Content uploaded by Stephen Seiler
Author content
All content in this area was uploaded by Stephen Seiler on Oct 15, 2018
Content may be subject to copyright.
Effect of work duration on physiological and rating scale of
perceived exertion responses during self-paced interval training
Stephen Seiler, Jarl Espen Sjursen
Department of Health and Sport, Agder University College, Kristiansand, Norway
Corresponding author: Stephen Seiler, PhD, Department of Health and Sport, Institute for Sport, Agder University College,
Service Box 422, 4604 Kristiansand, Norway. Tel: (47) 3814 1347, Fax: (47) 3814 1301, E-mail: Stephen.Seiler@hia.no
Accepted for publication 5 August 2003
This study compared running velocity, physiological
responses, and perceived exertion during self-paced interval
training bouts differing only in work bout duration. Twelve
well-trained runners (nine males, three females, 2875years,
VO
2 max
6576 mL min
1
kg
1
) performed preliminary
testing followed by four ‘‘high-intensity’’ interval sessions
(Latin squares, 1 session week
1
over 4 weeks) consisting
of 24 1, 12 2, 6 4, or 4 6-min running bouts with a
1:1 work-to-rest interval (total session duration 48 min).
The average running velocity decreased (93%, 88%, 86%,
84% vVO
2 max
,Po0.01) with increasing work duration.
Peak VO
2
averaged about 9274% of VO
2 max
for 2-, 4-,
and 6-min intervals compared with only 8275% for 1-min
bouts (Po0.001). Six of 12 athletes achieved their highest
average VO
2
and heart rate during 4-min intervals. The
average RPE
peak
(rating scale of perceived exertion) was
B1771 for all four interval sessions. RPE increased by
2–4 U during an interval training session. The mean lactate
concentration was similar across sessions (4.371.1–4.67
1.5 mmol L
1
). Under self-paced conditions, well-trained
runners perform ‘‘high-intensity’’ intervals at an RPE of
B17, independent of interval duration. The optimal
interval duration for eliciting a high physiological load is
3–5 min under these training conditions. Increases in RPE
during an interval bout are not associated with increasing
blood lactate concentration.
The assumed advantage of interval training, organiz-
ing exercise as intermittent bouts of higher and lower
intensity rather than one continuous bout, is the
accumulation of a greater training stimulus at high
exercise intensities (A
˚strand et al., 1960; Christensen
et al., 1960). Interval training has come to encompass
a broad array of training prescriptions. However,
for a given exercise modality, any interval training
prescription consists of five variables: (1) work inter-
val duration and (2) intensity, (3) recovery interval
duration and (4) intensity, and (5) total work
duration (interval number interval duration).
These variables can be manipulated to generate a
potentially endless number of specific interval train-
ing session prescriptions (Thibault & Marion, 1999).
Aerobic interval training for the purpose of enhancing
maximal aerobic capacity and endurance perfor-
mance typically involves a work interval duration
ranging from B1 to 8 min and a rest interval varying
from 30 s to 5 min (Fox et al., 1975). A work-to-
rest interval between 2:1 and 1:2 is common,
although work-to-rest ratios of up to 5:1 are used.
The prescribed exercise intensity for intervals in this
range is generally 85–105% of maximal aerobic
power (Thibault & Marion, 1999).
Little empirical data is available quantifying how
variation in the variables of the interval training
prescription impacts the athlete’s training response,
the achieved intensity, physiological responses, and
perceived exertion elicited. Published data quantify-
ing acute physiological responses associated with
different aerobic interval training conditions are
limited and almost entirely based on the measurement
of physiological responses to fixed work intensities
at a predefined percentage of maximal aerobic power
or velocity (Patterson et al., 1997; Billat et al., 1999,
2000a, b; Demarie et al., 2000; Stepto et al., 2001).
