Vous êtes sur la page 1sur 8

Eur J Appl Physiol (2000) 81: 214221 Springer-Verlag 2000

ORIGINAL ARTICLE

Fabien A. Basset Marcel R. Boulay

Specicity of treadmill and cycle ergometer tests


in triathletes, runners and cyclists

Received: 3 November 1998 / Accepted: 29 July 1999

Abstract The objective of this study was to evaluate the


viability of using a single test in which cardiorespiratory
Introduction
variables are measured, to establish training guidelines
Maximal oxygen uptake (V_ O2max) is generally accepted
in running and/or cycling training activities. Six triath-
to be the best indicator of endurance performance ca-
letes (two females and four males), six runners (two fe-
pacity (Mitchell et al. 1958; O'Toole and Douglas 1995;
males and four males) and six males cyclists, all with
Sleivert and Rowlands 1996; Taylor et al. 1955; Welt-
5.5 years of serious training and still involved in racing,
man et al. 1990). Consequently, this variable is fre-
were tested on a treadmill and cycle ergometer. Car-
quently used to determine training intensities in
diorespiratory variables [e.g., heart rate (HR), minute
numerous endurance sports. V_ O2max is assumed to be
ventilation, carbon dioxide output (V_ CO2)] were calcu-
highly dependent upon the mode of testing, with the
lated relative to xed percentages of maximal oxygen
highest values normally attained during treadmill run-
uptake (V_ O2max; from 50 to 100%). The entire group of
ning. Therefore, to optimize the eectiveness of a
subjects had signicantly (P < 0.05) higher values of
training program, training activities need some speci-
V_ O2max on the treadmill compared with the cycle ergo-
city with regard to mode, duration and intensity (Bou-
meter [mean (SEM) 4.7 (0.8) and 4.4 (0.9) l min)1,
chard et al. 1979; Kohrt et al. 1987). Training eects
respectively], and dierences between tests averaged
also appear to be specic to the mode of training used by
10.5% for runners, 6.1% for triathletes and 2.8% for
an athlete; therefore dierences between testing modes
cyclists. A three-way analysis of variance using a
vary with training (Pechar et al. 1974; Sharkey 1988).
3 2 6 design (groups tests intensities) demon-
Because of this specic adaptation, runners are generally
strated that all factors yielded highly signicant F-ratios
tested on a treadmill, and cyclists on a cycle ergometer.
(P < 0.05) for all variables between tests, even though
However, triathletes engage in three aerobic modes of
dierences in HR were only 4 beats min)1. When HR
exercise that use the same muscle groups. Therefore, to
was plotted against a xed percentage of V_ O2max, a high
take into account testing mode specicity, triathletes
correlation was found between tests. These results
should preferably use three dierent tests to monitor
demonstrate that for triathletes, cyclists and runners, the
their training program.
relationship between HR and percentage of V_ O2max,
However, there is a possibility that a cross-training
obtained in either a treadmill or a cycle ergometer test,
eect occurs when a high volume of training is per-
may be used independently of absolute V_ O2max to obtain
formed regularly with dierent modes of exercise (Flynn
reference HR values that can be used to monitor their
et al. 1998; Schneider and Pollack 1991). If cross-train-
running and/or cycling training bouts.
ing eects are real, then it may be possible to monitor
various training activities with a single indicator of work
Key words Performance test Specicity V_ O2max
intensity without having to undergo multiple testing
Monitoring of training
sessions using each training mode.
Numerous studies have shown that heart rate (HR)
and oxygen uptake (V_ O2) are strongly correlated in cy-
F.A. Basset M.R. Boulay (&) cling as well as running (Barbeau et al. 1993; Coast and
Laboratoire des Sciences de l'Activite Physique, Welch 1985; Foster et al. 1978; Van Handel et al. 1988;
Division de Kinesiologie, Faculte de Medecine,
Universite Laval, Ste-Foy, Quebec, Canada G1K 7P4 Veicsteinas et al. 1985). Since the availability of light-
e-mail: Marcel.Boulay@kin.msp.ulaval.ca weight monitors makes HR very easy to monitor con-
Fax: +1-418-6562441 tinuously during training bouts, HR is now the variable
215

