Content uploaded by Veronique Louise Billat
Author content
All content in this area was uploaded by Veronique Louise Billat on Mar 31, 2014
Content may be subject to copyright.
Interval Training for Performance: A
Scientific and Empirical Practice
Special Recommendations for Middle- and Long-Distance
Running. Part II: Anaerobic Interval Training
L. Véronique Billat
Faculty of Sport Science, University Lille 2, Lille, France
Abstract
Studies of anaerobic interval training can be divided into 2 categories. The
first category (the older studies) examined interval training at a fixed work-rate.
Theymeasuredthe time limit or thenumber of repetitions the individual was able
to sustain for different pause durations. The intensities used in these studies were
not maximal but wereat about 130 to 160%ofmaximaloxygenuptake(
V
.
O
2max
).
Moreover, they used work periods of 10 to 15 seconds interrupted by short rest
intervals (15 to 40 seconds). The second category(the more recent studies) asked
the participants to repeat maximal bouts with different pause durations (30 sec-
onds to 4to 5 minutes). These studies examined the changes in maximal dynamic
power during successive exercise periods and characterised the associated meta-
bolic changes in muscle.
Using short-interval training, it seems to be very difficult to elicit exclusively
anaerobicmetabolism. However, these studies have clearly demonstrated that the
contribution of glycogenolysis to the total energy demand was considerably less
than that if work of a similarintensity was performed continuously. However, the
latter studies used exercise intensities that cannot be described as maximal. This
is the main characteristic of the second category of interval training performed
above the minimal velocity associated with
V
.
O
2max
determinedinan incremental
test (v
V
.
O
2max
).
Many studies on the long term physiological effect of supramaximal intermit-
tent exercise havedemonstrated an improvement in
V
.
O
2max
or running economy.
LEADING ARTICLE
Sports Med 2001; 31 (2): 75-90
0112-1642/01/0002-0075/$22.00/0
© Adis International Limited. All rights reserved.
1. Anaerobic Interval Training
1.1 Acute Physiological Responses
Studies of anaerobic interval training can be di-
vided into 2 categories. The first category (the older
studies) examined interval trainingat a fixed work-
rate.
[1]
They measured the time limit or the number
of repetitions the athlete was able to sustain for dif-
ferent pause duration. The intensities used in these
studies were not maximal but were at about 130 to
160% of maximal oxygen uptake (V
.
O
2max
). More-
over, they used work periods of 10 to 15 seconds
interrupted by short rest intervals (15 to 40 sec-
onds).
The second category (the more recent studies)
askedtheparticipantstorepeatmaximalboutswith
different pause durations (30 secondsto 4 to5 min-
utes). These studies examined the changes in max-
imal dynamic power during successive exercise
periods and characterised the associated metabolic
changes in muscle.
[2]
However, the purpose of both ofthese approaches
was to understand the involvement of the different
types of metabolism (alactic, lactic anaerobic and
aerobic) by comparing the influence of pause du-
ration [time for creatine phosphate (CP) resynthe-
sis] on the decrease in the time limit and total work
performed and the decrease in the velocity or power
output. These decreases in total work performed or
ofthe work-rate were correlatedwith thesubstrates
(glycogen, CP) and the reactants of the different
metabolic pathways by biopsy techniques, blood
analysis and gas exchange measurements.
1.1.1 Fixed Work-Rate Studies
In the first category, Margaria et al.
[1]
demon-
strated that for supramaximal intermittent exercise
[about 160% of the minimal velocity associated with
V
.
O
2max
determined in an incremental test (vV
.
O
2max
)],
the total time leading to exhaustion depended on
the pause duration. Moreover, when the pause was
0 (uninterrupted running),10, 20 and 30 seconds, the
total times run at the supramaximal velocity were
32, 100 and 200secondsand indefinite, respective-
ly. When the pause was 10 seconds, the total run-
ning time could be increased about 3 times and when
the rest period was 20 seconds, 6 times. Doubling
the period of rest allowed the athlete to work twice
as long. Indeed, blood lactate accumulation was
equalto 11and 7.5mmol/Landstayedatthesteady
stateof 2 mmol/L,respectively,forthe10-,20- and
30-second pauses. This study demonstrates thatthe
pause duration was the determinant for the utilisa-
tion of anaerobic and aerobic pathways. By plot-
ting theblood lactate accumulation as afunction of
the length oftherest period, Margariaetal.
[1]
showed
that theminimal timeof restat which no lactic acid
accumulation took place was about 25 seconds. This
duration corresponds to the half-reaction time for
the alactic oxygen deficit payment (i.e. CP resyn-
thesis). These authors measured oxygen uptake
(V
.
O
2
)breathbybreathinarunnerperforming10
bouts of exercise with a 20-second pause and they
showed that there was a very fast increase in V
.
O
2
at the beginning of the exercise and there was no
appreciable difference during the first pause com-
pared with the firstrunning period. At steady state,
from the fifth repetition, V
.
O
2
was slightly lower
during the pause than during running (e.g. 3.5 vs 4
L/min for 1 participant). The first run took place
almost exclusively (90%) at the expense of phos-
phagen. During the following 25 seconds of rest,
the alactic pool was restored to about 45% by the
oxygen debt payment (fast post oxygen compo-
nent).
During the second run, energy was provided more
from the oxidative mechanism: 25 versus 10% of the
totalenergyrequirementinthefirstandsecondbouts,
respectively. This is too high to be sustained by the
alactic anaerobic metabolism, since the phospha-
gen pool is only equal to 45% of the initial resting
storage value. Consequently, anaerobic glycolysis
was involved and blood lactate accumulated. This
is whythe 30 seconds of pause, which is sufficient-
ly long to allow the phosphagen pool to be resyn-
thesised, also allows the athlete to run for a very
long time with a steady-state blood lactate below
2.5mmol/L. Withpausesofonly10 seconds,V
.
O
2max
is reached earlier and blood lactate accumulates.
This kind of interval training performed at supra-
maximal intensity can elicit a completely different
metabolism, depending upon the length of the pause.
If the oxygen deficit during the period of supra-
maximal work is higher than the alactic anaerobic
oxygenequivalent(0.5L),theintervaltrainingwill
involve lactic anaerobic processes. For instance, a
female runner who has a vV
.
O
2max
equalto16km/h
(4.44 m/sec) will run 10
× 4.44 × 1.6 = 71.1m dur-
ing a 10-second supramaximal run at 160% of
vV
.
O
2max
. If she weighs 50kg and has a gross oxy-
gen cost of running equal to 0.210 ml/kg/m, she
will need to consume a volume of oxygen equal to
71.1m
× 0.210 = 15ml of oxygen/kg, i.e. 750ml.
This is equal to 150% of the oxygen equivalent of
anaerobic metabolism. We can therefore calculate
76 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
that for the 10-second exercise bouts, the velocity
should be proportionately less, i.e. close to vV
.
O
2max
.
However, the pause should be at least 25 seconds
to be sure that the CP consumed during the work
interval is restored. As demonstrated by Margaria
et al.,
[1]
in these conditions interval training can be
performed for a very long time with no blood lac-
tate accumulation. Hence, the work
:
pause ratio
is equal to 1
:
3andV
.
O
2
is submaximal (82% of
V
.
O
2max
).
[3]
Using short–short interval training, it seems to
be very difficult to elicit exclusively anaerobic
metabolism. However, these studies have clearly
demonstrated that the contribution of glycogenol-
ysis to the total energy demand was considerably
less than that if work of a similar intensity was per-
formed continuously. However, the latter studies
used exercise intensities that cannot be described as
maximal. This is the main characteristic of the sec-
ond category of interval training performed above
vV
.
O
2max
.
More recently, Tabata et al.
[4]
have compared
the metabolic profiles of 2 different types of high
intensity interval training at a fixed percentage of
V
.
O
2max
equivalent. Participants cycled 6 bouts of
20 seconds at 170% V
.
O
2max
with a 10-second rest
between each bout. The second interval training
involved 4 bouts of 30 seconds at 200% of V
.
O
2max
with a 2-minute rest between each bout. The first
intervaltrainingprotocolelicitedboththemaximal
accumulated oxygen deficit and the maximal oxy-
gen consumption. Therefore, the authors concluded
that this supramaximal interval training might tax
both the anaerobic and aerobic energy releasing
systems at close to their maximal capacity.
1.1.2 Fixed Intensity Studies
It should be noted that we cannot really call these
exercises ‘interval training’, but rather ‘repeated
maximal sprints’. Maximal dynamic exercise has
been recently (for 10 years) used by physiologists
to study the regulation of metabolic pathways and
to shed light on the aetiology of fatigue (inability
to sustain a given power output) during high inten-
sityexercise. In thistypeof study,Bogdanisetal.
