Effect of Exercise Intensity, Duration and Mode on Post-Exercise Oxygen Consumption

Abstract

In the recovery period after exercise there is an increase in oxygen uptake termed the ‘excess post-exercise oxygen consumption’ (EPOC), consisting of a rapid and a prolonged component. While some studies have shown that EPOC may last for several hours after exercise, others have concluded that EPOC is transient and minimal. The conflicting results may be resolved if differences in exercise intensity and duration are considered, since this may affect the metabolic processes underlying EPOC. Accordingly, the absence of a sustained EPOC after exercise seems to be a consistent finding in studies with low exercise intensity and/or duration. The magnitude of EPOC after aerobic exercise clearly depends on both the duration and intensity of exercise. A curvilinear relationship between the magnitude of EPOC and the intensity of the exercise bout has been found, whereas the relationship between exercise duration and EPOC magnitude appears to be more linear, especially at higher intensities.
Differences in exercise mode may potentially contribute to the discrepant findings of EPOC magnitude and duration. Studies with sufficient exercise challenges are needed to determine whether various aerobic exercise modes affect EPOC differently. The relationships between the intensity and duration of resistance exercise and the magnitude and duration of EPOC have not been determined, but a more prolonged and substantial EPOC has been found after hardversus moderate-resistance exercise. Thus, the intensity of resistance exercise seems to be of importance for EPOC.
Lastly, training status and sex may also potentially influence EPOC magnitude, but this may be problematic to determine. Still, it appears that trained individuals have a more rapid return of post-exercise metabolism to resting levels after exercising at either the same relative or absolute work rate; however, studies after more strenuous exercise bouts are needed. It is not determined if there is a sex effect on EPOC.
Finally, while some of the mechanisms underlying the more rapid EPOC are well known (replenishment of oxygen stores, adenosine triphosphate/creatine phosphate resynthesis, lactate removal, and increased body temperature, circulation and ventilation), less is known about the mechanisms underlying the prolonged EPOC component. A sustained increased circulation, ventilation and body temperature may contribute, but the cost of this is low. An increased rate of triglyceride/fatty acid cycling and a shift from carbohydrate to fat as substrate source are of importance for the prolonged EPOC component after exhaustive aerobic exercise. Little is known about the mechanisms underlying EPOC after resistance exercise.

Full text

Sports Med 2003; 33 (14): 1037-1060
R
EVIEW
A
RTICLE
0112-1642/03/0014-1037/$30.00/0
 Adis Data Information BV 2003. All rights reserved.
Effect of Exercise Intensity, Duration
and Mode on Post-Exercise
Oxygen Consumption
Elisabet Børsheim and Roald Bahr
Norwegian University of Sport and Physical Education, Oslo, Norway
Contents
Abstract...................................................................................1037
1. Excess Post-Exercise Oxygen Consumption (EPOC) ........................................1038
2. Early Studies on EPOC ..................................................................1039
3. Methodological Considerations .........................................................1039
4. Effect of Intensity and Duration of Aerobic Exercise on EPOC ..............................1040
5. Effect of Split Exercise Sessions on EPOC ..................................................1047
6. Effect of Supramaximal Exercise on EPOC ................................................1048
7. Effect of Aerobic Exercise Mode on EPOC................................................1048
8. Effect of Resistance Exercise on EPOC ...................................................1049
9. Effect of Training Status on EPOC ........................................................1051
10. Effect of Sex on EPOC ..................................................................1052
11. Possible Mechanisms for the Rapid EPOC Component .....................................1053
12. Possible Mechanisms for the Prolonged EPOC Component.................................1053
13. Conclusions ...........................................................................1056
In the recovery period after exercise there is an increase in oxygen uptake
Abstract
termed the ‘excess post-exercise oxygen consumption’ (EPOC), consisting of a
rapid and a prolonged component. While some studies have shown that EPOC
may last for several hours after exercise, others have concluded that EPOC is
transient and minimal. The conflicting results may be resolved if differences in
exercise intensity and duration are considered, since this may affect the metabolic
processes underlying EPOC. Accordingly, the absence of a sustained EPOC after
exercise seems to be a consistent finding in studies with low exercise intensity
and/or duration. The magnitude of EPOC after aerobic exercise clearly depends
on both the duration and intensity of exercise. A curvilinear relationship between
the magnitude of EPOC and the intensity of the exercise bout has been found,
whereas the relationship between exercise duration and EPOC magnitude appears
to be more linear, especially at higher intensities.
Differences in exercise mode may potentially contribute to the discrepant
findings of EPOC magnitude and duration. Studies with sufficient exercise
challenges are needed to determine whether various aerobic exercise modes affect
EPOC differently. The relationships between the intensity and duration of resis-
tance exercise and the magnitude and duration of EPOC have not been deter-

1038 Børsheim & Bahr
mined, but a more prolonged and substantial EPOC has been found after hard-
versus moderate-resistance exercise. Thus, the intensity of resistance exercise
seems to be of importance for EPOC.
Lastly, training status and sex may also potentially influence EPOC magni-
tude, but this may be problematic to determine. Still, it appears that trained
individuals have a more rapid return of post-exercise metabolism to resting levels
after exercising at either the same relative or absolute work rate; however, studies
after more strenuous exercise bouts are needed. It is not determined if there is a
sex effect on EPOC.
Finally, while some of the mechanisms underlying the more rapid EPOC are
well known (replenishment of oxygen stores, adenosine triphosphate/creatine
phosphate resynthesis, lactate removal, and increased body temperature, circula-
tion and ventilation), less is known about the mechanisms underlying the pro-
longed EPOC component. A sustained increased circulation, ventilation and body
temperature may contribute, but the cost of this is low. An increased rate of
triglyceride/fatty acid cycling and a shift from carbohydrate to fat as substrate
source are of importance for the prolonged EPOC component after exhaustive
aerobic exercise. Little is known about the mechanisms underlying EPOC after
resistance exercise.
1. Excess Post-Exercise Oxygen neutral term ‘excess post-exercise oxygen consump-
Consumption (EPOC)
tion’ (EPOC), which also includes the more pro-
longed increase in
˙
VO
2
that may be observed for
During exercise, there is an increase in oxygen
hours after exercise.
uptake (
˙
VO
2
) to support the increased energy need.
EPOC consists of several components.
[6,7]
In this
After exercise,
˙
VO
2
does not return to resting levels
review, the term ‘rapid component’ will be used to
immediately, but may be elevated above resting
describe the sum of components that decays within
levels for some period of time. Originally, the in-
approximately 1 hour, whereas the prolonged com-
creased
˙
VO
2
after exercise was explained by the
ponent decays monoexponentially with a half-life in
oxygen debt hypothesis. The theoretical basis for
the order of several hours (figure 1). Therefore,
this was formulated by Hill et al.
[1-4]
They hypothe-
processes active also beyond the first hour post-
sised that the elevated
˙
VO
2
after exercise was neces-
sary for the repayment of the oxygen deficit incurred
exercise must be responsible for the prolonged
after the start of exercise, and ascribed the oxygen
EPOC component.
debt to the oxidative removal of lactate. Margaria et
Training (i.e. repetitive bouts of exercise) may
al.
[5]
modified the concept, and suggested that the
also have a more chronic effect on resting metabolic
oxygen debt consisted of a lactacid component
rate (RMR). In particular, this seems to be the case
caused by glycogen synthesis from lactate, and an
in trained compared with untrained individuals, es-
alactacid component related to other factors. The
pecially when combined with high/sufficient energy
lactacid component was considered to be the slower
intake, resulting in a high energy flux or turnover.
[8]
component. However, the causality implied by the
At times it may be difficult to separate this effect
term ‘oxygen debt’ is contrary to what is currently
from the EPOC effect. In this review, we will only
known about the biochemical mechanisms underly-
include studies of
˙
VO
2
after an acute bout of exer-
ing the increase in metabolism post-exercise. There-
cise.
fore, Gaesser and Brooks
[6]
introduced the causality
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1039
Later, more controlled studies have been per-
formed. Some studies have confirmed that there is
an increase in
˙
VO
2
after exercise that may last for
several hours.
[14-19]
However, other studies have
concluded that EPOC is transient and minimal after
exercise.
[20-24]
The conflicts in the results may be
resolved if differences in exercise intensity and du-
ration are taken into account, since this may be
expected to affect the metabolic processes underly-
ing EPOC. Also, differences in exercise mode, train-
ing status and sex may potentially contribute to the
discrepant findings.
3. Methodological Considerations
024681012
0
50
100
150
200
EPOC (mL/min)
Time after exercise (h)
Fig. 1. Time plot of excess post-exercise oxygen consumption
(EPOC) after exhaustive submaximal exercise (71–80 minutes at
69–78% of maximal oxygen uptake; n = 12). The solid line shows
the prolonged EPOC component (reproduced from Bahr,
[7]
with
permission
)
.
There are several methodological issues that are
important to consider when studying EPOC. Accu-
2. Early Studies on EPOC
rate control over the pre-experimental conditions,
and an excellent reproducibility in the indirect calo-
rimetry measures are prerequisites to be able to
The first report on an elevated RMR after physi-
detect small, but potentially important differences.
cal activity was published by Benedict and Carpen-
Only few authors report the precision of the indirect
ter in 1910.
[9]
They observed a mean increase in
calorimetry system used to measure
˙
VO
2
. The
RMR of 11.1% for their two study participants
Douglas bag method is generally considered to be
during sleep in a respiration calorimeter 7–13 hours
the most accurate method of expired gas analysis,
after severe exercise. Initially it was thought that
but few authors, especially of newer studies, have
post-exercise elevation in
˙
VO
2
contributed signifi-
used this technique. Instead, automated systems
cantly to the energy cost of exercise, and would be
have been used, often with unknown validity and
an important factor in daily energy expenditure.
reliability.
Herxheimer et al.
[10]
noted that the
˙
VO
2
of five
untrained individuals did not return to baseline until Furthermore, the pre-experimental conditions
36–48 hours after exercise, and Edwards et al.
[11]
have not always been well controlled. The study
reported a 25% elevation in metabolism 15 hours participants should have a stable weight, and food
after cessation of 2 hours of strenuous football. intake and exercise should be controlled. It is also
Also, Passmore and Johnson
[12]
found a 15% in- advisable for study participants to sleep overnight in
crease in RMR for 7 hours after a 16km walk at 6.4 the laboratory before a study to avoid exercise in the
km/hour in three males, and deVries and Gray
[13]
morning; however, an outpatient protocol may give
found a 10% increase in RMR for 6 hours after 1 no different values than an inpatient protocol when
hour of mixed aerobic exercise. However, in many the conditions are controlled and the study partici-
cases, the intensity and duration of exercise was not pants are transported to the laboratory.
[25,26]
Also,
quantified in these early studies, and they provided habituation of the study participants to testing proce-
minimal information about the controls. Also, they dures is of utmost importance. The experimental
did not account for other factors that may influence conditions both before and during measurements
RMR, such as time of day, prior uncontrolled exer- need to be strictly controlled. For female study
cise, food, temperature, caffeine intake, habituation participants, it may also be necessary to control for
and stress. menstrual cycle differences.
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1040 Børsheim & Bahr
When reviewing even the newer EPOC literature, tilatory threshold,
[21]
and beyond 40 minutes after 4
it is a problem that the methods for measuring
× 20 minutes of cycle ergometry at 35–55% of
baseline and EPOC duration are inconsistent among
˙
VO
2max
.
[22]
Only two males and two females took
investigations. In some studies, a separate control
part in the last study. Brehm and Gutin
[23]
found a
study has been used to control for time effects,
relationship between EPOC and the intensity of
whereas others have used only one pre-exercise
walking/running, but their intensity was still low,
value as baseline. In many studies, only 30 minutes
the highest being 11.3 km/hour in trained individu-
of rest in the morning has been used and the
˙
VO
2
als. After 3.2km running at this intensity, EPOC
during the final 10 minutes of this has been taken as
amounted to only 71kJ (~3.5L oxygen). Elliot et
baseline for EPOC. This can lead to falsely high
al.
[24]
also found a short lasting (<30 minutes) EPOC
˙
VO
2
baseline values in the morning, since a certain
after 10–30 minutes cycling at 80% of
˙
VO
2max
.
increase because of anticipation may be expected,
Finally, Maresh et al.
[27]
had male study participants
which subsequently leads to an underestimation of
cycle at 60% and 70% of
˙
VO
2max
, but only for 20
EPOC.
and 30 minutes, and found an EPOC duration less
In some cases, the baseline and recovery data
than 40 minutes.
have been collected with the individuals in a seated
Several studies on EPOC have their origin in the
position, in others in a recumbent position. The
late Lars Hermansen’s laboratory. He started the
RMR is lower in recumbent position, probably be-
series of studies himself by having a male study
cause it is difficult to avoid fidgeting and relax
participant cycle for 80 minutes at 75% of
completely when sitting for an extended period.
˙
VO
2max
.
[30]
After 12 hours, the size of EPOC was
This results in a greater measurement error and a
48L. After 24 hours of recovery,
˙
VO
2
was still
reduced ability to detect differences between the
increased by 5.9% compared with a control day.
baseline and recovery conditions.
