Sports Medicine 5: 57-68 (1988)

0112-1642/88/000 1-0057/$06.00/0
© ADIS Press Limited
All rights reserved.

Practical Considerations in the Testing of V0 2 max

in Runners
Timothy R. McConnell
Exercise Physiology Section, Department of Cardiology, Geisinger Medical Center,
Danville

Contents Summary ........................................................................................................................................ 57

I. Exercise Modalities ................................................................................................................... 59
2. Protocol Selection ..................................................................................................................... 61
2.1 Workstage Loading Schemes ............................................................................................. 61
2.2 Grade Versus Speed Increases ........................................................................................... 61
2.3 Warm-Ups ...................................................................................; ....................................... 62
2.4 Recommendations for Selection of Protocol ................................................................... 62
3. Laboratory Environment. ......................................................................................................... 63
4. Testing Schedule ....................................................................................................................... 63
5. Runners' Pretest Instructions and Preparation ...................................................................... 63
6. Test Termination ...................................................................................................................... 63
7. Reproducibility and Accuracy of V0 2 rna •.••••.••••••..•.•••.•.•••••.••••••...•••••.•.•••••.•.•••••.••••••.••••••.••••• 64
8. Conclusion ................................................................................................................................. 66

Summary Maximal oxygen intake d' 0:: /IIaJ may be the most physiologically sign(ficant and.
therefore, the most commonly measured parameter in the physiological assessment a/well
trained runners. To insure \'alid and reliable i' 0:: /IIax ralues. there are a number of prac-
tical concerns that must be addressed. An exercise modality must be selected that activates
the spec({ic muscle mass and duplicates the motion utilised in the runners' training pro-
gramme. Motorised treadmill running has been shown to allow the most efficient. valid
and reproducible (-0:: /II'" I'alues in a controlled testing em·ironment. A testing protocol
must be used that is easily administered. comfortablefor the runner. allows test completion
within IO minutes. uses grade increments at a constant speed, and is reproducible. Speeds
should be selected that approximate the runner's training pace. The criterion for the at-
tainment of i'o: /II"., include the lel'elling 0.0' 0.( C"O: with an increase in work stage. the
onset 0.( extreme exhaustion. the respiratory exchange ratio exceeding 1.1. blood lactate
lel'els approaching or exceeding 10 mmol/L. and ratings 0.( perceired exertion 0.( 19 or
lO. Qualit.\" control in terms 0.( standardisation 0.( procedures and calibration and main-
tenance 0.(equipment is essentialfor the minimisation o.(technological error..Heasurement
eqllipment must be a.O'ordable, dependable. easily calibrated. .\"ield reproducible data. and
hI.' I\'ithin the technical skill lerels 0.( the operating personnel.
v~ mu Testina in Runners
The measurement of maximal oxygen intake
(V02 max) determines an individual's respiratory,
cardiovascular, and skeletal muscular capabilities
for the uptake, transport, and utilisation of oxygen
(fig. 1). If this integration of physiological mech-
anisms is impaired due to pathology or as a result
of a sedentary lifestyle, V02 max will be reduced
from expected values.
Because of its functional sianificance, V02 max
provides a good estimate of the potential of the
cardiovascular system and endurance exercise ca-
pacity. In spite of the physiological significance of
V02 max, a wide variation in the relationship be-
tween V0 2 max and running performance (r - 0.08
to 0.91) bas been demonstrated (Astrand & Saltin
1961; Conley & Krahenbuhl 1980; Costill et aI.
1976; Daniels et al. 1978; Davies et JLI. 1984; Froe-
lieher et al. 1974; Gibson et al. 1979; Hammond
& Froelieher 1984; Londeree 1986; Moreira-Da-
Costa et aI. 1984; Shephard 1984; Stamford 1976;
Sjodin & Svedenhag 1985). This variability does
Muscle
activity
Circulation: Peripheral
PhysiOlogical OCOz Dilate
responses:
fig. 1. A scheme illustrating the gas transport mechanisms for coupling cellular (Internal) to pulmonary (external) respiration. WIth
pennIsaIon from Wauerman at aI. (1887).