[5]
demonstratedthat in asecond30 seconds of all-out
cyclingexerciseperformed 4 minutesafterthe first
bout, CP, which was at 17% of the resting value
after the first all-out cycling exercise, was resyn-
thesised to 79% of the resting value even though
the pH remained at a low level (pH 6.8). Despite
the 41% reduction in anaerobic energy, the total
work carried out during the second 30-second sprint
was reduced by only 18%. Aerobic metabolism pro-
vided a significant part(49%) of theenergy and the
rate of CP synthesis determined the work-rate per-
formed in the first 10 seconds of the second 30-
second all-out exercise.
In another major study on intermittent maximal
exercise training, 10
× 6-second maximal sprints
with 30 seconds of recovery between each sprint
were performed.
[6]
Needle biopsy samples were
taken from the vastus lateralis muscle before and
after the first sprint and 10 seconds before and im-
mediately after the tenth sprint. In this study, the
high mean power output was more than 3 timesthe
minimal power output which elicits V
.
O
2max
deter-
minedinanincrementaltest(pV
.
O
2max
). The peak
power (the maximum power output in 1 second)
was 5 times greater than pV
.
O
2max
. Mean power out-
puts significantly decreased from the fourth repe-
tition and the peak power decreased from the fifth.
Thelastrepetitionwas reduced to only 73% of power
generated during the first sprint. The energy re-
quired to sustain the high mean power output that
was generated over the first 6-second sprint was
provided by an equal contribution from CP degra-
dation (–57% of rest concentration) and anaerobic
glycolysis (muscle lactate had increased to 28.6
mmol/kg dry weight). In the tenth sprint, muscle
lactate concentration did not increase. The authors
suggested thatduring thelast sprint, the power out-
put (which was 73% of the first repetition) was
maintained by energy that was derived mainly from
CPdegradation and an increase in aerobic metabo-
lism.
For training purposes, because of the drop in the
power output and because of the high muscle lac-
tate accumulation, this configuration of intermit-
tent maximal sprint is not used. The purpose of
sprint training is to increase the maximal velocity
Anaerobic Interval Training and Performance 77
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
for 6 to 10 seconds (60 to 100m in running); in
order to achieve this, the intermittent maximal run
is separated by 4-minute pauses to allow the CP to
be restored so that each repetition can be run fast.
To increase the glycolysis pathways, which account
for40to50%oftheenergynecessarytocover
100m, the intermittent training consists of a series
of 100, 120 and 150m runs at 88 to 90% of the best
performance with a passive rest of 5 to 6 minutes
betweeneachbout.However, inthisprotocol,ithas
been demonstrated that during passive rest almost
all the CP was resynthesised after 4 minutes (from
19.8 to 36.9 mmol/kg dry muscle, instead of 39
±
3.2 mmol/kg).
[7]
The half time of CPresynthesis is
170 seconds.
[8]
However, thesestudies havea directapplication
for sprint training. Indeed, Balsöm et al.
[2]
have
compared the physiological responses to maximal
intensity intermittent running for 15m (in 2.6 sec-
onds), 30m (in 4.5 seconds) and 45m (in 5.6 sec-
onds). These sprints were performed every 30 sec-
ondsbutcovered thesamedistance(work)of600m
(40
× 15m, 20 × 30m, 15 × 40m). The time of the
last sprint of 15m was not significantly different
from that of the first (2.63
± 0.04 vs 2.62 ± 0.02
seconds). This was not the case for the 30m and
40m sprint times, which increased significantly. Ve-
locity at 40m decreased after the third sprint, asso-
ciated with a net loss of the adenine nucleotide pool.
Balsöm et al.
[9]
also compared the effect of recov-
ery duration (120, 60 and 30 seconds) on the 40m
maximal intensity intermittent exercise. The per-
formance (time) during the first 15m (acceleration
phase) was only affected by the shortest recovery
(30 seconds). Blood lactate reached 17 mmol/Lfor
this short recovery compared with the value regis-
tered using 60 and 120 seconds recovery (12.1
±
1.3and13.9±1.2mmol/L,respectively).Although
these work bouts could be classified primarily as
anaerobic exercise, oxygen uptake measured dur-
ing rest periods(Douglasbag method)increased to
52, 57 and 66% of V
.
O
2max
. However, to improve
performancein competitionslasting1 minute(1km
in track cycling or 100m in swimming and 400m
inrunning),generally performed at 150% of V
.
O
2max
,
it is important to practise aerobic interval training,
since the aerobic metabolism contribution to the
total energy is about 30%.
[2]
The increase in recovery time allowed the run-
ners to maintain their performance during moretri-
als; however, the increment of recovery duration
had no effect on the total adenine nucleotide pool
estimated by plasma hypoxanthine and uric acid
concentrations. When the rate of ATP hydrolysis
exceeds the rate at which it can be resynthesised,
AMPiseliminatedvia deaminationofAMPtoIMP
and subsequent oxidation to hypoxanthineand uric
acid.
It is well known that both anaerobic pathways –
lactic (glycolysis) and alactic (CP degradation) –
are activated instantaneously at the onset of maxi-
malactivity.
[10]
However, the abilityto repeatmax-
imal sprints depends on the duration of recovery,
which does not have the same effect on the 2 an-
aerobic pathways. The resynthesis of CP depends
on the endurance level of the participant.
[5,11]
The
percentage of CP resynthesis and the percentage
restoration of themean power output and pedalling
speedmeasuredonthefirst10secondsinthesec-
ond 30-second all-out exercise were highly corre-
lated (r = 0.84 and 0.91, respectively). Indeed, the
time course of CP resynthesis after a 30-second
sprint was found to be parallel with the time course
of the peak power output (PPO) restoration.
[12]
As described above, glycogenolysis is largely
elicited in these supramaximal intermittent exer-
cises.Intensityofworkintervaltrainingiswellabove
the crossover point: i.e. the power output at which
energy from carbohydrate-derived fuels predomi-
nates over energy from lipids.
[13]
Jenkins et al.
[14]
have demonstrated that a high (83% of the total
energy intake) and moderate (58% of the total en-
ergy intake) carbohydrate diet for 3 days preceding
supramaximal intermittent exercises would allow
performance to be maintained during 5 all-out
cycling bouts. Each of these 5 exercise periods,
at about 125% of pV
.
O
2max
, were separated by 5
minutes of passive recovery.
78 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
1.2 Long Term Physiological Effects
Many studies on the long term physiological ef-
fect of supramaximal intermittent exercises have
demonstrated an improvement in V
.
O
2max
.Inthe
1970s, Fox and his team reported that this kind of
interval training was efficient in inducing an im-
provement in V
.
O
2max
(+15%) in men
[15]
and in
women.
[16]
Indeed, Lesmes et al.
[16]
compared the
effects of 2 types of supramaximal interval training
over an8-week period on V
.
O
2max
(2 or 4 days/week)
in women with a vV
.
O
2max
equal to about 12.9 km/h,
using highintensity, short distance(50, 101, 201m)
at 170% of vV
.
O
2max
(22 km/h) or high intensity,
longdistance(604,805,1208m)at130%ofvV
.
O
2max
(15 km/h).
For both types of interval training, the relief in-
tervals consisted of walking 2 to 3 times the dura-
tion of the work intervals. The improvement was
the same for both short and long supramaximal
interval training and for both frequencies (2 or 4
training days/week). The authors concluded that
the change in aerobic power and submaximal heart
ratefor femaleswas independentof frequency,dis-
tanceand intensity. Incontrast,formen,ithasbeen
shown that training intensity, ratherthanfrequency
or distance, was the most important factor to im-
prove V
.
O
2max
.
[17]
The decrease in submaximal heart
rate during a run of 5 minutes at 50% of vV
.
O
2max
using high intensity interval training was due to a
decrease in sympathetic drive and probably due to
theincreaseofstrokevolume.Thefactthatthe
response was independent of intensity and frequen-
cies of interval training could mean that the maxi-
malresponsivenessto thisprogrammewasreached
for the women, who started with a V
.
O
2max
of only
40 ml/min/kg.
As demonstrated in rats,
[18,19]
anykindof training
programme would havethe same effect on V
.
O
2max
(+15%)and onpV
.
O
2max
.However,theproblem with
such longitudinal studies is the low degree of fit-
ness of the participants and the choice of so-called
‘supramaximal intermittent exercises’, which are
often less intensive than imagined by the authors.
Indeed, before drawing conclusions on the ineffi-
ciency of high intensity intermittent training in
improving anaerobic capacity, it is important to
check that the inputis well abovethe power output
associated with V
.
O
2max
(pV
.
O
2max
). For instance,
no increase was found in either the level of anaer-
obic metabolic enzymes or the percentage of type
IIb fibres due to the effect of training (only an in-
crease in the absolute number).
[20-22]
In fact, the
intensity of intermittent exercise (10- or 15-fold
× 15 to 30 seconds) was set at 60% of the absolute
maximal velocity, which corresponds to 110 to
120% of pV
.
