Eur J Appl Physiol

DOI 10.1007/s00421-017-3636-7

ORIGINAL ARTICLE

Acute effects of high‑intensity interval training

and moderate‑intensity continuous training sessions
on cardiorespiratory parameters in healthy young men
Gustavo Zaccaria Schaun1 · Cristine Lima Alberton1 · Diego Oliveira Ribeiro1 ·
Stephanie Santana Pinto1 

Received: 28 November 2016 / Accepted: 5 May 2017

© Springer-Verlag Berlin Heidelberg 2017

Abstract  compared to CONT (55.99 ± 10.20; 13.43 ± 10.45 kcal,

Purpose  The aim of the present study was to compare the respectively).
energy expenditure (EE) during and after two treadmill Conclusion These data suggest that supramaximal HIIT
protocols, high-intensity interval training (HIIT) and mod- has a higher impact on EE and EPOC in the early phase of
erate continuous training (CONT), in young adult men. recovery when compared to CONT.
Methods The sample was comprised by 26 physically
active men aged between 18 and 35 years engaged in aero- Keywords  Exercise · Interval training · Continuous
bic training programs. They were divided into two groups: training · Oxygen uptake · Energy expenditure · EPOC
HIIT (n = 14) which performed eight 20 s bouts at 130%
of the velocity associated with the maximal oxygen con- Abbreviations
sumption on a treadmill with 10 s of passive rest, or CONT ANOVA Analysis of variance
(n = 12) which performed 30 min running on a treadmill at CE Caloric equivalent
a submaximal velocity equivalent to 90–95% of the heart CONT Moderate-intensity continuous training
rate associated with the anaerobic threshold. Data related EE Energy expenditure
to oxygen consumption (V̇ O2) and EE were measured dur- EPOC Excess post-exercise oxygen consumption
ing the protocols and the excess post-exercise oxygen con- HIIT High-intensity interval training
sumption (EPOC) was calculated for both sessions. HR Heart rate
Results  No difference was found between groups HRmax Maximal heart rate
for mean V̇ O2 (HIIT: 2.84 ± 0.46 L min−1; CONT: iV̇ O2max Intensity associated with maximal oxygen
2.72  ± 0.43 L min−1) and EE per minute (HIIT: consumption
14.36 ± 2.34 kcal min−1; CONT: 13.21 ± 2.08 kcal min−1) V̇ O2 Maximal oxygen consumption
during protocols. Regarding total EE during session, VT2 Second ventilatory threshold
CONT resulted in higher values compared to HIIT
(390.45  ± 65.15; 55.20 ± 9.33 kcal, respectively). How-
ever, post-exercise EE and EPOC values were higher after Introduction
HIIT (69.31 ± 10.88; 26.27 ± 2.28 kcal, respectively)
Aerobic exercises play an important role to increase
energy expenditure (EE) during or immediately after
Communicated by Guido Ferretti.
exercise sessions (Borsheim and Bahr 2003). Therefore,
it is relevant to investigate the optimal exercise volume
* Gustavo Zaccaria Schaun and intensity to reach a time-efficient EE. In this con-
[email protected] text, EE derived from an exercise session is related to the
1 type of protocol employed, as there may be variations in
Neuromuscular Assessment Laboratory, Physical Education
School, Federal University of Pelotas, Rua Luís de Camões, EE based on manipulation of both its volume and inten-
625, Três Vendas, Pelotas, RS, Brazil sity (Borsheim and Bahr 2003; LaForgia et al. 2006) in