of choice used to control work intensity during training continuously with an automated open-circuit gas analysis system
bouts (Boulay 1995; Boulay et al. 1997; Gilman 1996; using O2 and CO2 analyzers (Model S-3A and Anarad AR-400,
Ametek, Pittsburgh, Pa., USA), and a turbine-driven digital
Weltman et al. 1990). In trained endurance athletes spirometer (Model S-430, Vacumetrics/Vacumed, Ventura, Calif.,
there is a possibility that HR values would be similar in USA) that had a 5.3-l mixing chamber. HR was recorded by
dierent exercise modalities when exercise intensities are electrocardiography in the CM5 position (Model M200, Burdick,
equivalent. Thus, the purpose of this study was to Milton, Wisc., USA). The criteria used to conrm that V_ O2max had
been reached were plateauing of V_ O2 in spite of an increase in
evaluate the feasibility of using a single exercise test to speed or work rate, and an R value greater than 1.1. HR at ex-
establish training guidelines in running and/or cycling haustion was taken as maximal heart rate (HRmax).
training activities.
Statistical analysis

Methods Simple linear regression analysis was used to determine the rela-
tionship between tests and cardiorespiratory values. A three-way
Subjects (three groups two tests six intensities) analysis of variance
(ANOVA) with repeated measures was used to test signicant
Six triathletes (two females and four males), six runners (two fe- dierences between dependent variables. Signicant F-ratios were
males and four males) and six male cyclists gave their written in- followed by post-hoc comparison using Newman-Keul's proce-
formed consent (in compliance with Laval University's Ethics dure. Homogeneity of variance was tested with the Levene proce-
Committee regulations) to participate in this study, which was dures (Steel and Torrie 1980). For all statistical tests, a probability
performed in the preparatory phase of their training year. All level of P < 0.05 was considered signicant. All values are ex-
subjects were active athletes and had been competing at the pro- pressed as means (SEM), unless specied otherwise.
vincial and national levels for periods ranging from 2 to 15 years;
some had participated in international events. Some of the physical
characteristics of these subjects are presented in Table 1.
Results

V_ O2max test The maximal power output values obtained at exhaus-


tion for both exercise tests are presented in Table 1, and
Subjects underwent continuous, incremental tests to volitional ex-
haustion on both a treadmill and a cycle ergometer. These two other measures of performance are given in Table 2. All
maximal tests were carried out randomly, with a minimal interval groups exhibited a signicantly (F1,15 22.79; P < 0.05)
between tests of 2 days, and a maximal interval of 7 days. The test higher V_ O2max on the treadmill than on the cycle ergo-
protocols were designed to yield similar test durations. The running meter, with cyclists exhibiting higher values than both
test was conducted on a motor-driven treadmill (Quinton Instru-
ments, Seattle, Wash., USA). After a 5-min warm-up at a speed of
the runners and the triathletes in both tests. Dierences
3.5 km h)1, the starting speed was set at 5.5 km h)1. The grade between tests averaged 2.8% for cyclists, 10.5% for
was set at 5%, and was increased every 2 min by 1.1 km h)1 until runners and 6.1% for triathletes. Subjects diered sub-
it reached 13.1 km h)1, after which it was raised by 3% every stantially in absolute V_ O2max, with values ranging from
2 min until exhaustion. The cycle ergometer test was performed on 2.3 to 6.1 l min)1 for the cycle ergometer, and from 3.0
an electromagnetically braked cycle ergometer (Warren E. Collins,
Braintree, Mass., USA). Subjects were asked to choose a familiar to 5.7 l min)1 for the treadmill. These large inter-in-
and comfortable pedaling rate equal or higher than 60 rpm and to dividual dierences were due mainly to gender. If the
maintain that rate throughout the test. After a 4-min warm-up values provided by the female runners and triathletes
period at 100 W, the test was initiated at an initial power output of were removed from the analysis, the average absolute
100 W. Increments of 25 W were made every min until 200 W was
reached; thereafter, 25-W increments were made every 2 min until V_ O2max values were 5.0 and 5.1 l min)1 for the runners
exhaustion. and 4.4 and 4.6 l min)1 for the triathletes in the cycle
ergometer and treadmill tests, respectively.
In separate groups, HRmax values were higher on
Physiological measurements
the treadmill than on the cycle ergometer. V_ E was higher
During the tests, V_ O2, minute ventilation (V_ E), carbon dioxide on the treadmill for the cyclists and runners, while V_ CO2
output (V_ CO2) and respiratory exchange ratio (R) were recorded was higher on the treadmill only for the cyclists.

