself 
or exercise-induced fat loss is associated with increases 
in adiponectin2,14). We therefore examined the effects of 
long-term HIIT on body fat distribution and changes in 
serum adiponectin levels. 
Material and Methods
 Twelve healthy male participants, free of any known 
chronic diseases and weight-stable for at least 3 months 
before enrollment, were assigned to one of two groups as 
described previously15). Clinical characteristics are shown 
in Table 1. Briefly, HIIT was performed using only a leg 
ergometer in one group (L-HIIT [n = 7]), and using both 
leg and arm ergometers in the other group (LA-HIIT [n 
= 5]), twice weekly for 16 weeks. The training programs 
consisted of 8-12 sets of >90% V・O2 peak for 1 min with a 
1-min very light active recovery. The number of the rep-
etitions was gradually increased from 8 to 12 sets during 
the first 4 weeks, and then it was adjusted depending on 
the participants’ physical condition and was a minimum 
of 8 sets, while the workload was progressively increased. 
HIIT using leg ergometer in LA-HIIT was performed by 
half sets (4-6 sets) of HIIT in L-HIIT, and the remaining 
4-6 sets of HIIT were performed using an arm ergom-
eter in LA-HIIT. The workload for HIIT using an arm 
ergometer was set at >90% of peak workload, which was 
determined from an incremental exercise test using an 
arm ergometer15). This study was approved by the ethics 
committee of Keio University (2011-098-2), and written 
informed consent was obtained from all participants.
Measurements. V・ O2 peak and peak work rate were 
measured during an incremental exercise test on a leg 
ergometer, as reported previously15), at baseline after 4 
and 16 weeks of HIIT. V・O2 at ventilatory threshold (VT) 
and work rate at VT were also used as indices of aerobic 
fitness. VT is the point at which pulmonary ventilation 
increases disproportionately with oxygen consumption 
during an incremental test. The increase in ventilation re-
sults from the body’s bicarbonate buffering of lactic acid 
accumulation from anaerobic metabolism and consequent 
exhalation of CO2.
 A 2 h, 75 g oral glucose-tolerance test (OGTT) with 
blood collection at 0, 30, 60, and 120 min., was per-
formed before and after the 16-week intervention. Blood 
samples were also collected in fasting state after4 weeks 
of HIIT, 24-48 h after completing the last training session. 
Blood samples for plasma collection were immediately 
placed on ice and subsequently centrifuged (3000 × g, 12 
min, 4°C). Samples for serum collection were left at room 
temperature for 30 min and centrifuged. All samples were 
stored at -80°C until analyses were performed. Plasma 
glucose and serum insulin were determined by enzyme-
linked immunosorbent assay (glucose: Glucoroder MAX, 
A & T Corp., Yokohama, Japan; insulin: AIA-2000 Au-
tomated Immunoassay Analyzer, Tosoh Bioscience Inc., 
South San Francisco, CA, USA) using an automated 
analyzer (LABOSPECT 008, Hitachi, Tokyo, Japan). 
Serum high-molecular-weight (HMW) adiponectin was 
measured by chemiluminescence enzyme immunoassay 
(Fujirebio Diagnostics, Tokyo, Japan). 
Body composition. Whole-body dual-energy X-ray ab-
sorptiometry (DEXA) scan (Lunar Prodigy® Advance, 
GE Healthcare Japan, Tokyo, Japan) was performed at 
Keio University Hospital using enCORE software, ver-
sion 9.2 (GE Healthcare). Standard scan mode was used 
for whole-body scans to measure total and regional (upper 
body, trunk, and legs) fat mass (FM) and lean body mass 
(LBM).
Table 1. HIIT-related improvements in aerobic fitness.
Combined HIIT, L-HIIT plus LA-HIIT groups; HIIT, high-intensity interval training; L-HIIT, leg-
ergometry group; LA-HIIT, leg- and arm-ergometry group; BMI, body mass index.
1 
 
Table 1 HIIT-related improvements in aerobic fitness. 