Mæhlum et al.
[15]
from the same laboratory, found a
Different methods have also been used to deter-
mean EPOC of 26L after 80 minutes cycling (in
mine when
˙
VO
2
has returned to resting levels. Some
periods of 10–30 minutes with 5-minute breaks be-
have measured
˙
VO
2
continuously, whereas others
tween) at an intensity of 70% of
˙
VO
2max
using eight
have measured at discrete time points. Furthermore,
study participants. In accordance with Hermansen et
some have measured
˙
VO
2
until it has returned to
al.’s case study,
[30]
they found that
˙
VO
2
was still
resting values, others only for a pre-determined time
increased by 5% as late as 24 hours after the end of
period.
exercise. We followed up with a series of exercise
Finally, because of inter-individual variability in
studies spanning from 20–80 minutes of cycling at
EPOC, it is important with a high enough number of
intensities ranging from 30–75% of
˙
VO
2max
[16,43]
. A
study participants to be able to detect differences.
clear relationship was found between the magnitude
of EPOC and both the intensity and duration of
4. Effect of Intensity and Duration of
exercise. A curvilinear relationship between the
Aerobic Exercise on EPOC
magnitude of EPOC and the intensity of the exercise
bout was observed (figure 2). No prolonged increase
Table I contains a review of studies on
˙
VO
2
after
in recovery
˙
VO
2
was found after 80 minutes at 29%
aerobic exercise. The absence of a sustained in-
of
˙
VO
2max
. Instead, it appeared that an intensity
crease in
˙
VO
2
after exercise seems to be a consistent
above 50–60% of
˙
VO
2max
was required in order to
finding in studies with low exercise intensity and/or
induce an EPOC lasting for several hours after exer-
low exercise duration. No EPOC was found beyond
cise. At exercise intensities at or above this level, a
35 minutes of recovery after 5 or 20 minutes cycling
linear relationship between the magnitude of EPOC
at 50%, 65%, and 80% of maximal oxygen uptake
and the duration of the exercise bout was observed
(
˙
VO
2max
),
[20]
beyond 40 minutes of recovery after
(figure 2). The size of EPOC was 11.1, 14.7 and
20–40 minutes of treadmill exercise around the ven-
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1041
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)
Table I. Studies on the effect of aerobic exercise on post-exercise
˙
VO
2
a
Study Year Study Exercise mode, duration and intensity EPOC size and duration Comments
c
participants
b
Passmore & Johnson
[12]
1960 3M TM: 16km, 6.4 km/h RMR 14–18%↑ for 6h, >7h Rest position not reported. NC
deVries & Gray
[13]
1963 2M Cycling, bench step, run/walk: 45 min ~1.9L (57 kcal) NC
(effective time 25 min)
Knuttgen
[28]
1970 5F, 7M Cycling: 45–98%, 15–55 min ≤5L, duration NA Bed rest. NC. EPOC magnitude
related to intensity and duration
Segal & Brooks
[29]
1979 11M Cycling: 55 and 95%, 2 min ≤4L, duration NA Seated. NC. EPOC magnitude
related to intensity
Hagberg et al.
[20]
1980 18M Cycling: 50%, 65% and 80%, 5 and 20 ≤5L, duration NA Moderate cycling (used as
min base-line). NC. EPOC
measured for 15 min
Hermansen et al.
[30]
1984 1M Cycling: 75%, 80 min 12h EPOC: 48L. At 24h:
˙
VO
2
↑ Bed rest. C
5.9%
Bielinski et al.
[14]
1985 10M (T) TM: 50%, 3h RMR↑ 9% for 4.5h. At 18h: Seated/respiratory chamber. C.
RMR↑ 4.7% Food given 30 min post-
exercise
Pacy et al.
[22]
1985 2F, 2M (T) Cycling: 35–55% for 20 min × 4 (40 <2.2L, <60 min Supine. C
min break)
Freedman-Akabas et 1985 13F, 10M (UT TM: Anaerobic threshold, 20 min <4.2L, <40 min Recumbent. C
al.
[21]
+ T)
Chad and Wenger
[31]
1985 (a) 3F, 3M; (b) Cycling: (a) 70%, 15 and 30 min; (b) Magnitude and duration not Seated. C. (a) EPOC↑ after 30
6M 70% for 30 min and 50% for 38 min given vs 15 min; (b) EPOC↑ after
(equal work) 50% vs 70%
Brehm & Gutin
[23]
1986 8F, 8M (UT + TM: 18–68%, 17–60 min. Incremental 1h EPOC: ≤8L, >1h after max Seated. NC. EPOC measured
T) max test test for 1h. EPOC curvilinear related
to intensity
Devlin & Horton
[32]
1986 6M Cycling: 85%, 71 min. Intermittent At 12–16h: RMR↑ 3–7% Rest position not reported. C
Mæhlum et al.
[15]
1986 4F, 4M Cycling: 70%, 80 min (10–30 min bouts 12h EPOC: 26L. At 24h: RMR↑ Bed rest. C
with 5 min break) 5%
Bahr et al.
[16]
1987 6M Cycling: 70%, 20, 40 and 76 min (20
˙
VO
2
↑ 5.1%, 6.8%, 14.4% over Bed rest. C. EPOC magnitude
min bouts with 5 min break) 12h. 76 min: 12h EPOC: 31.9L, linearly related to duration
≥12h
Elliot et al.
[24]
1988 3F, 3M Cycling: 80%, 10 and 30 min ~2.37L (11.4 kcal), 30 min. Supine. C
Chad and Wenger
[17]
1988 3F, 3M Cycling: 70%, 30, 45, 60 min 6.6, 14.9, 33L. 128, 204, 455 Seated. NC
min
Sedlock et al.
[33]
1989 10M (T) Cycling: (a) 75%, 20 min; (b) 50%, 30 (a) ~6.2L (29.4 kcal); (b) ~3.0L Seated. NC
min; (c) 50%, 60 min (14.3 kcal); (c) 2.5L (12.1 kcal)
Chad and Quigley
[34]
1989 5F TM: 55%, 90 min EPOC magnitude NA. At 1h: Seated. NC. EPOC measured
˙
VO
2
↑ 54%. >1h for 1h
Continued next pag
e

1042 Børsheim & Bahr
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)
Table I. Contd
Study Year Study Exercise mode, duration and intensity EPOC size and duration Comments
c
participants
b
Poehlman et al.
[35]
1989 6M Cycling: 50%, 90 min <24h C. RMR measured at 24 and
48h
Kaminsky et al.
[36]
1990 6F TM: 70%, 50 min vs 2 × 25 min 1.4L (continuous), <30 min. Seated. NC
3.1L (split), <30 min
Gore & Withers
[18,19]
1990 9M TM: 30%, 50% and 70%. 20, 50 and 80 8h EPOC: 30%: 1.0L, 1.4L, Bed rest. C. EPOC magnitude
min 1.0L, all <1h; 50%: 3.1L (<1h), linearly related to duration
5.2L (<1h), 6.1L (<2h); 70%: when intensity >50%. Intensity
5.7L (<1h); 10.0L (<4h), 14.6L major determinant of EPOC
(<8h)
Bahr et al.
[37]
1990 6M Cycling: 51%, 120 min 3.5h EPOC: 7.8L, >3.5h Bed rest. C. EPOC measured
for 3.5h
Chad and Quigley
[38]
1991 10F (5UT + 5T) Cycling: 50% and 70%, 30 min Unclear magnitude, but 50%↑ Seated. NC. EPOC measured
vs 70%, >3h for 3h
Sedlock
[39]
1991 4F, 4M Arm crank + cycling: 60% mode- Arm: ~1.9L (9.2 kcal), 23 min. Seated. NC
specific, 20 min Cycle: ~2.2L (10.4 kcal), 24 min
Sedlock
[40]
1991 7F Cycling: 40% and 60% until 850kJ 40%: ~6.2L (30kJ), 28 min. Seated. NC
60%: ~7.5L (36kJ), 18 min
Berg
[41]
1991 5F, 5M (T) TM: 40%, 30 min M: ~1.4L (6.7 kcal), ~1h. F: Seated. C. EPOC measured for
~2.2L (10.6 kcal), ~1h 1h
Withers et al.
[42]
1991 8M (T) TM: 70%, 164 min 24h EPOC: 32.4L, >8h, <24h Supine. C
Bahr & Sejersted
[43]
1991 6M Cycling: 29%, 50%, 75%. 80 min (20 1.3, 5.7, 30.1L; 0.3, 3.3, 10.5h Bed rest. C. EPOC curvilinear
min bouts with 5 min breaks) related to intensity
Bahr & Sejersted
[44]
1991 6M Cycling: 75%, 80 min 7h EPOC: 20.9L (fasted), 21.1L Bed rest. C
(fed), >7h
Bahr et al.
[45]
1992 6M Cycling: 108%, 1, 2, or 3 × 2 min with 3 5.6, 6.7, 16.3L; 30, 60 min, 4h Bed rest. C
min breaks
Elliot et al.
[46]
1992 5F, 4M Cycling: 80% of max heart rate. 40 min ~6.7L (32 kcal), <30 min Supine. C
Sedlock
[47]
1992 7M Cycling & TM: 60–65% of mode- Cycling: ~3.1L (15 kcal), 31 Seated. NC
specific, 30 min min; TM: ~3.5L (17 kcal), 34
min
Maresh et al.
[27]
1992 8M Cycling: 60% and 70%. 20 and 40 min Magnitude NA, <40 min Seated. NC. NS between
sessions
Donelly & 1992 6F Cycling: 55%, 90 min Magnitude NA, >60 min Rest position not reported. NC
McNaughton
[48]
Brockman et al.
[49]
1993 5F (T) TM: (a) 25%, 2h; (b) 81%, 10 min; (c) 1h mean
˙
VO
2
↑: (a) 12%, 40 Seated. C. EPOC measured for
89%, 7 × 2 min with 2 min break min; (b) 23%, >1h; (c) 44%, 1h: intermittent run >
>1h continuous. Run > walk
Kaminsky & Whaley
[50]
1993 10F (5 obese + TM: (a) Alternating 3 min bouts at 30% (a) ~3.6L (17.4 kcal), 37.5 min; Seated. NC. NS between
5 lean) and 90%, 36 min; (b) 60%, 36 min (b) ~1.9L (9.0 kcal), 16.5 min obese and lean
Continued next pag
e

EPOC and Exercise Intensity and Duration 1043
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)
Table I. Contd
Study Year Study Exercise mode, duration and intensity EPOC size and duration Comments
c
participants
b
Smith & McNaughton
[51]
1993 8F, 8M Cycling: 40%, 50%, 70%, 30 min M: 16.3, 22.1, 28.1L for 31.2, Rest position not reported. NC.
42.1, 47.6 min. F: 12.1, 20.8, EPOC duration increased with
24.3L for 26.9, 35.6, 39.1 min intensity
Neary et al.
[52]
1993 7M (T) Swimming & cycling: 65% of mode- Swim: 5.3, 5.6, 5.6L; 8.1, 10.1, Seated. NC. Relationship
specific, 30, 45, 60 min 9.4 min. Cycling: 8.2, 9.9, between core temperature and
10.0L; 18.3, 20.4, 22.9 min EPOC
Frey et al.
[53]
1993 13F (7UT + 6T) Cycling: 65% and 80%. 24–45 min UT: 4.0, 5.9L, >1h. T: 4.7L, 50 Seated. NC
(until 300 kcal) min; 5.6L, 40 min
Børsheim et al.
[54]
1994 6M Cycling: 78%, 60 min 6.5h EPOC: 14.4L, ≥6.5h Bed rest. C. EPOC measured
for 6.5h
Thomas et al.
[55]
1994 7M Cycling, jog and downhill jog: 60%, 60 Magnitude NA, <9h Semi-reclined. C. EPOC
min measured for 2h, and again at
9, 24 and 48h. NS between
modes
Sedlock
[56]
1994 10M (5UT + Cycling: 50%, ~300 kcal UT: ~2.5L (12.2 kcal), 20.4 min. Seated. NC
5T) T: ~2.5L (12.2 kcal), 16.6 min
Gilette et al.
[57]
1994 10M Cycling: 52%, 64 min ~5.6L (27 kcal), 5h Supine. C
Quinn et al.
[58]
1994 8F (T) TM: 70%. 20, 40, 60 min 3h EPOC: 8.6, 9.8, 15.2L; >3h Seated. C. EPOC measured for
3h
Harms et al.
[59]
1995 16M TM: 70%. 20 min Magnitude unclear, <30 min Supine. NC. EPOC negatively
related with body fat mass
Dawson et al.
[60]
1996 8F Cycling: (a) 67%, 34 min; (b) 55%, 41 (a) 3.6L, 13.9 min; (b) 2.6L, Bed rest. NC. Intensity effect
min; (c) 45%, 49 min (same work) 14.1 min; (c) 2.4L, 13.2 min on EPOC magnitude
Short et al.