Table 1 Physical and physiological characteristics of the subjects. Data are presented as the mean (SEM). Metabolic equivalents of energy
expended above metabolic rate (METS) = 3.5 mlO2 kg)1 min)1

Subjects Physical characteristics Test results

Age Mass Height Expertise Ergocycle Treadmill


(years) (kg) (m) (years)
Maximal Duration Metabolic Duration
power (W) (min) rate (METS) (min)

Cyclists 24.3 (7.5) 72.5 (3.7) 1.78 (0.06) 6.2 (4.1) 439.6 (42.3) 21.3 (1.4) 20.8 (1.4) 19.8 (1)
Runners 21.0 (2.4) 64.8 (13.8) 1.72 (0.1) 6.0 (2.2) 320.8 (78.1) 15.3 (6.3) 19.6 (1.6) 20.3 (1.9)
Triathletes 21.3 (1.6) 65.7 (5.6) 1.73 (0.05) 4.3 (2.8) 375 (31.6) 18.5 (3.2) 20.0 (1.5) 19.7 (1.4)
All groups 22.2 (5.0) 67.7 (9.1) 1.74 (0.07) 5.5 (3.2) 378.5 (71.5) 18.4 (4.7) 20.1 (1.5) 19.9 (1.4)
216

The dierences in HR between tests varied from

Table 2 Physiological measurements at exhaustion. Data are presented as the mean (SEM). (HR Heart rate, V_ O2 oxygen uptake, V_ O2max maximal V_ O2, V_ E minute ventilation, V_ CO2

(ml kg)1 min)1) (range)


3 beats min)1 for the cyclists and triathletes to

(3.8)* (70.580.3)
(4.1)* (57.772.4)
(3.7)* (59.573.3)
6 beats min)1 m for the runners. Figure 1 illustrates
the evolution of HR relative to absolute values of V_ O2.
There were large inter-individuals dierences, but in all
three graphs, individuals values for the treadmill test are

(4.3)*
almost superimposed upon those of the cycle ergometer
V_ O2max test, as demonstrated by the nearly identical regression

75.3
68.4
66.9
70.2
lines.
Figures 2 and 3 present graphs of V_ E and V_ CO2,
(l min)1)

(0.1)**
(0.5)
respectively, plotted against absolute values of V_ O2. The
(0.2)
(0.2)
inter-individual dierences were much smaller than
V_ CO2

those observed for HR, especially at lower V_ O2 values.


5.7
5.1
4.9
5.2
The dispersion of V_ E values at high V_ O2 values was
related to the inter-individual dierences in V_ O2max that
(0.4)*
(1.1)*
(0.4)*
(0.8)*
(lmin)1)
V_ O2max

were observed as subjects reached their ventilatory


threshold at dierent values of V_ O2. This was especially
5.3
4.5
4.4
4.7

evident in the runners' data, as the two female runners


had V_ O2max values of around 3.0 l min)1.
(14.6)*

(23.2)*
(7.4)*

(6.4)*

Figure 4 shows mean HR plotted against xed per-


(beatsmin)1) (lmin)1)

centages of V_ O2max (50100%) for the cyclists (left


163.6
148.4
147.9
152.2

panel), runners (middle panel) and triathletes (right


V_ E

panel). There were large dierences in individual HRs


(data not shown), but dierences between tests in indi-
Treadmill

(3.5)*
(2.9)*
(3.2)*
(2.2)*

vidual athletes were much smaller. Signicant dieren-


ces (F1,15 9,54; P < 0.05) were found between tests
HR

204
191
197
197

for mean HRmax, and ANOVA revealed signicantly


dierent mean submaximal HR values between tests for
all groups (F1,15 7,82; P < 0.05). Post-hoc analysis
(mlkg)1min)1) (range)