 L-HIIT LA-HIIT Combined HIIT 
 (n= 7) (n= 5) (n= 12) 
Age 34±4 37±9 35±6 
Height (cm) 173±5 175±5 174±5 
Body weight (kg) 72.2±6.8 77.3±6.1 74.3±6.7 
BMI 24.1±2.3 25.4±2.8 24.6±2.5 
Plasma Glucose (mg･ml-1) 96±5 86±5 92±7 
Serum Insulin (µU･ml-1) 8.6±4.0 6.0±2.7 7.5±3.6 
Combined HIIT, L-HIIT plus LA-HIIT groups; HIIT, high-intensity interval training; L-HIIT, leg-ergometry group; LA-HIIT, leg- and arm-
ergometry group; BMI, body mass index. 
 
105JPFSM : Regional fat loss after16-week HIIT
 Magnetic resonance imaging of abdominal and thigh 
adipose tissue (AT) was performed using a 1.5 T (Tesla) 
imager (Signa™ Excite 1.5 T TwinSpeed HD; GE Health-
care, Waukesha, WI, USA). A set of T2-weighted single-
shot fast-spin echo images (repetition time/echo time = 
∞/90 ms, field of view = 38 cm, slice thickness = 8 mm 
with 3-mm interval, matrix size = 288 × 192 pixels) was 
obtained for the abdomen, centered at the L3-L4 disc 
space level. Subsequently, a set of T2-weighted fast-spin 
echo images (repetition time/echo time = 4000/90 ms, 
field of view = 30 cm, slice thickness = 8 mm with 3-mm 
interval, matrix size = 320 × 224 pixels), taken at the 
midpoint of the femur (between the superior border of the 
patella and the greater trochanter) was used to measure the 
cross-sectional area of thigh subcutaneous AT (SAT). The 
obtained Digital Imaging and Communication in Medi-
cine (DICOM) images were analyzed with sliceOmatic 
software (TomoVision, Magog, Quebec, Canada). the pre-
intervention MRI scans of the thigh from one participant 
in L-HIIT group were missing; therefore, thigh SAT for six 
participants in the L-HIIT group were used for analyses. 
Statistical analyses. The effects of the intervention were 
determined using Friedman’s two-way ANOVA to test a 
linear time-trend in each intervention group. In the pres-
ent study, because we focused on the changes in body 
composition resulting from HIIT, the combined results 
of L-HIIT and LA-HIIT groups are presented, especially 
due to the increase in statistical power for the correla-
tion analyses. However, since HIIT using leg ergometer 
in LA-HIIT was performed by half sets (4-6 sets) of L-
HIIT, the effect of the training on adipose tissue distribu-
tion, especially thigh adiposity, as well as aerobic fitness 
measured using leg ergometer, can be different between 
the two groups. Therefore, results of each group were 
also presented. Post-hoc analyses were performed using a 
Bonferroni test to examine the significant improvements 
within each group. For the effects of the intervention on 
body composition and metabolic variables, the Wilcoxon 
signed-rank test was used in each group. A selective bi-
variate relationship was investigated using Spearman’s 
rank correlation coefficient. IBM SPSS for Windows, 
Version 19 (IBM Corp., Armonk, NY, USA) was used for 
all statistical analyses. The level of significance was set at 
p < 0.05, and all values are presented as mean ± standard 
deviation (SD).
Results
Exercise tolerance test. As shown in Table 2, a signifi-
cant linear increase in peak work rate was observed in 
the exercise tolerance tests of both the L-HIIT (p < 0.01) 
and LA-HIIT (p = 0.02) groups. Percent increases in V・O2 
peak after 4 weeks and 16 weeks were 10% and 16%, 
respectively, in the combined (L+LA-HIIT) group (both p 
< 0.01). Percent increases in V・O2 at VT after 4 weeks and 
16 weeks of HIIT were 17% and 30%, respectively (both 
p < 0.01).
Metabolic variables. There were no significant changes 
in glucose and insulin levels during the 75-g oral glucose-
tolerance test in the present participants, who were 
healthy, non-diabetic men (Fig. 1), but glucose and insu-
lin levels tended to be lower after 60 min in the L-HIIT 
group (both p = 0.06).
 There was no change in serum HMW adiponectin con-
Table 2. HIIT-related improvements in aerobic fitness.
1 
 
Table 2 HIIT-related improvements in aerobic fitness. 