O
2max
. The intensity of the longerhigh
intensity intermittent training (4- or 5-fold
× 60 to
90seconds)wassetat70%ofthemaximalvelocity
sustained for 90 seconds. All these intensities cor-
respond,effectively,to90to95% of vV
.
O
2max
,know-
ing that an exercise sustained for 1.5 minutes is at
about 130% of pV
.
O
2max
.
[23]
It should be noted that
the intensities of these exercises were probably too
low to enhance anaerobic metabolism or modify
thepercentageoftypeIIfibres.Theresponsewas
the same as in interval training for improving aer-
obic metabolism, using velocities between 90 and
120% of vV
.
O
2max
to give participants the time to
reach and therefore to elicit V
.
O
2max
.
[23]
In contrast, 1 study
[24]
has used truesupramaxi-
mal exercise: 2 series of 4 repetitions of 200m at
90% of the participants’ maximum velocity over
200m (which gave 29 seconds) with a recovery of
2 minutes (3 to 4 times per week) for 5 weeks. At
theend of the firstseriesthebloodlactatelevel was
14.5 mmol/L; after a rest of 10 minutes between
the 2 series, blood lactate was still 13 mmol/L, and
at the end of the second series of 200m blood lac-
tatewas 17.6mmol/L.Thistypeofintervaltraining
clearly elicited anaerobic glycolysis. Indeed, this high
intensity interval training with a 1
:
4work
:
rest
ratio resulted in an increase in the activity of key
enzymesinvolvedinglycogenolysisandanaerobic
glycolysis in skeletal muscle. The increase in suc-
cinate dehydrogenase (SDH) [+17.5%] was not sta-
tistically significant. Participants had improved their
time limit in a run at 16 km/h with a 15% slope
inclination (about 25 km/h equivalent on leveltread-
mill according to Margaria et al.
[25]
), i.e. at about
140to150%forthesemen.
Anaerobic Interval Training and Performance 79
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
Even very short, repeated, 5-second all-out sprints
followed by a recovery of 55 seconds induced an
increase in the proportion of slow twitch fibres, in
accordance with Simoneau’s studies.
[20-22]
More-
over, 7 weeksofthiskindofsprinttraining notonly
increasedlactateproductionafteratrainingsession
[associated with a 20% higher activity of both phos-
phofructokinase and lactate dehydrogenase (LDH)],
but also increased the proportion of slow twitch
fibres which was closely related to a concomitant
decrease in fast IIb fibres.
[26]
However, V
.
O
2max
did
not increase and the decrease in power output dur-
ing each sprint was not reported. Notably, 55 sec-
onds is too short toreconstitute the CP reserve. Thus,
intense and repetitive exercise could induce the con-
versionof intermediatefibres to theoxidativetype,
aerobic metabolism being involved in the recovery
and probably increasingly during the repetition of
the sprints.
[26]
Maximal power output measured with
the force-velocity test increased significantly after
sprint training (+13%). This was similar for the aver-
age power output in an all–out test of 30 seconds
(+27%) and for the maximal work-rate reached in
the Wingate test. These improvements in perfor-
mance were not related to a significant increase in
the mean resting muscle CP concentration, but to
an increase in the energy production from anaero-
bic glycolysis, as demonstrated by Nevill et al.
[27]
Because the training exercises were performed
intermittently, and because of their greater oxida-
tive capacity, type I fibres could have been more
extensively involved in the replenishment of the
depletedCPstoresduringrestperiodsthanfasttwitch
fibres. Moreover, type I fibres are more involved
in the removal of the lactate accumulated during
exercise periods (52.8 mmol/kg of dry mass) than
type II fibres. Indeed, in order to increase alactic
anaerobic metabolism, interval training acts: (i) by
increasing the ability to decrease CP as rapidly as
possible by sprint exercises separated by sufficient
rest (at least 4 minutes) to restore the CP reserve,
avoiding the involvement of anaerobic glycolysis;
and (ii) by increasing the ability to replenish as
quickly as possible the CP reserve. To accomplish
this, it is necessary to have muscle fibres with a
high oxidative capacity. However, with short pauses,
as in the study by Linossier et al.
[26]
(55 seconds),
anaerobic lactic metabolismis increasinglyinvolved.
Acidosis could impair CP production via mitochon-
drial creatine kinase during recovery.
Hargreaves et al.
[28]
studied6menduring4×
30-second all-out exercise bouts. The first 3 exer-
cise bouts were separated by 4 minutes of passive
recovery; after the third bout there was a 4-minute
rest period, followed by a 30-minute exercise pe-
riod at 30 to 35% peak V
.
O
2
. This was followed by
a further 60-minute rest period before completing
the fourth exercise bout. Theperformance was main-
tained in the fourth bout because CP resynthesis is
increased. Indeed, peak power and total work were
not significantly decreased in the fourth bout com-
pared with the first bout and were higher than in
the second and third bouts. Before the fourth bout,
muscle CPconcentration was above the resting level
(probably because of increased mitochondrial cre-
atinekinaseactivityduring recovery);IMP,lactate,
pH and sarcoplasmic reticulum calcium uptake were
thesameasbefore thefirstbout.However,ATPand
glycogen were still depleted. Since, in these meta-
bolic conditions, the performance was the same as
in the first bout, the authors concluded that the main-
tenance in performance observed during the fourth
bout does not appear to be related to a reduction in
muscle glycogen.
[28]
Heugas et al.
[29]
reported on elite 400m runners
(including an Olympic champion), in the precom-
petitive phase, who performed training sessions with
maximal all-out 30-second runs with 4 minutes of
rest in between, followed by a long rest (10 min-
utes) before a last 30-second maximal run. The ca-
pacity to sustain exercise at about 150% of vV
.
O
2max
with high acidosis and the ability torun a relayrace
(4
× 400m) a couple of hours later is improved by
using this type of intermittent high intensity train-
ing.
[29]
In a recent study
[30]
where the purpose was to
examine the effects of sprint interval training on
muscle glycolytic and oxidative enzyme activity
and exercise performance, it was reported that 4
highintervaltrainingsessionsperweek (consisting
80 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
of 4 × 30 seconds of all-out exercise spaced by a
recovery of 4 minutes, or 2 minutes 30 seconds in
weeks 1 and 2) enhanced PPO, total work over 30
seconds and V
.
O
2max
. Maximal enzyme activity of
hexokinase, phosphofructokinase (glycolysis en-
zymes), citrate synthase, SDH and malate dehy-
drogenase (oxidativeenzymes) wasenhanced after
thistypeoftraining.The participantswerestudents
in kinesiology, who were physically active but were
not athletes. From the seventh week of such train-
ing, they were able to perform 10 intervals with 2
minutes 30 seconds of recovery per session. This
type of high interval training induced an increase
in blood lactate of up to 32 mmol/L after the tenth
exercise interval. The power output ranged from
210%of V
.
O
2max
onthefirstintervalto140%V
.
O
2max
by the ninth andtenth intervals.At suchintensities,
the rate of production of pyruvate may be consid-
ered to be almost maximal and one would expect
major increases in the velocity of catalytic activ-
ity of the competingenzymespyruvatedehydroge-
nase(PDH)andLDH.However,the increaseof7%
in LDH after such training was not significantly
different; PDH was not measured, but the authors
speculatedthattheincreaseinPDHactivity(which
provokes an increased rate of pyruvate entry into
the mitochondria)could be the stimulusfor the up-
regulation of mitochondrial enzymes. This could ex-
plain the fact that 3 to 4 repetitions of 5 minutes of
effective high interval training per week (hence 20
minutes per week only) can result in an increase in
both glycolytic and oxidative muscle enzyme activ-
ity, maximum short term power output and V
.
O
2max
(+7%). The authors concluded that the increase in
power output (+24% on average after 4
× 30-second
bouts) may have been the result of an increase in the
maximal activity of the glycolytic enzymes (+49%
forphosphofructokinase)andinNa
+
/K
+
pumpcapac-
ity, whereas the increased mitochondrial enzyme
activity (+65% for SDH) may have been a result of
increased pyruvate flux rateduring thisintense sprint
interval training.
[30]
Thus, interval training using short all-out bouts
of exercise elicits the glycolytic pathway and can
be used to prepare long sprint (200 to 400m) and
middle distance (800 to 1500m) runners whose
anaerobic metabolism is a determinant for perfor-
mance. However, victory in longer races such as
5000 and 10 000m depends on the ability to cover
the last lap at a velocity well above vV
.
O
2max
,
[31]
although there are also 5000 and 10 000m races
with 1 runner 50 to 100m ahead when starting the
last lap, andthis runner can win at a modestlastlap
speed. The last 400m of a 10 000m racecan be run
in 52 seconds, i.e. 27.7 km/h (110% of vV
.