S8
not undermine the importance of V02 max but must
be interpreted in terms of the degree of homo-
geneity of the sample population and type of ath-
letes being tested. For example, the correlation be-
tween V0 2 max and marathon race pace (r = 0.78)
indicates a wide variation and poor predictability
among a group of elite marathon runners (SjOdin
& Svedenbag 1985). Even though there is a sub-
stantial spread in V02 max and poor correlations
among individuals in such a homogeneously trained
population, a good relationship is seen between
V02 max and performance when mean data are re-
viewed (table I) or a population more heteroge-
neous in terms of V02 max is tested. Even though
other variables have been shown to have as high
or higher relationship with running performance
(Conley & Krahenbuhl 1981; Costill et aI. 1973,
1976; Daniels 1985; Daniels et aI. 1978; Farrell et
al. 1979; Kenney & Hodgson 1985; Kumagi et aI.
1982; Londeree 1986), V0 2 max is felt to give an
indication of the runner's overall maximal cardio-
02/C02 Ventilation
delivery eVA + Vo = Ve.!
Pulmonary
SV Recruit
HR
V01 max Testing in Runners
Table I. V02 max and personal records (PR) of elite marathon
runners
V0 2max PR Reference
(ml/kg/min) (h:min)
70.9 2:23 Costill & Fox (1969)
74.1 2:16 Svedenhag & SjOdin (1984)
74.1 2:15 Pollock (1977)
79.0 2:13 Davies & Thompson (1979)
respiratory capabilities and sets the limits on the
amount of oxygen that can be utilised for the dur-
ation of an event (Snell & Mitchell 1984). Since
pacing is a function of both race distance and rel-
ative intensity, knowledge of V0 2 max assists the
runner and coach in determining the appropriate
fractional utilisation of aerobic capacity (%V0 2 max)
required to maintain a desired race pace .(Costill et
al. 1973; Snell & Mitchell 1984). If the V0 2 max of
a competitive runner is low relative to the antici-
pated need for a given event, a correction in train-
ing emphasis must then be addressed that will re-
sult in an enhancement of V0 2 max as well as the
ability to maintain exercise at high relative intens-
ities (Shephard 1978; Sjodin & Svedenhag 1985;
Williams et al. 1967).
Since the measurement of V0 2 max is para-
mount for the assessment of running performance
capabilities (Daniels et al. 1978), many articles have
discussed appropriate testing methodology. Inves-
tigations involving runners have used testing pro-
cedures that have been based on recommendations
from the literature, have duplicated others or have
devised their own protocols that meet the specific
objectives of their study. Concerns and practical
considerations that must be addressed when meas-
uring V0 2 max in runners include: (a) exercise mo-
dality; (b) protocol selection; (c) laboratory en-
vironment; (d) subject preparation and motivation;
and (e) accuracy of the measure.
Shephard (1984) published a detailed review of
tests of maximal oxygen intake. The intent and
scope of the present review is more limited and
will discuss specific practical concerns for testing
runners. Also. V0 2 max is only one physiological
parameter that is predictive of running perform-
59
ance. Testing protocols and principles for assessing
other performance parameters have been reviewed
by Londeree (1986).
1. Exercise Modalities
The specific exercise task employed for testing
has a definite influence on the V0 2 max response.
McConnell et al. (1984) found that the V0 2 max ob-
tained on a motorised treadmill was 10.2, 26.4, and
28.6% higher than that obtained during upright leg
ergometry, supine leg ergometry and arm ergo-
metry, respectively (table II). McArdle &. Magel
(1970) found that subjects had a 9.9% lower V0 2 max
during leg ergometry than during treadmill testing.
Of their subjects, the only 2 who had higher V0 2 max
values during leg ergometry were heavy recrea-
tional cyclists. Other investigators have shown
similar differences between treadmill and leg er-
gometry testing ranging from 5 to 11.2% (Astrand
1967; Astrand & Saltin 1961; Diaz et al. 1978; Fair-
shter et al. 1983; Faulkner et al. 1971; Glassford et
al. 1965; Hammond & Froelicher 1984; Harrison
et al. 1980; Hermansen & Saltin 1969; Hermansen
et al. 1970; Kamon & Pandolf 1972; McArdle et
al. 1973; McKay & Banister 1976; Miles et al. 1980;
Neiderberger et al. 1974; Shephard et al. 1968;
Smod1aka 1982; Stamford 1975, 1976).
When testing trained athletes, the cardiorespir-
atory training response is best demonstrated when
the athletes activate the specific muscle mass util-
ised in their training programmes; therefore,
Table II. Mean maximal exercise values
Treadmill Leg Supine Arm
ergometry ergometry ergometry
V0 2max 48.9 43.9 36.0 34.9
(ml/kg/min)
HRmax (beats/min) 187.1 181.2 164.6 175.8
SBP max (mm Hg) 174.3 180.0 178.4 148.8
RPP max (X 1(2) 328.4 326.8 294.0 261.3
Abbreviations: V02 max = maximal oxygen consumption; HRmax
= maximal heart rate; SBP max = maximal systolic blood pres-
sure; RPP max = maximal rate pressure prOduct.