In the normal training setting, precise external
control of intensity is unusual. In many sports, work
intensity is not a constant function of velocity due
to variable terrain, wind, water conditions, snow
conditions, etc. Coaches or athletes typically pre-
scribe interval training sessions using manipulation
of interval duration (distance or time), rest duration,
and number of work bouts. The actual exercise
intensity achieved can be defined as a dependent
variable constrained by several independent vari-
ables: goal intensity, work duration, recovery time,
total volume of work, and the perception of exertion
and response to that perception. The purpose of
this study was therefore to quantify how work
interval duration impacts physiological responses
and perceived exertion during training sessions
where well-trained endurance athletes are uniformly
Scand J Med Sci Sports 2004: 14: 318–325 COPYRIGHT &BLACKWELL MUNKSGAARD 2004
Printed in Denmark .All rights reserved
DOI: 10.1046/j.1600-0838.2003.00353.x
318
instructed to perform a ‘‘high-intensity interval
session’’.
Methods
Subjects and preliminary testing
Twelve athletes from a local running club (nine males, three
females) volunteered to participate in this investigation, which
was approved by the human subjects research review board of
the Department of Health and Sport, Agder University
College. Criteria for participation in the study were: (1)
regular training, (2) competition in running races, and (3)
regular use of interval type training sessions. Participants were
informed of the risks associated with the study and their
choice to terminate participation at any time. They had
been training for running events, either track and field
or orienteering, for Z4 years, and averaged two interval-
type training sessions per week out of an average of 5.4 total
training sessions per week (range 4–9) in their current
training. All subjects were comfortable running at high speeds
on a motorized treadmill. Throughout the study, subjects
maintained their pre-study volume of training.
Initially, a continuous ramp protocol run to exhaustion was
performed to quantify ventilatory thresholds (VT
1
and VT
2
),
maximal heart rate (HR
max
), maximal oxygen consumption
(VO
2 max
), running velocity at maximal oxygen consumption
(vVO
2 max
), and peak blood lactate concentration ([La
peak
]).
The maximal treadmill test was repeated following the data
collection period to control for possible significant changes in
physiological capacity. Both testing and subsequent interval
training bouts were performed on a motorized treadmill
(Woodway P55 Sport, Weil am Rhein, Germany) at 1%
elevation (Jones & Doust, 1996). After a 20–30-min warm-
up, the maximal treadmill test was initiated with a 2-min run
at 9 km h
1
(150 m min
1
) with subsequent increases of
0.4 km h
1
(6.7 m min
1
) every 30s until voluntary exhaustion.
Gas exchange measurements
Expired gas samples were measured continuously using a
Cortex Metamax gas analyser (Cortex Biophysik GmbH,
Liepzig, Germany) and a Triple-V turbine (Sensormedics BV,
Bilthoven, The Netherlands). Mixing chamber derived gas
exchange data were averaged over 10-s time intervals.
Calibration was performed prior to each test according to
the manufacturer’s instructions. This system has been
previously shown to be valid, and highly stable over up to
120 min of continuous measurement (Schulz et al., 1997). HR
responses were collected every 15 s using a telemetry system
(Polar Electro, Kempele, Finland). VO
2 max
was defined as the
highest 30-s average measurement recorded during the ramp
test. The running velocity at maximal oxygen consumption
(vVO
2 max
) was defined as the lowest treadmill velocity
corresponding to the 30-s period defining VO
2 max
.VT
1
was
defined as an increase in V
E
/VO
2
without an increase in V
E
/
VCO
2
.VT
2
was defined as the point where V
E
/VCO
2
also
started to rise.
Blood lactate measurements
At 1 and 3 min after voluntary exhaustion during the maximal
treadmill test, venous blood was sampled via finger stick to
identify peak lactate concentration. In all subjects, lactate
values at 3 min were equal to or already below the 1-min
values. Twenty microlitres capillary blood samples were
immediately analysed for whole, non-haemolysed blood
lactate concentration using a YSI 1500 Sport lactate analyser
calibrated with reference standards before and after each test.
Values were subsequently corrected for the B20% of total
blood lactate trapped in red blood cells using a multiplication
factor of 1.25 (Medbet al., 2000).
Interval training bouts
During the 4 consecutive weeks following preliminary testing,
each athlete replaced one regularly scheduled hard training
session with an interval bout performed in the laboratory. The
four interval workouts varied in interval duration: 1-, 2-, 4-, or
6-min work periods. The work-to-rest ratio was fixed at 1:1
and the total work at 24 min for each session (i.e., 24 1,
12 2, 6 4, and 4 6 min). The order of interval bout
exposure was randomized using a Latin squares design. Each
subject performed the four interval sessions at the same time
of day each week. Subjects were instructed to treat each
weekly training session as a ‘‘high-intensity interval session’’.