O
2max
for the best performers, who have a vV
.
O
2max
of 25
km/h). Also, athletes must enhance their anaerobic
capacity to be able to accelerate in championship
races. For this, they can use supramaximal interval
trainingoflongerduration:1minuteatabout130%
ofvV
.
O
2max
with arest of 5 minutes between. Eight
weeks of this type of training (3 times a week) has
been reported to efficiently enhance anaerobic ca-
pacity as measured as the y-intercept of the rela-
tionship between work rate and time limit.
[32]
The
critical power, i.e. the slope of this relationship,
was not modified. This critical power depends on
aerobic capacity and has been shown to be related
to V
.
O
2max
and lactate threshold.
[33]
This type of
high intermittent training has also been used to en-
hance performance in high level cyclists (see sec-
tion 3), in whom it is more difficult to obtain an
improvement in performance or physiological mod-
ifications.
Another type of anaerobic high interval training
is circuit training consisting of explosive jumping
exercises. Recently, with this type of circuit train-
ing, Paavolainen et al.
[34]
reportedanimprovement
of5kmtimeinwelltrainedenduranceathleteswith
nochangesin theirV
.
O
2max
.Thiswasduetoimprove-
ment of the running economy (RE), i.e. the oxygen
consumption at a given sublactate threshold veloc-
ity was decreased by 3% after 6 weeks of 3
× 20 to
30 minutes of circuit training per week with long
interval training or exhaustive distance training, but
not withshort interval training.
[35]
Thisimprove-
ment of RE was related to improved neuromuscu-
lar characteristics, allowing an increase of velocity
over 20m and over a distance covered in 5 jumps
that were transferred into improved RE. Improved
Anaerobic Interval Training and Performance 81
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
RE following intensive training was correlated
with reduced ventilatory demands (VE). This im-
provement in RE was related to a decrease in VE,
whereas muscle fibre composition and respiratory
exchange ratio, stride length and frequency during
running were unaltered by training.
If long interval training and long distance run-
ning induce a decrease in RE, circuit training com-
posed of jump exercises can yield the same result,
but probably by different mechanisms through the
stretch-shortening cycle.
[36]
In fact, it is probable
that the fitness level of participants plays a promi-
nent role in the results of the long term effect of
different types of training (continuous or intermit-
tent).
Finally, for not very well trained participants, the
intensity of exercise is more important than the type
ofexercise(cyclingandrunning).Flynnetal.
[37]
have
reported that in well trained runners, 6 weeks of high
intensity interval training (between 95 and 105%
pV
.
O
2max
) performed either with cross-training
(cycling and running) or specific running training
yielded the same performance improvement (–30
seconds over 18 minutes 15 seconds on a 5km race).
However, in all these studies, conclusions are based
on results obtained in low-level athletes. This un-
derliesthe fact that sports physiologists have a lim-
ited impact on the training practices of successful
competitors. Training programmes are more often
based on the experience of track and field coaches
and athletes.
[38,39]
1.3 Anaerobic Interval Training and
Creatine Ingestion
In a study of the effects of creatine supplemen-
tationon cyclingperformance(10
×6-second bouts
at 880Wand 140 rev/min,separated by passive 30-
second rest periods),performance couldbeenhanced
by creatine towards the end of each exercise bout,
as demonstrated by a smaller decline in work out-
put from a baseline throughout the 10 trials in the
last 2 seconds of the 6-second exercise periods.
[40]
Moreover, although more work was performed in
this intermittent maximal test after versus before
the administration period, blood lactate accumula-
tion decreased from 10.8 to 9.1 mmol/L. This im-
provementwasdue to a higheravailabilityofphos-
phate or to an increased rate of CP resynthesis dur-
ing the recovery periods. This could explain why
glycolysis is less involved in interval training at
maximal velocity.
High intensity intermittent training has been
shown to be a very effective way of increasing the
maximal oxygen uptake. Tabata et al.
[41]
reported
thataprotocolof6
×20seconds at 170% of V
.
O
2max
for 6 weeks increased V
.
O
2max
by 13% and anaero-
biccapacityby28%.Infact,itseemsthatprotocols
that elicit V
.
O
2max
, even if they are performed at
supra-vV
.
O
2max
velocities,canincreaseV
.
O
2max
and
anaerobic capacity. To do this, the recovery has to
be short, as in Tabata’s interval training (10 sec-
onds).
[41]
2. Interval Training for the Elderly and
the Young
2.1 The Elderly
Of the various methods used to train elderly peo-
ple, interval training has been reported to be effec-
tive in improving aerobic capacity (V
.
O
2max
and lac-
tate threshold).
In 1 study, interval training was more easily ac-
cepted (more than 70% of participants completed
both test and training sessions) than continuous
training (40%).
[42]
The interval training was per-
formed on the track, taking as a reference the heart
rate at the ventilatory threshold (HR
th
)of125± 5
beats/min, corresponding to 60% V
.
O
2max
.Thein-
terval training consisted of 1 minute exercise at HR
th
and 1 minute active recovery (20 beats/min below
HR
th
); 2 minutes exercise at HR
th
, 1 minute active
recovery; 4 minutes exercise at HR
th
,1minuteac-
tive recovery; and 10 minutes exercise at HR
th
,3
minutes active recovery.
From session to session, the order and duration
of the exercise/recovery sequences were varied in
such a way that the participants progressively in-
creasedtheir total exercise time from 30 to 60 min-
utes by the eighth week of training. This procedure
can be performed on the track and needs only a
82 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
cardiofrequencymeter.Asforyoungerrunners,the
elderly benefit from this interval training by in-
creasing oxidative enzyme activity,
[43]
V
.
O
2
at the
ventilatory threshold and cardiac output, since max-
imal oxygen pulse increased by 19% (as V
.
O
2max
).
Moreover, the attendance at these training sessions
duringperiodsaslongas3months(24sessionsand
12 interval training sessions) has been reported to
be more than 97.3% for the 73% of participants
who had been considered to have completed train-
ing and testprotocols.Intervaltraining significant-
ly increased V
.
O
2max
by 20% (from 25 ± 1.2 to 31
± 2.3 ml/min/kg) and the velocity associated with
the ventilatory threshold (vVT) by 26%.
However, as for younger participants, the train-
ing effect depends on the basal level of the partic-
ipants. In training camp, we often have senior run-
ners (>60 years) who are able to run for 6 minutes
at 18 km/h. They are generally very long distance
runners (marathon and 100km races). For those who
already have a high V
.
O
2max
(60 ml/min/kg at age
60 years) and maximal lactate steady-state ve-
locity (vMLSS = 85% of vV
.
O
2max
) and are not fa-
miliar with interval training, very short repetitions
are well supported.
[44]
Figure 1 shows the V
.
O
2
ki-
netics for a60-year-oldrunner(eighthinsenior100km
world championship 1999)during a very short inter-
val of 15 seconds at 90% and 15 seconds at 80% of
vV
.
O
2max
.The average velocity was at MLSS veloc-
ity (85% vV
.
O
2max
). It should be pointed out that
this very narrowamplitude training interval allows
him to remain for more than 10 minutes at V
.
O
2max
.
Blood lactate accumulation was only 8 mmol/L.
Moreresearchis neededtoallowtheseseniorswho
are already in good shape to improve their perfor-
mance, but by running fewer kilometres.
Nowadays, many runners participating in pop-
ular races over 5 to 100km are middle-aged (45 to
60 years). After several years ofslow long distance
training, they no longer improve their performance.
Following this type of training, these long distance
runners have a high endurance index, which is de-
0
00:55
0
01:50
02:45
03:40
04:40
05:35
06:30
07:25
08:20
09:15
10:10
11:05
12:00
12:55
13:50
14:45
15:40
16:35
17:30
18:25
19:20
20:15
21:10
22:05
23:00
500
1000
1500
2000
2500
3000
3500
4000
VO
2
(ml/min)
·
Time (min:sec)
0
20
40
60
80
100
120
140
160
180
200
Heart rate (beats/min)
VO
2
Heart rate
·
Fig. 1. Time course of oxygen uptake (V
.
O
2
) and heart rate during short interval training [15 seconds at 90% of vV
.
O
2max
(minimal
velocityassociated withmaximal oxygenuptake,
V
.
O
2max
)alternated with 15 secondsat 80%vV
.
O
2max
]inalongdistancerunneraged
60 years. His
V
.
O
2max
was 3350 ml/min.
Anaerobic Interval Training and Performance 83
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
fined as the ability to use a high fraction of maxi-
maloxygenconsumption(V
.
O
2max
)foragivenrun-
ning duration.
[45]
Therefore, to improve their per-
formance, they have to increase both V
.
O
2max
and
thevelocityassociatedwithV
.
O
2max
(vV
.
O
2max
).
[46-48]
For this purpose, interval training involves repeat-
ed bouts of work, each lasting from about 30 sec-
onds at vV
.