 L-HIIT LA-HIIT Combined HIIT 
 (n= 7) (n= 5) (n= 12) 
 0 week 4 week 16 week 0 week 4 week 16 week 0 week 4 week 16 week 
VO2peak (ml・min-1・kg-1) 41 ± 7 45 ± 5 48 ± 5# 38 ± 5 41 ± 4 43 ± 5 40 ± 6 44 ± 5 46 ± 5## 
 % change of VO2peak 11 ± 11 19 ± 11* 8 ± 10 13 ± 12 10 ± 10* 16 ± 11** 
VO2 at VT (ml・min-1・kg-1) 25 ± 7 29 ± 5 32 ± 4# 24 ± 5 26 ± 4 29 ± 5 24 ± 5 26 ± 4 29 ± 5## 
 % change of VO2 at VT 20 ± 18 33 ± 24* 13 ± 16 25 ± 28 17 ± 17* 30 ± 25** 
Peak work rate (W) 217 ± 30 237 ± 22 267 ± 20## 225 ± 35 243 ± 31 250 ± 26# 225 ± 35 243 ± 31 250 ± 26## 
 % change of Peak work rate 10 ± 7 25 ± 13** 9 ± 7 12 ± 11* 9 ± 7** 19 ± 13** 
Work rate at VT (W) 133 ± 29 148 ± 20 166 ± 23## 142 ± 39 152 ± 29 163 ± 24 142 ± 39 152 ± 29 163 ± 24## 
 % change of Work rate at VT 15 ± 18 28 ± 20** 10 ± 17 21 ± 31 13 ± 17** 25 ± 24** 
 
*p<0.05, **p<0.01 vs. baseline (0 weeks) by Bonferroni post hoc test. 
#p<0.05, ##p<0.01 by Friedman’s two-way ANOVA. 
*p<0.05, **p<0.01 vs. baseline (0 weeks) by Bonferroni post hoc test. 
#p<0.05, ##p<0.01 by Friedman’s two-way ANOVA.
Combined HIIT, L-HIIT plus LA-HIIT groups; HIIT, high-intensity interval training; L-HIIT, leg-ergometry group; LA-HIIT, leg- and 
arm-ergometry group; V・O2, oxygen consumption; VT, ventilatory threshold.
106 JPFSM : Azuma K, et al.
centration after the 16-week intervention in either group 
or the combined group (2.2 ± 1.4 vs. 2.0 ± 1.1 μg·mL-1); 
however, it tended to decrease temporarily after 4 weeks 
in both groups and was significant in the combined group 
(2.2 ± 1.4 vs. 1.7 ± 0.9 μg·mL-1, p = 0.01 by Wilcoxon’s 
signed-rank test) (Fig. 2).
Body composition (Table 3 and Fig. 3). As shown in 
Table 3, there was no change in FM (17.7 ± 5.0 vs. 17.1 
± 5.6 kg) or body weight after the 16-week intervention 
in either group or the combined group. However, leg 
FM slightly, but significantly, decreased in the combined 
group (5.4 ± 1.7 vs. 5.1 ± 1.7 kg, p = 0.02) with a slight, 
but significant, increase in LBM (53.6 ± 4.8 vs. 54.7 ± 4.6 
kg, p = 0.03). In the L-HIIT group, thigh SAT also tended 
to decrease (99 ± 31 vs. 89 ± 26 cm2, p = 0.06), and per-
cent changes in leg FM and thigh SAT were both nega-
tively correlated with percent changes in V・O2 peak (ρ = 
Fig. 1 Glucose and insulin during GTT. A, C, E: Plasma glucose curvesduring GTT. B, D, F: Serum insulin curve during GTT. A, B: 
L-HIIT; C, D: LA-HIIT; E, F: combined L- and LA-HIIT groups. Solid lines are baseline values; dotted lines are values after the 
16-week HIIT program. No significant change in glucose or insulin resulted from HIIT. GTT, glucose tolerance test; HIIT, high-
intensity interval training.
Fig. 2 Serum HMW adiponectin concentration. There was no significant change in serum HMW adiponectin after HIIT; however, 
when comparing baseline and 4-week time points only, serum HMW adiponectin was temporarily decreased after 4 weeks of 
HIIT (p = 0.01). *p < 0.05 by Wilcoxon’s signed-rank test. Combined HIIT, L-HIIT plus LA-HIIT groups; HIIT, high-intensity 
interval training; HMW, high-molecular-weight; L-HIIT, leg-ergometry group; LA-HIIT, leg- and arm-ergometry group.