(3.9) (63.876.4)

(2.6) (58.468.7)

indicated that HRs were not dierent between tests at


(5) (52.868.1)

values of under 90% V_ O2max for the runners and under


60% for the triathletes, while cyclists showed dierences
at each stage. Submaximal HRs were signicantly
(4.6)

higher on the treadmill than on the cycle ergometer,


V_ O2max

Signicantly dierent between tests;** signicantly dierent between groups

however these dierences were small, reaching about


71.2
61.7
64.6
65.8

4 beats min)1.
The statistical relationships between the HRs achieved
on the cycle ergometer and those achieved on the treadmill
(lmin)1)

(0.2)
(0.4)
(0.3)
(0.2)

at various percentages of V_ O2max in all athletes are pre-


V_ CO2

sented in Fig. 5. HRs achieved on the cycle ergometer


5.9
4.6
4.4
4.9

were highly correlated with those achieved on the tread-


mill when calculated at xed percentages of V_ O2max. The
(lmin)1)

(0.5)
(1.1)
(0.6)
(0.9)
V_ O2max

correlation coecients were similar for the three groups


and reached r 0.96 for the cyclists, r 0.97 for the
5.2
4.0
4.2
4.4

runners and r 0.98 for the triathletes.


(15.8)

(30.2)
(6.4)

(6.7)
(lmin)1)

Discussion
169.6
139.7
146.1
154.6
V_ E

Comparisons between exercise tests or between labora-


tory and eld tests have been conducted since the birth
(beatsmin)1)

of exercise physiology and have been the topic of nu-


carbon dioxide output)

Ergocycle

(3.2)
(4.1)
(3.4)
(2.5)

merous publications. Such comparisons generally used


protocols that were very similar or that generated very
HR

201
185
194
193

similar physiological responses. Nagel et al. (1971), for


example, compared the V_ O2 requirements of gradual
All groups
Triathletes

treadmill, cycle ergometer and step-device tests at seven


Runners
Subjects

Cyclists

dierent workloads. Even though the protocols of that


study were very similar from one test mode to another,
*
217

Fig. 1 Evolution of heart rate


(HR) relative to absolute oxy-
gen uptake (V_ O2) for the cycle
ergometer (closed circles) and
treadmill (open squares) tests
(individual values for the
cyclists, runners and triathletes)

Fig. 2 Evolution of minute


ventilation (V_ E) plotted against
absolute V_ O2 for the cycle
ergometer (closed circles) and
treadmill (open squares) tests
(individual values for the
cyclists, runners and triathletes)

Fig. 3 Evolution of carbon di-


oxide output (V_ CO2) plotted
against absolute maximal V_ O2
(V_ O2max) for the cycle ergome-
ter (closed circles) and treadmill
(open squares) tests (individual
values for the cyclists, runners
and triathletes)
218

Fig. 4 Mean HR plotted


against xed percentages of
V_ O2max for the cycle ergometer
(closed circles) and treadmill
(open squares) tests performed
by cyclists (left), runners (mid-
dle) and triathletes (right)