O
2max
to 5 minutes at 95% of vV
.
O
2max
.
[49]
Gorostiaga et al.
[50]
showed that interval training
with repetitions of 30 seconds of work at 100% of
vV
.
O
2max
,separatedby30secondsofrest,produced
largerincreasesin V
.
O
2max
thancontinuoustraining
at 50%vV
.
O
2max
. Recently, as seen above,it hasbeen
shown that the time spent specifically at V
.
O
2max
was much higher in 30/30-second light-heavy ex-
erciseintervalsthan in acontinuoussevere run per-
formedatanintermediatevelocity between the lac-
tate threshold and vV
.
O
2max
.
[51]
However, the types of interval training described
above seem to be too hard for middle-aged runners
who want to begin interval training. It may be pos-
sible to use shorter durations of less than 20 sec-
onds. Christensen et al.
[3]
showed previously that
very short intermittent runs alternating with heavy
intensity repetitions of 15 seconds at 100% vV
.
O
2max
with complete rests of 15 seconds also allowed the
runners to reach 90% V
.
O
2max
.Inthiscase,thein-
termittent exercise was continued for 30 minutes
with a blood lactate level of only 4 mmol/L until
the 20th minute (reaching 7 mmol/L at the end of
the exercise). This low blood lactate level was ob-
tained by the unloading of myoglobin oxygen and
CP stores and by their rapid recharging during the
passive recovery period. To elicit V
.
O
2
at its maxi-
mum with this ‘short–short’ interval training, ac-
tive pausemaybe preferabletopassiverecovery
[51]
and a small range of velocities between the high
and low velocity bouts has to be used. This differ-
ence in velocity is called the ‘amplitude’of the In-
terval training.
[52,53]
The amplitude describes the
degree to which the work intensity in the different
periods of the exercise varies from the average ve-
locity.
[51]
For instance, for an interval training us-
ing high and low velocity bouts of 100 and 50% of
vV
.
O
2max
, the average velocity is 75% vV
.
O
2max
,and
the amplitude of the interval training is equal to
[(100 – 75)/75]
× 100 = 33%.
For elderly runners, the average velocity of
interval training has been chosen to correspond to
the critical velocity, i.e. the vertical asymptote of
the velocity-time relationship.Thecriticalvelocity
is known to be sustained for 30 minutes
[54]
and to
be the velocity above which runners reach V
.
O
2max
with exercising time.
[55]
Moreover, in a group of
low level marathon runners (3 hours 51 minutes
±
27 minutes), the critical velocity was more closely
correlated with the marathon performance than
vV
.
O
2max
.
[56]
Therefore, the purpose of this study
[56]
was to
compare run time at V
.
O
2max
using a short–short in-
terval training protocol with the same average ve-
locity (the critical velocity) but with different am-
plitudes. We hypothesised that for middle-aged
runners, a very short interval training (15 seconds
of hard alternated with 15 seconds of easier run)
with asmallamplitudewould allow themto run for
a longer distance ata higher velocity and for alonger
time at V
.
O
2max
. In our group of middle-aged run-
ners, having an average critical velocity of 85.6
±
1.2% vV
.
O
2max
, this study compared the following
types of interval training:
• 90 to 80% vV
.
O
2max
(for hard bouts and active
recovery periods, respectively), amplitude equal
to 6%
• 100 to 70% vV
.
O
2max
(for hard bouts and active
recovery periods, respectively), amplitude equal
to 18%
• 110 to 60% vV
.
O
2max
(for hard bouts and active
recovery periods, respectively), amplitude equal
to 30%.
The results showed that short intermittent exer-
cise of 15 seconds at vV
.
O
2max
alternated with 15
seconds of active recovery, run at an averageveloc-
ity equal to the critical velocity (85% of vV
.
O
2max
in
this group), allowed middle-aged runners to reach
and sustain V
.
O
2max
. In fact, the result shows dis-
tinctly that the low and medium amplitudes in-
duced the same metabolic (blood lactate and time
spent at V
.
O
2max
) and performance (distance, time
limit) responses.
84 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
However, as underlined by Astrand and Ro-
dahl,
[57]
‘it is an important but unsolved question
which type of training is most effective: to main-
tain a level representing 90% of the maximal oxy-
gen uptake for 40 minutes, or to tax 100% of the
oxygenuptakecapacityforabout16minutes’.This
remains an open question. But before beginning
longitudinal studies to try to resolve this, it is first
necessary to identify the metabolic response during
different types of interval training (very short to
long; see Daniels & Scardina
[49]
)usedbytrainers,
especially for special (new) populations ofrunners
such as middle-aged runners.
Even if optimum improvement in cardiores-
piratory fitness is thought to occur following train-
ing at an intensity corresponding to 90 to 100% of
V
.
O
2max
,
[58]
this central factor is not the only one to
induce an improvement in performance.
If we consider the distance run at a high velocity,
the longer distance(2390
±815m)hasto be perform-
ed with the lower amplitude 90 to 80% vV
.
O
2max
and the shorter with the higher amplitude 110 to
60% vV
.
O
2max
(1048 ± 365m). However, the dis-
tance run was not significantly different between
the 90 to 80% vV
.
O
2max
and 100 to 70% vV
.
O
2max
interval training runs (2390 ±815m vs1980 ±611m,
p =0.23).When we considerthe distance run atthe
high velocity only, the difference was smaller be-
cause of the higher peakvelocity in the 100 to 70%
vV
.
O
2max
procedure compared with the 90 to 80%
vV
.
O
2max
procedure (1265 ±431 vs 1165m, p = 0.6).
However, the time spent at V
.
O
2max
is not the
only parameter to be taken into account to judge the
possibleefficiencyofagiventypeof interval train-
ing on the improvement of V
.
O
2max
. Even if the 90
to 80% of vV
.
O
2max
interval training elicited V
.
O
2max
for the same duration as the 100 to 70% of vV
.
O
2max
,
the difference in peak velocity is 10% of vV
.
O
2max
,
i.e. 1.6 km/h. This difference could reduce the in-
volvement of the fast IIa fibres, and prevent them
from enhancing their oxidative capacity. However,
evenif runningat90%ofvV
.
O
2max
insteadof100%
of vV
.
O
2max
elicits the type IIa fibres less, we must
emphasise that 90% of vV
.
O
2max
was just 1 km/h
above the lactate threshold velocity. Therefore we
can hypothesise that even at 90% of vV
.
O
2max
,type
IIa fibres are recruited.
In fact, central factors related to oxygen uptake
are not the only limiting factors, even in long dis-
tance running. In addition to aerobic processes, neu-
romuscular and anaerobic characteristics are also
involved.
[47,59]
Indeed, as suggestedby Noakes,
[47]
the benefits of training also depend on the distance
covered at a high velocity, which determines mus-
cular adaptation maximising the number of power-
ful muscle contractions, interval training at 100 to
70% of vV
.
O
2max
is preferable to that at 90 to 80%
of vV
.
O
2max
.
For elderly runnerswho areused to train at long
slow distances, a very short (15/15 seconds) inter-
val training at 100% of vV
.
O
2max
with a recovery at
70%ofvV
.
O
2max
seemsbeagoodproceduretocon-
ciliate central and peripheral adaptations.
[44]
How-
ever, longitudinal studies are now necessary to con-
firm this hypothesis.
2.2 The Young
Asfor elderlyrunners, few studieshavefocused
on interval training in children and adolescents.
The influence of pretraining condition on effec-
tiveness of the interval training has been tested in
15- to 19-year-olds.
[60]
It was found that the in-
crease in work rate at a heart rate of 170 beats/min
was significant for untrained participants versus
trained athletes (+15 vs +5%). Interval training con-
sisted of 4 to 8 repetitions of 200m (3 times per
week for 5 weeks).Velocity and recovery were not
specified.Giventhat the numberofrepetitions was
similar for trained and untrained participants, the
stimulus was perhaps insufficient for trained par-
ticipantseven if theyran at thesame relative work-
rate.
68 Nigerian schoolgirls aged 15.5 years perform-
ed aerobic interval training [4 minutes at 90% of
maximum heart rate(HR
max
) and 4 minutes of jog-
ging] 3 times a week for 8weeks.
[61]
Anothergroup
of young girls (of the same level, i.e. V
.
O
2max
39 ±
5 ml/min/kg) performed continuous work (4.8km
at 80 to 85% HR
max
). Improvement in V
.
O
2max
was
similar for both groups (11%) and was significantly
Anaerobic Interval Training and Performance 85
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
higher than in the control group. It should be pointed
out that the intensity was not very different (5% of
HR
max
) and it is probable that at the end of the
severe run of 4.8km the girls were at V
.
O
2max
be-
cause of the slow component of V
.
O
2
. The end blood
lactate level for both types of training was not re-
ported in this study.