[61]
1996 5M, 5F Arm crank ergometer: (a) 35%, 15 min; (a) 0.46L, 5.7 min; (b) 0.5L, 5.5 Seated. NC. Intensity more
(b) 35%, 30 min; (c) 70%, 15 min min; (c) 1.4L, 14 min important than duration
Trost et al.
[62]
1997 5M (T) Cycling: 65%, 60 min 1h EPOC: 5.5L, >1h Seated. NC
Phelain et al.
[63]
1997 8F (T) Cycling: (a) 50%, 78 min; (b) 75%, 51 3h EPOC: (a) 4.8L, <3h; (b) Seated + bed rest. C. EPOC
min (500 kcal) 9.0L, >3h measured for 3h
Laforgia et al.
[64]
1997 8M (T) TM: (a) 70%, 30 min; (b) 105%, 20 × 1 (a) 6.9L, 1h; (b) 15L, 8h Bed rest. C
min with 2 min break (equal work)
Short & Sedlock
[65]
1997 11F, 11M Cycling: (a) 70%, 30 min; (b) 1.5L (a) UT: 3.5L, 50 min; T: 3.2L, Seated. NC. EPOC duration
(10UT + 12T) oxygen/min, 30 min 40 min; (b) UT: 2.4L, 39 min; T: shorter in T
1.5L, 21 min
Burleson et al.
[66]
1998 15M TM: 45%, 27 min ~3.4L, <30 min Supine. NC
Almuzaini et al.
[67]
1998 10M Cycling: 70%, 30 vs 2 × 15 min 5.3L (cont); 7.4L (split), Supine. C. EPOC measured for
duration NA 40/2 × 20 min
Børsheim et al.
[68]
1998 7M Cycling: 58%, 90 min 4.5h EPOC: 7.5L, >4.5h Bed rest. C. EPOC measured
for 4.5h
Continued next pag
e

1044 Børsheim & Bahr
31.9L of oxygen after 20, 40 and 76 minutes at 70%
of
˙
VO
2max
, respectively. After the longest bout,
EPOC duration was at least 10 hours.
Similar comprehensive studies have been done
by Gore and Withers.
[18,19]
Their treatments ranged
from a 20-minute walk at 30% of
˙
VO
2max
to an
80-minute run at 70% of
˙
VO
2max
. The maximal
EPOC was 14.6L oxygen (297kJ) after the longest
and hardest bout. In accordance with our stud-
ies,
[16,43]
they found an exponential relationship be-
tween exercise intensity and the magnitude of
EPOC. They calculated that the intensity explained
five times more of EPOC than either exercise dura-
tion or total work completed.
[18]
While intensity
accounted for 45.5% of the variation in EPOC, the
duration of exercise, and the interaction between
intensity and duration accounted only for 8.9% and
7.7%, respectively.
In our studies,
[15,16,43]
we found much higher
EPOC values, and also more prolonged duration of
EPOC, compared with Gore and Withers.
[18,19]
In the
study by Mæhlum et al.,
[15]
the response was about
twice the response of that observed in the study by
Gore and Withers.
[18]
Even though the measurement
period was 4 hours shorter in the latter study, this
can still not explain the difference. While we used
cycling in our experiments, Gore and Withers used
running, and the different exercise modes may be a
possible explanation of the discrepancy in results.
Another possible explanation, perhaps more likely,
is the different training status of the study partici-
pants. In the studies by Gore and Withers,
[18,19]
the
participants were well trained, whereas in our stud-
ies they were not.
In support of a significant EPOC after high-
intensity exercise, a 3–7% increase in energy expen-
diture was found as late as 12–16 hours after inter-
mittent exercise for a total of 71 minutes at 85% of
˙
VO
2max
.
[32]
The most exhaustive bout for any of the
EPOC studies was a 35km road run (intensity about
70% of
˙
VO
2max
for about 160 minutes).
[42]
A total
EPOC of 32.4L of oxygen, lasting for a total of 8
hours, was found. Furthermore, an EPOC lasting for
3 hours was observed after only 30 minutes tread-
mill walking at 70% of
˙
VO
2max
in young trained
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)
Table I. Contd
Study Year Study Exercise mode, duration and intensity EPOC size and duration Comments
c
participants
b
Børsheim et al.
[69]
1998 8M Cycling: 56%, 90 min 4.5h EPOC: 8.1L, >4.5h Bed rest. C. EPOC measured
for 4.5h
Matsuo et al.
[70]
1999 7F Cycling: 60%, 60 min 6h EPOC: 8.4L (follicular); Rest position not reported. C.
12.3L (luteal), duration NA EPOC measured for 6h
Lee et al.
[71]
1999 10M TM: (a) 40%, 40 min; (b) 85%, 20 min Magnitude NA. (a) 10 min; (b) Supine. NC
30 min
Fukuba et al.
[72]
2000 5F Cycling: 70%, 60 min 7h EPOC: 8.6L (follicular); 8.9L Bed rest. C
(luteal), duration NA
a All studies are in young people. Exercise is shown in percentage of
˙
VO
2max
or
˙
VO
2peak
. EPOC size is given in litres of oxygen.
b If nothing is mentioned, individuals were untrained/medium trained.
c Individuals’ position during measurements of resting
˙
VO
2
is given first.
C = separate rest control experiment; cont = continuous; EPOC = excess post-exercise oxygen consumption; F = females; M = males; max = maximum; NA = data not available
for estimate; NC = no rest control experiment, but pre-exercise values used as baseline instead; NS = not significant; RMR = resting metabolic rate; T = trained; TM = treadmill;
UT = untrained;
˙
VO
2
= oxygen uptake;
˙
VO
2max
= maximal oxygen uptake;
˙
VO
2peak
= peak oxygen uptake; ↑ indicates increase.

EPOC and Exercise Intensity and Duration 1045
Exercise duration (min)
0204060
80
0
10
−10
20
30
40
Exercise intensity (%)
0 20406080
EPOC (L)
0
10
20
30
40
50
ab
Fig. 2. (a) Plot of excess post-exercise oxygen consumption (EPOC) magnitude versus exercise intensity (constant duration of 80 minutes).
(b) Plot of EPOC magnitude versus exercise duration (constant intensity of 70% of maximal oxygen uptake). Different symbols are used for
individual stud
y
participants
(
reproduced from Bahr,
[7]
with permission
)
.
women.
[58]
Also, a comparison of interval-type exer- protocols have supported the conclusion that exer-
cise intensity is more important than total work for
cise alternating between 30% and 90% of
˙
VO
2max
,
EPOC magnitude.
[51,60]
with continuous exercise of equal duration (36 min-
utes) at 60% of
˙
VO
2max
(equal work) showed a
There are also some studies of the effect of
longer EPOC (38 versus 17 minutes) and higher
intensity on EPOC, where EPOC has been measured
EPOC magnitude (~3.6 versus ~1.9L oxygen) after
for a specific pre-determined time period. Phelain et
the interval exercise.
[50]
al.
[63]
found a significantly higher 3-hour EPOC
after cycling at 75% versus 50% of
˙
VO
2max
(equal
Sedlock et al.
[33]
also investigated the effect of
work, 500 kcal) in women. Brockman et al.
[49]
inves-
exercise intensity and duration on EPOC, but used a
tigated 1-hour EPOC after intense intermittent run-
more narrow spectrum of exercise conditions. In one
ning (7 × 2 minutes at 90% of
˙
VO
2max
, 2 minutes of
series of experiments, caloric output was kept con-
rest between intervals), continuous running for 10
stant (300 kcal), whereas intensity was varied (50%
minutes at 81% of
˙
VO
2max
, and continuous walking
versus 75% of
˙
VO
2max
). It took only 30 and 20
for 2 hours at 24.5% of
˙
VO
2max
. EPOC was higher
minutes, respectively, for the trained male study
after running versus walking and highest after the
participants to finish these work bouts. In another
highest intensity running. However, since EPOC
series, intensity was kept constant (50% of
˙
VO
2max
),
was still increased after 1 hour, no definite conclu-
whereas duration was varied (30 versus 60 minutes).
sion can be drawn.
Hence, no strenuous exercise was performed, and no
substantial EPOC duration or magnitude was found.
In contrast to the findings of an intensity-related
Still, it was concluded that the intensity of exercise
EPOC, Chad and Wenger
[31]
did not find any rela-
influenced both the magnitude and duration of
tionship between exercise intensity and EPOC. They
EPOC, whereas exercise duration only influenced
had one group of individuals cycling for 15 and 30
EPOC duration. It was also concluded that the dura-
minutes at 70% of
˙
VO
2max
. Another group cycled
tion of EPOC and the subsequent caloric output was
for 38 minutes at 50% of
˙
VO
2max
and 30 minutes at
not necessarily related, since high-intensity exercise
70% of
˙
VO
2max
(equal total work). They found that
of short duration produced a higher EPOC (~6L
EPOC was highest after the longest duration in both
oxygen) than lower intensity exercise of long dura-
groups. In another study, Chad and Quigley
[38]
had
tion (~2.5L oxygen), even though no difference in
trained and untrained women cycle for 30 minutes at
EPOC duration (33 versus 28 minutes, respectively)
50% and 70% of
˙
VO
2max
, and found a higher EPOC
was found. Similar studies with modest exercise
after the lowest intensity bout. The results of these
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1046 Børsheim & Bahr
0
10
20
30
40
50
EPOC (L)
0
20
40
60
80
100
120
140
Exercise duration (min)
0
20
40
60
80
100
120
Exercise intensity
(% of V
O
2max
or VO
2peak
)
•
•
Fig. 3. Relationship between exercise intensity, exercise duration and excess post-exercise oxygen consumption (EPOC) magnitude. Plot
shows mean EPOC values from studies that have used cycling exercise, and where EPOC values are reported or possible to esti-
mate.
[15-17,22,24,30,33,39,40,43-47,51-54,56,57,60,63,65,67-70,72]
Note that recovery oxygen uptake was not measured until it had returned to resting control
values in all the included studies (for more details on studies, see table I).
˙
VO
2max
= maximal oxygen uptake;
˙
VO
2peak
= peak oxygen
uptake.
two studies are somewhat puzzling when compared was increased from 30 minutes to 45 and 60 min-
with other findings. The authors explain the results utes, respectively. After the most exhaustive bout,
with a lower respiratory exchange ratio (R-value) EPOC lasted for 7.5 hours. Sedlock,
[40]
on the other
after the low intensity bout. It should also be men- hand, did not find any effect of exercise duration on
tioned that post-exercise
˙
VO
2
was still elevated at EPOC, but the intensities used (40% versus 60% of
the end of the experiments 3 hours post-exercise in
˙
VO
2max
) were low.
the second study, whereas EPOC duration is unclear
The interaction between exercise intensity and
in the first study. Also, the experimental conditions
duration is not completely understood, and it is
were less rigorously controlled (menstrual cycle
difficult to separate the effect of each of these fac-
does not appear to be controlled for in the female
tors. However, the fact that exercise has to be of a
participants, participants were sitting during rest and
certain intensity for the linear relationship between
were allowed to write, watch television and read, the
exercise duration and EPOC magnitude to become
time of the day for a repeated trial appears to be
apparent, indicates that the interaction between them
slightly different, and
˙
VO
2
was recorded in short
is synergistic rather than additive. This is illustrated
periods).
in figure 3, which shows mean EPOC values from
Chad and Wenger
[17]
also investigated the effect studies that have used cycling exercise, and where
of duration (30, 45 and 60 minutes) on EPOC after EPOC magnitude is presented or can be estimat-
cycling at 70% of
˙
VO
2max
. They found that EPOC ed.
[15-17,22,24,30,33,39,40,43-47,51-54,56,57,60,63,65,67-70,72]
It
increased 2.3- and 5.3-fold when exercise duration should be noted that some of the studies included
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1047
probably did not capture the entire EPOC, since ≥70% of
˙
VO
2max
), the corresponding values can be
estimated to approximately 700kJ after the exercise
post-exercise
˙
VO
2
was still elevated compared with
bout, and about 109 200kJ per year (training three
resting control values at the end of the measurement
times per week), equivalent to about 2.9kg of fat.
period (for further details on these studies see table
However, such strenuous exercise cannot be expec-
I). Gore and Withers
[19]
analysed the interaction
ted to be undertaken by overweight or untrained
effect between exercise intensity and duration on
individuals. Although overweight may be the result
EPOC after treadmill exercise ranging from 20 min-
of a small positive energy balance over a long time,
utes at 30% of
˙
VO
2max
to 80 minutes at 70% of
and EPOC may contribute to the opposite when
˙
VO
2max
. They found that the interaction between
strenuous exercise is undertaken regularly, it must
exercise intensity and duration accounted for a much
be concluded that EPOC is negligible in relation to
smaller part (7.7%) of the variation in EPOC, com-
weight loss in the overweight or obese person. It
pared with exercise intensity alone (45.5%). The
must be noted that exercise per se has an important
studies illustrated in figure 3 appear to support the
role for weight regulation. Furthermore, EPOC may
hypothesis that exercise intensity has to be of a
have greater implications for elite athletes, who
certain size to achieve a significant EPOC. In addi-
often have two bouts of exercise on the same day as
tion, they show that the highest EPOC values have
a normal training routine. If
˙
VO
2
is elevated from a
been found when both exercise intensity and dura-
previous exercise bout, the mechanical efficiency
tion are substantial.
may be reduced during the following bout.