C
D
E
F
Pl
as
am
gl
uc
os
e 
(m
g䡡
m
l-1
)
time (min)
Se
ru
m
 In
su
lin
 (ȝ
8
䡡m
l-1
)
time (min)
Pl
as
am
gl
uc
os
e 
(m
g
䡡m
l-1
)
Se
ru
m
 In
su
lin
 (ȝ
8
䡡m
l-1
)
Pl
as
am
gl
uc
os
e 
(m
g
䡡m
l-1
)
Se
ru
m
 In
su
lin
 (ȝ
8
䡡m
l-1
)
A
B
time (min)
time (min)
time (min)
time (min)
0
50
100
150
200
0 30 60 120
0 week
16 week
0
40
80
120
160
0 30 60 120
0 week
16 week
0
50
100
150
200
0 30 60 120
0
40
80
120
160
0 30 60 120
0
50
100
150
200
0 30 60 120
0
40
80
120
160
0 30 60 120
Se
ru
m
 H
M
W
 a
di
po
ne
ct
in
 (
ȝJ
䡡m
l-1
) * 
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
L-HIIT LA-HIIT Combined HIIT
Baseline
4-week
16-week
107JPFSM : Regional fat loss after16-week HIIT
-0.59, p = 0.04; and ρ = -0.71, p = 0.02, respectively) and 
with percent changes in V・O2 at VT (ρ = -0.67, p = 0.02; 
and ρ = -0.73, p = 0.01, respectively) in the combined 
group. Though visceral AT (VAT) also decreased in the 
L-HIIT group (115 ± 45 vs 100 ± 47 cm2, p < 0.05) and 
tended to decrease in the combined group (117 ± 38 vs 
107 ± 49 cm2, p = 0.07), abdominal SAT was unchanged 
in either group and there was no correlation between 
percent changes in abdominal SAT or VAT and percent 
change in V・O2 peak or V
・O2 at VT.
Association of serum HMW adiponectin (Figs. 3 and 4). 
 As shown in Figs. 3C and D, percent changes in serum 
adiponectin were not associated with percent changes in 
V・O2 peak or peak work rate, or V
・O2 at VT or work rate at 
VT in either group. However, percent changes in serum 
adiponectin were associated with percent changes in AT, 
especially VAT in the combined group (ρ = -0.72, p < 0.01, 
Fig. 4). 
Discussion
 In the present study, we observed that thigh AT, rather 
than abdominal AT, was preferentially lost during 16-
week HIIT. We also observed that a decrease in thigh AT 
was associated with an aerobic fitness gain, suggesting 
a loss of regional body fat near the main active muscles 
during the training.
 We did not, however, observe an increase in the serum 
adiponectin level due to HIIT, which was associated with 
changes in fat loss, especially loss of VAT, rather than 
aerobic fitness gains owing to HIIT.
 Energy consumption during HIIT is relatively smaller 
than during conventional moderate-intensity, continuous 
training; because of its high intensity, HIIT cannot be 
continuously performed1). Nevertheless, HIIT has been 
reported to have a greater effect on fat loss, as well as 
favorable metabolic changes, than continuous training, 
independent of exercise volume2,3). Tremblay et al. were 
the first to report that better preferential fat loss could be 
achieved with HIIT than with conventional endurance 
training despite less than half the energy expenditure with 
HIIT3). Coker et al. showed a significant reduction in VAT, 
with no significant change in body weight, in the high-
intensity continuous exercise group only, with no change 
observed in the moderate-intensity and non-exercising 
groups16). Because visceral adiposity is a sensitive marker 
for energy excess of the whole body17), high-intensity 
exercise may indeed modulate whole-body energy bal-
ance independent of exercise-induced energy expenditure. 