V~mu Testing in Runners
V02 max is dependent upon the mode of exercise
and the specific trainilll of the athlete (Snell &
Mitchell 1984). The more closely a laboratory can
simulate the specific muscular action involved in
training during the test, the more objective and
valuable the V02 max assessment (Davies et al.
1984). Corry & Powers (1982) demonstrated the
importance of using specifically trained muscle
groups by comparilll trained runners to trained
swimmers. They found that the runners had sig-
nificantly higher V0 2 max values during treadmill
running (70.2 ml/kg/min) when compared with
swimmers (60.0 ml/kg/min). But, when an arm
pulling motion similar to that of swimming was
used, the swimmers obtained significantly greater
values (47.1 ml/kg/min vs 37.4 ml/kg/min).
McArdle et al. (1978) found that following a nlD-
ning programme individuals improved their tread-
mill-obtained V02 miX by 6.3%, while their swim-
ming V02 mix improved by only 2.6%. Earlier, they
had demonstrated that swim training did not sig-
nificantly improve V0 2 max values obtained during
a running test (Magel et al. 1975).
Moreira-Da-Costa et al. (1984) investigated
V02 max in trained runners, cyclists and untrained
individuals to compare trained to untrained muscle
groups. The non-athletes and runners had II % and
12% greater V02 max values during treadmill ex-
ercise when compared with leg ergometry. On the
other hand, the cyclists obtained 7% greater V02 max
values during leg ergometry. They concluded that
an ergometer which requires approximately the
same muscular activity as is usually performed by
the subject should always be employed to evaluate
the quantitative effects of training on cardiovas-
cular and respiratory functions.
Others have also demonstrated that the differ-
ence in V0 2 max values and the physiological re-
sponse to submaximal exercise can be due to the
utilisation of specifically trained versus untrained
muscle groups (Jacobs & Sjodin 1985; Kamon &
Pandolf 1972; Pelchar et al. 1974; StrBDlme et al.
1977).
In terms of different running modalities,
McMiken and Daniels (1976) found no differences
in the V0 2 max measured between treadmill and
60
track running. Davies et al. (1984) tested trained
runners on motorised and non-motorised tread-
mills and found no differences in the V0 2 max
values.
Running must be employed to obtain valid
V02 max values when testing trained runners. Even
though a few studies have demonstrated similar
V02 max values when motorised treadmills were
compared to track running or non-motorised
treadmills, the vast majority of the literature sup-
ports the use of motorised treadmill running as the
most effective modality for the laboratory analysis
of V0 2 max in runners. When using a motorised
treadmill, running on a grade has been shown to
elicit greater and more reliable V02 mo values when
compared with that obtained with flat treadmill
running and speed increments. It is felt that run-
ners will eventually be unable to maintain higher
speeds without having reached their actual V0 2 mo.
V02 values during submaximal and maximal track
running have been shown to be highly correlated
and not significantly different from that obtained
during same speed motorised treadmill running
(McMiken & Daniels 1976). But the difficulty of
measuring V0 2 during track running and other
confounding variables (i.e. temperature, humidity,
and wind velocity) makes motorised treadmill run-
ning the more reliable measurement procedure.
Leger and Mercier (1984) reinvestigated published
data and derived an average equation for the pre-
diction of the gross energy cost of treadmill and
track running:
Treadmill: V02 (ml/kg/min) = 2.209 + 3.163 speed
(km/h)
Track: V0 2 (ml/kg/min) = 2.209 + 3.163 speed +
0.000525542 speed3
They concluded that there are wide variations
in the energy costs of running and specific equa-
tions need to be developed for male and female
and good and bad runners. If the energy cost of
running is needed for research purposes, it should
be measured because of the wide variations dem-
onstrated between individuals.
V0 2 max Testing in Runners
2. Protocol Selection
Investigators who have assessed the effects of
different protocols for testing well trained runners
have shown that VOl max values are essentially re-
silient to minor changes in testing protocol. The
following discussion will review the literature that
has compared different stage loading schemes,
treadmill speeds and grade selection, and protocols
that have been used when well trained runners were
investigated.
