Scand J Med Sci Sports 2015: ••: ••–•• © 2015 John Wiley & Sons A/S.

doi: 10.1111/sms.12540 Published by John Wiley & Sons Ltd

Short-term low-intensity blood flow restricted interval training

improves both aerobic fitness and muscle strength
M. F. M. de Oliveira1,2,3, F. Caputo2, R. B. Corvino2, B. S. Denadai1,3
1
Physical Effort Laboratory, Sports Center, UFSC, Florianópolis, Brazil, 2Human Performance Research Group, Center for Health
and Exercise Science, UDESC, Florianopolis, Brazil, 3Human Performance Laboratory, UNESP, Rio Claro, Brazil
Corresponding author: Mariana F. M. Oliveira, Human Performance Research Group, Santa Catarina State University,Rua Pascoal
Simone, 358, Coqueiros, Florianópolis, SC CEP 88080-350, Brasil. Tel: (55) 48 3321 8641, Fax: (55) 48 3321 8607, E-mail:
[email protected]
Accepted for publication 19 July 2015

The present study aimed to analyze and compare the strength were measured for all subjects. All training
effects of four different interval-training protocols on groups were able to improve OBLA (BFR, 16%; HIT,
aerobic fitness and muscle strength. Thirty-seven subjects 25%; HIT + BFR, 22%; LOW, 6%), with no difference
(23.8 ± 4 years; 171.7 ± 9.5 cm; 70 ± 11 kg) were assigned between groups. However, VO2max and Pmax improved only
to one of four groups: low-intensity interval training with for BFR (6%, 12%), HIT (9%, 15%) and HIT + BFR
(BFR, n = 10) or without (LOW, n = 7) blood flow restric- (6%, 11%), with no difference between groups. Muscle
tion, high-intensity interval training (HIT, n = 10), and strength gains were only observed after BFR training
combined HIT and BFR (BFR + HIT, n = 10, every (11%). This study demonstrates the advantage of short-
session performed 50% as BFR and 50% as HIT). Before term low-intensity interval BFR training as the single
and after 4 weeks training (3 days a week), the maximal mode of training able to simultaneously improve aerobic
oxygen uptake (VO2max), maximal power output (Pmax), fitness and muscular strength.
onset blood lactate accumulation (OBLA), and muscle

The use of multiple training components is common- aerobic fitness improvements, and between the total BFR
place to address different functional parameters, e.g., training duration with hypertrophy and, perhaps, muscle
neuromuscular performance and aerobic fitness (Garber strength gains. However, none of the aforementioned
et al, 2011). However, to perform strength and endurance studies, or, any other single mode of short-term training
training in the same exercise session, or even on different (up to 4 weeks) have yet demonstrated simultaneous
days, can be both physically demanding and time con- improvements in functional aspects such as aerobic
suming. Therefore, there is interest in identifying effi- fitness and muscle strength.
cient training methods, which can simultaneously The magnitude of BFR seems to influence directly the
improve cardiovascular and neuromuscular performance exercise intensity performed during the aerobic training
but using a single training mode. Combining low- sessions (i.e., %VO2max; Sundberg, 1994; Iida et al.,
intensity aerobic exercise with blood flow restriction 2007). In addition, a severe BFR (i.e., lower O2 supply
(BFR) has been proposed as an adapted single training and disposal of metabolites), which induces fatigue and
method, which brings about multiple benefits including concomitant ability to generate force, could recruit fast-
peripheral (e.g., muscle strength and hypertrophy, twitch motor units and thus promote a favorable milieu
increase in oxidative enzymes, angiogenesis, glycogen for muscle adaptations (strength and hypertrophy) at low
content) and central aspects (e.g., stroke volume; exercise intensities (Moritani et al., 1992; Moore et al.,
Sundberg, 1994; Abe et al., 2006, 2010a,b). While 2004). To date, low-intensity BFR training has been per-
increases in maximal oxygen uptake (VO2max) have only formed restricting the blood flow throughout the total
been observed following aerobic BFR training per- exercise session (Sundberg, 1994; Moore et al., 2004;
formed at or above 40% of VO2max (Sundberg, 1994; Abe Abe et al., 2006, 2010b; Park et al., 2010), which may
et al., 2010b; Park et al., 2010), muscle hypertrophy, have induced fatigue and limited the training stimulus
which has (Abe et al., 2006) and has not (Abe et al., (i.e., tolerable combination of exercise intensity, dura-
2010b) been associated with strength gains, were only tion, and BFR), and could explain some of the divergent
reported for aerobic BFR training lasting 3 weeks or results obtained after “aerobic” BFR training (Sundberg,
more. Taken together, these results seem to indicate an 1994; Abe et al., 2006, 2010a,b; Park et al., 2010).
association between the intensity of BFR training and Based on these suggestions, an intermittent low-intensity