In the present study VAT did not change, whereas thigh 
adiposity significantly decreased. Goodpaster et al.18) re-
ported that thigh subfascial AT has similar characteristics 
to those of VAT, which was not separated from thigh SAT 
in the present study. In fact, a similar reduction in lower-
body vs. upper-body fat after a 1-year lifestyle interven-
tion was reported by Albu et al.19), who observed prefer-
ential loss of AT in deeper (i.e., VAT and deep SAT in the 
abdomen and subfascial AT in the thigh) than in more su-
perficial locations. Indeed, in the present study there was 
Table 3. HIIT-related changes in body composition seen using DEXA and MRI.Table 3 HIIT-related changes in body composition seen using DEXA and MRI 
 L-HIIT LA-HIIT Combined HIIT 
 (n= 7) (n= 5) (n= 12) 
DEXA 0 week 16 week 0 week 16 week 0 week 16 week 
Body weight (kg) 72.2 ± 6.8 72.0 ± 7.0 77.3 ± 6.1 78.0 ± 8.7 74.3 ± 6.7 74.5 ± 8.0 
Fat mass (kg) 16.2 ± 4.8 15.1 ± 4.4 19.8 ± 4.9 19.9 ± 6.5 17.7 ± 5.0 17.1 ± 5.6 
Lean body mass (kg) 53.1 ± 4.8 54.1 ± 4.4 54.4 ± 5.3 55.5 ± 5.2 53.6 ± 4.8 54.7 ± 4.6* 
Leg fat mass (kg) 5.0 ± 1.5 4.6 ± 1.3 5.9 ± 1.9 5.8 ± 2.0 5.4 ± 1.7 5.1 ± 1.7* 
Trunk fat mass (kg) 9.2 ± 3.3 8.6 ± 2.8 11.6 ± 2.5 11.7 ± 3.9 10.2 ± 3.2 9.9 ± 3.6 
MRI 
Thigh SAT (cm2) 99 ± 31 89 ± 26 113 ± 44 110 ± 55 105 ± 36 98 ± 41 
Abdominal SAT (cm2) 126 ± 44 114 ± 39 155 ± 56 161 ± 67 138 ± 49 134 ± 55 
VAT (cm2) 115 ± 45 100 ± 47* 119 ± 30 116 ± 56 117 ± 38 107 ± 49 
 
* p<0.05 by Wilcoxon’s signed-rank test. 
* p<0.05 by Wilcoxon’s signed-rank test.
Combined HIIT, L-HIIT plus LA-HIIT groups; DEXA, dual-energy X-ray absorptiometry; HIIT, high-intensity interval training; 
L-HIIT, leg-ergometry group; LA-HIIT, leg- and arm-ergometry group; MRI, magnetic resonance imaging; SAT, subcutaneous 
adipose tissue; VAT, visceral adipose tissue.
108 JPFSM : Azuma K, et al.
Fig. 3 Association of percent changes in aerobic fitness with adiposity and serum HWM adiponectin. (A) Significant negative associa-
tion between percent change in V・O2 peak and percent change in thigh SAT (ρ = -0.71, p = 0.02). (B) No significant association 
between percent change in V・O2 peak with percent change in VAT. (C, D) No association of percent change in serum HMW 
adiponectin with aerobic fitness gain, a marker of response to HIIT intervention. Squares indicate L-HIIT participants; circles 
indicate LA-HIIT participants. HIIT, high-intensity interval training; HMW, high-molecular-weight; L-HIIT, leg-ergometry 
group; LA-HIIT, leg- and arm-ergometry group; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; V・O2, oxygen 
consumption; WR, work rate.
-40
-30
-20
-10
0
10
20
30
-20 -10 0 10 20 30 40
-50
-40
-30
-20
-10
0
10
20
30
40
-20 -10 0 10 20 30 40
-60
-40
-20
0
20
40
60
80
-20 -10 0 10 20 30 40
-60
-40
-20
0
20
40
60
80
-20 -10 0 10 20 30 40 50
% changes in VO2 peak
%
 c
ha
ng
es
 in
 th
ig
h 
SA
T 
% changes in VO2 peak
%
 c
ha
ng
es
 in
 V
AT
% changes in VO2 peak
%
 c
ha
ng
es
 in
 se
ru
m
 H
M
W
 a
di
po
ne
ct
in
 
% changes in peak work rate
%
 c
ha
ng
es
 in
 se
ru
m
 H
M
W
 a
di
po
ne
ct
in
 
B
D
A
C
L- HIIT(n= 7)
LA- HIIT(n= 5)
ρ= -0.71, p= 0.02
Fig. 4 Association of percentchanges in serum HWM adiponectin with adiposity. (A) No significant association of percent change in 
serum HMW adiponectin with thigh SAT. (B) Significant negative association between percent change in serum HMW adipo-
nectin and percent change in VAT (ρ = -0.72, p < 0.01). HIIT, high-intensity interval training; HMW, high-molecular-weight; L-
HIIT, leg-ergometry group; LA-HIIT, leg- and arm-ergometry group; SAT, subcutaneous adipose tissue; VAT, visceral adipose 
tissue.