[74]
It appears that the post-exercise increase in ener-
In summary, the magnitude of EPOC after aero-
gy expenditure per se after exercise bouts spanning
bic exercise is clearly dependent on both the dura-
from 20 minutes at 30%
˙
VO
2max
to 80 minutes at
tion and intensity of exercise. There is a curvilinear
70%
˙
VO
2max
may be negligible in relation to energy
relationship between the magnitude of EPOC and
balance and weight loss. However, this may be a
the intensity of the exercise bout. At least for cy-
speculative conclusion, since there are no available
cling, it appears that an intensity above 50–60% of
data from prolonged studies of the effect of EPOC
˙
VO
2max
is required in order to induce an EPOC
on energy balance. Also, it should be noted that
lasting for several hours. The relationship between
there may be a high inter-individual variability in
exercise duration and EPOC magnitude appears to
EPOC in response to the same relative exercise
be more linear, especially at higher intensities.
stimulus, as can be seen from figure 2. Thus, it can
be hypothesised that there are high, medium and low
5. Effect of Split Exercise Sessions
responders, corresponding to what has been found
on EPOC
for other parameters in response to exercise train-
ing.
[73]
However, there are no studies where EPOC
A couple of studies have shown a higher EPOC
has been measured repeatedly in the same individu-
after split exercise sessions compared with a contin-
als under the same conditions. Therefore, it is not
uous bout. Kaminsky et al.
[36]
investigated women
known if the observed variability is the result of
during 50 minutes of continuous running versus two
biological variation related to differences in, for
25-minute sessions at 70% of
˙
VO
2max
. The com-
example, body composition or training status, or
bined EPOC after the two split sessions correspond-
whether it is caused by measurement errors. Still, if
ed to ~3.1L versus 1.4L of oxygen after the continu-
we estimate that EPOC amounts to 50–100kJ after
ous session. Almuzaini et al.
[67]
compared EPOC
moderate exercise (≥1 hour, approximately 50% of
after 30 minutes of continuous cycling versus two
˙
VO
2max
), this would result in an extra energy loss of
15-minute sessions (separated by 6 hours) at 70% of
about 11 700kJ per year if training is undertaken
˙
VO
2max
. In this study, the sum of EPOC after the
three times per week. This represents only about
split sessions was 7.4L versus 5.3L of oxygen after
311g of fat. After more strenuous exercise (≥1 hour,
the continuous bout. Hence, total EPOC was ap-
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1048 Børsheim & Bahr
proximately 120% and 40% greater after split ses-
˙
VO
2max
, very similar 1-hour EPOCs were found;
sions compared with a continuous session in these 7.6L (submaximal exercise) and 7.8L (supramax-
two studies. It should be noted that even though imal exercise), respectively. Of this, about 4.5L
EPOC is higher after split sessions, the extra EPOC (submaximal) and 2.0L (supramaximal) could be
is small in relation to the exercise energy expendi- attributed to the prolonged curve component, and
ture. Thus, prolonging the exercise bout by a few about 3.1L (submaximal) and 5.8L (supramaximal)
minutes may make up for the increase in EPOC after to the rapid component. Hence, it appears that high-
splitting the session. Furthermore, in split sessions, intensity short exercise affects mainly the rapid
there will be one oxygen deficit for each session. EPOC component, whereas more prolonged ex-
This means that the relative difference in energy hausting exercise stimulates mechanisms also pre-
expenditure during recovery after split versus con- sent beyond the first hour of recovery. However, this
tinuous sessions will be smaller than the difference relationship remains to be fully elucidated.
in EPOC. This was not taken into account in the two
studies mentioned.
[36,67]
7. Effect of Aerobic Exercise Mode
on EPOC
6. Effect of Supramaximal Exercise
A curvilinear relationship between exercise in-
on EPOC
tensity and EPOC, and a linear relationship between
Given the relationship between exercise intensity exercise duration (when intensity is higher than the
and EPOC, it comes as no surprise that supramax- break-point) are both shown for cycle exercise
[16,43]
imal exercise stimulates EPOC. We found that brief and for treadmill exercise.
[18,19]
However, the abso-
intermittent bouts of exhaustive supramaximal exer- lute EPOC magnitude for a certain exercise intensity
cise (1, 2, or 3 × 2-minute bouts of cycling at 108% and duration may differ depending on exercise
of
˙
VO
2max
, separated by 3-minute rest periods) ele- mode.
vated post-exercise energy expenditure for 4
Muscle damage is more likely to occur after
hours.
[45]
eccentric-type exercise than after concentric exer-
Laforgia et al.
[64]
measured EPOC after cise,
[75-77]
and it is possible that this influences
supramaximal (20 × 1 minute runs at 105% of EPOC. However, in studies where concentric and
˙
VO
2max
with 2-minute recovery periods between) eccentric exercise have been compared, no differ-
versus submaximal running (30 minutes at 70% of ences have been found in EPOC
[47,55]
or RMR for
˙
VO
2max
). A significantly higher 9-hour EPOC was several days after exercise.
[78]
Sedlock
[47]
compared
observed after the supramaximal exercise. Interest- 30 minutes of cycling and treadmill exercise at an
ingly, the EPOC values in this study were similar to intensity of 60–65% of mode-specific peak oxygen
the values found after 80 minutes of running at 70% uptake (
˙
VO
2peak
). No differences could be detected
˙
VO
2max
in the study by Gore and Withers,
[18,19]
even between modes, but the exercise was moderate, and
though more than twice the total work was per- hence EPOC was small (15–17 kcal; ~3.5L oxygen).
formed in the latter study. EPOC has also been compared after 60 minutes of
either jogging, downhill jogging or cycling at 60%
Few authors have tried to divide EPOC into
of mode-specific
˙
VO
2max
to induce different de-
different components. This is unfortunate, since ex-
grees of eccentric muscular activity.
[55]
No increase
ercise intensity and duration may affect the short
in recovery energy expenditure was found after ec-
and prolonged components to a different extent. We
centric exercise, but it may be that the detection
calculated the short EPOC component as the differ-
level was not sufficient (power = 0.40 for medium
ence between the observed 1-hour EPOC and the
effect size).
prolonged curve component.
[7]
By comparing EPOC
after 70–80 minutes of cycling at 69–78% of To investigate the effect of relative metabolic
˙
VO
2max
with 3 × 2 minutes at 104–117% of rate of the active musculature on EPOC, Sedlock
[39]
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1049
compared the effect of 20 minutes of arm crank 14.5 hours after the resistance exercise compared
exercise versus cycling at 60% of mode-specific
with a resting control experiment.
˙
VO
2peak
.
˙
VO
2
during arm crank exercise was about
Burleson et al.
[66]
compared weight training exer-
72% of uptake during cycling in absolute terms. No
cise (two circuit sets of eight exercises at 60% of
difference in EPOC was found between modes, but
1RM for 8–12 repetitions) and treadmill exercise
again the exercise was short and not very strenuous,
(27 minutes at 45% of
˙
VO
2max
). The weight training
and consequently EPOC lasted less than 25 minutes.
exercise was performed first, and the average
˙
VO
2
Short et al.
[61]
have investigated the effect of
was used to determine the intensity for the treadmill
intensity and duration of upper body exercise alone
workout. EPOC was found to be higher the first 30
on EPOC, and found a similar pattern as for the
minutes after resistance exercise, but not at 60 and
lower body. EPOC was measured after 15 and 30
90 minutes. Even though the
˙
VO
2
volumes were
minutes of arm crank exercise at 35% of
˙
VO
2peak
,
equated, the resistance exercise was considered to
and after 15 minutes at 70% of
˙
VO
2peak
. The intensi-
be of higher intensity activity than the aerobic exer-
ty was found to have a greater effect than duration,
cise, and this may explain the higher EPOC after
but durations were low and within a narrow spec-
resistance exercise. Hence, it is difficult to compare
trum in this study. Since exercise duration and inten-
resistance exercise to steady-state exercise, since it
sity did not vary over a range of values, it was not
is not easy to precisely quantify the energy cost of
possible to fully describe the relationships with
resistance exercise with indirect calorimetry.
EPOC.
The relationship between the intensity and dura-
In summary, it is not clear whether various
tion of resistance exercise, and the magnitude and
modes of aerobic exercise affect EPOC differently.
duration of EPOC has not been determined. Table II
Further studies with a sufficient exercise challenge
shows a review of studies on the effect of resistance
and adequate statistical power are needed.
exercise on EPOC.
8. Effect of Resistance Exercise on EPOC
Williamson and Kirwan
[89]
and Dolezal et al.
[84]
found that RMR remained elevated for 48 hours
EPOC has also been compared between aerobic
after an acute moderate- to high-intensity bout of
and resistance exercise. Elliot et al.
[46]
compared
resistance exercise. This was hypothesised to be due
aerobic cycling (40 minutes at 80% of maximal
largely to protein turnover and tissue repair. It
heart rate), circuit weight training exercise (4 sets, 8
should be noted that to avoid eccentric muscle work,
exercises, 15 repetitions at 50% of one repetition
only the concentric phase was used in the study of
maximum [1RM]), and heavy resistance exercise (3
Williamson and Kirwan,
[89]
and still they found a
sets, 8 exercises, to exhaustion at 80–90% of 1RM).
prolonged effect on recovery metabolic rate. Their
They found that heavy resistance exercise produced
study was done in 59- to 77-year-old men. Similarly,
the biggest EPOC, but it is unclear how the work
Melby et al.
[81]
found that
˙
VO
2
was still increased by
volumes related to each other.
9.4% and 4.7% as late as 15 hours after two hard
Studies in which similar estimated exercise ener-
resistance exercise work bouts (each 90 minutes of
gy cost
[57]
or similar exercising
˙
VO
2
[66]
have been
weight lifting, six sets of ten exercises, 8–12 repeti-
used to equate continuous aerobic exercise and in-
tions at 70% of 1RM).
termittent resistance exercise, have indicated that
After more moderate resistance exercise (three
resistance exercise produces a greater EPOC res-
sets of seven exercises, ten repetitions at 12RM),
[79]
ponse. A higher EPOC was found after hard resis-
a smaller EPOC was found, but the
˙
VO
2
was mea-
tance exercise (50 sets of 8–12 repetitions at 70% of
sured for only 1 hour after exercise, and was still
1RM, 2 minutes of rest between sets) compared with
elevated at this time point. Binzen et al.
[86]
found an
an equated work bout of aerobic cycling (50% of
EPOC shorter than 2 hours after three sets of ten
˙
VO
2max
for 60 minutes).
[57]
RMR was still elevated
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1050 Børsheim & Bahr
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)
Table II. Studies on the effect of resistance exercise on post-exercise
˙
VO
2
a
Study Year Study Exercise EPOC size and duration Comments
b
participants
Melby et al.
[79]
1992 6M 4 sets of upper and 3 sets of lower body. 7–10 1h EPOC: ~4L (20 kcal), ≥1h Supine. C. EPOC measured for 1h only
reps. 12RM
Elliot et al.
[46]
1992 5F, 4M 8 exercises, upper and lower body. (a) 4 sets, 15 (a) ~10.2L (48 kcal); (b) ~10.6L Supine. C. EPOC duration unclear
reps, 50% of 1RM, 30 sec rest; (b) 3 sets, 3–8 (51 kcal). 30 min
reps, 80–90% of 1RM, 30 sec rest
Murphy & 1992 10M 6 exercises, upper and lower body. Circuit: 3 Circuit: 5L, 20 min. Standard: Rest position not reported. NC.
Schwarzkopf
[80]
times, 8–12 reps, 50% of 1RM, 30 sec rest. 2.7L, 15 min Experiments in the afternoon. 5 min
Standard: 3 sets, reps to exhaustion, 80% of baseline measurement
1RM, 120 sec rest
Melby et al.
[81]
1993 (a) 6M; (b) (a) 6 sets of 10 exercises, upper and lower body, (a) 2h EPOC: 7L, ≥15h; (b) 2h Supine. (a) NC; (b) C. EPOC measured
6M ~8–12 reps, 70% of 1RM, 90 min; (b) as (a), but EPOC: 7L, ≥15h for 2h, minus first 5 min post-exercise.
5 sets, and longer rest, 90 min Measured again at 15h
Olds & 1993 7M 2 sets of 7 exercises, upper and lower body, 3.5 (a) 8.2L, <1h; (b) 6.5L, <1h Seated. C
Abernethy
[82]
min rest. (a) 12 reps, 75% of 1RM; (b) 15 reps,
60% of 1RM
Gilette et al.
[57]
1994 10M 5 sets of 10 exercises, upper and lower body, 5h EPOC: 12.6L (51 kcal), Supine. C. Fed study participants. EPOC
8–12 reps, 70% of 1RM, 2 min rest ≥14.5h measured for 5h and again at 14.5h
Burleson et 1998 15M Circuit, 2 times, 8 exercises, upper and lower Magnitude NA, ≥90 min Supine. NC
al.