1
Oliveira et al.
“aerobic” BFR training protocol, but with BFR applied would be the first single mode of training able to evoke
only during the exercise (deflating the cuffs during rest simultaneous gains in aerobic fitness parameters and
periods), could provide a greater training stimulus, i.e., a muscle strength. If so, a third purpose was to investigate
higher tolerable training intensity and/or duration and whether a combination of HIT + BFR, i.e., every session
occlusion pressures leading to high order fibers recruit- performed 50% as BFR and 50% as HIT, in random
ment. Additionally, the brief episodes of low and high O2 order, could result in aerobic fitness improvements to a
exposure in skeletal muscle, generated by the cycles of similar magnitude as HIT alone, as well as induce
cuffs inflation and deflating (i.e., ischemia-reperfusion), strength gains by including BFR training.
could also induce additional training-effects on oxidative
metabolism and angiogenesis (Clanton & Klawitter,
2001; Zhao et al., 2011). Therefore, the application of an Methods
Subjects
intermittent compared with continuous BFR may be
advantageous in trying to induce simultaneous gains in Thirty-seven young adults (23.8 ± 4 years; 171.7 ± 9.5 cm;
70 ± 11 years.) volunteered and gave written informed consent to
aerobic fitness and muscle strength.
participate in the study. Following the completion of an initial
We have designed an intermittent BFR training proto- incremental exercise test (see below), the subjects were assigned
col in an attempt to concurrently improve aerobic fitness to one of the four training groups: high-intensity interval training
and muscle strength, and which combines the minimal (HIT, n = 10, seven males and three females), low-intensity inter-
exercise intensity (40% VO2max) that seems to evoke val training with blood flow restriction (BFR, n = 10, eight males
and two females), low-intensity interval training without blood
aerobic adaptations (Sundberg, 1994; Abe et al.,
flow restriction (LOW, n = 7, four males and three females) or
2010a,b; Park et al., 2010) with a severe intermittent combined HIT and BFR (HIT + BFR, n = 10, three males and
BFR (140–200 mmHg in 18 cm cuffs; Loenneke et al., seven females). Physical characteristics of the subjects are pre-
2012), spread over a 4-week training period. Thus, the sented in Table 1. All subjects were healthy with no known
primary purpose of the present study was to verify the musculo-skeletal or cardiorespiratory disease, and none were
taking medications known to affect the cardiorespiratory system.
effects of this low-intensity cycling interval training,
Subjects were all recreationally active, but not currently involved
with BFR applied only during the exercise bout, on in either resistance or endurance training programs and were
parameters of aerobic fitness and maximal isometric instructed to continue normal daily activities and to refrain from
knee extension strength. Since previous aerobic BFR beginning any other training until the completion of the study. All
training studies have not compared the magnitude of participants were informed of the methods, procedures, and risks,
and signed an informed consent document before participating in
aerobic fitness improvements with those induced by dif-
the study. Participants were also instructed to maintain their
ferent types of endurance training, a secondary purpose normal diets over the course of the study. The study was approved
was to compare the magnitude of the training-induced by the university’s ethics committee.
changes between BFR and high-intensity interval train-
ing (HIT). HIT was chosen for this comparison because
it is a short-term low-volume endurance training method, Experimental design
able to promote the greatest central and peripheral adap- The total duration of the present study was 6 weeks. All subjects
tations linked to O2 delivery and utilization (Laursen & (BFR, HIT, BFR + HIT, and LOW group) visited the laboratory on
Jenkins, 2002; Burgomaster et al., 2008). If the magni- two occasions before and after the 4 weeks of training and per-
formed two exercise protocols: (1) anthropometric measures,
tude of the improvement in the aerobic fitness is higher familiarization with the maximal isometric muscle strength test
for HIT compared with BFR training, there would be no and an incremental exercise test to determine measures of aerobic
reasons to prescribe low-intensity BFR training in fitness [onset blood lactate accumulation (OBLA), VO2max, and
healthy and non-injured subjects as a replacement to maximal power output (Pmax)] and (2) isometric muscle strength
HIT. Conversely, if 4 weeks of intermittent BFR training (Fig. 1). All subjects were instructed to arrive at the laboratory in
a fully rested and hydrated state and to avoid strenuous exercise in
performed at ∼40% of VO2max could induce improve- the 48 h preceding a test session. The first experimental trial
ments in parameters of aerobic fitness, to a smaller (pretraining tests) was performed on different days with at least
extent than HIT, but also improve muscle strength, it 2 days separating the incremental exercise test and isometric

Table 1. Subject’s physical characteristics at pre- and post-4 weeks of the high-intensity interval training (HIT), combined training (HIT + BFR), low-
intensity interval training with blood flow restriction (BFR), and low-intensity interval training (LOW)

Age (years) Height (cm) Body mass (kg) Body fat (%)

Pre Post Pre Post

HIT 22 ± 7 173 ± 9 70 ± 9 70 ± 9.3 21 ± 2 21 ± 4

HIT + BFR 24 ± 5 169 ± 8 68 ± 11 70 ± 11 23 ± 6 23 ± 6
BFR 26 ± 5 173 ± 9 68 ± 12 68 ± 13 20 ± 2 20 ± 3
LOW 24 ± 3 170 ± 12 70 ± 12 70 ± 13 21 ± 3 21 ± 2

Values are means ± SD.

2
 Functional gains and short-term training
4-Wks Training
Pre-Training Post - Training
2 Sets of 5-8 reps of 2-min +1-min passive
Between Sets: 3-min at 30%Pmax +
2-min passive rest
Visit 1 Visit 2 Visit 1 Visit 2
Week 4
Week 3 3x/wk
Anthropometry Week 2 3x/wk Anthropometry
Maximal Isometric Maximal Isometric
Strength test Week 1 3x/wk Strength test
Familiarization Familiarization
Maximal isometric
3x/wk
2x7 2x8 Maximal isometric
strength
2 x 5-s MVC
2x5 2x6 reps strength
2 x 5-s MVC
3min rest
reps reps 3min rest
Incremental test reps Incremental test

120
Power output (%Pmax) 110
Peak Torque 100 HIT (n= 10): 110/105/100/95% Pmax
90
Body Mass 80
% Body fat 70
60 HIT + BFR (n=10): Random order
50 1st set HIT or BFR and 2nd set HIT or BFR
VO2max 40
Pmax 30
OBLA 20 LOW (n= 7): 30%Pmax
10 BFR (n=10): 30%Pmax + 140-200mmHg
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Time (min)

Fig. 1. Experimental design.

muscle strength test. The second experimental trial (post-training highest 15 s VO2 value reached during the incremental test. Pmax
tests) was performed at least 4 days after the final exercise training was determined according to the equation: Pmax (W) = power at
session and was identical in all respects to the first experimental last stage completed (W) + [t (s)/step duration (s) × step increment
trial. All tests were conducted at the same time of day for each (W)]; “t” is the time of the uncompleted stage. OBLA was deter-
subject. mined by linear interpolation, and considered to occur at a fixed
lactate concentration of 3.5 mM (Heck et al., 1985).