Bar-Or
[62]
considered that children achieved
steady state during intense exercise and recovered
more quickly than adults following intense exercise.
A practical implication is that, during high inten-
sity interval training, children may need shorter
resting periods than adults. It can be added that the
aerobic contribution is higher. Fleck and Kraemer
[63]
underlined the fact that children should not copy
elite athletic programmes.
11-year-old children carried out interval train-
ing at 25 and 50% above their anaerobic threshold
(measured with a ventilatory method) [50 minutes
persession,5 timesperweekfor6weeks].
[64]
They
improved their anaerobic threshold (expressed in
%V
.
O
2max
)by22%.
As for adults,variationoftraining velocity from
lactate (or ventilatory) velocity to vV
.
O
2max
might
be the best stimulus to improve aerobic capacity.
Indeed, a combination of continuous steady-state
running (2 days per week) and interval training (2
days per week) was used for 8 weeks in boys aged
from 10 to 14 years.
[65]
However, the biological
age, which is acritical factor in this age group, was
not reported in these studies. Continuous steady
state was run at the ventilatory threshold velocity
that the boys were able to sustain for 15 minutes at
the beginning of the training period to twice this
duration (30 minutes) in the eighth and final week
of training. Interval training was run at 90 to 100%
V
.
O
2max
and was equivalent to 135% of the ventila-
tory threshold V
.
O
2
. Interval training consisted of
repeated bouts of set distances in the range of 100
to 800m.Thetotal distance run in thistype of train-
ing ranged from 1.5km at the beginning of the 8
weeksoftrainingto2.5kmattheendofthetraining
period. V
.
O
2max
wasimprovedby8%(45.9to49.4
ml/min/kg). During the same time, the ventilatory
thresholdwas improvedby 19%. Therefore,before
training the V
.
O
2
at ventilatory threshold appeared
at 66.6% of V
.
O
2max
and after training at 73.8% of
V
.
O
2max
. We suggest that the higher efficiency of
training for improving ventilatory threshold rather
than V
.
O
2max
may come from the fact that the inter-
val training increased both ventilatory threshold and
V
.
O
2max
. Therefore, the children had 4 training ses-
sions per week for ventilatory threshold improve-
ment and 2training sessionsper week (interval train-
ing only) for V
.
O
2max
improvement. This means that
young untrained boys need to have high intensity
training to improve V
.
O
2max
.Duringthe8weeks,
height and bodyweight had not changed. Some au-
thors have reported that children have higher lac-
tate levels or ventilatory thresholds relative to their
V
.
O
2max
than adults.
[65]
However, it is probably only
a question of oxygen kinetics, and when a steady-
state protocol is used to determine MLSS,children
have the same MLSS values as adults when ex-
pressed as a percentage of V
.
O
2max
.
[66]
However, children have,or at least usedtohave,
a style of life that is more comparable to interval
training at or above vV
.
O
2max
. Therefore, to improve
V
.
O
2max
, this would imply a high intensity of work
with very short intermittent training (10 to 15 sec-
onds)above V
.
O
2max
or long intervals, long enough
(3 minutes of work at vV
.
O
2max
,3minutesat50%
vV
.
O
2max
) to elicitV
.
O
2max
for a long time.Children
have the same time limit at vV
.
O
2max
as adults
[67]
and the improvement in vV
.
O
2max
is not systemati-
cally, asforadults,
[67,68]
accompaniedbyanimprove-
ment in the time limit at vV
.
O
2max
.
[66]
In 10- to 11-year-old boys, 9 weeks of interval
training was found to improve both aerobic (V
.
O
2max
per kilogram of bodyweight: +8%) and anaerobic
(mean power and peak power on Wingate test: +10
and 14%, respectively) capacity. These improve-
ments were correlated with those of a 1200m run.
[69]
Docherty et al.
[70]
showed that changes in aerobic
and anaerobic functions wereindependentof phys-
iological maturity as determined by serum testos-
terone levels in young boys.
86 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
3. Studies Performed on Top Athletes
Acevedo and Goldfarb’s study
[71]
remains one
of the most cited papers, since they demonstrated
that increased training intensity improved athletic
performanceinpreviouslytrainedrunners (V
.
O
2max
65.8 ± 2.4 ml/min/kg). The training consisted of 8
weeks of increased intensity running 3 days per
week at 48-hour intervals while continuing to run
8to19km(5to12miles)perdayontheotherdays
of the week. One day was dedicated to interval
training at 90 to 95% of HR
max
with limited recov-
ery times between runs (distance was not speci-
fied). As soon as heart rate returned to 120 beats/min
they started again. The other 2 runs were fartlek
workouts covering 10 to 16km at the specific pace
of a 10km run, i.e. about 92% vV
.
O
2max
. The total
distance of the week was maintained; therefore the
modifications could be attributed to an increase in
training intensity. The performanceover10kmwas
improved in relation to the decrease in blood lac-
tate level at 85 and 90% vV
.
O
2max
and with no in-
crease in V
.
O
2max
. The authors hypothesised that
lactate production and clearance (as already reported
in rats by Denis et al.
[72]
) could be improved. This
assumption has been recently confirmed by Fukuba
et al.
[73]
in triathletes. An improvement in perfor-
mancewith no increasein V
.
O
2max
hasalreadybeen
reported in well trained participants by Daniels et
al.
[74]
A series of longitudinal studies performed by a
SouthAfricangrouponwelltrainedendurancecy-
clists examined metabolic and performance adap-
tation to interval training.
[75]
They replaced a por-
tion (15%) of their 300km per week base endurance
training with high intensityintervaltraining(HIT).
HIT consisted of 6 to 9
× 5-minute rides at 80% of
PPO reached and sustained for 1 minute in an in-
cremental test as defined by Noakes et al.
[59]
(This
can be a little higher than vV
.
O
2max
,sinceV
.
O
2max
can be reached before the highest work-rate; see
Billat and Koralsztein.
[48]
) Recovery time was 1
minute. The programme lasted 6 weeks and the
cyclists carried out 2 HIT sessions per week. In all
these studies, an increase in performance during
the 40km trial (reported as being a reproducible
test) was obtained with an increasein the peak work-
rate values. Cyclists were generally able to sustain
a higher absolute and relative work-rate during the
time trials, which lasted about 1 hour.
[76,77]
This
HIT programme allowed time to fatigue to be in-
creasedby150%atpeak(1minute)poweroutput
without any change in oxidative and glycolytic en-
zyme metabolism. The skeletal muscle buffering
capacity was improvedafter this HITprogramme.
[76]
This programme improved performance but it was
notspecifiedwhetherV
.
O
2max
wasincreased.There
was, as in many longitudinal studies, no control
group.
To determine whether a short-fast or a longer-
slower interval training programme produced bet-
ter improvements in performance, Septo et al.
[78]
compared 5 types of interval training sessions: (i)
12
×30 seconds(restinterval 4.5 minutes)at 175%
of PPO according to the definition by Noakes et
al.;
[59]
(ii) 12 × 60 seconds at 100% PPO (rest 4.0
minutes); (iii) 12
× 2 minutes at 90% PPO (rest 3.0
minutes); (iv) 8
× 4 minutes at 85% PPO (rest 1.5
minutes); or (v)4
×8 minutesat80% PPO (rest 1.0
minutes). Cyclists completed 6 sessions over a 3-
week period in addition to their usual aerobic base
training.The percentageimprovement inthe 40km
time trial was modelled as a polynomial function
of the rank order of the intensity of the training
intervals. The cubic trend was strong and statisti-
cally significant and predicted the greatest enhance-
ment for the intervals performed at 85% PPO and
at175%PPO.PPO was correlatedwithperformance,
i.e. the 40km time trial. Intervals performed at 100%
PPO and 80% PPO didnot produceanystatistically
significant enhancements in performance. The au-
thors concluded that interval training with work
bouts of 3 to 6 minutes at an intensity of 85% of
PPO gave the maximal enhancement of performance
for this exercise over 40km, lasting about 1 hour at
this power output. However, according to the prin-
ciple of specificity, the 30-second work bouts, which
would have been achieved by a substantial contri-
bution from oxygen-independent glycolysis, would
not enhance performance over a 40km time trial,
which depends almost entirely on power provided
Anaerobic Interval Training and Performance 87
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
by the aerobic system. However, 12 × 30 seconds
also involves aerobic metabolism, and it is prob-
able that buffering capacity plays a part in an all-
out exercise of 1 hour where glycolysis is at a high
rate.
Interval training should be practisedby cyclists,
since Palmer et al.
[79]
showedthat elite cyclists rac-
ing in a pack randomly vary their work rates from
around 50% to almost 100% of the peak sustained
power output, independently of the track.
Elite runners train at a higher level of intensity
(relative to their vV
.
O
2max
) than lower level ath-
letes.
[59,80]
The authors of these studies suggested
that V
.