They were also instructed to attempt to maintain the highest
average running velocity they could across all the work bouts
of each interval session.
The end of the 20–30-min warm-up was used to determine
the starting velocity for the first interval. Thereafter, the
treadmill velocity could be increased or decreased at any time
via a hand signal to the test administrator controlling the
treadmill via PC. At regular, frequent intervals during each
work bout, subjects were verbally encouraged and asked
whether they desired ‘‘more or less speed’’. Between work
bouts, subjects were also free to select their own recovery
intensity. About 15 s prior to each new work bout, subjects
straddled the treadmill belt as it was accelerated up to the
athlete’s desired starting velocity, using the velocity of the
previous interval as a reference. At the conclusion of a
running bout, athletes held the handrails and hopped to the
sides of the treadmill surface as the velocity was rapidly
reduced. They were then free to choose their activity intensity
during the recovery period. Two small, motorized fans,
positioned at each side and chest high, were used to ensure
effective evaporative cooling. The laboratory temperature
during the training sessions was 18–21 1C. Treadmill calibra-
tion demonstrated that treadmill velocity was highly accurate
over a broad range of velocities.
Physiological measurements during interval training
Running velocity, gas exchange data, and heart rate were
collected continuously during each interval training session.
The total workout duration varied slightly (42–47 min) across
the four conditions, since the last time period was a rest
interval of varying length. Two blood samples were collected
during each interval training session: 60 s after the 12th
minute of interval training, and 60 s after the completion of
the last work interval.
Perceived exertion measurements
The 15-point Borg rating scale of perceived exertion (RPE)
(Borg, 1970) was used to quantify global perceived exertion
during each interval session. Subjects were provided with
standard, written instructions translated into Norwegian.
RPE was determined during the final 10 s of every other
Responses to self-paced interval training
319
interval bout for the 1- and 2-min interval sessions and each of
the 4- and 6-min interval bouts.
Training control
This study was carried out during the early pre-competition
preparation phase of training (February–early March).
Training control was confirmed via a training diary. Subjects
were instructed to abstain from hard training the day prior to
laboratory sessions. No special attempt was made to control
the diet of the athletes; they were merely reminded to come to
the training session well hydrated and non-fasted.
Statistical analyses
Physiological and RPE responses during the four different
interval sessions were compared using the General Linear
Model with a repeated measures design (SPSS 10.0, Chicago,
Illinois, USA). An alevel of 0.05, after Bonferroni adjustment
to control for the increased risk of Type I error associated
with multiple comparisons, was considered to be statistical-
ly significant. Where appropriate, pairwise comparisons of
physiological data were made using the paired samples t-test
with the alevel set at 0.05.
Results
Maximal treadmill testing
Subject characteristics and results from preliminary
testing are presented in Table 1. The average VO
2max
of the three females in the study was 60 and
67 mL
min
1
kg
1
among the nine males, suggest-
ing that the males and females were of similar
relative capacity. The combination of VO
2max
,
ventilatory threshold determinations, and self-report
of current training confirmed that the subjects were
all representative of a regularly training, non-elite
athlete population. The intermittent treadmill test to
exhaustion was repeated following the period of
interval training data collection, which lasted 4–5 weeks.
The treadmill distance at exhaustion (34567441 vs.
36317543 m) improved by 5% (Po0.001) in the
follow-up test. HR
max
also declined significantly
(190710 vs. 187710 beats min
1
(bpm), Po0,05).
However, VO
2max
(65.375 vs. 65.676 mL min
1
kg
1
), ventilatory thresholds, and [La
]
peak
(8.472 vs. 8.672 mmol L
1
) did not change sig-
nificantly over the course of the study. The relative
intensity of the interval training sessions was there-
fore calculated based on the results of the prelimin-
ary treadmill test.
Physiological responses during interval sessions
Running velocity
Figure 1 depicts the average running velocity for
each intermittent bout of all four interval sessions.