Materials
The incremental exercise tests and the training were performed on Strength measurements
a cycle ergometer (Lode Excalibur Sport; Lode Medical Technol- The subjects were carefully familiarized with the testing proce-
ogy, Groningen, the Netherlands) and during all tests and training dures of maximal voluntary contraction. Maximal isometric quad-
sessions the cadence was maintained at 70 rpm. Throughout the riceps contractions were performed during static knee extension at
tests, the respiratory and pulmonary gas-exchange variables were a knee joint angle of 60° (0° full extension). After 5 min of
measured using a breath-by-breath gas analyzer (Quark PFTergo, warm-up at 50% Pmax, the subject was seated in the test chair and
Cosmed Srl, Rome, Italy). Before each test, the O2 and CO2 analy- firmly strapped at the hip and distal thigh. Two maximal efforts of
sis systems were calibrated using ambient air and a gas of known 5 s were performed with both legs to determine bilateral maximal
O2 and CO2 concentration according to the manufacturer’s instruc- voluntary contraction (MVC). A 5-min rest period was given
tions, while the turbine flow-meter was calibrated using a 3-L between MVC measurements. All participants were encouraged to
syringe (Quark PFTergo, Cosmed Srl). Heart rate (HR) was also provide a maximal effort by stronger verbal encouragement. The
monitored throughout the tests (Quark PFTergo, Cosmed Srl). MVC was measured by the load cell. Data were collected via a
Breath-by-breath VO2 was averaged every 15 s in the incremental four-channel acquisition system with an analogic digital converter
test for VO2max determination (Data Management Software, (14-bit) with sampling rate set at 2000 Hz (Miootol 400, Miotec
Cosmed). Analyses of the blood lactate concentration ([La]) were Biomedical Equipment LTDA, Porto Alegre, Brazil). The signal
undertaken by an electrochemical method (YSI 1500 Sport, was subsequently converted to newtons and multiplied by the lever
Yellow Springs Instrument, Yellow Springs, Ohio, USA), which arm length to calculate the moment of force (“torque”). All
was calibrated with standards of (5 mmol/L) lactate concentra- recorded moments were corrected for the effect of gravity on the
tions. Muscle strength was assessed by isometric knee extension in lower limb according to procedures described previously.
a test chair (knee extension – Tonus®; São Paulo, Brazil) and were
measured using a load cell (SDS 200 kg) adapted for the test chair.
A high reproducibility has been observed in our laboratory for the Training protocols (see Fig. 1)
isometric actions (intra-class correlation = 0.89).
The principal criteria for the assignment of subjects to one of the
four groups were VO2max. For all groups (BFR, HIT, HIT + BFR,
and LOW) the training program consisted of three exercise ses-
Incremental exercise sions per week on a stationary cycle ergometer for a total duration
The incremental exercise test began at 0.5W/Kg, followed by an of 4 weeks. For all training groups, every training session was
increase of 35 W for male and 25 W for female every 3 min until preceded by 5-min warm-up at 30% of Pmax. The training power
voluntary exhaustion, to provide the maximal power output (Pmax) output was 30% of Pmax for LOW and BFR training groups. Each
and maximal oxygen uptake (VO2max). VO2max was defined as the training session consisted of two sets of five repetitions for the first

3
Oliveira et al.
Table 2. Description of training

Parameters HIT HIT + BFR BFR LOW

Exercise protocol 2 sets (5–8 repeats × 120-s, 60-s rest between repetitions)
Total session time (s) 1200–1920
Exercise intensity ∼102%PMAX ∼66%PMAX 30% PMAX
(∼236 W) (∼138 W)* (∼66 W)*‡
Perceived exertion 7.6 ± 1.4 5.1 ± 1.4 7.3 ± 1.4 1.3 ± 0.5*†‡
Peak [La] (mM) 10.1 ± 2† 9.1 ± 1.9† 5.3 ± 1.7*‡ 1.6 ± 0.4*†‡
Peak HR (%HRmax) 96 ± 2.9†‡ 84 ± 8.6*† 70 ± 9.2*‡ 60 ± 5.1*†‡
Peak VO2 (%VO2max) 98 ± 6† 70 ± 6.5*† 42 ± 4.7*‡ 45 ± 2.4*‡
Session volume (kJ) 284–454 165–264 83–133
Training volume (kJ) 4424 2583 1300

Perceived exertion, [La], Peak HR and Peak VO2 represent the mean of averaging exercise values obtained during the first and 12th training session. Rating
of perceived exertion measured by 0 to 10 scale.
*Significantly different from HIT.
†
Significantly different from BFR.
‡
Significantly different from HIT + BFR (P < 0.05).
BFR, low-intensity interval training with blood flow restriction; HIT, high-intensity interval training; HR, heart rate; LOW, low-intensity interval training
without blood flow restriction; VO2max, maximal oxygen uptake.