[66]
body, 8–12 reps, 60% of 1RM, 1 min rest
Haltom et 1999 7M (a) Circuit, 2 times, 8 exercises, upper and lower 1h EPOC: (a) 10.3L; (b): 7.4L Supine. NC. EPOC measured for 1h.
al.
[83]
body, 20 reps, 75% of 20RM, 20 sec rest; (b) EPOC↑ for first 5 min after (a) vs (b)
same as (a), but 60 sec rest
Dolezal et 2000 18M (9UT 8 sets of leg press, 6 reps, 6RM, 3 min rest At 24h: RMR↑ 18%. At 48h: Supine. NC
al.
[84]
+ 9T) RMR↑ 11%. UT > T
Osterberg & 2000 7F 5 sets of 10 exercises, upper and lower body, At 3h:
˙
VO
2
↑ 13%. At 16h: Supine. NC. EPOC measured for 3h and
Melby
[85]
10–15 reps, 12RM or 70% of 1RM, ~2 min rest RMR↑ 4.2% again at 16h
Binzen et 2001 10F 3 sets of 10 exercises, upper and lower body, 10 6.2L, 1h Seated. C
al.
[86]
reps, 70% of 1RM, 1 min rest
Thornton & 2002 14F 2 sets of 9 exercises, upper and lower body. (a) (a) ~1.1L (5.5 kcal); (b) ~2.3L Seated. C
Potteiger
[87]
15 reps, 45% of 8RM, 1 min rest; (b) 8 reps, (11 kcal). <2h
85% of 8RM, 1 min rest
Schuenke et 2002 7M Circuit, 4 times, 3 exercises, upper and lower Mean RMR↑ ~20% for two Supine. C.
˙
VO
2
measured 3×/day.
al.
[88]
body, 8–12 reps, 10RM, 2 min rest days, ≥38h Exercise in evening. Unclear control
conditions between measurements
a All studies are in young people. In all studies, except UT, the individuals participated in resistance exercise several times per week. EPOC size is given in litres of oxygen.
b Under Comments, individuals’ position during measurements of resting
˙
VO
2
is given first.
C = separate rest control experiment; EPOC = excess post-exercise oxygen consumption; F = females; M = males; NA = data not available for estimate; NC = no rest control
experiment, but pre-exercise values used as baseline instead; reps = repetitions; RM = repetition maximum; RMR = resting metabolic rate; T = trained (resistance exercise several
times per week); UT = untrained;
˙
VO
2
= oxygen uptake; ↑ indicates increase.

EPOC and Exercise Intensity and Duration 1051
exercises with ten repetitions at 10RM (1 minute of multiple sets), weights, sets, repetitions, and length
rest between sets) in resistance-trained women. of rest periods. These factors will all influence the
energy cost of the exercise, but this effect is difficult
Murphy and Swartzkopf
[80]
compared two differ-
to quantify precisely. Furthermore, resistance train-
ent protocols of resistance exercise (three sets of six
ing is similar to interval training and split sessions of
exercises, repetitions to exhaustion at 80% of 1RM
aerobic exercise in that each session or set of resis-
with 120-second rest periods versus three circuit sets
tance exercise will have an EPOC of its own during
of the six same exercises, 10–12 repetitions at 50%
the recovery period between exercises. To deter-
of 1RM with 30-second rest periods). Work volume
mine the total energy expenditure, this has to be
of each session was similar, but intensity (weight
included in the calculations.
lifted per unit of time) was greater for the circuit
All this taken into consideration, it seems likely
exercise. The higher intensity session produced
that EPOC after resistance exercise is influenced by
higher EPOC than the standard set exercise (4.9
the intensity of the exercise, since a more prolonged
versus 2.7L oxygen), but the duration of EPOC was
and substantial EPOC has been found after hard
only 20 minutes.
versus more moderate exercise. More research is
Olds and Abernethy
[82]
compared high- and low-
still needed to elucidate the effect of intensity and
intensity resistance exercise with equated work vol-
duration of resistance exercise on the magnitude and
ume, and found no difference in EPOC. However,
duration of EPOC.
their range of intensities was narrow (12 repetitions
at 75% of 1RM and 15 repetitions at 60% of 1RM),
9. Effect of Training Status on EPOC
and may not have been large enough to elicit a
treatment effect. Accordingly,
˙
VO
2
returned to
Individuals of different fitness levels have been
baseline within 1 hour after exercise. Also, the age
used in EPOC studies, and this may also potentially
range was wide (22–55 years) and there was a high
explain some of the differences observed in magni-
inter-individual difference in EPOC (0.7–27L).
tude and duration of EPOC. The effect of training
In a recent study, Thornton and Potteiger
[87]
in-
status is not easy to study, since comparing groups
vestigated the effect of high- and low-intensity resis-
of different fitness levels at the same absolute exer-
tance exercise of similar work volume on EPOC.
cise intensity level means that trained are working at
The exercise bouts consisted of two sets of nine
a lower relative intensity, which has been shown to
exercises. In the high-intensity bout, eight repeti-
influence EPOC. Furthermore, if trained and un-
tions at 85% of 8RM of each exercise were per-
trained are compared at the same relative exercise
formed, whereas 15 repetitions at 45% of 8RM were
intensity, when total work is equal, the untrained
performed in the low-intensity bout. The rest period
have to work for a longer duration, which may also
between sets was 2 minutes. The duration was 23
influence the results. In other words, there is no
minutes for the high-intensity bout and 26 minutes
study design available to provide a definite answer
for the low-intensity bout, whereas the actual exer-
to whether training status affects EPOC.
cise time was 6.9 minutes (high) and 8.3 minutes
Brehm and Gutin
[23]
found similar values for
(low), respectively.
˙
VO
2
was measured between
EPOC in runners versus non-exercisers following a
0–20, 45–60, and 105–120 minutes post-exercise.
3.2km walk at 6.4 km/hour (i.e. same absolute inten-
The magnitude of EPOC was higher during each
sity), but the exercise intensity may have been too
measurement period after high- versus low-intensity
low to detect a difference. Sedlock
[56]
compared fit
resistance exercise.
and unfit males after cycling at 50% of
˙
VO
2peak
,
Understanding the effect of resistance exercise until 300 kcal were used (27 and 35 minutes to
on EPOC is confounded by the considerable diversi- finish, respectively). No difference in EPOC dura-
ty in the protocols commonly employed in this re- tion or magnitude was found, but again, the exercise
search, for example, type (i.e. circuit training or challenge was low. Finally, in a study of treadmill
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1052 Børsheim & Bahr
exercise at approximately the anaerobic threshold, it more, the magnitude of EPOC was not different
between groups after exercise at the same relative
was concluded that fitness level does not significant-
intensity (3.2L oxygen in trained versus 3.5L oxy-
ly alter magnitude and duration of EPOC, but both
gen in untrained). The trained group had higher
˙
VO
2
the separation of study participants into groups and
at the end of the exercise bout, and if the data were
the exercise intensity were somewhat unclear in that
normalised to percentage change afterwards, the
study.
[21]
trained had a more rapid fall in post-exercise
˙
VO
2
.
In contrast to these studies, Chad and Quigley
[38]
After exercise of similar absolute intensity, the un-
found a higher 3-hour EPOC in trained female cy-
trained group had higher EPOC (2.4L oxygen) ver-
clists versus untrained women after 30 minutes cy-
sus trained (1.5L oxygen). Longitudinal training
cling at 50% or 70% of
˙
VO
2max
. The difference was
studies also support the findings of a faster recovery
most apparent immediately after exercise, which
in
˙
VO
2
in trained individuals.
[90,91]
was explained by higher absolute exercise intensity,
In summary, it appears that trained individuals
and thereby a higher
˙
VO
2
in trained versus un-
have a more rapid return of post-exercise metabol-
trained.
˙
VO
2
was still increased 3 hours after exer-
ism to resting levels when exercising at either the
cise, and no definite EPOC magnitude could be
same relative or same absolute work rate, but studies
determined.
after more strenuous exercise bouts should be done.
Frey et al.
[53]
measured
˙
VO
2
for 1 hour after
cycling at ~65% and ~80% of
˙
VO
2max
until 300 kcal
10. Effect of Sex on EPOC
had been expended in trained and untrained women.
No difference in EPOC magnitude was found after
Sex is also a factor that can potentially influence
cycling at the highest intensity, but EPOC was sig-
EPOC, as well as have implications for study de-
nificantly smaller in the untrained (4.0L oxygen)
sign. Energy expenditure at rest or during exercise
versus the trained group (4.7L oxygen) after the
may vary with menstrual phase,
[92-95]
and this has
lowest intensity. EPOC decreased rapidly during the
not always been taken into consideration when stud-
first 10 minutes post-exercise in both groups. Fol-
ying EPOC. Basal metabolic rate has been shown to
lowing both intensities, the initial EPOC (first 10
be at its lowest level 1 week before ovulation.
[92]
minutes post-exercise) was greater in trained versus
Webb
[96]
found an 8–16% increase in 24-hour ener-
untrained, most likely because of higher
˙
VO
2
during
gy expenditure during the 14-day luteal phase fol-
exercise in this group. While there still was a signif-
lowing ovulation. Accordingly, Matsuo et al.
[70]
icant EPOC at the end of the experiment in the
found a higher EPOC after 60 minutes cycling at
untrained group, post-exercise
˙
VO
2
was elevated for
60% of
˙
VO
2max
in the luteal versus the follicular
only 50 and 40 minutes after low- and high-intensi-
phase in seven healthy women. On the other hand,
ty, respectively, in the trained group. Thus, EPOC
Fukuba et al.
[72]
did not detect any significant effect
duration was shorter in the trained.
of menstrual cycle on EPOC in five young women
after 60 minutes cycling at 70% of
˙
VO
2max
.
Short and Sedlock
[65]
compared EPOC in trained
(
˙
VO
2max
: 53 mL/kg/min) versus untrained males
When comparing EPOC between men and
(37 mL/kg/min) after 30 minutes cycling on both
women, the same question as for trained versus
similar absolute (1.5L oxygen/min) and relative
untrained must be addressed: should EPOC be com-
(70% of
˙
VO
2peak
) intensities. At a
˙
VO
2
of 1.5 L/
pared after absolute or relative workloads? Berg
[41]
min, the trained group exercised at 45% of
˙
VO
2peak
,
measured
˙
VO
2
for 1 hour after 30 minutes of exer-
while the untrained worked at 61% of
˙
VO
2peak
. The
cise at 40% of
˙
VO
2max
in active men and women. A
results showed that the trained group had a signifi-
higher post-exercise energy expenditure was found
cantly shorter duration of EPOC whether compared
among men. However, the calculations are some-
at the same absolute or relative intensity, but in all
what unclear, since the data are not compared with
situations EPOC lasted less than 1 hour. Further-
control resting values, and hence it is the absolute
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1053
levels rather than the increase in energy expenditure ture may contribute; however, the cost of this is low
after exercise that is compared. (<1L oxygen).
[7]
Smith and McNaughton
[51]
compared trained
It has been shown that there is an increase in the
men and women after 30 minutes of exercise at
rate of the energy-requiring triglyceride/fatty acid
40%, 50% and 70% of
˙
VO
2max
. In all three condi-
(TG/FA) cycle after prolonged exhausting exer-
tions, the men had a longer EPOC duration than the
cise.
[37,68,69,98]
In the TG/FA cycle, FA released
women; however, the longest duration observed was
during the process of lipolysis are subsequently re-
only 47 minutes, at least partly because of the short
esterified into TG rather than oxidised. ATP is
exercise duration. In all conditions, EPOC magni-
needed for the re-esterification, which can occur
tude in absolute terms was also higher in men versus
within the adipocyte (intracellular recycling), or the
women, but this difference disappeared when EPOC
FA can be released and re-esterified elsewhere, e.g.
was adjusted for body mass. Furthermore, there
in the liver (extracellular recycling). The TG/FA
were no differences between sexes when EPOC was
cycle is under both hormonal
[99]
and substrate con-
expressed as a percentage of total energy expended.
trol.
[100]
The energy cost associated with the increase
In summary, few studies have been conducted to
in TG/FA cycling can account for a significant part
compare EPOC in men and women. Thus, the sex
of EPOC after prolonged steady-state exercise.
[7]
effect on EPOC is not fully clarified, but controlling
The existence of gluconeogenic-glycolytic sub-
for changes in energy expenditure during the men-
strate cycles have also been demonstrated in vivo in
strual cycle appears to be important in studies with
humans,
[101-104]
but no increase in the rates of these
women.
cycles have been found during recovery from exer-
cise.
[104,105]
11. Possible Mechanisms for the Rapid
A relative shift from carbohydrate to fat as sub-
EPOC Component
strate source is a consistent finding after prolonged
EPOC is the sum of many underlying mechan-
exhausting exercise.
[37,43,68,69]
Since the energy equi-
isms, some of which are still not known. Hence, the
valent of oxygen is lower with fat as the substrate
influence of individual factors on EPOC is via an
compared with carbohydrates (free fatty acids: ~4.7
effect on the underlying mechanisms. Most of the
mol ATP/mol oxygen; glucose: ~5.1 mol ATP/mol
studies on mechanisms causing EPOC are done after
oxygen), part of EPOC can be explained by this
cycling protocols, and little is known about EPOC
substrate shift. The substrate shift after exhaustive
after resistance exercise.
submaximal exercise has been calculated to account
Some of the metabolic processes believed to be
for 10–15% of the observed EPOC.