13
 Eur J Appl Physiol

addition to exercise mode performed (Borsheim and Methods

Bahr 2003). In this context, sessions EE can be split into
the following: (1) energy expended during the session Subjects
itself and (2) energy demanded after it, which is directly
influenced by excess post-exercise oxygen consumption Twenty-six healthy physically active men
(EPOC; Gaesser and Brooks 1984) as both contribute (23.3  ± 2.9 years; 76.7 ± 12.6 kg; 1.76 ± 0.06 cm; and
to a considerable volume to the maintenance of a nega- V̇ O2max = 45.9 ± 5.8 ml kg−1 min−1) volunteered to par-
tive daily energetic balance (Borsheim and Bahr 2003). ticipate in the study. This study was part of a larger research
Researchers suggest that the increased metabolic demand in which participants were already randomly assigned
during recovery, expressed by EPOC, seems to be higher (simple randomization) to one of two experimental groups:
after intermittent activities when compared to lower HIIT (n = 14) or CONT (n = 12). Therefore, we opted not
intensity continuous isocaloric ones (Cunha et al. 2016). to use a cross-over design to avoid bias, since participants
Nevertheless, there are still conflicting results regarding were already accustomed to one protocol. Specifically, sub-
the influence of exercise mode, intensity, volume, and jects were engaged in training, three times per week, for
duration in EPOC (Greer et al. 2015). less than 30 days at the time of the assessments. All were
This way, two training methods have been highlighted nonsmokers and did not present previous history of cardio-
and compared in the literature: high-intensity interval vascular, endocrine, metabolic, or neuromuscular diseases
training (HIIT), in which exercise volume may be low as well as could not be using any type of medication that
with exercises performed at high intensity interspersed could interfere in the cardiorespiratory or neuromuscular
with recovery intervals (Buchheit and Laursen 2013) and systems.
moderate-intensity continuous training (CONT), in which Prior to any procedure, all participants read, agreed, and
a higher exercise volume is used with a relatively lower signed an informed consent which was previously approved
intensity when compared to HIIT. by the Federal University of Pelotas Research Ethics Com-
Previously, sprint interval exercise (i.e., 4 × 30 s Win- mittee (CAAE 49499415.0.0000.5313).
gate-based) was shown to elicit similar oxygen consump-
tion (V̇ O2) and EE post-exercise when compared to 30 min Procedures
(~70% iV̇ O2max) and 60 min (~65% iV̇ O2peak) of CONT
along 30 min (Williams et al. 2013) and 24 h post-exer- For the present study, participants attended three separate
cise (Hazell et al. 2012). This is supported by Skelly et al. sessions at the university laboratory.
(2014) who also found the same result when comparing
an HIIT protocol (10 × 60 s at 90% H ­ Rmax interspersed
Physical and anthropometrical measures
with 60 s rest) to a CONT one (50 min at ~70% H ­ Rmax)
up to 24 h post-exercise. On the other hand, CONT tended
On the first day, weight and height were assessed using a
to present a higher EE during the session itself in all the
digital scale (FILIZOLA, São Paulo, Brazil) and stadiom-
above-mentioned studies (Hazell et al. 2012; Skelly et al.
eter, respectively, to calculate subject’s body mass index
2014; Williams et al. 2013).
(BMI). In addition, seven-site skinfold thickness (chest,
Despite this fact, submaximal or maximal HIIT proto-
axilla, triceps, subscapular, abdominal, suprailium, and
cols examined up to the present generally have similar
thigh) were measured with a skinfold caliper (CESCORF,
duration or workload compared to CONT (LaForgia et al.
Porto Alegre, Brazil). Based on the sum of skinfolds, body
1997) and investigated supramaximal protocols commonly
density was estimated (Jackson and Pollock 1978) and
employed relatively long recovery periods (Hazell et al.
body fat calculated (Siri 1993). Skinfolds were measured
2012; Williams et al. 2013). Within this context, evaluation
in circuit, totaling a maximum of three attempts per skin-
of time-efficient HIIT protocols may highlight interesting
fold. Finally, subjects were familiarized with the labora-
facts in understanding EE and EPOC. Having said that, the
tory, treadmill, and masks which they would perform the
aim of the present study was to compare the acute effects
following procedures.
of a very low-volume supramaximal HIIT protocol and a
standard 30-min moderate-intensity continuous session on
EE and EPOC along 30 min in young healthy men, which Incremental test measures
was not performed yet. It was hypothesized that total EE
and EE during session would be greater following CONT On a second occasion, separated by at least 48 h from the
compared with HIIT, whereas session EE per minute, post- previous one, subjects performed a maximal incremental
exercise EE and EPOC would be higher following HIIT test on a treadmill to determine V̇ O2max and the corre-
compared to CONT. sponding intensities for the HIIT and CONT protocols. The