An inspection of the individual data showed that
three of the 12 athletes underestimated their sustain-
able velocity during the initial bouts and increased
during the following bouts (as much as 12%), while
the remainder appeared to target their average
velocity more accurately from the start. When the
average velocity of the first 4 min and the last 4 min
of each session were compared, the upward adjust-
ment in velocity averaged 5%, 4%, 4%, and 1.5%
for 1-, 2-, 4-, and 6-min work durations, respectively,
and was statistically significant for the 1-, 2-, and 4-min
interval sessions (Po0.005). Figure 1 also reveals
that the average running velocity declined with
each increase in work bout duration, from 9372%
vVO
2max
for the 24 1-min interval session to
8372.5% during the 4 6-min interval session
(Po0.005 for each duration comparison). Expressed
as a percentage of the velocity corresponding to the
second ventilatory threshold (vVT
2
), the interval
sessions were performed at 11377%, 10775%,
10576%, and 10274% of vVT
2
for the 1-, 2-, 4-,
and 6-min work durations, respectively.
VO
2
and heart rate responses
When athletes self-selected their running velocity, the
peak oxygen consumption (92–93% of VO
2max
) and
oxygen consumption during recovery (26–31%
VO
2max
) were not statistically different for work
durations of 2, 4, or 6 min (Table 2). Individual
responses to the four training prescriptions are
presented in Fig. 2. Six of 12 subjects achieved their
highest oxygen consumption when performing 4-min
work intervals. Peak VO
2
during 1-min intervals was
significantly lower (8275% VO
2max
vs. 92–93%
VO
2max
for 2-, 4-, and 6-min intervals Po0.001),
and recovery oxygen consumption significantly higher
(4675% VO
2max
vs. 26–31%, Po0.001) compared
with the 2-, 4-, and 6-min work interval duration
sessions. However, the average oxygen consumption
Table 1. Characteristics of the subjects (
N
512, nine males, three females)
Age (years)
Height (cm)
Weight (kg)
VO
2 max
(ml kg
1
min
1
)
HF
max
(beats min
1
)
VT
1
(%VO
2 max
)
VT
2
(%VO
2 max
)
n
VO
2 max
(km h
1
)
2875 176786576 190710 68768474 19.771
68710
N
512 values are presented as mean7SD. The velocity at VO
2 max
was determined at a constant incline of 1%.
Seiler and Sjursen
320
for the entire session (even numbers of work and rest
periods) was virtually identical for all four condi-
tions (59.4–61.4% VO
2max
). Separate repeated mea-
sures analyses were performed to examine whether a
testing order effect on exercise intensity was present.
Whether measured as peak HR or peak VO
2
, the
order of testing did not significantly influence self-
selected exercise intensity.
Figure 3 displays the averaged heart rate responses
throughout the four interval training sessions. The
peak heart rate (Fig. 4) was slightly but significantly
lower during the 1-min interval, compared with the
4- and 6-min intervals (92% for 1-min intervals vs.
95% of HR
max
for both 4- and 6-min intervals,
Po0.005). Peak heart rate responses during the 2-
min intervals were only significantly different from
the 4-min interval responses (P50.05). Overall, the
differences in peak heart rate responses between the
1-min intervals and the longer interval durations
were smaller than the differences in peak oxygen
consumption.
Blood lactate concentration
Blood lactate concentration, measured immediately
after the 12th and 24th minute of interval running
during each interval session, was similar across
interval duration and approximated 4.5 mM (Fig.
5). Blood lactate concentration tended to rise slightly
from midway to the end of the interval sessions
under all conditions. However, this increase was
only statistically significant during 1-min intervals
(5.071.4 vs. 3.971.1 mM, P50.02).
Ratings of perceived exertion
RPE increased steadily throughout an interval
session (Fig. 6) under all work duration conditions.
The increasing perception of exertion during the
interval sessions was well fit by a linear regression
line (RPE 514.570.11 (accumulated work dura-
tion in minutes)). RPE tended to be lower at the end
of the initial 1-min bouts compared with the 4- and
6-min interval bouts, consistent with the greater
degree of velocity adjustment observed for the
shortest interval duration. However, peak RPE was
virtually identical under all four work-duration
conditions (ranging from 16.871 to 17.271).