[7]
responsible for the rapid EPOC component are well
The energy cost of glycogen resynthesis has also
defined: replenishment of oxygen stores in blood
been suggested as a mechanism for the prolonged
and muscle, resynthesis of adenosine triphosphate
EPOC component after aerobic exercise.
[30]
How-
(ATP) and creatine phosphate, lactate removal, and
ever, glycogen resynthesis is low during fasting, and
increased body temperature, circulation and ventila-
we did not find any difference in the magnitude of
tion.
[6,7,43,97]
Thus, the classic oxygen debt hypo-
EPOC in the fed versus the fasted state after 80
thesis is one of several factors explaining the rapid
minutes of cycling at 75% of
˙
VO
2max
.
[44]
Further-
component.
more, on a biochemical basis, it can be argued that
the oxygen cost of glycogen resynthesis should not
12. Possible Mechanisms for the
be included as a part of EPOC.
[7,44]
The rationale for
Prolonged EPOC Component
this is that in situations when food is given, the
The mechanisms for the prolonged EPOC com- energy cost of carbohydrate storage is less after
ponent are less well understood. A sustained in- exercise when the carbohydrates are used for replen-
creased ventilation, circulation and body tempera- ishment of muscle and liver glycogen stores com-
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1054 Børsheim & Bahr
pared with rest where more of the carbohydrates nephrine and norepinephrine are potent stimulators
may be stored as fat. The energy cost of storing
of the energy metabolism.
[118,119]
Their calorigenic
dietary carbohydrates as fat requires 23–24% of the
effect seems to be mediated through the β-adre-
ingested energy, whereas storage as glycogen re-
noceptors.
[120,121]
Secondly, the sympathoadrenal
quires only 5.3%.
[106]
On the other hand, emerging
system is activated during exercise with elevated
data
[107,108]
suggest that hepatic de novo lipogenesis
concentrations of plasma catecholamines as a result.
is quantitatively insignificant under most conditions
During dynamic exercise, the plasma concentrations
of carbohydrate overfeeding.
of catecholamines increase linearly with the exer-
cise duration and exponentially with the exercise
It has been suggested that when food is given in
intensity,
[111,122]
a similar relationship to that observ-
the recovery period there is a synergistic interaction
ed between EPOC and exercise duration and intensi-
of food and exercise on energy expenditure. The
ty, respectively (figure 2). Thirdly, catecholamines
difference in opinion as to whether this interaction
are important regulators of TG/FA cycling and FA
exists
[109,110]
may be caused by the large intra- and
oxidation through stimulation of lipolysis via β-
inter-individual variations in the thermogenic effect
adrenoceptors. Both processes are increased after
of a standard meal. We
[44]
could not detect any
exercise and may account for a significant part of
major interaction effects between food and previous
EPOC.
aerobic exercise on the
˙
VO
2
.
In early studies on the effect of β-adrenoceptor
Several hormones are potential stimulants of en-
blockade on post-exercise
˙
VO
2
, antagonists were
ergy expenditure, including insulin, cortisol, thyroid
administered in dogs before an exercise bout.
[114,123]
hormones, growth hormone, adrenocorticotropic
The results showed that short-term EPOC decreased
hormone (ACTH) and catecholamines. Plasma con-
after administration of the non-selective β-adre-
centrations of growth hormone and ACTH may in-
noceptor antagonist, propranolol. The results have
crease during exercise, but no sustained increase in
been used as an indication of the importance of the
secretion has been found after exercise.
[111,112]
sympathoadrenal system for EPOC, but since
Mæhlum et al.
[15]
found no changes in plasma insu-
propranolol was administered before the start of the
lin and free thyroxine during the recovery period
exercise, the physiological effects of the exercise
after exhausting endurance exercise, and only a tran-
bout were different between the control and the
sient increase in the plasma concentration of cor-
propranolol situation. We have shown that both
tisol. However, in this experiment, food was given
propranolol and also the selective β
1
-adrenoceptor
in the recovery period, which may have influenced
antagonist, atenolol, reduced
˙
VO
2
to a similar extent
the hormonal response. Although we found that
when given intravenously during rest post-exercise
venous plasma insulin concentration rapidly re-
and during rest without previous exercise in
turned to resting levels after aerobic exercise.
[68,69]
humans.
[68]
Hence, there was no effect of β-adre-
we observed a prolonged depression in arterial plas-
noceptor blockade on EPOC after aerobic exercise,
ma insulin concentrations after similar exercise.
[113]
and the results do not support the hypothesis that the
This may be important for the increase in fat
prolonged EPOC component is caused by increased
mobilisation during the recovery period. We also
sympathoadrenal activity. A possible β
3
-adre-
found an increased hormonal response to a repeated
noceptor effect on EPOC could not be excluded, but
bout of endurance exercise for catecholamines, AC-
the β
3
-adrenoceptor does not seem to be of impor-
TH, cortisol and growth hormone.
[112]
tance for the sympathoadrenal-mediated thermogen-
Many authors have suggested that an increased
esis,
[121]
and thus it seems unlikely that there is any
sympathoadrenal activity may be one of the mech-
β
3
-adrenoceptor effect on EPOC.
anisms underlying EPOC, and of the prolonged
An increased sensitivity to catecholamines in the
component in particular.
[6,15,30,37,114-117]
This hypo-
post-exercise period has also been proposed.
[7,124]
thesis was based on several findings. Firstly, epi-
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1055
The hypothesis was built on observations in vitro of expensive, and thus it seems reasonable to speculate
that the energy cost associated with an accelerated
an increased lipolytic response to catecholamine
rate of protein synthesis in the post-exercise state
stimulation in human gluteal adipocytes
[125]
and in
can contribute to higher energy expenditure. Further
suprailiac adipocytes
[126]
removed immediately after
studies on the importance of increased protein turn-
an exercise bout compared with resting samples.
over and adaptive protein synthesis on EPOC after
However, we have shown
[69]
that isoprenaline (β-
different types of exercise are necessary. To our
adrenoceptor agonist) stimulated whole body
˙
VO
2
knowledge, EPOC and protein turnover have not
to the same extent during rest with and without
been measured in the same study.
previous aerobic exercise. Hence, no increased sen-
sitivity to catecholamines was detected in the post-
Both whole body (pulmonary) and muscle
˙
VO
2
exercise period. Also, the lipolytic effect of isopren-
should be measured to determine the quantitative
aline in abdominal adipose tissue was not increased
contribution of muscle metabolism to EPOC. In a
after a prolonged moderate exercise bout, but in-
study by Bangsbo et al.,
[134]
leg
˙
VO
2
accounted for
stead a desensitisation was seen.
[113]
only one-third of EPOC in the 60-minute recovery
period after exhaustive knee extensor exercise last-
Another potential effect of the sympathoadrenal
ing 3 minutes. Only a minor fraction (26%) of the
system on EPOC is by stimulation of various pro-
excess leg
˙
VO
2
could be attributed to the oxygen
cesses during the exercise bout, which are reversed
requirements for resynthesis of substrate. Hence, a
slowly after the end of the exercise bout, even if
large part of the increase in muscle
˙
VO
2
after exer-
there is no increased sympathoadrenal activity in
cise remained to be explained. Non-exercised mus-
this period. During an exercise bout, catecholamines
cles may also contribute to EPOC, as both in vitro
stimulate both the heart rate, the contractility of the
and in vivo studies have shown an increased
˙
VO
2
in
heart, glycogenolysis, gluconeogenesis, and lipoly-
inactive muscles perfused with blood high in lac-
sis in adipose tissue and in muscles.
[127]
The cat-
tate.
[135,136]
echolamines also influence the release of other hor-
mones, e.g. insulin and renin. During exercise,
The energy efficiency may change during exer-
blood flow in some tissues is decreased through α-
cise,
[97,137]
and this may be the case also during
adrenoceptors, which may be of importance for
recovery from exercise. One potential mechanism
blood flow and
˙
VO
2
in the tissues after exercise.
for reduced energy efficiency is the activity of un-
Hence, the influence of catecholamines on various
coupling proteins (UCP). The expression of UCP3
processes during exercise may in turn influence
in various organs is consistent with a role in adaptive
EPOC. This is in agreement with the finding of a
thermogenesis.
[138-140]
It remains to be determined if
reduced EPOC when propranolol was administered
the cellular milieu during exercise stimulates the
in dogs before exercise.
[114,123]
activity of UCP, and if changes in energy efficiency
during and after exercise may explain part of EPOC.
An elevated rate of both protein breakdown and
protein synthesis has been demonstrated in the re-
In summary, whereas several of the metabolic
covery period after exercise.
[128]
An increased whole
processes believed to be responsible for the rapid
body protein synthesis has been demonstrated after
EPOC component are well known (replenishment of
3.75 hours of treadmill running at 50% of
oxygen stores in blood and muscle, resynthesis of
˙
VO
2max
,
[129]
and after 1 hour of cycling at 75% of
ATP and creatine phosphate, lactate removal, and
˙
VO
2max
.
[130]
Also, an increased muscle protein syn-
increased body temperature, circulation and ventila-
thesis has been shown after prolonged exercise,
[131]
tion) the mechanisms underlying the prolonged
whereas studies of the breakdown of muscle protein
component are less well understood. An increased
after such exercise is lacking. Resistance exercise
TG/FA cycling, and a shift from carbohydrate to fat
also stimulates protein synthesis in the post-exercise
as substrate source, may explain a substantial part of
period.
[132,133]
Synthesis of protein is energetically
the prolonged EPOC component after exhaustive
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1056 Børsheim & Bahr
submaximal exercise. A minor part may be ex- Acknowledgements
plained by a sustained elevation in circulation, ven-
No sources of funding were used to assist in the prepara-
tilation and temperature. No increased
tion of this manuscript. The authors have no conflicts of
sympathoadrenal activity has been found after such
interest that are directly relevant to the content of this manu-
exercise. Little is known about the mechanisms un-
script.
derlying EPOC after resistance exercise.
References
1. Hill AV, Lupton H. Muscular exercise, lactic acid, and the
13. Conclusions
supply and utilization of oxygen. Q J Med 1923; 16: 135-71
2. Hill AV, Long CNH, Lupton H. Muscular exercise, lactic acid,
and the supply and utilisation of oxygen: parts I-III. Proc R Soc
Much of the conflicting reports of EPOC magni-
Lond (Biol) 1924; 96: 438-75
tude and duration may be resolved if differences in
3. Hill AV, Long CNH, Lupton H. Muscular exercise, lactic acid,
and the supply and utilisation of oxygen: parts IV-VI. Proc R
exercise intensity and duration are taken into ac-
Soc Lond (Biol) 1924; 97: 84-138
count. Thus, a small and short-lasting EPOC has
4. Hill AV, Long CNH, Lupton H. Muscular exercise, lactic acid,
typically been observed in studies with low exercise
and the supply and utilitsation of oxygen: parts VII-VIII. Proc
R Soc Lond (Biol) 1924; 97: 155-76
intensity and/or low exercise duration, whereas a
5. Margaria R, Edwards HT, Dill OB. The possible mechanisms of
more prolonged and substantial EPOC can be ob-
contracting and paying the oxygen debt and the role of lactic
acid in muscular contraction. Am J Physiol 1933; 106:
served after more strenuous exercise. The relation-
689-715
ship between exercise intensity and EPOC magni-
6. Gaesser GA, Brooks GA. Metabolic bases of excess post-
tude is curvilinear, whereas at higher intensities
exercise oxygen consumption: a review. Med Sci Sports Exerc
1984; 16 (1): 29-43
there is a linear relationship between exercise dura-
7. Bahr R. Excess postexercise oxygen consumption: magnitude,
tion and EPOC. For resistance exercise, the relation-
mechanisms and practical implications. Acta Physiol Scand
1992; 144 Suppl. 605: 1-70
ship between exercise intensity, duration and EPOC
8. Bullough RC, Gillette CA, Harris MA, et al. Interaction of acute
is less clear, but a more prolonged and substantial
changes in exercise energy expenditure and energy intake on
EPOC has been found after hard versus moderate
resting metabolic rate. Am J Clin Nutr 1995 Mar; 61 (3):
473-81
resistance exercise. Furthermore, the individual ef-
9. Benedict FG, Carpenter TM. The metabolism and energy trans-
fects of exercise mode, training status and sex on
formations of healthy man during rest. Washington, DC: The
Carnegie Institute, 1910
EPOC are unclear, partly because of difficulties in
10. Herxheimer H, Wissing E, Wolff E. Sp
¨
atwirkungen er-
designing appropriate protocols to investigate these
sch
¨
opfender Muskelarbeit auf den Sauerstoffverbrauch. Z
effects. Still, it appears that trained individuals have
Gesamte Exp Med 1926; 51: 916-28
11. Edwards HT, Thorndike A, Dill DB. The energy requirements
a more rapid return of
˙
VO
2
to resting levels after
in strenuous muscular exercise. N Engl J Med 1935; 213:
exercising at either the same relative or the same
532-5
absolute work rate, but studies after more strenuous
12. Passmore R, Johnson RE. Some metabolic changes following
prolonged moderate exercise. Metabolism 1960; 9: 452-5
exercise bouts should be done. Finally, while some
13. deVries HA, Gray DE. After effects of exercise upon resting
of the mechanisms underlying the more rapid EPOC
metabolic rate. Res Q 1963; 34 (3): 314-21
14. Bielinski R, Schutz Y, Jequier E. Energy metabolism during the
after strenuous aerobic exercise are well known, the
postexercise recovery in man. Am J Clin Nutr 1985 Jul; 42 (1):
mechanisms underlying the prolonged EPOC com-
69-82
ponent are more unclear. The sum of the increased
15. Mæhlum S, Grandmontagne M, Newsholme EA, et al. Magni-
tude and duration of excess postexercise oxygen consumption
rate of TG/FA cycling and the shift from carbohy-
in healthy young subjects. Metabolism 1986; 35 (5): 425-9
drate to fat as substrate source has been suggested to
16. Bahr R, Ingnes I, Vaage O, et al. Effect of duration of exercise
on excess postexercise O
2
consumption. J Appl Physiol 1987
account for more than half of the prolonged EPOC
Feb; 62 (2): 485-90
component, but there is still a significant part that
17. Chad KE, Wenger HA. The effect of exercise duration on the
remains to be explained. It is hypothesised that
exercise and post-exercise oxygen consumption. Can J Sport
Sci 1988; 13: 204-7
increased protein turnover and/or a change in energy
18. Gore CJ, Withers RT. Effect of exercise intensity and duration
efficiency in the recovery period may explain some
on postexercise metabolism. J Appl Physiol 1990; 68 (6):
2362-8
of this.