13
Eur J Appl Physiol

test started at 6 km h−1 for 5 min (warm-up) which was fol- 2005). In this regard, for both protocols (CONT or HIIT)
lowed by subsequent 1 km h−1 increments every min until V̇ O2, area was calculated (in liters) according to the spe-
volitional exhaustion. cific moment (i.e., 30-min pre-; CONT, HIIT; 30-min post)
Throughout the test, expiratory gases were directly col- and was multiplied by the corresponding CE.
lected and analyzed using a breath-by-breath portable
spirometer with a frequency of acquisition set for the aver- Exercise protocols
age of every three breaths (VO2000, MedGraphics, Ann
Arbor, USA), which was calibrated according to manu- HIIT group performed a standard warm-up corresponding
facturer’s specifications before every test. The maximum to 90–95% ­VT2 for 4 min. Following a 3-min passive inter-
V̇ O2 (ml kg−1 min−1) was considered as the mean 30 s of val, participants performed eight 20 s bouts on a motorized
the last completed stage and second ventilatory threshold treadmill (KIKOS KX 9000, São Paulo, Brazil) at 130% of
­(VT2) was determined by the ventilation vs. intensity curve the velocity associated with V̇ O2max interspersed with 10 s
and confirmed by the ventilatory C ­ O2 equivalent curve (V̇E passive recovery between bouts. As for the CONT group,
/V̇ CO2; Wasserman et al. 1973). Regarding threshold detec- subjects exercised on the same treadmill for 30 min with
tion, it was performed by visual inspection by two experi- an intensity corresponding to 90–95% of the HR associated
enced physiologists in a blinded and independently fashion with ­VT2.
(Reinhard et al. 1979) as previously described (Alberton
et al. 2014). When there was no agreement on the deter-
mination of the break point, a third physiologist was Statistical analysis
employed. In addition, heart rate (HR) was assessed using a
­Polar® monitor (RS800CX, Finland). Descriptive and inferential statistics were used for data
Incremental tests were considered valid when at least analysis. After testing the normality of the data with Sha-
two of the following criterions were achieved: (1) plateau piro–Wilk’s test, data were presented as mean and stand-
in V̇ O2 despite an increase in exercise intensity; (2) res- ard deviations (±SD). For V̇ O2, EE and EPOC comparison
piratory exchange ratio higher than 1.10; and (3) heart rate between HIIT and CONT groups, a t test for independent
predicted by age (220 age) achieved (Ferreti 2014; Howley measures was applied. Moreover, an ANOVA two-way for
et al. 1995). In addition, the day before test participants repeated measures followed by Bonferroni post hoc was
were asked not to perform vigorous physical activity, not applied to evaluate the EE between moments and protocols.
to consume any stimulant beverages and to sleep at least In addition, a value of α = 0.05 was adopted and all tests
8 h. were performed in the SPSS 20.0 program.

Exercise measures
Results
On a third occasion, participants performed their assigned
protocol. First, they remained seated for 30-min pre-exer- Initially, no significant difference was identified between
cise. After, subjects performed their respective protocol groups regarding V̇ O2 at rest (p > 0.05), which corre-
and, again, remained seated for 30-min post-exercise. Dur- sponded to 0.29 ± 0.07 and 0.30 ± 0.04 L min−1 for
ing all the procedures, expiratory gases were collected and CONT and HIIT groups, respectively, totaling 8.78 ± 2.02
V̇ O2 (L min−1) was measured for posterior determination and 9.21 ± 1.61 L of ­O2 over the 30-min period. Similar
of the EE in each protocol, as well as before and after them. behavior has been verified for all other descriptive variables
Furthermore, pre-exercise EE was subtracted from post- (Table 1). All t tests for independent variables revealed that
exercise EE to determine EPOC. there were no statistical differences between these variables
Different caloric equivalents (CE) were used for the (all p > 0.05) at the onset of the study. In addition, it should
determination of EE based on the moment and protocol be noted that all subjects reached at least two out of the
performed (Wilmore et al. 1978). For the CONT group, a three criteria for V̇ O2max attainment during the incremen-
CE of 4.85 kcal L−1 of ­O2 was used on the pre-exercise, tal test.
exercise, and post-exercise moments. As for the HIIT Results regarding exercise and post-exercise compari-
group, the same CE was used on the pre- and post-exercise sons between sessions are presented in Table 2. Differences
(i.e., 4.85 kcal L−1); however, during the exercise, a CE of between protocols were observed for absolute V̇ O2 and EE
5.05 kcal L−1 was adopted due to the high intensity arising during and after session, as well as for V̇ O2 and EE per
from the session. In addition, we chose to perform the cal- minute after the session. Absolute V̇ O2 and EE during exer-
culation of the area corresponding to the V̇ O2 (L) through cise were higher in CONT when compared to HIIT, while
integration, as already frequently employed (Jacobsen et al. HIIT presented higher absolute and per minute V̇ O2 and EE

13
 Eur J Appl Physiol

Table 1  Physical and anthropometrical characteristics of a healthy

and physically active men’s sample according to group (n = 26)

CONT (n = 12) HIIT (n = 14)

Mean ±s Mean ±s

Age (years) 23.3 ±3.1 23.3 ±2.9

Body weight (kg) 79.9 ±12.6 73.8 ±12.6
Height (m) 1.79 ±5.0 1.75 ±4.8
BMI (kg/m2) 25.0 ±3.3 24.0 ±3.4
BF (%) 17.0 ±6.7 15.9 ±5.3
∑Skinfolds (mm) 127.5 ±51.5 108.5 ±39.3
V̇ O2max (ml kg−1 min−1) 47.5 ±6.9 44.5 ±4.4
Fig. 1  Energy expenditure between moments according to protocol
CONT moderate-intensity continuous training session, HIIT high- performed. CONT moderate-intensity continuous training session
intensity interval training session, BMI body mass index, BF percent- (n = 12), HIIT high-intensity interval training session (n = 14), Pre
age of body fat, ∑Skinfolds sum of seven-site skinfold thickness, twenty-fifth minute at rest, Exercise last minute of exercise session,
V̇ O2max maximal oxygen consumption. No statistical difference Post 5 fifth minute post-exercise, Post 25 twenty-fifth minute post-
between groups was identified exercise. *Significant difference between protocols (p ≤ 0.01), †Sig-
nificant difference from exercise to the other moments (Pre, Post 5,
and Post 25) in both protocols (p < 0.05), ‡Significant difference from
Post 5 to the other moments in the HIIT protocol (p < 0.05). Values
Table 2  Comparisons of oxygen consumption (V̇ O2), energy expend- represent the mean from all the subjects’ instantaneous values over
iture (EE), and excess post-exercise oxygen consumption (EPOC) the corresponding minute
according to protocol in healthy young man (n = 26)