exercise tests in adults and children. Therefore, this pro-


cedure ensures that the results of both tests are not only
comparable, but are also applicable to eld situations in
mimicking what athletes do in the eld, because habitually
they do not adhere to the same standardized work pro-
tocols in dierent exercise modes.
Test durations were not dierent between tests
(Table 1), even though there was a 5-min dierence for
the runners that was caused by the presence of two fe-
male runners with a small body frame (42 and 57 kg)
who could attain cycle ergometer maximal work loads of
200 and 250 W, respectively. Although the presence of
these two female runners complicated the analysis of the
data, their results have been retained because they are an
expression of the biological variation that is likely to be
encountered in the sports world.
The results of our investigation have shown that, as
frequently reported by others, V_ O2max was higher on the
treadmill than on the cycle ergometer (Bouchard et al.
1979; Hermansen and Saltin 1969; Kohrt et al. 1987;
Fig. 5 Correlation between HR values attained during the cycle
ergometer and treadmill tests at dierent percentages of V_ O2max for Martinez et al. 1993; McArdle and Magel 1970;
the entire group (triathletes circles, runners squares, cyclists McConnell et al. 1984; Schneider et al. 1990). These
triangles) dierences were small for cyclists (2.8%), but higher for
triathletes (6.1%) and runners (10.5%). These data thus
the V_ O2 results, presented in l min)1, showed small but conrmed that absolute V_ O2max values are dependent
signicant dierences between tests since: ``no single upon the mode of exercise and the specic training of an
line could describe the metabolic responses to the work athlete (McConnell 1988). Moreira-Da-Costa et al.
loads on the three instruments.'' These results, as well as (1984, 1989) also reported that V_ O2max values were sig-
many others, have revealed that adaptation in an exer- nicantly dierent between testing modes for runners,
cise test appears to be specic to the mode of testing. but not for cyclists. Conversely, Miura et al. (1997),
The tests used in the present protocol were neither Medelli et al. (1993) and Zhou et al. (1997) did not nd
designed to generate similar V_ O2 values at each stage, nor dierent values for V_ O2max between tests with male tri-
to give similar rates of increase between stages. The major athletes, while Schneider and Pollack (1991) found
means used to ensure that tests results would be compa- similar results with female triathletes. One can postulate
rable were based on attaining a roughly similar test du- that these ndings might be related to the training spe-
ration, and on reporting results against xed percentages cicity that causes local adaptation in active muscles,
of V_ O2max. The expression of work intensity in a relative which then may induce a specic enhancement of me-
form has denitive merits because it provides the possi- chanical work eciency. The small dierence observed
bility of comparing groups with very dierent character- in the cyclists of the present study might be related to the
istics. For example Turley and Wilmore (1997) used that fact that as a group, these athletes were older and had a
procedure to compare the physiological stress induced by far greater and more diverse training and racing expe-
219