Discussion
The key finding of this study is that when prescribed
interval training sessions vary in work period
duration from 1 to 6 min, athletes adjust their
intensity such that blood lactate and perceived
exertion responses throughout each session are
essentially identical. To our knowledge, this study
is unique in that subjects self-selected their exercise
intensity throughout each session in response to a
standardized work prescription, conditions that are
similar to how endurance athletes normally train.
240
250
260
270
280
290
300
310
1
3
5
7
9
11
13
15
17
19
21
23
25
Velocity (m min-1)
Accumulated work duration (min)
1 min
2 min
4 min
6 min
(a)
(b)
(c)
Fig. 1. Mean velocity (m min
1
) for each work bout
performed during 1-, 2-, 4-, and 6-min work duration
interval sessions. Error bars are omitted for clarity. The
mean running velocity was significantly different for each
work duration condition. (a, b, c): The initial running
velocity was significantly lower than subsequent running
bouts (48 min accumulated work duration, Po0.05).
Table 2. Oxygen consumption responses to four interval bouts differing
in work duration
Work bout
(and recovery)
duration (min)
Peak VO
2
(%VO
2 max
)
VO
2
at end of
recovery
periods
(%VO
2 max
)
Average VO
2
for entire
session
(%VO
2 max
)
1 81.875.2
n
46.073.4
n
61.373.9
2 92.474.4 27.575.3 60.774.7
4 93.374.8 25.678.1 59.474.2
6 91.773.0 30.6711.6 61.473.3
Values 5mean7SD.
n
P
o0.01 vs. 2-, 4-, and 6-min work duration trials.
60
65
70
75
80
85
90
95
100
105
1 min 2 min 4 min 6 min
Interval work duration
Percent VO2 max
Fig. 2. Individual peak VO
2
during interval sessions with 1-,
2-, 4-, or 6-min work duration. The bold line indicates the
averaged peak VO
2
response under each condition.
Responses to self-paced interval training
321
The results also suggest that under self-paced
conditions with a typical 1:1 work-to-rest ratio, peak
physiological responses are very similar for interval
durations between 2 and 6 min and significantly
higher than those obtained during an equal duration
of work performed as 1-min work intervals.
A work duration of 4 min appears to approximate
an optimal duration for achieving peak cardiovas-
cular responses under self-paced conditions. Six of
12 athletes achieved their highest oxygen consump-
tion during 4-min work intervals. Our findings are
interesting in the light of those of Stepto et al. (1999),
who compared the cycling performance enhancing
effects of five different interval training programmes
built around work bouts of 0.5-, 1-, 2-, 4-, or 8-min
duration. Power output was controlled for the
85
95
105
115
125
135
145
155
165
175
185
Heart Rate
85
95
105
115
125
135
145
155
165
175
185
Heart Rate
85
95
105
115
125
135
145
155
165
175
185
Heart Rate
85
95
105
115
125
135
145
155
165
175
185
0
12
24
36
48
Heart Rate
Accumulated duration (min)
a
b
c
d
Fig. 3. Mean heart rate responses during (a) 1-, (b) 2-, (c) 4-,
and (d) 6-min work bout duration conditions. Data presented
are based on results from 10 of 12 subjects because continuous
telemetry recordings were partially lost for two subjects.
86
88
90
92
94
96
98
100
1 min 2 min 4 min 6 min
Percent HFpeak
**
∗
Interval Duration
Fig. 4. Group means (7SD) for peak heart rate (% HF
peak) during 1-, 2-, 4-, and 6-min work bout duration
sessions.
*
Po0.05 vs. 4- and 6-min conditions.
**
P50.05 vs.
4-min condition.
0
1
2
3
4
5
6
7
1 min 2 min 4 min 6 min
Interval Duration
[Blood Lactate] mmol l-1
MID
END
*
Fig. 5. Group means (7SD) for blood lactate measure-
ments performed after 12 min (MID, open bars) and after
24 min (END, solid bars) of accumulated work duration
under the four different interval training conditions.
*
Po0.05 vs. END.
Seiler and Sjursen
322
different interval training groups, and equal to 175%,
100%, 90%, 85%, and 80% of peak power output,
respectively. Both peak sustained power output and
40-km time trial performance were improved most in
the group of cyclists performing 4-min work bouts at
85% of PPO.