CONT (n = 12) HIIT (n = 14)

Mean ±s Mean ±s descriptive V̇ O2 (a) and EE (b) pattern during the 30-min
post-exercise period according to protocol performed.
V̇ O2 exercise (L O
­ 2) 80.50 13.43 10.45* 1.48
V̇ O2 exercise (L O2 min ) −1 2.68 0.45 2.73 0.46
V̇ O2 post (L ­O2) 11.55 2.10 14.29* 2.25
V̇ O2 post (L ­O2 min−1) 0.39 0.07 0.48* 0.08 Discussion
EE exercise (kcal) 390.45 65.15 52.78* 7.50
EE exercise (kcal min−1) 13.02 2.17 13.20 1.87 The main findings of the present study were the higher
EE post (kcal) 55.99 10.20 69.31* 10.88 V̇ O2, EE, and EPOC values during the 30 min immedi-
EE post (kcal min−1) 1.87 0.34 2.31* 0.36 ately after HIIT in comparison with CONT session, despite
EPOC (L ­O2) 2.77 2.15 5.41* 1.30 its extremely short volume. In addition, it was observed a
EPOC (kcal) 13.43 10.45 26.27* 6.28
greater total EE during CONT session, prescribed by %HR
corresponding to ­VT2, when compared to supramaximal
CONT moderate-intensity continuous training session, HIIT high- HIIT. Moreover, although exercise intensity is indicated
intensity interval training session
as a critical factor for a prolonged component of EPOC
* Significant difference between protocols (p ≤ 0.005) (Bahr et al. 1992), in the present study, such behavior was
not observed after our HIIT protocol, since both groups
post-exercise. Moreover, a greater EPOC was verified after returned to their resting values within 30-min post-session.
HIIT when compared to CONT. This fact demonstrates that in extremely short supramaxi-
When considering EE in relation to values at rest mal protocols, the volume can indeed play an important
(Fig.  1), it was observed that CONT already had similar role in EPOC duration.
values to the resting condition on the fifth minute after the Regarding EE during exercise sessions, values found
protocol (p > 0.05). However, HIIT had not returned to are in agreement with previous studies. Smith and Mc
resting values at this point (p = 0.001), but only in the 25th Naughton (1993) assessed young men and women who
minute (p > 0.05). Although no difference was identified in pedaled at 70% V̇ O2max during 30 min and found an EE of
the mean EE between protocols (HIIT 52.78 ± 7.50 kcal; approximately 379 kcal (78.2 L ­O2). Similar values were
CONT 390.45 ± 65.15 kcal) during the sessions (Table 1), also presented by Matsuo et al. (2012a), which reported
Fig.  1 shows that on the last minute of the exercise ses- an EE of 351 kcal (71.1 L O ­ 2) after 30-min cycling at
sion, HIIT was promoting significantly greater EE when 60–65% iV̇ O2max. Regarding HIIT, the more similar
compared to CONT (p  = 0.011). Figure 2 presents the investigations identified in the literature were of Matsuo

13
Eur J Appl Physiol

Fig. 2  Descriptive oxygen
consumption (a) and energy
expenditure (b) results
(mean ± s) during the 30-min
post-exercise period between
moderate-intensity continu-
ous training session (CONT;
n = 12) and high-intensity
interval training session (HIIT;
n = 14) protocols. Rest value is
represented by a continuous line
corresponding to the mean value
of both protocols, since there
were no differences between
them during rest