2.1 Workstage Loading Scheme
Protocol selection should be based on the ob-
jectives of the test, subject comfort, and data ac-
quisition. If a physiological steady-state at each ex-
ercise increment or a rest period between each stage
is important, then a discontinuous protocol with
longer stages may be more appropriate. But if the
attainment ofV0 2 max is the desired outcome, then
continuous protocols with shorter exercise incre-
ments may be more efficient with regard to testing
time and may be perceived as being more favour-
able by the runner.
A number of investigators have studied whether
discontinuous protocols that allow rest periods be-
tween each exercise increment result in different
V0 2 max values when compared with continuous
protocols (no rest periods between stages). Taylor
et al. (1955) presented a comprehensive discussion
concerning the measurement and implications of
V0 2 max' Their testing protocol consisted of inter-
mittent stages that were performed one each on
successive days. A constant running speed was used
and each stage was increased by 2.5% grade incre-
ments. The subjects returned each day until the
2.5% grade increment resulted in a VOl increase
ofless than 150 ml/min (2.1 ml/kgjmin). This 'lev-
elling otT' of V0 2 was considered to be the indi-
vidual's true physiological V0 2 max' Since then,
many others have assessed discontinuous versus
continuous protocols for measuring V0 2 max' Be-
cause of the impracticality of Taylor's (1955) pro-
tocoL Mitchell et al. (1958) used a modification
that could be completed in one laboratory visit and
61
found similar V0 2 max reproducibility. McArdle et
al. (1973) found no difference in the VOl max ob-
tained between a continuous or discontinuous pro-
tocol or the protocol suggested by Mitchell et al.
(1958). Others have obtained similar results (Mak-
sud & Coutts 1971; Stamford 1976) and have con-
cluded that reliable V0 2 max values can be ob-
tained with shorter protocols of a continuous
nature.
Clark and McConnell (1986) compared proto-
cols with 1- versus 2-minute increments for assess-
ing V0 2 max in well trained runners and found that
there was no significant difference between pro-
tocols in the V0 2 max obtained. Also, the protocol
with the I-minute stages had significantly shorter
testing times and lower rating of perceived exertion
values at peak exercise.
Fairshter et al. (1983) studied the use of short
duration exercise increments and found that 15-
second stages elicited the same VOl max values as
protocols that had I-minute stages. Whipp et al.
(1981) used ramp testing on a cycle ergometer
(gradual I-minute increments in exercise intensity)
and demonstrated similar V0 2 max values when
compared with protocols of longer duration.
These findings suggest that a continuous pro-
tocol with workstage durations of I minute or less
may be most efficient, in terms of testing time, for
obtaining VOl max in runners and may be per-
ceived as being less difficult by the runner.
2.2 Grade J'ersus Speed Increases
Gibson et al. (1979) suggested that either speed
or grade be increased throughout the test, but not
both. Taylor et al. (1955) felt that grade alone
should be incremented because subjects are unable
to maintain their running skill at higher speeds.
They demonstrated that a greater percentage of
runners attained a 'plateau' of VOl when grade was
increased at a constant speed. Saltin and Astrand
(1967) suggested raising the treadmill by 2.6% every
3 minutes from a predetermined value (based on
estimated V0 2 rna,) at a constant speed of 10
km/h.
In 1968 the International Biological Programme
V02 mall Testing in Runners
(IBP) recommended a standard protocol using a
constant speed and 2.S% grade increments every 2
minutes (lshiko 1978; Shephard et a1. 1968). Others
have agreed with using gradient increases and con-
stant running speeds throughout the test (Gass et
al. 1981; Harrison et al. 1980; Pollock 1977; Saltin
& Astrand 1967).
When runners were investigated, testing proto-
cols were selected that use the same speed for all
runners or different speeds and stage increments
according to the subject's predicted V0 2 max
(Astrand & Rodahl 1977; Costill & Fox 1969).
Taylor et al. (19SS) and Gibson et a1. (I979) sug-
gested constant testing speeds of 4.3 km/h and 6.2
km/h, respectively, because it was the slowest
speeds at which all subjects had to maintain a run-
ning stride. Running V0 2 max is also interrelated
with speed of running and the state of training.
Harrison et al. (1980) agreed that a constant speed
of 10 km/h was an optimal speed for those unac-
customed to running and 12 km/h was more suit-
able for experienced runners. Pollock (1977) sug-
gested that a speed be selected that results in a
testing time of 7 to 10 minutes and felt that IO:S
to 12 km/h for untrained and 16.1 to 19.4 km/h
for trained runners were appropriate. Davies et al.