-40
-30
-20
-10
0
10
20
30
-60 -40 -20 0 20 40 60 80
-50
-40
-30
-20
-10
0
10
20
30
40
-60 -40 -20 0 20 40 60 80
%
 c
ha
ng
es
 in
 th
ig
h 
SA
T 
%
 c
ha
ng
es
 in
 V
AT
% changes in serum HMW adiponectin 
BA
% changes in serum HMW adiponectin 
L- HIIT(n=7)
LA- HIIT(n=5)
ρ = -0.72, p < 0.01
109JPFSM : Regional fat loss after16-week HIIT
physical activity levels were changed by supervised aero-
bic exercise, that the HMW adiponectin level was indeed 
reduced by aerobic exercise training. Therefore, one can 
assume that fat loss associated with training counteracts 
or overrides the decreasing effect of exercise per se lead-
ing to no change or even an increase in adiponectin level 
in exercise training programs, which are long enough to 
influence changes in body composition.
 We were not able to observe improvements in metabolic 
variables along with improvements in aerobic fitness 
of the male participants in the present study. Consider-
ing a robust improvement in aerobic fitness and well-
maintained non-training physical activity level (data not 
shown), a larger sample size that includes participants 
with metabolic abnormalities would be expected to reveal 
favorable metabolic changes. 
 The present study has several limitations. First, the 
study consisted of men with a relatively small range of 
ages. As age and sex might effect changes in aerobic fit-
ness, physical activity, and body composition, a study 
of female and/or older participants is needed to confirm 
that the observed decreases in regional adiposity are ap-
plicable to a wider range of individuals. Second, the pres-
ent study had no control group and a small sample size 
because of budget and facility constraints, which weakens 
our findings and may reflect the failure to detect differ-
ences resulting from the HIIT intervention. Moreover, 
assessment of arm fat is needed to observe changes in re-
gional adiposity by HIIT using an arm ergometer in LA-
HIIT.
 In conclusion, regional fat loss near main active muscles 
was observed during a 16-week HIIT program. Changes 
in serum adiponectin concentration were more strongly 
associated with changes in visceral adiposity than with 
changes in aerobic fitness resulting from HIIT. 
Conflict of Interests
 The authors declare no conflict of interests.
greater loss of VAT than of SAT (9% vs. 3%). However, 
the association of decreased thigh SAT with increased 
aerobic fitness strengthens our finding that high-energy 
demand in the thigh during training may cause regional 
fat loss concomitant with muscular hypertrophy, and that 
the regional effect may outweigh the whole-body effect 
on fat loss. The reason for the discrepancy with prefer-
ential loss in truncal adiposity by HIIT3,4,7) is unclear, but 
training frequency of twice a week in the current study 
may not be enough for maintaining exercise-induced 
appetite suppression, one of the possible factors explain-
ing preferential fat loss in HIIT. However, this does not 
mitigate the importance of HIIT on obese people, since 
a recent meta-analysis confirmed that fitness is a main 
determinant of all-cause mortality, regardless of obesity 
status20).
 L-HIIT group had non-significant, but ~10% loss of 
cross-sectional area of thigh and abdominal adipose tis-
sue, whereas LA-HIIT group had minimal change (< 3%). 
Since exercise volume of HIIT using leg ergometer was 
twice in L-HIIT, more regional fat loss in the thigh was 
expected in L-HIIT vs. LA-HIIT, which was not clear 
in the current study. It is possible that the absolute, not 
relative, exercise intensity was less in HIIT using an arm 
ergometer vs leg ergometer, and therefore, total exercise 
volume may have been less in LA-HIIT, resulting in the 
tendency of less fat loss irrespective of the part of the 
body. 
 The association of adiponectin with exercise has re-
cently been more frequently examined12-14) because of 
its action in muscle11,21). Thus far, however, there has 
been no consensus as to the effect of exercise on serum 
adiponectin concentration, which has been reported to 
decrease22,23), increase5,6,24), and remain unchanged7,25) with 
exercise.
 Boyd et al.22) reported a decreased serum adiponectin 
concentration after a 3-week HIIT program; the same was 
shown in another short-term (2 week) study of HIIT23). In 
the present study, serum HMW adiponectin was tempo-
rarily decreased after a 4-week HIIT program, suggesting 
that short-term HIIT may temporarily decrease the level 
of serum adiponectin.