O
2max
and endurance performance may be
limited not only by central cardiovascular factors
related to V
.
O
2
but also by the so-called ‘muscle
power’ factors affected by the interaction of both
neuromuscular and anaerobic characteristics. There-
fore, supramaximal interval training can also be
included in a programme for middle distance run-
ners.
Evenif more than 75% of the energyisprovided
by aerobic metabolism,
[81,82]
anaerobic capacity is
very important for 1500m runners. They must be
able to run the first 800m in 1 minute 50 seconds,
as seen in the last world athletic championship in
1999 (26.2 km/h, which corresponds to a pace of 3
minutes 25 seconds, faster than the world record).
Even for the 10 000m event, the last 1000m and
300m in the last world championship were run in
2 minutes 25.2 seconds (24.8 km/h) and 39.8 sec-
onds (27.1 km/h), respectively. This means that if
the world champion Gebreselassie has a vV
.
O
2max
of 25.5 km/h estimated from his best performance
at 3000m, he is able to run at 106% of vV
.
O
2max
during the last (25th) lap of the 10 000m race.
4. Conclusion
Muchmorestudyisneededtounderstandthe
physiological basis of interval training, especially
in elite athletes. However, sports records are per-
formed with variable velocity, even in long dis-
tance running. Considering the last 3 world records
in middle and long distance running (1500, 3000,
5000 and 10 000m), it can be observed that the
range of the coefficient of variation of velocity is
1to5%(inthe10000mof1997byGebreselassie
and in the 1500m of 1995 by Morcelli, respectively).
This could mean that interval training could be in-
dividualised taking into account this individual sto-
chastic pace, allowing athletes to reach their best
performance.
Acknowledgements
This study was supported by grants from la Caisse Cen-
trale de Activités Sociales d’Electricité et Gaz de France.
References
1. Margaria R, Oliva RD, di Prampero PE, et al. Energy utilization
in intermittent exercise of supramaximalintensity. J ApplPhys-
iol 1969; 26: 752-6
2. Balsöm PD, Seger JY, Sjodin B, et al. Physiological responses
tomaximal intensity intermittent exercise.EurJ Appl Physiol
1992; 65: 144-9
3. Christensen EH, Hedman R, Saltin B. Intermittent and contin-
uous running. Acta Physiol Scand 1960; 50: 269-6
4. Tabata I, Irisawa K, Kouzaki M, et al. Metabolic profile of high
intensity intermittent exercises. Med Sci Sports Exerc 1997;
29: 390-5
5. Bogdanis GC, Nevill ME, Boobis LH, et al. Contribution of
phosphocreatine and aerobic metabolism to energy supply
during repeated sprint exercise. J Appl Physiol 1996; 80:
876-84
6. Gaitanos GC, Williams C, Boobis LH, et al. Human muscle
metabolism during intermittent maximalexercise. J ApplPhys-
iol 1993; 75: 712-9
7. Spriet LL, Lindinger MI, McKelvie RS, et al. Muscle glycoge-
nolysisand H+ concentrationduringmaximalintermittentcy-
cling. J Appl Physiol 1989; 66: 8-13
8. Hirvonen J, Rehunen S, Rusko H, et al. Breakdown of high-
energy phosphate compounds and lactate accumulation dur-
ing short supramaximal exercise. Eur J Appl Physiol 1987;
56: 253-9
9. Balsom PD, Seger JY, Sjodin B, et al. Maximal-intensity inter-
mittent exercise: effect of recovery duration. Int J Sports Med
1992, 13: 528-33
10. SprietLL.Anaerobic metabolism inhumanskeletalmuscledur-
ing short-term, intense activity. Can J Physiol Pharmacol
1992; 70: 157-65
11. Yoshida T, Watari H.Metabolic consequences of repeated exer-
cise in long distance runners. Eur J Appl Physiol 1993; 67:
261-5
12. Bogdanis GC, NevillME, Boobis LH, et al. Recovery of power
output and muscle metabolites following 30s of maximal
sprint cycling in man. J Physiol (Lond) 1995; 15: 467-80
13. Brooks GA, Mercier J. Balance of carbohydrate and lipid utili-
zationduringexercise:the‘crossover’concept.JApplPhysiol
1994; 76: 2253-61
14. Jenkins DG, Palmer J, Spillman D. The influence of dietary
carbohydrate on performance of supramaximal intermittent
exercise. Eur J Appl Physiol 1993; 67: 309-14
15. Fox EL, Bartels RL, Billing CE, et al. Frequency and duration
of interval trainingprograms and changes in aerobic power. J
Appl Physiol 1975; 38: 481-4
88 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
16. Lesmes GR, Fox EL, Stevens C, et al. Metabolic responses of
females to high intensity interval training of different fre-
quencies. Med Sci Sport Exerc 1978; 10: 229-32
17. Fox EL, Bartels RL, Billings CE, et al. Intensity and distance
of interval training programs and changes in aerobic power.
Med Sci Sports Exerc 1973; 5: 18-22
18. Davies KJA, Packer L, Brooks GA. Biochemical adaptation of
mitochondria, muscle, and whole-animal respiration to en-
durance training. Arch Biochem Biophys 1981; 209: 539-54
19. Davies KJA, Packer L, Brooks GA. Exercise bioenergetics fol-
lowing sprint training. Arch Biochem Biophys 1982; 215:
260-5
20. Simoneau JA,LortieG, BoulayMR,et al. Inheritanceofhuman
skeletal muscle and anaerobic capacity adaptation to high-in-
tensity intermittent training. Human skeletal muscle fiber
type alteration with high-intensity intermittent training. Int J
Sports Med 1986; 7: 167-71
21. Simoneau JA, LortieG, Boulay MR, et al.Human skeletalmus-
cle fiber type alterationwith high-intensityintermittenttrain-
ing. Eur J Appl Physiol 1985; 54: 250-3
22. Simoneau JA, Lortie G, Boulay MR, et al. Effects of two high-
intensity intermittent training programs interspersed by de-
training on human skeletal muscle and performance. Eur J
Appl Physiol 1987; 56: 516-21
23. 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: 159-61
24. Roberts AD, Billeter R, Howald H. Anaerobic muscle enzyme
changes after interval training. Int J Sports Med 1982; 3:
18-21
25. Margaria R, Cerretelli P, Aghemo P, et al. Energy cost of run-
ning. J Appl Physiol 1963; 18: 367-70
26. Linossier MT, Denis C, Dormois D, et al. Ergometric and me-
tabolic adaptation to a 5-s sprint training programme. Eur J
Appl Physiol 1993; 67: 408-14
27. Nevill ME, Boobis LH, Brooks S, et al. Effect of training on
musclemetabolismduringtreadmillsprinting.JApplPhysiol
1989; 67: 2376-82
28. Hargreaves M, McKenna MJ, Jenkins DG, et al. Muscles me-
tabolites and performance during high-intensity, intermittent
exercise. J Appl Physiol 1998; 84: 1687-91
29. Heugas AM, Brisswalter J, Vallier JM. Effet d’une période
d’entraînement de trois mois sur le déficit maximal en
oxygène chez des sprinters de haut niveau de performance.
Can J Appl Physiol 1997; 22: 171-81
30. MacDougall DJ, Hicks AL, MacDonald JR, et al. Muscle per-
formance and enzymatic adaptations to sprint interval train-
ing. J Appl Physiol 1998; 84: 2138-42
31. Houmard JA, Costill DL, Mitchell JB, et al. The role of anaer-
obic ability in middle distance running performance. Eur J
Appl Physiol 1991; 62: 40-3
32. Jenkins DG, Quigley BM. The influence of high-intensity ex-
ercise training on the Wlim-Tlimrelationship.Med Sci Sports
Exerc 1993; 25: 275-82
33. Lechevalier JM, Vandewalle H, Chatard JC, et al. Relationship
between the 4 mMol running velocity, the time-distance rela-
tionship and the Leger-Boucher test. Arch Int Physiol Bio-
chem 1989; 97: 355-60
34. Paavolainen L, Hakkinen K, Hamalainen I, et al. Explosive-
strength training improves 5-km running time by improving
running economy and muscle power.J ApplPhysiol1999; 86:
1527-33
35. Franch J, Madsen K, Djurhuus MS, et al. Improved running
economy following intensified training correlates with reduces
ventilatorydemands.Med Sports SciExerc 1998; 30:1250-6
36. Komi PV, Kyröläiinen H. Mechanical efficiency of stretch-
shortening cycle exercise. In: Marconnet P, Saltin B, Komi P,
et al., editors. Human muscular function during dynamic ex-
ercise. Basel: Karger, 1996: 44-56
37. Flynn MG, Fahlman M, Braun WA, et al. Effects of resistance
training on selected indices of immune function in elderly
women. J Appl Physiol 1999; 86: 1905-13
38. Wells CL, Pate RR. Training for performance of prolonged ex-
ercise.In:LambDR,MurrayR,editors.Perspectivesinexercise
scienceandsportsmedicine.Vol.1.Indianapolis(IN):Bench-
mark Press, 1988: 357-91
39. Tanaka H, Swensen T. Impact of resistance training on endur-
ance performance. Sports Med 1985; 6: 266-70
40. Balsom PD,EkblomB,SöderlundK,etal.Creatinesupplemen-
tation and dynamic high-intensity intermittent exercise. Scand
J Med Sci Sports 1993; 3: 143-9
41. Tabata I, Nishimura K, Kouzaki M, et al. Effects of moderate-
intensity endurance and high-intensity intermittent training
on anaerobic capacity and
V
.