three sessions, after which one repetition per set was added each A paired Student t-test was used to compare pre- to post-training in
week. Therefore, in the fourth training week, the session consisted each group. When a significant time by group interaction was
of two sets of eight repetitions. Each repetition lasted 2 min, inter- found, one-way ANOVAs were performed to locate the difference
spersed by 1-min passive rest. The rest interval between sets was followed by Bonferroni post-hoc tests. The threshold values for
5 min (3-min active recovery at 30% Pmax followed by 2-min Cohen’s effect size (i.e., post-training – pretraining standardized
passive rest). difference) were assessed using the method proposed by Hopkins
The BFR group wore pressure cuff belts (18 cm wide, Missouri, et al. (2009); (Yasuda et al., 2011). The magnitude was considered
São Paulo, Brazil) on the proximal portion of both thighs during < 0.2 trivial; 0.2–0.6 small; 0.6–1.2 moderate; 1.2–2 large; 2–4
all training sessions. In the first week, cuff belts were inflated to very large. Statistical significance was declared when P < 0.05.
140 mmHg during the 2-min repetitions and deflated during the
1-min rest periods. The pressure was progressed (Abe et al., 2006;
Park et al., 2010) by 20 mmHg after three completed sessions,
thus, in the last week, the pressure applied was 200 mmHg. Results
In the HIT group, the subjects completed a variable power
output training protocol. Each repetition began at 110% Pmax with Physical characteristics were similar among the groups
a progressively 5% decrease in the intensity every 30 s (110%, (Table 1). No significant differences were apparent in
105%, 100%, and 95% Pmax, respectively). This training protocol body mass, OBLA, VO2max, Pmax, or strength measured
was designed to increase the average power output of the training,
pretraining, allowing us to compare the training effects
as fast-start protocols have shown faster VO2p kinetics and higher
exercise tolerance compared with constant work rate exercise between and within training groups. Body mass and
(Turnes et al., 2014). body fat remained unchanged after all training programs.
For BFR + HIT one set was performed as BFR and the other as A descriptive summary of the acute response to the
HIT. The order of the sets was alternated at every session and the training modes used in this study are listed in Table 2.
total exercise time was the same for all training protocols.
Despite an intentional similar total exercise session dura-
Cardiorespiratory variables were measured throughout the first
and the last training session for all groups. Ratings of perceived tion for the four groups, the work session volume and
exertion were reported after each repetition using the 0–10 total work training volume in HIT was ∼340% higher
category-ratio scale and capillary blood samples (25 μL) were than BFR and LOW and was ∼171% higher than
taken from the earlobe immediately after the end of each set for HIT + BFR. The perceived exertion was similar between
analysis of blood lactate concentration ([La]) in the first and the
HIT, HIT + BFR and BFR, but all were higher than
last training session.
LOW. Peak [La] was similar between HIT and
HIT + BFR and both were higher than BFR and LOW,
Statistics however BFR values were higher compared with LOW.
Data are presented as mean ± standard deviation (SD). For each The HR responses were significantly different between
set of data, normal distribution (Shapiro–Wilks test) and homoge- the four groups. VO2 during the training session was
neity of variance were checked. Comparisons between parameters similar between BFR and LOW, both were lower than
obtained during training sessions were made by one-way repeated- HIT + BFR and HIT; however, VO2 during HIT was
measures analysis of variance (ANOVA). A 4 × 2 two-way facto- higher than HIT + BFR.
rial analysis of variance (group × time), with repeated measures
for the training factor (pre- and post-test) was initially performed. Parameters of aerobic fitness and isometric knee
In order to interpret statistical significance of time and extension strength at pre- and post-training periods are
group × time interactions, the follow-up analyses were performed. shown in Table 3.

4
 Functional gains and short-term training
Table 3. Aerobic parameters and isometric knee extension strength before and after 4 weeks of high-intensity interval training (HIT), combined training
(HIT + BFR), low-intensity interval training with blood flow restriction (BFR), low-intensity interval training without blood flow restriction (LOW)

Parameter Group

HIT HIT + BFR BFR LOW

VO2max Pre 45.1 ± 4.2 44.3 ± 8.3 47.6 ± 7.8 44.6 ± 4.0
(mL/kg/min) Post 49.2 ± 4.9* 47.2 ± 9.0* 50.2 ± 7.7* 44.8 ± 4.7
Pmax Pre 3.26 ± 0.36 3.06 ± 0.79 3.34 ± 0.85 3.18 ± 0.40
(W/kg) Post 3.73 ± 0.29* 3.38 ± 0.81* 3.72 ± 0.90* 3.23 ± 0.36
OBLA Pre 2.02 ± 0.25 2.06 ± 0.55 2.19 ± 0.86 2.04 ± 0.33
(W/kg) Post 2.52 ± 0.30* 2.49 ± 0.67* 2.57 ± 0.83* 2.17 ± 0.27*
Strength Pre 450 ± 104 399 ± 112 452 ± 97 387 ± 102
(N/m) Post 447 ± 83 385 ± 100† 497 ± 89* 377 ± 91

Values are means ± SD.

*Significantly different from pretraining (P < 0.05).
†
Significantly different from BFR at post (P < 0.05).
BFR, low-intensity interval training with blood flow restriction; HIT, high-intensity interval training; LOW, low-intensity interval training without blood flow
restriction; OBLA, onset blood lactate accumulation; Pmax, maximal power output; VO2max, maximal oxygen uptake.

OBLA no difference was demonstrated for the other groups

There was no significant two-way interaction (group × (HIT, −0.7 ± 9.9%, P = 0.88; HIT + BFR, −3.5 ± 6.8%,
time, F = 2.58, P = 0.072, statistical power = 0.58) and P = 0.32; LOW, −2.6 ± 6.7%, P = 0.82).
no main effect for group (F = 0.32, P = 0.8), but there
was a significant main effect for time (F = 59.8;
Discussion
P < 0.001). OBLA increased from pre- to post-training
periods for all groups (BFR = 16 ± 13%, ES = 0.43; The present study aimed to analyze and compare the
HIT = 25 ± 13%, ES = 1.8; HIT + BFR = 22 ± 12%, effects of four different interval-training protocols on
ES = 0.7; LOW = 6 ± 4%). aerobic fitness and muscle strength. The main and origi-
nal finding of the present study was that 4 weeks (12
sessions) of low-intensity interval BFR training resulted
VO2max
in a significant improvement of all selected variables
There was no significant two-way interaction (group × (VO2max, Pmax, OBLA, and muscle strength). However,
time, F = 3.49, P = 0.09, statistical power = 0.54) and no 4 weeks of HIT and BFR + HIT training only induced
main effect for group (F = 0.76, P = 0.5), but there was a improvements for aerobic variables, although HIT pro-
significant main effect for time (F = 2.4; P < 0.001). vided a higher effect size compared with BFR and
VO2max increased from pre- to post-training periods BFR + HIT training. On the other hand, the low-
for BFR (5.6 ± 4.2%, P = 0.006, ES = 0.33), HIT intensity interval training on its own without occlusion
(9.2 ± 6.5%, P = 0.002, ES = 0.9), and HIT + BFR (LOW) was not sufficient to improve VO2max, Pmax, or
(6.5 ± 5.5%, P = 0.03, ES = 0.33), but was unchanged in muscular strength, but was able to improve OBLA.
LOW (0.4 ± 4.7%, P = 0.75). Therefore, we have now demonstrated the advantage of
short-term low-intensity interval BFR training compared
with other training modes studied because only BFR
Pmax training was able to concurrently improve aerobic
There was a significant two-way interaction (group × parameters and muscular strength. Furthermore, the time
time, F = 9.61, P < 0.001, statistical power = 0.99). Like exposed to an increased overload induced by BFR,
VO2max, Pmax increased from pre- to post-training periods seemed to determine the adaptive responses associated
for BFR (11.7 ± 4.7%, P < 0.001, ES = 0.44), HIT with muscle strength gains, since BFR + HIT was not
(15.0 ± 4.5%, P < 0.001, ES = 1.5), and HIT + BFR sufficient to induce an increase in muscle strength. From
(10.9 ± 4.5%, P < 0.001, ES = 0.39), but was unchanged a practical point of view, BFR training provided a sig-
in LOW (1.6 ± 3.9%, P = 0.34). nificant functional improvement for our active subjects
by using a shorter exercise training session (30–45 min,
including warm-up) and a low work rate (∼66 W or 40%
Isometric strength VO2max). The improvements were similar in magnitude to
There was a significant two-way interaction (group × those which have been shown by combining two differ-
time, F = 5.18, P = 0.005, statistical power = 0.89). In ent modes of training (∼1 h session), i.e., concurrent
this case, t-test showed a significant increase only for the endurance and strength training (Mikkola et al., 2012; de
BFR group (11.4 ± 7.3%, P < 0.001, ES = 0.66), while Souza et al., 2013).