Another important aspect of the interval training
description is the intensity of work during recovery
periods. This intensity typically ranges from com-
plete rest up to 50% of VO
2max
(Billat et al., 2000a).
The intensity of work during rest can be expected to
influence the rate of recovery of phosphocreatine, the
rate of lactate elimination, and the oxygen kinetics at
the onset of the next exercise bout. What strategy do
athletes actually choose when prescribed a specific
interval training task? In this study, the athletes
appeared to choose recovery intensity sufficient to
maintain oxygen consumption at 20–35% VO
2 max.
This intensity range is somewhat lower than that
found to be optimal for lactate elimination after
high-intensity work (McLellan & Skinner, 1982) and
also lower than the B50% VO
2max
, which has been
suggested as optimal for achieving and maintaining
the highest possible oxygen consumption levels
during the work periods (Billat et al., 2000b). One-
minute intervals actually resulted in lower VO
2
during the work portion and higher VO
2
during
recovery, such that the ‘‘interval training’’ effect was
truncated. In contrast, 4-min interval bouts were
associated with the highest VO
2
during work and the
lowest VO
2
during recovery (Table 2), although these
differences were not statistically significant when
compared with the 2- and 6-min work duration
conditions. Very little research has been reported on
the impact of recovery duration and/or recovery
work intensity on the overall intensity maintained
during interval training bouts. This is currently a
topic of investigation in our laboratory.
As expected, heart rate responses tracked well with
oxygen consumption curves. The peak heart rate was
similar for the 4- and 6-min work bouts, equalling
95% of maximal heart rate. However, the difference
in peak heart rate response (B91% HR
max
vs. 95%
HR
max
) was smaller than the difference in peak
oxygen consumption (82% vs. 93% VO
2max
). In a
training setting, heart rate may give a somewhat
misleading picture of the relative oxygen consump-
tion achieved during short intervals. One explana-
tion for this discordance is the break from linearity
in the heart rate-intensity relationship occurring
around the second ventilatory threshold (Hoffmann
et al., 1994, 1997). During preliminary testing, a clear
negative heart rate inflection point was identified for
10 of the 12 athletes. This would result in a lower
HR/VO
2
ratio at intensities above the inflection
point.
The peak heart rate during each work bout
increased over the course of an interval session. If
we ignore the first interval bout that was often
followed by velocity adjustments, peak heart rate
drifted upwards B5 bpm over the course of an
interval session. More striking however was the
increase in end-recovery heart rate, or ‘‘recovery
heart-rate drift’’ in an interval training context,
observed over the course of the sessions (Fig. 2).
Using the 2-min interval bouts as an example, we
observed a B25 bpm increase in end- recovery heart
rate from the end of the first work bout to the end of
the next to last work bout. At the same time, end-
recovery oxygen consumption remained stable (mean
of first three bouts 1.1770.34 L min
1
vs. mean of
last three bouts 1.1970.34 L min
1
). The increase in
end-recovery heart rate was greatest for the 1- and
2-min intervals, suggesting that primarily the fast
phase of heart rate recovery was slowed over the
course of the training session. Peak VO
2
, recovery
VO
2
, and blood lactate concentration were stable
over most of the session, suggesting attainment of a
quasi-steady-state in and around the active muscle.
This suggests that the fast phase of heart rate
recovery is largely controlled by central mechanisms.
If this is the case (see e.g., Savin et al., 1982), then
using heart rate recovery as a guide for determining
when to begin subsequent interval bouts would
appear to have little direct connection to recovery
processes at the muscular level.
Blood lactate responses were surprisingly constant
for the four different interval training sessions. In
isolation, the similarity in blood lactate responses
suggests a similar overall contribution of anaerobic
13
14
15
16
17
18
04812162024
RPE
1 min
2 min
4 min
6 min
Accumulated interval duration (min)
Fig. 6. Group means for rating scale of perceived exertion
(RPE) recorded throughout the four different interval
training sessions. Error bars are omitted for clarity. SD
averaged B0.8 RPE units for all four conditions. RPE
changes over time were well fit by a linear regression equa-
tion: RPE 514.510.11 accumulated work duration (min).