rience than the other groups. These observations con- The results of studies performed with triathletes are
rm that athletes with a substantial previous cycling equivocal for V_ Emax. In some studies subjects achieved a
experience as a form of exercise training exhibit cycle higher V_ Emax on the cycle ergometer than on the tread-
ergometer V_ O2max values that are equal to or approach mill (Dengel et al. 1989; Kreider et al. 1988; Nagle et al.
those attained on a treadmill (Pechar et al. 1974). 1971; Schneider et al. 1990), while in other studies re-
At maximal work rate, signicant dierences were sults were inconsistent because they used either un-
observed in HRmax between tests for all groups, with trained subjects or single-sport athletes (McArdle and
higher values being attained on the treadmill. These Magel 1970; Pannier et al. 1980; Pechar et al. 1974). In
results are in agreement with those of previous reports the present study, runners displayed the highest tread-
(Martinez et al. 1993; McArdle and Magel 1970; Me- mill V_ Emax values; the cyclists achieved their V_ Emax on
delli et al. 1993; Schneider and Pollack 1991; Zhou et al. the cycle ergometer, while the triathletes displayed sim-
1997), but dier from others in which no signicant ilar values in both tests. The cyclists displayed the
dierences are reported (Hermansen and Saltin 1969; highest V_ Emax values on both tests probably because as a
Kohrt et al. 1987; Moreira-Da-Costa et al. 1989; Schn- group they were taller and heavier than the two other
eider et al. 1990). As in the present study, the former groups. The statistical analysis also revealed that the
studies were performed with male and female athletes, cyclists and runners exhibited dierent V_ E values at each
while the latter studies included only male athletes. In percentage of V_ O2max tested, whereas the triathletes
the present study, without the female runners, the showed no dierence at maximal values, but dierences
dierences in HRmax between tests decreased to at sub-maximal work intensities. The higher V_ E values
3 beats min)1 (190 beats min)1 for the treadmill and observed in runners during treadmill running was a
187 beats min)1 for the cycle ergometer), and conse- consequence of their greater mechanical eciency on the
quently reduced the mean HRmax for all groups. This treadmill (calculated at 10.0 km h)1). This translated
might be an indication that in an exercise restricted to into a longer duration of their run in spite of V_ O2max
localized muscle groups such as cycling, the female values smaller than those observed in cyclists. As a
runners of the present study were probably limited by motor activity, running is more complex than cycling
their small muscle mass and/or by the specicity of such that a really good runner is able to minimize the
movement patterns required on the cycle ergometer, energetically expensive vertical variations of his or her
while the two female triathletes (being taller and heavier center of gravity, thus permitting a more ecient use of
in addition to being more familiar with the activity) his or her aerobic power. Cyclists, on the other hand,
were probably more ecient on the cycle ergometer. displayed a higher V_ Emax during the cycle ergometer test
Statistical analysis revealed that when expressed as they were more ecient on this instrument (gross
against percentages of V_ O2max, the runners' HRs were not mechanical eciency calculated at 200 W), while being
signicantly dierent between tests until 90% of V_ O2max, limited by a less ecient running technique. The im-
while the cyclists and triathletes exhibited distinct HR portance of mechanical eciency on the perception of
proles throughout the tests, since, with the exception of physiological strain appears to be conrmed by the
triathletes at 50% of V_ O2max, signicant dierences were similar V_ E values observed in triathletes. Since they de-
observed at each percentage of V_ O2max. These results vote about the same amount of time in each of the
might have been dependent upon the training period, training modes studied, the triathletes probably had
however, because when athletes were tested in their more similar mechanical eciencies in both testing
competitive phase, Martinez et al. (1993) showed no HR modes than did the other two groups.
dierences between cycle ergometer and treadmill tests at Although all groups exhibited a higher V_ O2max and
work intensities expressed as percentages of V_ O2max. HRmax on the treadmill than on the cycle ergometer, the
Moreover, unpublished data from our laboratory show triathletes exhibited a lower V_ O2max in the treadmill test,
that experienced triathletes obtain similar submaximal and the runners exhibited the lowest HR for both tests
HRs when performing cycle ergometer and treadmill tests than both of the other groups, HR values achieved on
at the same percentage of V_ O2max in the competitive phase the treadmill were highly correlated to those attained on
of their training year, corroborating the results of the cycle ergometer when expressed at a xed percentage
McArdle and Magel (1970), who showed similar HRs in of V_ O2max (Fig. 5). Our results were congruent with
both tests with trained subjects. On the other hand, Her- those of other studies. Similar observations have been
mansen and Saltin (1969) reported a dierent pattern; reported by Hermansen and Saltin (1969) who found a
their subjects achieved higher HRs on the cycle ergometer high correlation between HRs attained on a cycle er-
than on the treadmill at the same metabolic rates. They gometer and those attained on a treadmill when ex-
assumed that a position eect was responsible, with a re- pressed at absolute V_ O2max. Arts and Kuipers (1994)
duced venous return and a lower stroke volume occurring demonstrated that the relationship between power out-
when the subjects were in the upright position. However, put, V_ O2max and HR were linear for absolute values as
the trend for higher HRs on the cycle ergometer was re- well as for percentages of maximum power output,
duced with training, since trained subjects demonstrated V_ O2max and HRmax, and could be used interchangeably
smaller dierences (4 beats min)1) than untrained sub- to monitor intensity bouts. On the other hand, Weltman
jects (11 beats min)1). et al. (1990) reported that HR data provide guidelines
220