5
Oliveira et al.
The present results show that only endurance interval findings of McCarthy et al. (1995) who reported that
training with BFR improved maximum isometric knee cycle exercise training alone did not significantly change
extension strength by 11%. It has long been known that isometric strength. On the other hand, contrary to one of
muscle cross-sectional area (CSA), together with the our hypotheses, isometric muscle strength gains were
ability to activate maximal numbers of motor neurons, is not observed after HIT + BFR training. The reason for
the major determinant of maximal contractile force in the lack of strength gains for this group, which also
human skeletal muscle (Komi, 1986). Unfortunately, we performed BFR training, is unclear, but might be related
were unable to measure these determinants of contractile to the differences in the gender distribution and/or the
force in this study. Another limitation was the lack of volume of BFR training. While there are differences in
dynamic strength measurements, which could provide both the acute and chronic responses between males and
further information concerning the nature of the strength females to certain modes of exercise such as sprint
adaptations after BFR training. However, aerobic train- cycling (Esbjornsson Liljedahl et al., 1996; Billaut &
ing with BFR led to similar gains (∼10%) in both Bishop, 2009; Esbjornsson et al., 2012), females gener-
maximal isometric and dynamic strength (Abe et al., ally respond quite similarly to males following strength
2006, 2010b). These data indicate that for cycling train- training and endurance training (Abe et al., 2000; Carter
ing with BFR, the strength adaptations seems to be less et al., 2001; Kraemer et al., 2004; Astorino et al., 2011).
dependent of training specificity compared with resis- With regard to BFR training, there is currently a paucity
tance BFR training, which has presented the higher gains of research on strength and muscle adaptations in males
for dynamic strength measurements (Moore et al., 2004; vs females. Thus, it cannot be ruled out that some of the
Yasuda et al., 2011; Farup et al., 2015). Thus, it is pos- differences between the BFR and HIT + BFR groups
sible that the strength adaptations induced by the inter- could have been in part due to gender-specific responses
mittent low-intensity BFR training used in the present to the various exercise stimuli. It has been postulated that
study may also have evoked strength gains throughout alternating exercise modes during concurrent training
the range of motion. reduces the capacity for the simultaneous acquisition of
Increases in the muscle CSA have also been seen after hypertrophy and/or mitochondrial training-induced
aerobic training with BFR in active (Abe et al., 2006, adaptation responses, compared with single-mode train-
2010a) and elderly subjects (Abe et al., 2010b). Despite ing (Hawley, 2009). Indeed, Chtara et al. (2008) showed
of the lack of assessment of muscle CSA in the present that HIT compromised strength adaptations when con-
study, these previous results suggest that the strength current training was performed in the same session, irre-
improvements observed in the present study could in part spective of the sequence of training modes. However,
be explained by greater muscle CSA. In addition, previ- recent findings have challenge this general belief
ous studies have reported greater muscle activation showing that intense aerobic exercise can be executed
during low-intensity BFR resistance training (Moritani prior to resistance exercise without compromising per-
et al., 1992; Takarada et al., 2000). Since the availability formance outcome and potentiate the hypertrophic
of oxygen was severely reduced during our BFR proto- stimulus to short-term resistance training (Fernandez-
col (Crenshaw et al., 1988; Loenneke et al., 2012), a Gonzalo et al., 2013; Lundberg et al., 2013). Further-
progressive recruitment of additional motor units (type II more, it is important to note that, contrary to our study,
muscle fiber) may have taken place to compensate the no differences in the volume (total work performed)
deficit in force development (Moritani et al., 1992). This during the resistance training were observed for Chtara
is an important aspect because type II fibers have a larger et al. (2008), while our HIT + BFR group performed
capacity for hypertrophy than that of type I fibers. Thus, 50% of the BFR training volume (which supposedly
it could be suggested that, in spite of the low level of could induce strength gain). It was also demonstrated
force generation, BFR could evoke the activation of a that the magnitude of the muscle mass and strength gains
sufficient number of fast-twitch fibers, which would be following once-daily walking was approximately 50%
one of the requirements for gaining muscular size and that reported for twice-daily walking BFR training over
strength. However, it is unlikely that such increased a 3-week period (Abe et al., 2009). Because the training
recruitment was sufficient to induce neuromuscular groups in the present study were matched by a similar
adaptations in our current study because the motor unit training duration, the HIT + BFR group performed 50%
activation did not change following the low-load resis- of the BFR training. This lower BFR training volume
tance training (< 50% of 1 RM) with BFR (Moore et al., seems to explain the lack of strength gains after
2004; Kubo et al., 2006). Further studies addressing the HIT + BFR, since HIT has apparently minimal or
rate of force development, dynamic maximal strength, null effect on the hypertrophy and strength gains
EMG, and muscle CSA analyses are needed to investi- (Fernandez-Gonzalo et al., 2013; Lundberg et al., 2013).
gate the nature of the strength adaptations following While total training volume of HIT was substantially
intermittent aerobic BFR training. higher than HIT + BFR and BFR (4424 vs 2583 kJ vs
As expected, both the HIT and the LOW training 1300 kJ, respectively), the improvements on VO2max,
groups showed no increase in strength, confirming the Pmax, and OBLA were not statistically different between