 Several studies have shown significant increases in 
serum adiponectin concentration after high-intensity 
training5,6,24). However, in those studies, increases in adi-
ponectin level were almost always accompanied by loss 
of fat mass5,6,24). Boutcher et al.2) found that changes in 
adiponectin were associated with changes in fat mass 
rather than with changes in fitness gains7), which is also 
consistent with the present study. In fact, it has been well 
known that the circulating adiponectin level is inversely 
associated with obesity, especially visceral fat accumula-
tion26), though adiponectin is an adipokine, and predomi-
nantly synthesized and secreted by adipocytes.
 Recently, Gastebois et al.27) showed in their ancillary 
study, in which body weight was clamped and only the 
Acknowledgments
 The authors would like to thank Yoshinobu Nunokawa, Yasu-
nari Hisashi, Koshi Okabe, Yasutomi Shimada, Yasuko Kamitaki, 
Yujiro Nakamura, and the staff of the radiology department for 
their support with the MRI and DEXA measurements. We also are 
indebted to the staff of the Institute for Integrated Sports Medi-
cine for their assistance with data collection and maintenance of 
training and evaluation equipment. Most importantly, we express 
our appreciation to the research volunteers who participated in 
this study.
Funding
 This work was supported by the Nateglinide Memorial Toyo-
shima Research and Education Fund (2011) and a Grant-in-Aid 
for Young Scientists (B; no. 22700700).
110 JPFSM : Azuma K, et al.
References
 1) Gibala MJ, Little JP, Macdonald MJ and Hawley JA. 2012. 
Physiological adaptations to low-volume, high-intensity in-
terval training in health and disease. J Physiol 590 (Pt 5): 
1077-1084.
 2) Boutcher SH. 2011. High-intensity intermittent exercise and 
fat loss. J Obes 2011: 868305.
 3) Tremblay A, Simoneau JA and Bouchard C. 1994. Impact of 
exercise intensity on body fatness and skeletal muscle me-
tabolism. Metabolism 43: 814-818.
 4) Heydari M, Freund J and Boutcher SH. 2012. The effect of 
high-intensity intermittent exercise on body composition of 
overweight young males. J Obes 2012: 480467.
 5) Moghadasi M, Mohebbi H, Rahmani-Nia F, Hassan-Nia S, 
Noroozi H and Pirooznia N. 2012. High-intensity endurance 
training improves adiponectin mRNA and plasma concentra-
tions. Eur J Appl Physiol 112: 1207-1214.
 6) Racil G, Ben Ounis O, Hammouda O, Kallel A, Zouhal H, 
Chamari K and Amri M. 2013. Effects of high vs. moderate 
exercise intensity during interval training on lipids and adi-
ponectin levels in obese young females. Eur J Appl Physiol 
113: 2531-2540.
 7) Trapp EG, Chisholm DJ, Freund J and Boutcher SH. 2008. 
The effects of high-intensity intermittentexercise training on 
fat loss and fasting insulin levels of young women. Int J Obes 
(Lond) 32: 684-691.
 8) Yamauchi T and Kadowaki T. 2008. Physiological and patho-
physiological roles of adiponectin and adiponectin receptors 
in the integrated regulation of metabolic and cardiovascular 
diseases. Int J Obes (Lond) 32 Suppl 7: S13-S18.
 9) Dai Y, Pang J, Gong H, Fan W and Zhang TM. 2013. Roles 
and tissue source of adiponectin involved in lifestyle modifi-
cations. J Gerontol A Biol Sci Med Sci 68: 117-128.
10) Krause MP, Liu Y, Vu V, Chan L, Xu A, Riddell MC, Swee-
ney G and Hawke TJ. 2008. Adiponectin is expressed by 
skeletal muscle fibers and influences muscle phenotype and 
function. Am J Physiol Cell Physiol 295: C203-C212.
11) Liu Y and Sweeney G. 2014. Adiponectin action in skeletal 
muscle. Best Pract Res Clin Endocrinol Metab 28: 33-41.
12) Bouassida A, Chamari K, Zaouali M, Feki Y, Zbidi A and 
Tabka Z. 2010. Review on leptin and adiponectin responses 
and adaptations to acute and chronic exercise. Br J Sports 
Med 44: 620-630.