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1057
19. Gore CJ, Withers RT. The effect of exercise intensity and 39. Sedlock DA. Postexercise energy expenditure following upper
duration on the oxygen deficit and excess post-exercise oxygen body exercise. Res Q Exerc Sport 1991 Jun; 62 (2): 213-6
consumption. Eur J Appl Physiol 1990; 60: 169-74
40. Sedlock DA. Effect of exercise intensity on postexercise energy
20. Hagberg JM, Mullin JP, Nagle FJ. Effect of work intensity and
expenditure in women. Br J Sports Med 1991 Mar; 25 (1):
duration on recovery O
2
. J Appl Physiol 1980; 48: 540-4
38-40
21. Freedman-Akabas S, Colt E, Kissileff HR, et al. Lack of
41. Berg KE. Comparison of energy expenditure in men and women
sustained increase in
˙
VO
2
following exercise in fit and unfit
at rest and during exercise recovery. J Sports Med Phys Fitness
subjects. Am J Clin Nutr 1985; 41: 545-9
1991; 31 (3): 351-6
22. Pacy PJ, Barton N, Webster JD, et al. The energy cost of aerobic
42. Withers RT, Gore CJ, Mackay MH, et al. Some aspect of
exercise in fed and fasted normal subjects. Am J Clin Nutr
metabolism following a 35 km road run. Eur J Appl Physiol
1985 Nov; 42 (5): 764-8
1991; 63: 436-43
23. Brehm BA, Gutin B. Recovery energy expenditure for steady
43. Bahr R, Sejersted OM. Effect of intensity of exercise on excess
state exercise in runners and nonexercisers. Med Sci Sports
postexercise O
2
consumption. Metabolism 1991 Aug; 40 (8):
Exerc 1986; 18: 205-10
836-41
24. Elliot DL, Goldberg L, Kuehl KS. Does aerobic conditioning
44. Bahr R, Sejersted OM. Effect of feeding and fasting on excess
cause a sustained increase in the metabolic rate? Am J Med Sci
postexercise oxygen consumption. J Appl Physiol 1991 Dec;
1988; 296: 249-51
71 (6): 2088-93
25. Turley KR, McBride PJ, Wilmore JH. Resting metabolic rate
45. Bahr R, Grønnerød O, Sejersted OM. Effect of supramaximal
measured after subjects spent the night at home vs at a clinic.
exercise on excess postexercise O
2
consumption. Med Sci
Am J Clin Nutr 1993 Aug; 58 (2): 141-4
Sports Exerc 1992 Jan; 24 (1): 66-71
26. Bullough RC, Melby CL. Effect of inpatient versus outpatient
46. Elliot DL, Goldberg L, Kuehl KS. Effect of resistance training
measurement protocol on resting metabolic rate and respira-
on excess post-exercise oxygen consumption. J Appl Sport Sci
tory exchange ratio. Ann Nutr Metab 1993; 37 (1): 24-32
Res 1992; 6 (2): 77-81
27. Maresh CM, Abraham A, De Souza MJ, et al. Oxygen con-
47. Sedlock DA. Post-exercise energy expenditure after cycle er-
sumption following exercise of moderate intensity and dura-
gometer and treadmill exercise. J Appl Sport Sci Res 1992; 6
tion. Eur J Appl Physiol 1992; 65: 421-6
(1): 19-23
28. Knuttgen HG. Oxygen debt after submaximal physical exercise.
48. Donelly K, McNaughton L. The effects of two levels of caffeine
J Appl Physiol 1970; Nov 29 (5): 651-7
ingestion on excess postexercise oxygen consumption in un-
29. Segal SS, Brooks GA. Effects of glycogen depletion and work
trained women. Eur J Appl Physiol Occup Physiol 1992; 65
load on postexercise O
2
consumption and blood lactate. J Appl
(5): 459-63
Physiol 1979; 47: 514-21
49. Brockman L, Berg K, Latin R. Oxygen uptake during recovery
30. Hermansen L, Grandmontagne M, Mæhlum S, et al. Post-
from intense intermittent running and prolonged walking. J
exercise elevation of resting oxygen uptake: possible mechan-
Sports Med Phys Fitness 1993 Dec; 33 (4): 330-6
isms and physiological significance. In: Marconnet P,
50. Kaminsky LA, Whaley MH. Effect of interval-type exercise on
Poortmans J, Hermansen L, editors. Medicine and sport
excess postexercise oxygen consumption (EPOC) in obese and
science. Vol. 17. Basel: Karger, 1984: 119-29
normal-weight women. Med Exerc Nutr Health 1993; 2:
31. Chad KE, Wenger HA. The effects of duration and intensity on
106-11
the exercise and post-exercise metabolic rate. Aust J Sci Med
51. Smith J, McNaughton L. The effects of intensity of exercise on
Sport 1985; 17: 14-8
excess postexercise oxygen consumption and energy expendi-
32. Devlin JT, Horton ES. Potentiation of the thermic effect of
ture in moderately trained men and women. Eur J Appl Physiol
insulin by exercise: differences between lean, obese, and
1993; 67: 420-5
noninsulin-dependent diabetic men. Am J Clin Nutr 1986; 43:
52. Neary JP, Docherty D, Wenger HA. Post-exercise metabolic
884-90
rate is influenced by elevated core temperature. Aust J Sci Med
33. Sedlock D, Fissinger JA, Melby CL. Effect of exercise intensity
Sport 1993; 25 (2): 43-7
and duration on postexercise energy expenditure. Med Sci
53. Frey GC, Byrnes WC, Mazzeo RS. Factors influencing excess
Sports Exerc 1989; 21: 662-6
postexercise oxygen consumption in trained and untrained
34. Chad K, Quigley B. The effects of substrate utilization, manipu-
women. Metabolism 1993; 42 (7): 822-8
lated by caffeine, on post-exercise oxygen consumption in
54. Børsheim E, Bahr R, Hansson P, et al. Effect of β-adrenoceptor
untrained female subjects. Eur J Appl Physiol 1989; 59: 48-54
blockade on post-exercise oxygen consumption. Metabolism
35. Poehlman ET, LaChance P, Tremblay A, et al. The effect of
1994; 43 (5): 565-71
prior exercise and caffeine ingestion on metabolic rate and
55. Thomas TR, Londeree BR, Lawson DA. Prolonged recovery
hormones in young adult males. Can J Physiol Pharmacol 1989
from eccentric versus concentric exercise. Can J Appl Physiol
Jan; 67 (1): 10-6
1994; 19 (4): 441-50
36. Kaminsky LA, Padjen S, LaHam-Saeger J. Effect of split exer-
56. Sedlock DA. Fitness level and postexercise energy expenditure.
cise sessions on excess post-exercise oxygen consumption. Br
J Sports Med Phys Fitness 1994 Dec; 34 (4): 336-42
J Sports Med 1990; 24 (2): 95-8
57. Gilette CA, Bullough RC, Melby C. Postexercise energy expen-
37. Bahr R, Hansson P, Sejersted OM. Triglyceride/fatty acid cy-
diture in response to acute aerobic or resistive exercise. Int J
cling is increased after exercise. Metabolism 1990; 39 (9):
Sport Nutr 1994; 4: 347-60
993-9
38. Chad KE, Quigley BM. Exercise intensity: effect on post- 58. Quinn TJ, Vroman NB, Kertzer R. Postexercise oxygen con-
exercise O
2
uptake in trained and untrained women. J Appl sumption in trained females: effect of exercise duration. Med
Physiol 1991; 70: 1713-9 Sci Sports Exerc 1994 Jul; 26 (7): 908-13
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1058 Børsheim & Bahr
59. Harms CA, Cordain L, Stager JM, et al. Body fat mass affects 77. Frid
´
en J, Sj
¨
ostr
¨
om M, Ekblom B. Myofibrillar damage follow-
postexercise metabolism in males of similar lean body mass. ing intense eccentric exercise in man. Int J Sports Med 1983; 4:
Med Exerc Nutr Health 1995; 4: 33-9 170-6
78. Kolkhorst FW, Londeree BR, Thomas TR. Effect of consecutive
60. Dawson B, Straton S, Randall N. Oxygen consumption during
exercise days of jogging or cycling on the resting metabolic
recovery from prolonged submaximal cycling below the anaer-
rate and nitrogen balance. J Sports Med Phys Fitness 1994; 34:
obic threshold. J Sports Med Phys Fitness 1996 Jun; 36 (2):
343-50
77-84
79. Melby CL, Tincknell T, Schmidt WD. Energy expenditure
61. Short KR, Wiest JM, Sedlock DA. The effect of upper body
following a bout of non-steady state resistance exercise. J
exercise intensity and duration on post-exercise oxygen con-
Sports Med Phys Fitness 1992; 32: 128-35
sumption. Int J Sports Med 1996 Nov; 17 (8): 559-63
80. Murphy E, Schwarzkopf R. Effects of standard set and circuit
62. Trost S, Wilcox A, Gillis D. The effect of substrate utilization,
weight training on excess post-exercise oxygen consumption. J
manipulated by nicotinic acid, on excess postexercise oxygen
Appl Sport Sci Res 1992; 6 (2): 88-91
consumption. Int J Sports Med 1997; 18 (2): 83-8
81. Melby C, Scholl C, Edwards G, et al. Effect of acute resistance
63. Phelain JF, Reinke E, Harris MA, et al. Postexercise energy
exercise on postexercise energy expenditure and resting meta-
expenditure and substrate oxidation in young women resulting
bolic rate. J Appl Physiol 1993 Oct; 75 (4): 1847-53
from exercise bouts of different intensity. J Am Coll Nutr 1997
82. Olds TS, Abernethy PJ. Postexercise oxygen consumption fol-
Apr; 16 (2): 140-6
lowing heavy and light resistance exercise. J Strength Cond
64. Laforgia J, Withers RT, Shipp NJ, et al. Comparison of energy
Res 1993; 7: 147-52
expenditure elevations after submaximal and supramaximal
83. Haltom RW, Kraemer RR, Sloan RA, et al. Circuit weight
running. J Appl Physiol 1997; 82 (2): 661-6
training and its effects on excess postexercise oxygen con-
65. Short KR, Sedlock DA. Excess postexercise oxygen consump-
sumption. Med Sci Sports Exerc 1999 Nov; 31 (11): 1613-8
tion and recovery rate in trained and untrained subjects. J Appl
84. Dolezal BA, Potteiger JA, Jacobsen DJ, et al. Muscle damage
Physiol 1997; 83 (1): 153-9
and resting metabolic rate after acute resistance exercise with
66. Burleson Jr MA, O’Bryant HS, Stone MH, et al. Effect of
an eccentric overload. Med Sci Sports Exerc 2000 Jul; 32 (7):
weight training exercise and treadmill exercise on post-exer-
1202-7
cise oxygen consumption. Med Sci Sports Exerc 1998 Apr; 30
85. Osterberg KL, Melby CL. Effect of acute resistance exercise on
(4): 518-22
postexercise oxygen consumption and resting metabolic rate in
67. Almuzaini KS, Potteiger JA, Green SB. Effects of split exercise
young women. Int J Sport Nutr Exerc Metab 2000 Mar; 10 (1):
sessions on excess postexercise oxygen consumption and rest-
71-81
ing metabolic rate. Can J Appl Physiol 1998 Oct; 23 (5):
86. Binzen CA, Swan PD, Manore MM. Postexercise oxygen con-
433-43
sumption and substrate use after resistance exercise in women.