6
 Functional gains and short-term training
these training groups. Thus, the hypothesis that the mag- tion capacity in humans (as an index of capillarity),
nitude of training-induced aerobic enhancement is while Suzuki et al. (2000) showed evidence of aug-
higher in HIT compared with BFR training would be mented capillarization in rats. Therefore, our intermit-
rejected. However, the statistical power for the two-way tent BFR training, similar to HIT, seems to have been
factorial analysis of variance for VO2max and OBLA did able to promote adaptive responses responsible for
not reach 0.60. Furthermore, the effect size was always increasing the oxidative ATP production.
higher in HIT compared with BFR and HIT + BFR for A novel aspect of this study was the application of
all aerobic indexes, suggesting possibly larger training low-intensity interval training with blood flow restriction
effects on aerobic fitness following short-term high- applied only during the exercise, i.e., deflating the cuffs
intensity training. Anyway, other studies have also during passive rest periods. Aside from enabling longer
shown that either short-term moderate or high-intensity exercise durations and a higher occlusion pressure (up to
aerobic training improved OBLA and VO2max to the same 16 total reps and 200 mmHg, respectively), the ischemia
magnitude (Berger et al., 2006; Burgomaster et al., and reperfusion per se or the intermittent hypoxia could
2008; Zoladz et al., 2013). This is the first study to dem- have also caused additional effects on some adaptive
onstrate increases in OBLA and Pmax after training with responses. Zhao et al. (2011) showed that physiologic
BFR, although improvements in VO2max have been ischemic-reperfusion training (3 min of cuff inflation
already reported (Sundberg, 1994; Abe et al., 2010a; followed by 5 min of deflation repeated three times) in
Park et al., 2010). Compared with low-volume sprint rabbits promoted angiogenesis, increase distribution of
interval training (225 kJ per week; Burgomaster et al., type I fibers, and endurance performance in the patho-
2008), high-volume moderate intensity training logic ischemic skeletal muscle. The increased fluid shear
(∼2250 kJ; Berger et al., 2006; Burgomaster et al., 2008) stress by ischemia and reperfusion could have to lead to
and HIT with different volumes (Berger et al., 2006; endothelial activation and up-regulation of expression of
Denadai et al., 2006), our findings showed that low- angiogenesis-related factors such as vascular endothelial
intensity interval training with BFR (250 kJ per week) growth factor and endothelial NO synthase (Hudlicka &
were also able to provide similar increases in aerobic Brown, 2009). In addition, reactive oxygen species
parameters. (ROS) formation is likely amplified by conditions of
The possible mechanisms involved in enhanced repeated hypoxia-reoxygenation exposure (Zuo &
aerobic parameters throughout different exercise training Clanton, 2005). Previous results have supported the
programs are complex and likely arise from an integra- notion that contraction-induced ROS are critical signal-
tion of both enhanced metabolic control and O2 delivery ing molecules for exercise-induced adaptations in vivo
(Daussin et al., 2008). It is well known that HIT pro- and that peroxisome proliferator-activated receptor
motes rapid adaptations in VO2max, work capacity, gamma coativators (PGC-1α) is under redox control
muscle oxidative capacity, and vascular function (Ristow et al., 2009). Thus, the episodes of ischemia and
(Denadai et al., 2006; Jacobs et al., 2013). Reducing reperfusion during intermittent BFR training could have
blood flow to exercising limbs has also been shown to also triggered important cellular messengers in signal
produce beneficial adaptive responses. Supine one- transduction involved in activation of transcription
legged cycling training with 50 mmHg chamber pressure factors signaling molecules for mitochondrial biogenesis
(reduced leg blood flow by 16%) for 4 weeks (four and angiogenesis. However, direct approaches to
sessions/week) resulted in an increase in muscle oxida- confirm whether the episodes of ischemia and
tive enzyme activity and capillary density (Esbjornsson reperfusion induced by cuff inflation/deflation during
et al., 1993). It has been shown that the application of BFR training could provide additional aerobic adaptive
external pressure on the working muscle itself reduces responses are warranted.
blood flow with a dose-dependent response (Sundberg,
1994; Iida et al., 2007). Unfortunately, we are not able to Perspectives
quantify the magnitude of blood flow reduction in the
present study. However, it is likely that the external pres- The present study has now demonstrated the advantage
sure (140–200 mmHg) applied in our study, using a of short-term low-intensity interval BFR training com-
18-cm wide cuff (Crenshaw et al., 1988; Loenneke et al., pared with other single conditioning modes because only
2012), was sufficient to provide a stimulus capable of BFR training was able to concurrently improve aerobic
improving the skeletal muscle respiratory capacity and parameters and muscular strength. No significant
promote angiogenesis. Indeed, cycling (Abe et al., changes in VO2max and Pmax were observed for the LOW
2010a) and walking (Park et al., 2010) low-intensity group, who performed identical training to the BFR
continuous training with BFR (∼190 mmHg in 5 cm group, but without cuff restriction of blood flow.
cuffs) have been able to enhance aerobic metabolism, Therefore, the increased metabolic and physiologic
reflected by an increase in VO2max. Although using resis- strains induced by blood flow restriction seem to have
tance training with blood flow occlusion, Evans et al. been responsible to trigger the adaptive responses linked
(2010) demonstrated an enhanced microvascular filtra- to increased aerobic power and muscle strength.

7
Oliveira et al.
Additionally, BFR training volume seems to be impor- Key words: Short-term interval training, blood flow
tant to determine the adaptive responses associated with restriction, high-intensity exercise, cycling, VO2max, iso-
muscle strength gains, since the 50% BFR training metric knee extension torque.
volume performed by BFR + HIT was not sufficient to
induce increases in muscle strength. Future research is
needed to examine the mechanisms underlying the con- Acknowledgements
current adaptation evoked by low-intensity intermittent
BFR training. These findings support the additional fea- We thank the subjects for participating in this study, and CNPq and
sibility of intermittent BFR training method for rehabili- FAPESC for financial support. Mariana Oliveira was supported by
a CAPES PhD fellowship. We are grateful to Ms. Anna Wittekind
tation in injured athletes or other groups where high for reviewing the paper.
mechanical loading is contraindicated or impractical.