et al. (2012a, b), in which subjects performed seven 30 s related to faster V̇ O2 adjustment (Poole and Jones 2012),
bouts at 120% iV̇ O2max interspersed by 15 s passive rest which could be extremely important in view of the short
on a cycle ergometer, resulting in an EE equal to 77.8 kcal duration of these protocols. In this regard, these factors
(14.5 L ­O2) during exercise session, while our HIIT group could overlap the effect of the higher intensity employed
presented an EE corresponding to 52.8 kcal (10.5 L O ­ 2). in our HIIT protocol (i.e., 130 vs. 120% iV̇ O2max).
Such difference in EE between studies, although small in With respect to the comparison between protocols, a
absolute terms, represents a ~32% higher EE when com- higher total EE during CONT when compared to HIIT
pared to our results. was already expected. Previous studies have shown that
Three possible explanations could be: (1) difference when session volume (i.e., session duration and/or effort
in total exercise volume performed, since Matsuo et al. time) is similar between workouts, those performed at
(2012a, b) protocol was composed by 210 s of effort higher intensity generally induce higher EE (Borsheim
(3.5 min) compared to 160 s (2.7 min) in the present and Bahr 2003; McGarvey et al. 2005; Smith and Mc
study; (2) a greater prevalence of eccentric actions dur- Naughton 1993). However, when high-intensity sessions,
ing running compared to cycling, which could also result HIIT in particular, are conducted with reduced volume
in elastic energy storage due to stretch-shortening cycle in relation to CONT, absolute EE from CONT tends to
(Poole and Jones 2012); (3) difference between V̇ O2max be greater (Matsuo et al. 2012a, b; Skelly et al. 2014).
of both study groups (47.5 vs. 52.0 ml kg−1 min−1), con- In contrast to absolute values, when comparing mean val-
sidering that a higher cardiorespiratory conditioning is ues of both V̇ O2 and EE per minute between protocols

13
 Eur J Appl Physiol

(Table  2), significant differences were not observed. fifth minutes, while HIIT remained a little bit higher for
This result is conflicting with values reported in Fig. 1, a longer period. As previously explained, intensity has
which demonstrated difference between groups for EE been shown to be able to explain approximately 45.5% of
during the last minute of exercise. Thus, considering the variation in EPOC compared to only 8.9% for effort dura-
higher intensity employed during HIIT protocol, it should tion (Gore and Withers 1990). Thus, intensity employed
be expected that this would result in a greater absolute in the present study during HIIT (i.e., 130% iV̇ O2max)
V̇ O2 and hence higher V̇ O2 and EE per minute when when compared to CONT (90–95%HR corresponding to
compared to CONT protocol. However, extremely low ­VT2) may have been determinant to this result (LaForgia
volume employed in the HIIT protocol (less than 4 min) et al. 1997). Some previous investigations also presented
may have prevented subjects from reaching and maintain- reduced EPOC durations after CONT. Olmedo (2011)
ing a higher V̇ O2 throughout a great portion of the ses- and Dawson et al. (1996) found EPOC to be only 9 and
sion. This notion is supported by previous investigation 14 min long after 30 min at 65% iV̇ O2max, respectively.
comparing the performance of supramaximal protocols A third study (Townsend et al. 2013) also showed simi-
(Muniz-Pumares et al. 2016). lar results, ~6 min after 30 min of running at 60% HRre-
In regards of the post-session period, greater impact serve. Such differences may be associated with the differ-
imposed by HIIT protocol when compared to CONT, ent prescriptions of these protocols. As in Dawson et al.
both in absolute and relative terms, was observed which (1996) study, ­VT2 of the subjects in our CONT group was