13) Lee S and Kwak HB. 2014. Effects of interventions on adipo-
nectin and adiponectin receptors. J Exerc Rehabil 10: 60-68.
14) Passos MC and Goncalves MC. 2014. Regulation of insu-
lin sensitivity by adiponectin and its receptors in response to 
physical exercise. Horm Metab Res 46: 603-608.
15) Osawa Y, Azuma K, Tabata S, Katsukawa F, Ishida H, Oguma 
Y, Kawai T, Itoh H, Okuda S and Matsumoto H. 2014. Effects 
of 16-week high-intensity interval training using upper and 
lower body ergometers on aerobic fitness and morphological 
changes in healthy men: a preliminary study. Open Access J 
Sports Med 5: 257-265.
16) Coker RH, Williams RH, Kortebein PM, Sullivan DH and 
Evans WJ. 2009. Influence of exercise intensity on abdomi-
nal fat and adiponectin in elderly adults. Metab Syndr Relat 
Disord 7: 363-368.
17) Ross R, Dagnone D, Jones PJ, Smith H, Paddags A, Hudson 
R and Janssen I. 2000. Reduction in obesity and related co-
morbid conditions after diet-induced weight loss or exercise-
induced weight loss in men. A randomized, controlled trial. 
Ann Intern Med 133: 92-103.
18) Goodpaster BH, Thaete FL and Kelley DE. 2000. Thigh adi-
pose tissue distribution is associated with insulin resistance 
in obesity and in type 2 diabetes mellitus. Am J Clin Nutr 71: 
885-892.
19) Albu JB, Heilbronn LK, Kelley DE, Smith SR, Azuma K, 
Berk ES, Pi-Sunyer FX and Ravussin E. 2010. Metabolic 
changes following a 1-year diet and exercise intervention in 
patients with type 2 diabetes. Diabetes 59: 627-633.
20) Barry VW, Baruth M, Beets MW, Durstine JL, Liu J and Blair 
SN. 2014. Fitness vs. fatness on all-cause mortality: a meta-
analysis. Prog Cardiovasc Dis 56: 382-390.
21) Iwabu M, Yamauchi T, Okada-Iwabu M, Sato K, Nakagawa 
T, Funata M, Yamaguchi M, Namiki S, Nakayama R, Tabata 
M, Ogata H, Kubota N, Takamoto I, Hayashi YK, Yamauchi 
N, Waki H, Fukayama M, Nishino I, Tokuyama K and Ueki 
K. et al. 2010. Adiponectin and AdipoR1 regulate PGC-1al-
pha and mitochondria by Ca(2+) and AMPK/SIRT1. Nature 
464: 1313-1319.
22) Boyd JC, Simpson CA, Jung ME and Gurd BJ. 2013. Reduc-
ing the intensity and volume of interval training diminishes 
cardiovascular adaptation but not mitochondrial biogenesis 
in overweight/obese men. PLoS One 8: e68091.
23) Leggate M, Carter WG, Evans MJ, Vennard RA, Sribala-Sun-
daram S and Nimmo MA. 2012. Determination of inflamma-
tory and prominent proteomic changes in plasma and adipose 
tissue after high-intensity intermittent training in overweight 
and obese males. J Appl Physiol 112: 1353-1360.
24) Shing CM, Webb JJ, Driller MW, Williams AD and Fell JW. 
2013. Circulating adiponectin concentration and body com-
position are altered in response to high-intensity interval 
training. J Strength Cond Res 27: 2213-2218.
25) Boudou P, Sobngwi E, Mauvais-Jarvis F, Vexiau P and 
Gautier JF. 2003. Absence of exercise-induced variations in 
adiponectin levels despite decreased abdominal adiposity and 
improved insulin sensitivity in type 2 diabetic men. Eur J En-
docrinol 149: 421-424.
26) Kishida K, Funahashi T and Shimomura I. 2014. Adiponectin 
as a routine clinical biomarker. Best Pract Res Clin Endocri-
nol Metab 28: 119-130.
27) Gastebois C, Villars C, Drai J, Canet-Soulas E, Blanc S, Ber-
gouignan A, Lefai E and Simon C. 2016. Effects of train-
ing and detraining on adiponectin plasma concentration and 
muscle sensitivity in lean and overweight men. Eur J Appl 
Physiol 116: 2135-2144.