References
Abe T, DeHoyos DV, Pollock ML, McGee SL, Gibala MJ. Similar Esbjornsson Liljedahl M, Holm I, Sylven
Garzarella L. Time course for strength metabolic adaptations during exercise C, Jansson E. Different responses of
and muscle thickness changes after low volume sprint interval and skeletal muscle following sprint
following upper and lower body traditional endurance training in training in men and women. Eur J Appl
resistance training in men and women. humans. J Physiol 2008: 586: 151–160. Physiol 1996: 74: 375–383.
Eur J Appl Physiol 2000: 81: 174–180. Carter SL, Rennie CD, Hamilton SJ, Esbjornsson M, Jansson E, Sundberg CJ,
Abe T, Fujita S, Nakajima T, Sakamaki Tarnopolsky. Changes in skeletal Sylven C, Eiken O, Nygren A, Kaijser
M, Ozaki H, Ogasawara R, Sugaya M, muscle in males and females following L. Muscle-fiber types and
Kudo M, Kurano M, Yasuda T, Sato Y, endurance training. Can J Physiol enzyme-activities after training with
Ohshima H, Mukai C, Ishii N. Effects Pharmacol 2001: 79: 386–392. local leg ischemia in man. Acta Physiol
of low-intensity cycle training with Chtara M, Chaouachi A, Levin GT, Scand 1993: 148: 233–241.
restricted leg blood flow on thigh Chaouachi M, Chamari K, Amri M, Esbjornsson M, Rundqvist HC, Mascher
muscle volume and VO2max in young Laursen PB. Effect of concurrent H, Osterlund T, Rooyackers O,
men. J Sport Sci Med 2010a: 9: endurance and circuit resistance Blomstrand E, Jansson E. Sprint
452–458. training sequence on muscular strength exercise enhances skeletal muscle
Abe T, Kearns CF, Fujita S, Sakamaki M, and power development. J Strength p70S6k phosphorylation and more so in
Sato Y, Brechue WF. Skeletal muscle Cond Res 2008: 22: 1037–1045. women than in men. Acta Physiol
size and strength are increased Clanton TL, Klawitter PF. Physiological (Oxf) 2012: 205: 411–422.
following walk training with restricted and genomic consequences of Evans C, Vance S, Brown M. Short-term
leg muscle blood flow: implications for intermittent hypoxia – invited review: resistance training with blood flow
training duration and frequency. Int J adaptive responses of skeletal muscle to restriction enhances microvascular
KAATSU Train Res 2009: 5: 9–15. intermittent hypoxia: the known and the filtration capacity of human calf
Abe T, Kearns CF, Sato Y. Muscle size unknown. J Appl Physiol 2001: 90: muscles. J Sport Sci 2010: 28: 999–1007.
and strength are increased following 2476–2487. Farup J, de Paoli F, Bjerg K, Riis S,
walk training with restricted venous Crenshaw AG, Hargens AR, Gershuni Ringgard S, Vissing K. Blood flow
blood flow from the leg muscle, DH, Rydevik B. Wide tourniquet cuffs restricted and traditional resistance
Kaatsu-walk training. J Appl Physiol more effective at lower inflation training performed to fatigue produce
(1985) 2006: 100: 1460–1466. pressures. Acta Orthop Scand 1988: 59: equal muscle hypertrophy. Scand J
Abe T, Sakamaki M, Fujita S, Ozaki H, 447–451. Med Sci Sports 2015: doi:
Sugaya M, Sato Y, Nakajima T. Effects de Souza EO, Tricoli V, Roschel H, Brum 10.1111/sms.12396.
of low-intensity walk training with PC, Bacurau AVN, Ferreira JCB, Aoki Fernandez-Gonzalo R, Lundberg TR,
restricted leg blood flow on muscle MS, Neves M, Aihara AY, Fernandes Tesch PA. Acute molecular responses
strength and aerobic capacity in older ADC, Ugrinowitsch C. Molecular in untrained and trained muscle
adults. J Geriatr Phys Ther 2010b: 33: adaptations to concurrent training. Int J subjected to aerobic and resistance
34–40. Sports Med 2013: 34: 207–213. exercise training versus resistance
Astorino TA, Allen RP, Roberson DW, Daussin FN, Zoll J, Dufour SP, Ponsot E, training alone. Acta Physiol 2013: 209:
Jurancich M, Lewis R, McCarthy K, Lonsdorfer-Wolf E, Doutreleau S, 283–294.
Trost E. Adaptations to high-intensity Mettauer B, Piquard F, Geny B, Garber CE, Blissmer B, Deschenes MR,
training are independent of gender. Eur Richard R. Effect of interval versus Franklin BA, Lamonte MJ, Lee IM,
J Appl Physiol 2011: 111: 1279–1286. continuous training on cardiorespiratory Nieman DC, Swain DP, American
Berger NJA, Tolfrey K, Williams AG, and mitochondrial functions: College of Sports M. American College
Jones AM. Influence of continuous and relationship to aerobic performance of Sports Medicine position stand.
interval training on oxygen uptake improvements in sedentary subjects. Quantity and quality of exercise for
on-kinetics. Med Sci Sports Exerc Am J Physiol Regul Integr Comp developing and maintaining
2006: 38: 504–512. Physiol 2008: 295: R264–R272. cardiorespiratory, musculoskeletal, and
Billaut F, Bishop D. Muscle fatigue in Denadai BS, Ortiz MJ, Greco CC, de neuromotor fitness in apparently
males and females during Mello MT. Interval training at 95% and healthy adults: guidance for prescribing
multiple-sprint exercise. Sports Med 100% of the velocity at VO2max: effects exercise. Med Sci Sports Exerc 2011:
2009: 39: 257–278. on aerobic physiological indexes and 43: 1334–1359.
Burgomaster KA, Howarth KR, Phillips running performance. Appl Physiol Hawley JA. Molecular responses to
SM, Rakobowchuk M, MacDonald MJ, Nutr Metab 2006: 31: 737–743. strength and endurance training: are