for exercise prescription for runners of similar ability for Flynn MG, Carroll KK, Hall HL, Bushman BA, Brolinson PG,
durations as short as 15)30 min. Moreover, Boulay Weideman CA (1998) Cross-training: indices of training stress
and performance. Med Sci Sports Exerc 30:294300
et al. (1997) observed that HR was a more stable Foster C, Costill DL, Daniels JT, Fink WJ (1978) Skeletal muscle
physiological variable than either V_ E or blood lactate in enzyme activity, ber composition and VO2max in relation to
an endurance performance test lasting 90 min, and re- distance running performance. Med Sport Sci 11:338344
vealed that establishing training intensities with HR re- Gilman MB (1996) The use of heart rate to monitor the intensity of
endurance training. Sports Med 21:7379
sults and the corresponding V_ O2 obtained from a Hermansen L, Saltin B (1969) Oxygen uptake during maximal
common laboratory exercise test remains the most treadmill and bicycle exercise. J Appl Physiol 26:3137
practical approach for controlling exercise intensity Kohrt WM, Morgan DW, Bates B, Skinner J (1987) Physiological
during a prolonged workout. responses of triathletes to maximal swimming, cycling, and
The results of the present study have shown that the running. Med Sci Sports Exerc 19:5155
Kreider RB, Boone T, Thompson WR, Burke S, Cortes CW (1988)
HR dierences generated by using the HR attained Cardiovascular and thermal responses of triathlon perfor-
during a cycle ergometer test to predict the HR that mance. Med Sci Sports Exerc 20:385390
would be attained while running were smaller than the Martinez ML, Modrego A, Ibanez Santos J, Grijalba A, San-
habitual intra-individual variations observed while testeban MD, Gorostiaga EM (1993) Physiological comparison
of roller skating, treadmill running and ergometer cycling. Int J
carrying out prolonged training activities. For example, Sport Med 14:7277
at 70% of V_ O2max these dierences were 1 beat min)1 McArdle WD, Magel JR (1970) Physical work capacity and max-
for runners and 2 beats min)1 for triathletes. At a imum oxygen uptake in treadmill and bicycle exercise. Med Sci
higher training intensity (at V_ O2max) the dierences Sports Exerc 2:118123
McConnell TR (1988) Practical considerations in testing of VO2max
reached 3 beats min)1 for cyclists and triathletes and in runners. Sports Med 5:5768
6 beats min)1 for runners. Therefore, to the athletes McConnell TR, Swett DD, Jeresaty RM, Missiri JC, Al-Hani AJ
and coaches trying to monitor more precisely the den- (1984) The hemodynamic and physiologic dierences between
sity of training, these small statistical dierences would exercise modalities. J Sports Med Phys Fitness 24:238245
appear to be rather non-signicant and may very well Medelli J, Maingourd Y, Bouferrache B, Bach V, Freville M,
Libert JP (1993) Maximal oxygen uptake and aerobic-anaerobic
justify the use of a single progressive test to provide transition on treadmill and bicycle in triathletes. Jpn J Physiol
information about training zones in endurance training 43:347360
activities. Mitchell JH, Sproule BJ, Chapman CB (1958) The physiological
Taken together, these results have demonstrated that meaning of the maximal oxygen intake test. J Clin Invest
37:538547
the relationship between HR and V_ O2 expressed as a Miura H, Kitagawa K, Ishiko T (1997) Economy during a simu-
percentage of V_ O2max, gathered during cycle ergometer lated laboratory test triathlon is highly related to Olympic
or treadmill tests, are interchangeable and can be used to distance triathlon. Int J Sport Med 18:276280
monitor the intensity of activities performed using either Moreira-Da-Costa M, Russo AK, Picarro IC, Silva AC, Leite-De-
Barros-Neto JK (1984) Maximal oxygen uptake during exercise
mode of exercise. Although trained cyclists would using trained or untrained muscles. Braz J Med Biol Res
probably prefer a cycle ergometer test and runners a 17:197202
treadmill test, these results are especially interesting for Moreira-Da-Costa M, Russo AK, Picarro IC, Barros Neto TL,
triathletes and other athletes who use a variety of Silva AC, Tarasantchi J (1989) Oxygen consumption and ven-
training modes; those athletes may then be able to use tilation during constant-load exercise in runners and cyclists.
J Sports Med Phys Fitness 29:3644
only one mode of testing to obtain their training Nagel F, Balke B, Bapista G, Alleyia J and Howley E (1971)
guidelines. Compatibility of progressive treadmill, bicycle and step tests
based on oxygen uptake responses. Med Sci Sports 3:149
154
O'Toole ML, Douglas PS (1995) Applied physiology of triathlon.
References Sports Med 19:251267
Pannier JL, Vrijens J, Van Cauter C (1980) Cardiorespiratory re-
Arts FJP, Kuipers H (1994) The relation between power output, sponses to treadmill and bicycle exercise in runners. Eur J Appl
oxygen uptake and heart rate in male athletes. Int J Sports Med Physiol 43:243251
15:228231 Pechar GS, McArdle WD, Katch FI, Magel JR, De Luca J (1974)
Barbeau P, Serresse O, Boulay MR (1993) Using maximal and Specicity of cardiorespiratory adaptation to bicycle and
submaximal aerobic variables to monitor elite cyclists during a treadmill training. J Appl Physiol 36:753756
season. Med Sci Sports Exerc 25:10621069 Schneider DA, Pollack P (1991) Ventilatory threshold and maximal
Bouchard C, Godbout P, Mondor JC, Leblanc C (1979) Specicity oxygen uptake during cycling and running in female triathletes.
of maximal aerobic power. Eur J Appl Physiol 40:8593 Int J Sport Med 12:379383
Boulay MR (1995) Physiological monitoring of elite cyclists. Sports Schneider DA, Lacroix KA, Atkinson GR, Troped PJ, Pollack J
Med 20:111 (1990) Ventilatory threshold and maximal oxygen uptake dur-
Boulay MR, Simoneau JA, Lortie G, Bouchard C (1997) Moni- ing cycling and running in triathletes. Med Sci Sports Exerc
toring high-intensity endurance exercise with heart rate and 22:257264
thresholds. Med Sci Sports Exerc 29:125132 Sharkey BJ (1988) Specicity of testing. In: Grana WA, Lombardo
Coast JR, Welch HG (1985) Linear increase in optimal pedal rate JA, Sharley BJ, Stone JA (eds) Advances in sports medicine and
increased power output in cycle ergometry. Eur J Appl Physiol tness. Years Book Medical Publishers, Chicago, pp 2543
53:339342 Sleivert GG, Rowlands DS (1996) Physical and physiological fac-
Dengel DR, Flynn MG, Costill DL (1989) Determinants of tors associated with success in the triathlon. Sports Med 22:818
success during triathlon competition. Res Q Exerc Sport Steel RGD, Torrie JH (1980) Principles and procedures of statis-
60:234238 tics: a biometrical approach. McGraw-Hill, New-York
221