reflected in greater EPOC after HIIT. These results are assessed and exercise intensity was prescribed based on a
consistent to a number of investigations that compared percentage below it. It had been previously demonstrated
EPOC for these two training protocols (Frey et al. 1993; that this type of prescription is a more robust method,
Greer et al. 2015; LaForgia et al. 1997; Larsen et al. which allows proper individualization of exercise load
2014; Malatesta et al. 2009; Matsuo et al. 2012b; Met- when compared to % iV̇ O2max, since the latter does
calfe et al. 2015). Larsen et al. (2014) found EPOC along not permit differentiation between moderate and intense
~48 min with twice the magnitude (2.9 vs. 1.4 L ­O2) domains (Tschakert and Hofmann 2013; Wolpern et al.
after HIIT (4 × 4:3 min 85–95%HRmax and 70%HRmax) 2015). Therefore, in the studies previously mentioned, it
when compared to a CONT session (47 min 70%HRmax). is not possible to state that all subjects were training at a
Such results are in agreement with Matsuo et al. (2012b), moderate domain and this, in turn, may have biased and
who observed a 6.8 L ­O2 EPOC along 180 min follow- actually prolonged subjects mean EPOC duration.
ing a sprint-based HIIT workout on a cycle ergometer Regarding EPOC duration after HIIT, studies that
(7× 30:15 s 120% iV̇ O2max), significantly higher than reported EPOC lasting from 17 min (Olmedo 2011) up
a longer HIIT protocol (4.5 L O ­ 2; 3× 3:2 min 85–90% to 12 h (Greer et al. 2015) were found. In this sense, the
iV̇ O2max and 60% iV̇ O2max), and both higher than majority of studies that had active exercise volume and/
CONT (2.9 L O ­ 2; 40 min 60–65% iV̇ O2max). In terms of or duration greater than the present one (i.e., 2 min and
EE, values observed in our protocol are also very similar 40 s) resulted in longer EPOC (Bahr et al. 1992; Fried-
to those verified by Matsuo et al. (2012b), with 32 kcal man et al. 2012; Greer et al. 2015; LaForgia et al. 1997;
for the sprint-based HIIT and 13 kcal for CONT group. Larsen et al. 2014; McGarvey et al. 2005; Metcalfe et al.
Nevertheless, one study found no difference between 2015; Skelly et al. 2014). Therefore, investigations of
HIIT (7× 2:3 min 90% iV̇ O2max and 30% iV̇ O2max ) extremely low volume or reduced effort duration (i.e.,
and CONT (30 min 60% iV̇ O2max), with EPOC val- more similar to the one employed in our protocol) were
ues of 7.6 vs. 7.0 L ­O2, respectively (McGarvey et al. the ones that showed the lowest values. This is the case of
2005). However, HIIT average intensity corresponded to seven 30-s cycling bouts at 120% iV̇ O2max interspersed
60.5% V̇ O2max, and it was previously shown that EPOC by 15 s rest (Matsuo et al. 2012a, b) or two sets of three
increased exponentially with increasing intensity (Gore 20-s all-out bouts interspersed by 2 min of recovery
and Withers 1990), especially when it approaches or (Friedman et al. 2012) leading to EPOC duration close
exceeds 100% of iV̇ O2max (Hagberg et al. 1980). to 30 and ~49 min, respectively. Furthermore, Bahr et al.
Concerning EPOC duration, as can be observed in (1992) had already demonstrated a linear relationship
Fig. 1, CONT group had already returned to basal levels between EPOC duration after comparing supramaximal
within 5 min after the end of the session, while in HIIT, HIIT protocols composed by one, two, or three 2-min
EE returned to baseline values between the 6th and 25th bouts at 108% iV̇ O2max (i.e., 30 min, 60 min, and 4 h,
minutes. These behaviors can also be observed in Fig. 2, respectively), whereas LaForgia et al. (1997) found
which demonstrate a, descriptively only, faster EPOC EPOC close to 9 h after completion of 20 sets of 1:2 min
reduction for CONT at approximately the fourth and at 105% iV̇ O2max.