8
 Functional gains and short-term training
they incompatible? Appl Physiol Nutr Laursen PB, Jenkins DG. The scientific prevent health-promoting effects of
Metab 2009: 34: 355–361. basis for high-intensity interval training physical exercise in humans. Proc Natl
Heck H, Mader A, Hess G, Mucke S, – optimising training programmes and Acad Sci U S A 2009: 106: 8665–8670.
Muller R, Hollmann W. Justification of maximising performance in highly Sundberg CJ. Exercise and training during
the 4-mmol/I lactate threshold. Int J trained endurance athletes. Sports Med graded leg ischemia in healthy man
Sports Med 1985: 6: 117–130. 2002: 32: 53–73. with special reference to effects on
Hopkins WG, Marshall SW, Batterham Loenneke JP, Fahs CA, Rossow LM, skeletal-muscle. Acta Physiol Scand
AM, Hanin J. Progressive statistics for Sherk VD, Thiebaud RS, Abe T, 1994: 150: 2–50.
studies in sports medicine and exercise Bemben DA, Bemben MG. Effects of Suzuki J, Kobayashi T, Uruma T,
science. Med Sci Sports Exerc 2009: cuff width on arterial occlusion: Koyama T. Strength training with
41: 3–13. implications for blood flow restricted partial ischaemia stimulates
Hudlicka O, Brown MD. Adaptation of exercise. Eur J Appl Physiol 2012: 112: microvascular remodelling in rat calf
skeletal muscle microvasculature to 2903–2912. muscles. Eur J Appl Physiol 2000: 82:
increased or decreased blood flow: role Lundberg TR, Fernandez-Gonzalo R, 215–222.
of shear stress, nitric oxide and Gustafsson T, Tesch PA. Aerobic Takarada Y, Takazawa H, Sato Y,
vascular endothelial growth factor. exercise does not compromise muscle Takebayashi S, Tanaka Y, Ishii N.
J Vasc Res 2009: 46: 504–512. hypertrophy response to short-term Effects of resistance exercise combined
Iida H, Kurano M, Takano H, Kubota N, resistance training. J Appl Physiol with moderate vascular occlusion on
Morita T, Meguro K, Sato Y, Abe T, 2013: 114: 81–89. muscular function in humans. J Appl
Yamazaki Y, Uno K, Takenaka K, McCarthy JP, Agre JC, Graf BK, Pozniak Physiol 2000: 88: 2097–2106.
Hirose K, Nakajima T. Hemodynamic MA, Vailas AC. Compatibility of Turnes T, Salvador AF, Lisboa FD, de
and neurohumoral responses to the adaptive responses with combining Aguiar RA, Cruz RS, Caputo F. A
restriction of femoral blood flow by strength and endurance training. Med fast-start pacing strategy speeds
KAATSU in healthy subjects. Eur J Sci Sports Exerc 1995: 27: 429–436. pulmonary oxygen uptake kinetics and
Appl Physiol 2007: 100: 275–285. Mikkola J, Rusko H, Izquierdo M, improves supramaximal running
Jacobs RA, Fluck D, Bonne TC, Burgi S, Gorostiaga EM, Hakkinen K. performance. PLoS ONE 2014: 9:
Christensen PM, Toigo M, Lundby C. Neuromuscular and cardiovascular e111621.
Improvements in exercise performance adaptations during concurrent strength Yasuda T, Ogasawara R, Sakamaki M,
with high-intensity interval training and endurance training in untrained Ozaki H, Sato Y, Abe T. Combined
coincide with an increase in skeletal men. Int J Sports Med 2012: 33: effects of low-intensity blood flow
muscle mitochondrial content and 702–710. restriction training and high-intensity
function. J Appl Physiol 2013: 115: Moore DR, Burgomaster KA, Schofield resistance training on muscle strength
785–793. LM, Gibala MJ, Sale DG, Phillips SM. and size. Eur J Appl Physiol 2011: 111:
Komi PV. Training of muscle strength Neuromuscular adaptations in human 2525–2533.
and power – interaction of muscle following low intensity Zhao Y, Li JN, Lin AC, Xiao MY, Xiao
neuromotoric, hypertrophic, and resistance training with vascular B, Wan CX. Improving angiogenesis
mechanical factors. Int J Sports Med occlusion. Eur J Appl Physiol 2004: 92: and muscle performance in the
1986: 7: 10–15. 399–406. ischemic limb model by physiological
Kraemer WJ, Nindl BC, Ratamess NA, Moritani T, Sherman WM, Shibata M, ischemic training in rabbits. Am J Phys
Gotshalk LA, Volek JS, Fleck SJ, Matsumoto T, Shinohara M. Oxygen Med Rehabil 2011: 90: 1020–1029.
Newton RU, Hakkinen K. Changes in availability and motor unit-activity in Zoladz JA, Grassi B, Majerczak J,
muscle hypertrophy in women with humans. Eur J Appl Physiol 1992: 64: Szkutnik Z, Korostynski M, Karasinski
periodized resistance training. Med Sci 552–556. J, Kilarski W, Korzeniewski B.
Sports Exerc 2004: 36: 697–708. Park S, Kim JK, Choi HM, Kim HG, Training-induced acceleration of O2
Kubo K, Komuro T, Ishiguro N, Beekley MD, Nho H. Increase in uptake on-kinetics precedes muscle
Tsunoda N, Sato Y, Ishii N, maximal oxygen uptake following mitochondrial biogenesis in humans.
Kanehisa H, Fukunaga T. Effects of 2-week walk training with blood flow Exp Physiol 2013: 98: 883–898.
low-load resistance training with occlusion in athletes. Eur J Appl Zuo L, Clanton TL. Reactive oxygen
vascular occlusion on the mechanical Physiol 2010: 109: 591–600. species formation in the transition to
properties of muscle and tendon. Ristow M, Zarse K, Oberbach A, Kloting hypoxia in skeletal muscle. Am J
J Appl Biomech 2006: 22: N, Birringer M, Kiehntopf M, Stumvoll Physiol Cell Physiol 2005: 289:
112–119. M, Kahn CR, Bluher M. Antioxidants C207–C216.