Taylor HL, Buskirk ER, Henschel A (1955) Maximal oxygen in- Veicsteinas A, Feroldi P, Dotti A (1985) Ventilatory response
take as an objective measure of cardiorespiratory performance. during incremental exercise tests in weight lifters and endurance
J Appl Physiol 8:7380 cyclists. Eur J Appl Physiol 53:322329
Turley KR, Wilmore JH (1997) Cardiovascular responses to Weltman A, Snead D, Seip R, Schurrer R, Weltman J, Ruttand R,
treadmill and cycle ergometer exercise in children and adults. Rogol A (1990) Percentages of maximal heart rate, heart rate
J Appl Physio 83:948957 reserve and VO2max for determining endurance training inten-
Van Handel PJ, Baldwin C, Puhl J, Katz A, Dantine S, Bradley PW sity in male runners. Int J Sport Med 11:218222
(1988) Measurement and interpretation of physiological pa- Zhou S, Robson SJ, Davie AJ (1997) Correlations between short-
rameters associated with cycling performance. In: Burke ER, course triathlon performance and physiological variables de-
Newsom M (eds) Medical and scientic aspects of cycling, termined in laboratory cycle and treadmill tests. J Sports Med
Human Kinetics, Champaign, Ill., pp 4772 Phys Fitness 37:22130