13
Eur J Appl Physiol

These results agree with Borsheim and Bahr (2003), Conclusions

who argue that exercise intensity and session duration
are synergistic and not additive, both being important Based on the results presented, our data suggest that a time-
to determine magnitude and duration of EPOC. Moreo- efficient supramaximal HIIT protocol (~10-min total ses-
ver, differences in the position used to evaluate subjects sion duration) has a higher impact on EE and EPOC in the
(e.g., seated or supine position) hinder comparisons of early phase of recovery when compared to CONT protocol.
EPOC values between studies. Another important aspect
is related to investigations that found EPOC with longer Acknowledgements The authors would like to thank all subjects
durations, i.e., >9 h (for example, LaForgia et al. 1997). who took part in the study for their genuine effort.
During these studies, subjects were allowed to feed them- Compliance with ethical standards 
selves during post-exercise resting period, a factor that
could overestimate EPOC due to the thermic effect of Conflict of interest  The authors attest that they have no conflict of
consumed food. interest.
In addition, the ways in which resting values were pro-
cessed and analyzed could also interfere in EPOC duration. Ethical approval All procedures performed in studies involving
McGarvey et al. (2005) when comparing CONT and HIIT human participants were in accordance with the ethical standards of
the institutional and/or national research committee and with the 1964
protocols found EPOC durations corresponding to 50.8 and
Helsinki declaration and its later amendments or comparable ethical
53 min, respectively, when using the one standard devia- standards.
tion technique. However, when the same analysis was per-
formed based on the 5-min technique, EPOC was 88.8 and Informed consent  Informed consent was obtained from all individual
94.9 min for the same data, a difference of almost 50%. participants included in the study.
This could potentially explain why in CONT studies with
lower intensities than those of the present study greater
EPOC durations were found.
Thereby, the present study demonstrated that, although
References
a prolonged EPOC after supramaximal HIIT was expected
when compared to CONT, there seems to be a kind of Alberton CL, Pinto SS, Antunes AH, Cadore EL, Finatto P, Tartaruga
threshold in volume required for the development of a MP, Kruel LF (2014) Maximal and ventilatory thresholds cardi-
longer component of EPOC (Borsheim and Bahr 2003), orespiratory responses to three water aerobic exercises compared
with treadmill on land. J Strength Cond Res 28:1679–1687.
since it had already returned to baseline levels within
doi:10.1519/JSC.0000000000000304
25 min. In addition, our results agree with the literature: Bahr R, Gronnerod O, Sejersted OM (1992) Effect of supramaximal
(1) as to the magnitude of EPOC, nearly twice in the HIIT exercise on excess postexercise O
­ 2 consumption. Med Sci Sports
group (5.41 vs. 2.77 L O ­ 2), although an extremely short Exerc 24:66–71
Borsheim E, Bahr R (2003) Effect of exercise intensity, duration
duration was employed and (2) as also for the fact that this
and mode on post-exercise oxygen consumption. Sports Med
result caused by intensity was more expressive during the 33:1037–1060. doi:10.2165/00007256-200333140-00002
fast component of the curve (i.e., within the first hour; Bahr Buchheit M, Laursen PB (2013) High-intensity interval train-
et al. 1992). Therefore, our results indicate that when the ing, solutions to the programming puzzle: part I: cardiopul-
monary emphasis. Sports Med 43:313–338. doi:10.1007/
main objective is to produce a higher total EE, EE corre-
s40279-013-0029-x
sponding to EPOC, even in supramaximal session, has Cunha FA, Midgley AW, McNaughton LR, Farinatti PT (2016) Effect
a relatively small magnitude (HIIT: 26.27 kcal; CONT: of continuous and intermittent bouts of isocaloric cycling and
13.43 kcal) and emphasis should be directed into spend- running exercise on excess postexercise oxygen consumption. J
Sci Med Sport 19:187–192. doi:10.1016/j.jsams.2015.02.004
ing energy during the exercise session itself and not after it
Dawson B, Straton S, Randall N (1996) Oxygen consumption dur-
(LaForgia et al. 2006). ing recovery from prolonged submaximal cycling below the
Finally, we suggest that future investigations evaluate anaerobic threshold. J Sports Med Phys Fit 36:77–84
different supramaximal HIIT protocols with similar inten- Ferreti G (2014) Maximal oxygen consumption in healthy humans:
theories and facts. Eur J Appl Physiol 114:2007–2036.
sities but differing in terms of volume (i.e., number of
doi:10.1007/s00421-014-2911-0
sets) to determine if in fact, there is a duration threshold Frey GC, Byrnes WC, Mazzeo RS (1993) Factors influ-
that results in a substantially longer EPOC after extremely encing excess postexercise oxygen consumption in
short duration supramaximal HIIT protocols. Furthermore, trained and untrained women. Metabolism 42:822–828.
doi:10.1016/0026-0495(93)90053-Q
future studies should address possible gender differences
Friedman RA, Navalta JW, Fedor EA, Kell HB, Lyons TS, Arnett SW,
between EPOC in these protocols while also assessing the Schafer MA (2012) Repeated high-intensity Wingate cycle bouts
prolonged EPOC component, as it was not performed in influence markers of lymphocyte migration but not apoptosis.
the present investigation. Appl Physiol Nutr Metab 37:241–246. doi:10.1139/h11-156