Responses to self-paced interval training
323
metabolism across the four different conditions,
since averaged blood concentrations measured under
the four work-duration conditions were not signifi-
cantly different (4.2–4.5 mM). The data also suggest
a diminishing reliance on lactate metabolism over the
course of a given interval session, since blood lactate
measurements changed little from midway to the end
of the interval training sessions (Fig. 5).
The observation that most of the increase in blood
lactate concentration occurs after the initial bouts of
intermittent exercise is not novel. Stepto and collea-
gues measured physiological responses to repeat 5-
min cycling bouts at 82.5% of the maximal aerobic
power separated by 1-min recovery periods. In that
study, blood lactate concentration increased to
4.7 mM after the first bout and remained between 5
and 6 mM over the remaining seven bouts of
exercise. Oxygen consumption was 5–7% lower in
that initial bout compared with the third, fifth, and
seventh bouts (only these were reported), suggesting
that the initially greater oxygen deficit accounts for
the early rise in blood lactate, and that subsequent
work was performed with a closer balance between
lactate production and elimination, in part due to
reduced reliance on glycolytic metabolism (Stepto
et al., 2001). In the present study, we also observed
nearly constant blood lactate concentration from the
middle of the session to the end, suggesting that
lactate production and elimination were in balance
despite running velocities averaging 102–113% of
vVT
2
. This interpretation is made with caution,
however. The concentration of a substance in a fluid
compartment is by definition the amount of sub-
stance divided by the fluid volume in the compart-
ment being measured. Direct comparisons of blood
lactate concentrations at the beginning, middle, and
end of an interval session are complicated by the fact
that the distribution volume for lactate will progres-
sively increase over a period of 20–40 min, since
lactate ions distribute slowly in the extracellular
space (Medb& Toska, 2001). Over the course of a
short, intense interval training workout, blood
lactate concentration could theoretically stabilize
despite a continued mismatch between production
and elimination.
Perhaps the most common measure of exercise
intensity in the normal endurance training setting
remains perception of effort. Across a spectrum of
endurance sports, the power output for a given
velocity changes with weather, wind, hill elevation,
snow conditions, etc. The heart rate at a given
intensity drifts over time, changes with different
environmental conditions, and lags behind during
short high-intensity exercise such as interval training.
Therefore, the athlete’s perception of effort remains a
critical tool for matching their actual work intensity
with session goals. Our results suggest that over a
range of interval training prescriptions, endurance
athletes are able to calibrate their exercise intensity
accurately so that perceived exertion is held con-
stant. We found that at an average lactate of about
4.5 mM, peak RPE was B17 across all four interval
conditions. Our data are consistent with those of
Seip et al. (1991), who reported an average RPE of
16.5 in trained runners running at a fixed blood
lactate concentration of 4.0 mM. There are few data
available describing RPE responses during interval-
type training sessions. Among our subjects, the
perception of effort increased linearly with time over
the course of all four interval training sessions
(Fig. 5). This was true even when running velocity,
ventilation, and blood lactate remained constant. It
appears reasonable to recommend that a typical
high-intensity ‘‘aerobic’’ interval session should feel
‘‘hard’’ (RPE 15–16) initially, and will be perceived
as ‘‘very hard’’ (RPE 17–18) by the end of the
workout.
Perspectives
In recent years, research has increasingly focused on
the physiological responses to intermittent exercise.
However, to our knowledge, this study is the first to
compare the physiological and perceptual responses
of athletes during different high-intensity aerobic
interval training prescriptions under the free-range
conditions typical of endurance training. We con-
clude that within a reasonable range of prescriptions
typical of ‘‘aerobic’’ interval training, trained en-
durance athletes automatically manipulate work
intensity appropriately to maintain similar levels of
peak exertion (an RPE of about 17) and blood
lactate. Blood lactate measurements suggest that a
quasi-steady-state is achieved, while RPE climbs
linearly at a rate that suggests an appropriate upper
limit for total work bout duration of B30 min. Four
minutes may represent a median optimal work
duration for eliciting a high cardiovascular stress to
self-paced VO
2max
intervals when the work-to-rest
ratio is 1:1. Future studies aim to investigate the
physiological and perceptual impact of manipulating
the recovery duration component of the interval
training prescription.
Key words: endurance, athletes, intermittent exercise,
heart rate, maximal oxygen consumption, perceived
exertion.