68. Børsheim E, Bahr R, Høstmark AT, et al. Effect of β-adre-
Med Sci Sports Exerc 2001 Jun; 33 (6): 932-8
noceptor blockade on post-exercise oxygen consumption and
87. Thornton MK, Potteiger JA. Effects of resistance exercise bouts
triglyceride/fatty acid cycling. Metabolism 1998; 47 (4):
of different intensities but equal work on EPOC. Med Sci
439-48
Sports Exerc 2002 Apr; 34 (4): 715-22
69. Børsheim E, Bahr R, Knardahl S. Effect of β-adrenoceptor
88. Schuenke MD, Mikat RP, McBride JM. Effect of an acute
stimulation on oxygen consumption and triglyceride/fatty acid
period of resistance exercise on excess post-exercise oxygen
cycling after exercise. Acta Physiol Scand 1998; 164: 157-66
consumption: implications for body mass management. Eur J
70. Matsuo T, Saitoh S, Suzuki M. Effects of the menstrual cycle on
Appl Physiol 2002 Mar; 86 (5): 411-7
excess postexercise oxygen consumption in healthy young
89. Williamson DL, Kirwan JP. A single bout of concentric resis-
women. Metabolism 1999 Mar; 48 (3): 275-7
tance exercise increases basal metabolic rate 48 hours after
71. Lee YS, Ha MS, Lee YJ. The effects of various intensities and
exercise in healthy 59-77-year-old men. J Gerontol A Biol Sci
durations of exercise with and without glucose in milk inges-
Med Sci 1997 Nov; 52 (6): M352-5
tion on postexercise oxygen consumption. J Sports Med Phys
90. Hagberg JM, Hickson RC, Ehsani AA, et al. Faster adjustment
Fitness 1999 Dec; 39 (4): 341-7
to and recovery from submaximal exercise in the trained state.
72. Fukuba Y, Yano Y, Murakami H, et al. The effect of dietary
J Appl Physiol 1980 Feb; 48 (2): 218-24
restriction and menstrual cycle on excess post-exercise oxygen
91. Girandola RN, Katch FI. Effects of physical conditioning on
consumption (EPOC) in young women. Clin Physiol 2000
changes in exercise and recovery O
2
uptake and efficiency
Mar; 20 (2): 165-9
during constant-load ergometer exercise. Med Sci Sports 1973;
73. Bouchard C, Rankinen T. Individual differences in response to
5 (4): 242-7
regular physical activity. Med Sci Sports Exerc 2001 Jun; 33 (6
92. Solomon SJ, Kurzer MS, Calloway DH. Menstrual cycle and
Suppl.): S446-51
basal metabolic rate in women. Am J Clin Nutr 1982 Oct; 36
74. Bahr R, Opstad PK, Medbø JI, et al. Strenuous prolonged
(4): 611-6
exercise elevates resting metabolic rate and causes reduced
93. Meijer GA, Westerterp KR, Saris WH, et al. Sleeping metabolic
mechanical efficiency. Acta Physiol Scand 1991 Apr; 141 (4):
rate in relation to body composition and the menstrual cycle.
555-63
Am J Clin Nutr 1992 Mar; 55 (3): 637-40
75. Newham DJ, McPhail G, Mills KR, et al. Ultrastructural
94. Bisdee JT, James WP, Shaw MA. Changes in energy expendi-
changes after concentric and eccentric contractions of human
ture during the menstrual cycle. Br J Nutr 1989 Mar; 61 (2):
muscle. J Neurol Sci 1983 Sep; 61 (1): 109-22
187-99
76. Friden J, Lieber RL. Eccentric exercise-induced injuries to 95. Hessemer V, Bruck K. Influence of menstrual cycle on thermo-
contractile and cytoskeletal muscle fibre components. Acta regulatory, metabolic, and heart rate responses to exercise at
Physiol Scand 2001 Mar; 171 (3): 321-6 night. J Appl Physiol 1985 Dec; 59 (6): 1911-7
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

EPOC and Exercise Intensity and Duration 1059
96. Webb P. 24-hour energy expenditure and the menstrual cycle. 117. Sagnol M, Claustre J, Pequignot JM, et al. Catecholamines and
Am J Clin Nutr 1986 Nov; 44 (5): 614-9 fuels after an ultralong run: persistent changes after 24-h
recovery. Int J Sports Med 1989 Jun; 10 (3): 202-6
97. Bangsbo J, Gollnick PD, Graham TE, et al. Anaerobic energy
118. Cori CF, Buchwald KW. Effect of continuous injection of
production and O
2
deficit-debt relationship during exhaustive
epinephrine on the carbohydrate metabolism, basal metabol-
exercise in humans. J Physiol 1990; 422: 539-59
ism and vascular system of normal man. Am J Physiol 1930;
98. Wolfe RR, Klein S, Carraro F, et al. Role of triglyceride-fatty
95: 71-8
acid cycle in controlling fat metabolism in humans during and
119. Webber J, Macdonald IA. Metabolic actions of catecholamines
after exercise. Am J Physiol 1990; 258 (21): E382-9
in man. Bailli
´
eres Clin Endocrinol Metab 1993; 7 (2): 393-413
99. Miyoshi H, Schulman GI, Peters EJ, et al. Hormonal control of
120. Blaak EE, Saris WHM, van Baak MA. Adrenoceptor subtypes
substrate cycling in humans. J Clin Invest 1988; 81: 1545-55
mediating catecholamine-induced thermogenesis in man. Int J
100. Wolfe RR, Peters EJ. Lipolytic response to glucose infusion in
Obes 1993; 17 (3): S78-81
human subjects. Am J Physiol 1987; 252 (15): E218-23
121. Blaak EE, van Baak MA, Kempen KPG, et al. Role of α- and β-
101. Chaliss RAJ, Arch JRS, Newsholme EA. The rate of substrate
adrenoceptors in sympathetically mediated thermogenesis. Am
cycling between fructose 6-phosphate and fructose
J Physiol 1993; 264 (27): E11-7
1,6-bisphosphate in skeletal muscle. Biochem J 1984; 221:
122. Kjær M, Secher NH, Galbo H. Physical stress and catecho-
153-61
lamine release. Bailli
´
eres Clin Endocrinol Metab 1987; 1:
102. Wolfe RR, Herndon DN, Jahoor F, et al. Effect of severe burn
279-89
injury on substrate cycling by glucose and fatty acids. N Engl J
123. Cain SM. Exercise O2 debts of dogs at ground level and at
Med 1987; 317: 403-8
altitude with and without β-block. J Appl Physiol 1971; 30:
103. Stein TP, Rumpler WV, Leskiw MJ, et al. Effect of reduced
838-43
dietary intake on energy expenditure, protein turnover, and
124. Bahr R, Høstmark AT, Newsholme EA, et al. Effect of exercise
glucose cycling in man. Metabolism 1991; 40 (5): 478-83
on recovery changes in plasma levels of FFA, glycerol, glu-
104. Weber JM, Klein SE, Wolfe RR. Role of the glucose cycle in
cose and catecholamines. Acta Physiol Scand 1991 Sep; 143
control of net glucose flux in exercising humans. J Appl
(1): 105-15
Physiol 1990; 68: 1815-9
125. Wahrenberg H, Engfeldt P, Bolinder J, et al. Acute adaptation
105. Wolfe RR, Klein S, Herndon DN, et al. Substrate cycling in
in adrenergic control of lipolysis during physical exercise in
thermogenesis and amplification of net substrate flux in human
humans. Am J Physiol 1987; 253 (16): E383-90
volunteers and burned patients. J Trauma 1990; 30 (12): 6-9
126. Savard R, Despr
´
es JP, Marcotte M, et al. Acute effects of
endurance exercise on human adipose tissue metabolism. Met-
106. Flatt JP. The biochemistry of energy expenditure. In: Bray GA,
abolism 1987; 36 (5): 480-5
editor. Recent advances in obesity research II. London: New-
man, 1978: 211-28 127. Landsberg L, Young JB. Catecholamines and the adrenal me-
dulla. In: Wilson JD, Foster DW, editors. Textbook of endocri-
107. Hellerstein MK, Schwarz JM, Neese RA. Regulation of hepatic
nology. 8th ed. Philadelphia (PA): WB Saunders Company,
de novo lipogenesis in humans. Annu Rev Nutr 1996; 16:
1992: 621-81
523-57
128. Viru A. Postexercise recovery period: carbohydrate and protein
108. McDevitt RM, Bott SJ, Harding M, et al. De novo lipogenesis
metabolism. Scand J Med Sci Sports 1996; 6: 2-14
during controlled overfeeding with sucrose or glucose in lean
129. Rennie MJ, Edwards RHT, Krywawych S, et al. Effect of
and obese women. Am J Clin Nutr 2001 Dec; 74 (6): 737-46
exercise on protein turnover in man. Clin Sci 1981; 61: 627-39
109. Richard D, Rivest S. The role of exercise in thermogenesis and
130. Devlin JT, Brodsky I, Scrimgeour A, et al. Amino acid metabol-
energy balance. Can J Physiol Pharmacol 1989; 67 (4): 402-9
ism after intense exercise. Am J Physiol 1990; 258: E249-55
110. Poehlman ET, Horton ES. The impact of food intake and
131. Carraro F, Stuart CA, Hartl WH, et al. Effects of exercise and
exercise on energy expenditure. Nutr Rev 1989; 47: 129-37
recovery on muscle protein synthesis in human subjects. Am J
111. Galbo H. Hormonal and metabolic adaptation to exercise. Stutt-
Physiol 1990; 259: E470-6
gart: Georg Thieme Verlag, 1983
132. Biolo G, Maggi SP, Williams BD, et al. Increased rates of
112. Rønsen O, Haug E, Pedersen BK, et al. Increased neuroendo-
muscle protein turnover and amino acid transport after resis-
crine response to a repeated bout of endurance exercise. Med
tance exercise in humans. Am J Physiol 1995; 268 (31):
Sci Sports Exerc 2001 Apr; 33 (4): 568-75
E514-20
113. Børsheim E, L
¨
onnroth P, Knardahl S, et al. No difference in the
133. Phillips SM, Tipton KD, Aarsland A, et al. Mixed muscle
lipolytic response to β-adrenoceptor stimulation in situ but a
protein synthesis and breakdown after resistance exercise in
delayed increase in adipose tissue blood flow in moderately
humans. Am J Physiol 1997 Jul; 273 (1 Pt 1): E99-107
obese compared with lean men in the postexercise period.
134. Bangsbo J, Gollnick PD, Graham TE, et al. Substrates for
Metabolism 2000 May; 49 (5): 579-87
muscle glycogen synthesis in recovery from intense exercise in
114. Barnard RJ, Foss ML. Oxygen debt: effect of beta-adrenergic
man. J Physiol 1991; 434: 423-40
blockade on the lactacid and alactacid components. J Appl
135. Chin ER, Lindinger MI, Heigenhauser GJF. Lactate metabolism
Physiol 1969; 27: 813-6
in inactive skeletal muscle during lactacidosis. Am J Physiol
115. Chapler CK, Stainsby WN, Gladden LB. Effect of changes in 1991; 261: R98-105
blood flow, norepinephrine and pH on oxygen uptake by
136. Bangsbo J, Graham T, Johansen L, et al. Muscle lactate
resting skeletal muscle. Can J Physiol Pharmacol 1980; 58:
metabolism in recovery from intense exhaustive exercise: im-
93-6
pact of light exercise. J Appl Physiol 1994; 77 (4): 1890-5
116. Gladden LB, Stainsby WN, MacIntosh BR. Norepinephrine 137. Scott CB. Re-interpreting anaerobic metabolism: an argument
increases canine skeletal muscle
˙
VO
2
during recovery. Med for the application of both anaerobic glycolysis and excess
Sci Sports Exerc 1982; 14 (6): 471-6 post-exercise oxygen consumption (EPOC) as independent
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)

1060 Børsheim & Bahr
sources of energy expenditure. Eur J Appl Physiol 1998; 77:
200-5
Correspondence and offprints: Dr Elisabet Børsheim, De-
138. Vidal-Puig A, Solanes G, Grujic D, et al. UCP3: an uncoupling
partment of Surgery, Metabolism Unit, Shriners Hospitals
protein homologue expressed preferentially and abundantly in
skeletal muscle and brown adipose tissue. Biochem Biophys
for Children/UTMB, 815 Market Street, Galveston, TX
Res Commun 1997; 235: 79-82
77550, USA.
139. Fleury C, Neverova M, Collins S, et al. Uncoupling protein-2: a
E-mail: elboersh@utmb.edu
novel gene linked to obesity and hyperinsulinemia. Nat Genet
1997; 15 (3): 269-72
140. Boss O, Samec S, Paoloni-Giacobino A, et al. Uncoupling
protein-3: a new member of the mitochondrial carrier family
with tissue-specific expression. FEBS Lett 1997; 408: 39-42
 Adis Data Information BV 2003. All rights reserved. Sports Med 2003; 33 (14)