13
 Eur J Appl Physiol

Gaesser GA, Brooks GA (1984) Metabolic bases of excess post- exercise. Int J Sport Nutr Exerc Metab 15:28–37. doi:10.1123/
exercise oxygen consumption: a review. Med Sci Sports Exerc ijsnem.15.1.28
16:29–43. doi:10.1249/00005768-198401000-00008 Metcalfe RS, Koumanov F, Ruffino JS, Stokes KA, Holman GD,
Gore CJ, Withers RT (1990) The effect of exercise intensity and Thompson D, Vollaard NB (2015) Physiological and molecular
duration on the oxygen deficit and excess post-exercise oxygen responses to an acute bout of reduced-exertion high-intensity
consumption. Eur J Appl Physiol 60:169–174. doi:10.1007/ interval training (REHIT). Eur J Appl Physiol 115:2321–2334.
BF00713508 doi:10.1007/s00421-015-3217-6
Greer BK, Sirithienthad P, Moffatt RJ, Marcello RT, Panton LB Muniz-Pumares D, Pedlar C, Godfrey RJ, Glaister M (2016) Accumu-
(2015) EPOC comparison between isocaloric bouts of steady- lated oxygen deficit during exercise to exhaustion determined at
state aerobic, intermittent aerobic, and resistance training. Res different supramaximal work-rates. Int J Sports Physiol Perform
Q Exerc Sport 86:190–195. doi:10.1080/02701367.2014.999190 24:1–17. doi:10.1123/ijspp.2015-0343
Hagberg JM, Mullin JP, Nagle FJ (1980) Effect of work intensity and Olmedo AC (2011) Post-exercise oxygen consumption after con-
duration on recovery ­O2. J Appl Physiol Respir Environ Exerc tinuous and interval exercise on a treadmill. Apunts Educación
Physiol 48:540–544 Física y Deportes 104:21–27. doi:10.5672/apunts.2014-0983.
Hazell TJ, Olver TD, Hamilton CD, Lemon PW (2012) Two min- es(2011/2).104.02
utes of sprint-interval exercise elicits 24-hr oxygen consump- Poole DC, Jones AM (2012) Oxygen uptake kinetics. Compr Physiol
tion similar to that of 30 min of continuous endurance exer- 2:933–996. doi:10.1002/cphy.c100072
cise. Int J Sport Nutr Exerc Metab 22:276–283. doi:10.1123/ Reinhard U, Muller PH, Schmulling RM (1979) Determination of
ijsnem.22.4.276 anaerobic threshold by the ventilation equivalent in normal indi-
Howley ET, Bassett DR Jr, Welch HG (1995) Criteria for maximal viduals. Respiration 38(1):36–42. doi:10.1159/000194056
oxygen uptake: review and commentary. Med Sci Sports Exerc Siri WE (1993) Body composition from fluid spaces and density:
27:1292–1301 analysis of methods. Nutrition 9:480–491
Jackson AS, Pollock ML (1978) Generalized equations for predicting Skelly LE, Andrews PC, Gillen JB, Martin BJ, Percival ME, Gibala
body density of men. Br J Nutr 40:497–504 MJ (2014) High-intensity interval exercise induces 24-h energy
Jacobsen DJ, Bailey BW, LeCheminant JD, Hill JO, Mayo MS, Don- expenditure similar to traditional endurance exercise despite
nelly JE (2005) A comparison of three methods of analyzing reduced time commitment. Appl Physiol Nutr Metab 39:845–
post-exercise oxygen consumption. Int J Sports Med 26(1):34– 848. doi:10.1139/apnm-2013-0562
38. doi:10.1055/s-2004-815819 Smith J, Mc Naughton L (1993) The effects of intensity of exercise
LaForgia J, Withers RT, Shipp NJ, Gore CJ (1997) Comparison of on excess postexercise oxygen consumption and energy expendi-
energy expenditure elevations after submaximal and supramaxi- ture in moderately trained men and women. Eur J Appl Physiol
mal running. J Appl Physiol 82:661–666 Occup Physiol 67:420–425. doi:10.1007/BF00376458
LaForgia J, Withers RT, Gore CJ (2006) Effects of exercise intensity Townsend JR et al (2013) Excess post-exercise oxygen consumption
and duration on the excess post-exercise oxygen consumption. J (EPOC) following multiple effort sprint and moderate aerobic
Sports Sci 24:1247–1264. doi:10.1080/02640410600552064 exercise. Kinesiology 45:16–21
Larsen I, Welde B, Martins C, Tjonna AE (2014) High- and moder- Tschakert G, Hofmann P (2013) High-intensity intermittent exercise:
ate-intensity aerobic exercise and excess post-exercise oxygen methodological and physiological aspects. Int J Sports Physiol
consumption in men with metabolic syndrome. Scand J Med Sci Perform 8:600–610
Sports 24:e174–e179. doi:10.1111/sms.12132 Wasserman K, Whipp BJ, Koyl SN, Beaver WL (1973) Anaerobic
Malatesta D, Werlen C, Bulfaro S, Cheneviere X, Borrani F (2009) threshold and respiratory gas exchange during exercise. J Appl
Effect of high-intensity interval exercise on lipid oxidation dur- Physiol 35:236–243
ing postexercise recovery. Med Sci Sports Exerc 41:364–374. Williams CB, Zelt JG, Castellani LN, Little JP, Jung ME, Wright DC,
doi:10.1249/MSS.0b013e3181857edo Tschakovsky ME, Gurd BJ (2013) Changes in mechanisms pro-
Matsuo T, Ohkawara K, Seino S, Shimojo N, Yamada S, Ohshima posed to mediate fat loss following an acute bout of high-inten-
H, Tanaka K, Mukai C (2012a) Cardiorespiratory fitness level sity interval and endurance exercise. Appl Physiol Nutr Metab
correlates inversely with excess post-exercise oxygen consump- 38:1236–1244. doi:10.1139/apnm-2013-0101
tion after aerobic-type interval training. BMC Res Notes 5:646. Wilmore JH, Parr RB, Ward P, Volak PA, Barstow TJ, Pipes TV,
doi:10.1186/1756-0500-5-646 Grimditch G, Leslie P (1978) Energy cost of circuit weight train-
Matsuo T, Ohkawara K, Seino S, Shimojo N, Yamada S, Ohshima H, ing. Med Sci Sport Exerc 10:75–78
Tanaka K, Mukai C (2012b) An exercise protocol designed to Wolpern AE, Burgos DJ, Janot JM, Dalleck LC (2015) Is a threshold-
control energy expenditure for long-term space missions. Aviat based model a superior method to the relative percent concept
Space Environ Med 83:783–789. doi:10.3357/ASEM.3298.2012 for establishing individual exercise intensity? a randomized con-
McGarvey W, Jones R, Petersen S (2005) Excess post-exercise oxy- trolled trial. BMC Sports Sci Med Rehabil 7:16. doi:10.1186/
gen consumption following continuous and interval cycling s13102-015-0011-z

13