Seiler and Sjursen
324
References
A
˚strand I, A
˚strand PO, Christensen EH,
Hedman R. Intermittent muscular work.
Acta Physiol Scand 1960: 48: 448–453.
Billat V, Blondel N, Berthoin N.
Determination of the velocity
associated with the longest time to
exhaustion at maximal oxygen uptake.
Eur J Appl Physiol 1999: 80: 159–161.
Billat VL, Bocquet V, Slawinski J, Laffite
L, Demarle A, Chassaing P,
Koralsztein JP. Effect of a prior
intermittent run at vVO
2 max
on oxygen
kinetics during an all-out severe run in
humans. J Sports Med Phys Fitness
2000a: 4: 185–194.
Billat VL, Slawinski J, Bocquet V,
Demarle A, Lafitte L, Chassaing P,
Koralsztein JP. Intermittent runs at the
velocity associated with maximal
oxygen uptake enables subjects to
remain at maximal oxygen uptake for
a longer time than intense but
submaximal runs. Eur J Appl Physiol
2000b: 81: 188–196.
Borg G. Perceived exertion as an indicator
of somatic stress. Scand J Rehab Med
1970: 2(2–3): 92–98.
Christensen EH, Hedman R, Saltin B.
Intermittent and continuous running.
Acta Physiol Scand 1960: 50: 269–286.
Demarie S, Koralsztein JP, Billat V. Time
limit and time at VO
2 max
during a
continuous and an intermittent run.
J Sports Med Phys Fitness 2000: 40:
96–102.
Fox EL, Bartels RL, Billings CE, O’Brien
R, Bason R, Matthews DK. Frequency
and duration of interval training and
changes in aerobic power. J Appl
Physiol 1975: 38: 481–484.
Hoffmann P, Pokan R, Von Duvillard SP,
Seibert FJ, Zweiker R, Schmid P. Heart
rate performance curve during
incremental cycle ergometer exercise in
healthy young male subjects. Med Sci
Sports Exerc 1997: 29: 762–768.
Hoffmann P, Rokan R, Preidler K,
Leitner H, Szolar D, Eber B,
Schwaberger G. Relationship between
heart rate threshold, lactate turn point
and myocardial function. Int J Sports
Med 1994: 15: 232–237.
Jones AM, Doust JH. A 1% treadmill
grade most accurately reflects the
energy cost of outdoor running.
J Sports Sci 1996: 321–327.
McLellan TM, Skinner JS. Blood lactate
removal during active recovery related
to anaerobic threshold. Int J Sports
Med 1982: 3: 224–231.
MedbJI, Hanem S, Noddeland H,
Jebens E. Arterio-venous differences of
blood acid–base status and plasma
sodium caused by intense bicycling.
Acta Physiol Scand 2000: 168: 311–326.
MedbJI, Toska K. Lactate release,
concentration in blood, and apparent
distribution volume after intense
bicycling. Jpn J Physiol 2001: 51:
303–312.
Patterson CRM, Leary JP, Wenger HA.
Effect of different exercise intervals and
work:recovery ratios on oxygen uptake.
Sports Med Training Rehab 1997: 7:
185–192.
Savin WM, Davidson DM, Haskell WL.
Autonomic contribution to heart rate
recovery from exercise in humans.
J Appl Physiol 1982: 6: 1572–1575.
Schulz H, Helle S, Heck H. The validity of
the telemetric system Cortex X1 in the
ventilatory and gas exchange
measurement of exercise. Int J Sports
Med 1997: 18: 1–4.
Seip RL, Snead D, Pierce EF, Stein P,
Weltman A. Perceptual responses and
blood lactate concentration: effect of
training state. Med Sci Sports Exerc
1991: 23: 80–87.
Stepto NK, Hawley JA, Dennis SC,
Hopkins WG. Effects of different
interval training programs on cycling
time-trial performance. Med Sci Sports
Exerc 1999: 31: 736–741.
Stepto NK, Martin DT, Fallon KE,
Hawley JA. Metabolic demands of
intense aerobic interval training in
competitive cyclists. Med Sci Sports
Exerc 2001: 33: 303–310.
Thibault G, Marion A. Interval training-a
practical model. Coaches Rep 1999: 6:
16–20.
Responses to self-paced interval training
325