O
2
max. Med Sci Sports Exerc
1996; 28: 1327-30
42. Ahmaidi S, Masse-Biron J, Adam B, et al. Effects of interval
training at the ventilatory threshold on clinical and cardiore-
spiratory responses in elderly humans. Eur J Appl Physiol
1998; 78: 170-6
43. Suominen H, Heikkinen E, Liesen H, et al. Effect of 8-weeks
endurance training on skeletal muscle metabolism in 56-70
year old sedentary men. Eur J Appl Physiol 1977;37: 173-80
44. Billat V, Slawinski J, Bocquet V, et al. Very short interval train-
ing (15s-15s) around the critical velocity allows middle-aged
runners to maintain
V
.
O
2max
for 14 min. Int J Sports Med.
In press
45. Péronnet F, Thibault G. Mathematical analysis of running per-
formance and world running records. J Appl Physiol 1989;
67: 453-65
46. Daniels J. Jacks Daniels’formula. Champaign (IL): Human Ki-
netics Publishers, 1998
47. Noakes T. Lore of running. Champaign (IL): Human Kinetics
Publishers, 1991
48. BillatV, KoralszteinJP. Significanceof the velocity at
V
.
O
2
max
and its time to exhaustion at this velocity. Sports Med 1996;
22: 90-108
49. Daniels J, Scardina N. Interval training and performance. Sports
Med 1984; 1: 327-34
50. Gorostiaga EM, Walter CB, Foster C, et al. Uniqueness of in-
terval and continuous training at the same maintained exer-
cise intensity. Eur J Appl Physiol 1991; 63: 101-7
51. Billat V, Slawinski J, Bocquet V, et al. Intermittent runs at
v
V
.
O
2
max enables subjects to remain at V
.
O
2
max for a longer
time than submaximal runs. Eur J Appl Physiol 2000; 81:
188-96
52. Saltin B, Essen B. Muscle glycogen, lactate, ATP, and CP in
intermittent exercise. In: Pernow B, Saltin B, editors. Muscle
metabolism during exercise. Advances in experimental med-
icine and biology. Vol. 11. New York: Plenum Press, 1971:
419-24
53. Saltin B,Essen B, PedersenPK. Intermittent exercise: its phys-
iology and some practical applications. In: Joekle E, Anand
RL, Stoboy H, editors. Advances in exercise physiology. Ba-
sel: Karger, 1976: 9: 23-51 (Medicine Sport; 5)
Anaerobic Interval Training and Performance 89
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
54. Pepper ML, Housh TJ,JohnsonGO.Theaccuracy of thecritical
velocitytestforpredicting timetoexhaustionduringtreadmill
running. Int J Sports Med 1992; 13: 121-4
55. Gaesser GA, Poole DC. The slow component ofoxygen uptake
kinetics in humans. Exerc Sports Sci Rev, 1996; 24: 35-70
56. Florence SL, Weir JP. Relationship of critical velocity to mara-
thon running performance. Eur J Appl Physiol 1997; 75: 274-8
57. Astrand PO, Rodahl K. Textbook of work physiology. 3nd ed.
New York: McGraw Hill, 1986
58. Robinson DM, Robinson SM, Hume PA, et al. Trainingintensity
of elite male distance runners. Med Sci Sports Exerc 1991; 23:
1078-82
59. Noakes TD, Myburgh KH, Schall R. Peak treadmill running
velocity during the
V
.
O
2
max test predicts running performance.
J Sports Sci 1990; 8: 35-45
60. Watson AWS, O’Donovan DJ. Theeffects of five weeks of con-
trolled intervaltrainingon youthsofdiverse pre-trainingcon-
dition. J Sports Med Phys Fitness 1977; 17: 139-46
61. Adeniran SA, Toriola AL. Effects of continuous and interval
running programmes on aerobic and anaerobic capacities in
schoolgirls aged 13 to 17 years. J Sports Med Phys Fitness
1988; 28: 260-6
62. Bar-Or O. The young athlete: some physiological considera-
tions. J Sports Sci 1995; 13: 531-3
63. Fleck SJ, Kraemer WJ. Designing resistance training programs.
Champaign (IL): Human Kinetics Publishers, 1997
64. Haffor AA, Harrison AC, Catledge-Kirk PA. Anaerobic thresh-
old alterations caused by interval training in 11-years olds. J
Sports Med Phys Fitness 1990; 30: 53-6
65. Mahon AD, Vaccaro P. Ventilatory threshold and
V
.
O
2
max
changes in children following endurance training. Med Sci
Sports Exerc 1989; 4: 425-31
66. Billat V, Gratas-Delamarche A, Monnier M, et al. Test to ap-
proach maximal lactate steady-state in 12-year old boys and
girls. Arch Biochem Physiol 1995; 103: 65-72
67. Berthoin S, Manteca F, Gerbeaux M, et al. Effect of a 12-week
training programme on maximal aerobic speed (MAS) and
running time to exhaustion at 100% of MAS for students aged
14 to 17 years. J Sports Med Phys Fitness 1995; 35: 251-6
68. Billat V, Flechet B, Petit B, et al. Interval training at
V
.
O
2
max:
effects on aerobic performance and overtraining markers.
Med Sci Sport Exerc 1999; 31: 156-63
69. Rostein A, Dotan R, Bar-Or O, et al. Effect of training on an-
aerobic threshold, maximalaerobic powerand anaerobicper-
formance of preadolescent boys. Int J Sports Med Phys 1986;
7: 281-6
70. Docherty D, Wenger HA, Collis ML. The effects of resistance
training on aerobic power of young boys. A test to approach
maximal lactate steady-state in 12-year old boys and girls.
Med Sci Sport Exerc 1987; 19: 389-92
71. Acevedo EO, Goldfarb AH. Increased training intensity effects
on plasma lactate, ventilatory threshold, and endurance. Med
Sci Sports Exerc 1989; 21: 563-8
72. Denis C, Fouquet R, Poty P, et al. Effect of 40 weeks of endur-
ance training on the anaerobic threshold. Int J Sports Med
1982; 3: 208-13
73. Fukuba Y, Walsh ML, Morton RH, et al. Effect of endurance
training on blood lactate clearance after maximal exercise. J
Sports Sci 1999; 17: 239-48
74. Daniels JT, Yarbough RA, Foster C. Changes in
V
.
O
2
max and
running performance with training. Eur J Appl Physiol 1978;
39: 249-54
75. Westgarth-Taylor C, HawleyJA, RichardS,et al.Metabolicand
performance adaptations to interval training in endurance-
trained cyclists. Eur J Appl Physiol 1997; 75: 298-304
76. Weston AH, Myburgh KH, Lindsay FH, et al. Skeletal muscle
buffering capacity and endurance performance after high-
intensity interval-training well-trained cyclists. Eur J Appl
Physiol 1997; 75: 7-13
77. Lindsay FH, Hawley JA, Myburgh KH, et al. Improved athletic
performance in highly trained cyclists after interval training.
Med Sci Sports Exerc 1996; 28: 1427-37
78. Septo NK, Hawley JA, Dennis SC, et al. Effects of different
interval training programs on cycling time-trial performance.
Med Sci Sports Exerc 1999; 31: 736-41
79. Palmer GS, Hawley JA, Dennis SC, et al. Heart rate responses
during a 4-d cycle stage race. Med Sci Sports Exerc 1994; 26:
1278-83
80. Green H, Patla M. Maximalaerobic power: neuro-muscular and
metabolic considerations. Med Sci Sports Exerc 1992; 24:
38-46
81. Spencer MR, Gastin PB, Payne WR. Energy system contribu-
tion during 400 to 1500 metres. New Studies Athletics 1996; 4:
59-65
82. Weyand PG, Cureton KJ, Conley DS, et al. Peak oxygen deficit
predicts sprint and middle-distance track performance. Med
Sci Sports Exerc 1994; 9: 1174-80
Correspondence and offprints: Véronique Billat, Centre de
Médecine du Sport CCAS, 2 av. Richerand, 75010 Paris,
France.
90 Billat
Adis International Limited. All rights reserved. Sports Med 2001; 31 (2)