(1984) stated that it is essential for laboratories to
simulate as closely as is practically possible the
training and competitive environments of the ath-
letes during the test. Clark and McConnell (1986)
attempted to duplicate the training environment
by comparing protocols that allowed the runners
to run at their training pace and at manually self-
selected speeds to protocols that used the same
speed for all runners. Although there was no sig-
nificant difference in the V0 2 max values obtained,
the runners had lower heart ~tes and lower rating
of perceived exertion values at every level of V0 2
when their training pace (P3) was used.
The consensus of the literature recommends a
constant running speed throughout the test with
increments in elevation. Speeds should be selected
according to the runners' level of training with faster
speeds (approximating training pace) being more
appropriate for well trained runners. Also, jf the
speed selected is too slow, the runner may obtain
62
elevations (greater than 15%) that may result in a
loss of running efficiency or posterior leg or lower
back muscle soreness.
2.3 Warm-Ups
Warming up for 15 minutes at the testing speed
resulted in higher heart rate, V0 2 and muscle tem-
peratures, and lower lactic acid levels when com-
pared with tests that did not use a warm-up (Mar-
tin et a1. 1975). Even a brief (2- to 3-minute) warm-
up of a moderate intensity « 70% V0 2 mu) has
been shown to minimise the risks of ECG and other
cardiovascular abnormalities (Bernard et al. 1973;
Foster et al. 1982) and musculoskeletal injuries (de
Vries 1980). Astrand and Rodahl (1977) agreed and
suggested a warm-up of 5 to 10 minutes at SO% of
the estimated V0 2 max.
2.4 Recommendations for Selection
of Protocol
Valid and reliable V0 2 max values are not af-
fected by minor adaptations in protocols. This al-
lows the investigator some latitude for selecting a
protocol that meets the objectives of the investi-
gation and ensures adequate data acquisition. The
following recommendations summarise the above
discussion:
I. A speed should be selected that is comfort-
able and requires all subjects to run.
2. Speed should be held constant with gradual
increments in elevation.
3. Total testing time should be from 7 to 10
minutes.
4. Continuous protocols result in equivalent
V02 mu values when compared to discontinuous
protocols and result in shorter testing times.
S. Protocols that employ short work stages of I
minute, 15 seconds or continuously incremented
intensities have been shown to elicit reliable
V02 max values, are time efficient and may be more
desirable to the runner.
6. The test should be preceded by a warm-up
at a relative intensity of at least 50% of VOl max.
V0 2 rna. Testing in Runners
3. Laboratory Environment
Since the perception of work difficulty may af-
fect the runner's willingness to push to maximal
levels of exertion (Morgan 1985). ambient temper-
ature should be maintained so as to prevent ex-
cessive discomfort due to overheating (Hammond
& Froelicher 1984). Taylor et al. (1955) found that
a room temperature of 32.2T resulted in signifi-
cantly lower V0 2 max values than an ambient tem-
perature of 25.6°C, while others have shown
V0 2 max to be unaffected by temperature (Rowell
et al. 1964). V0 2 max measured in a hot environ-
ment is felt to be a poor indicator of aerobic ca-
pacity because of the extreme peripheral vasodila-
tation that occurs. Rowell et al. (1964) suggests that
the optimal laboratory temperature is 17°C. Others
have reported laboratory temperature and humid-
ity values ranging from 18 to 2rC and 30 to 60%.
respectively (Conley & Krahenbuhl 1980; Dimri et
al. 1980; Falls & Humphrey 1973; Hammond &
Froelicher 1984; McMiken & Daniels 1976; Mor-
eira-Da-Costa et al. 1984: Taylor et al. 1955). The
American Heart Association recommends an am-
bient temperature of 20T with approximately 50%
humidity (American Heart Association 1979). In a
discussion concerning environmental variables af-
fecting fitness testing, Strydom (1978) recom-
mends dry bulb temperatures within the range of
19 to 24°C and a wet bulb temperature below 24°C.
The testing area should also be well ventilated and
the use of fans may be considered to enhance the
evaporative cooling process.
4. Testing Schedule
While a number of articles have demonstrated
no significant circardian rhythm effect on VO~ max
(Faria et al. 1982; Reilly & Brooks 1982a.b). others
have presented some inconsistent results. Wojtczak-
Jaroszowa and Banaskiewicz (1974) reported that
VO~ rna, was 5% greater in the morning. while Ya-
maji et al. (1981) found 3% greater values at 6pm.
These inconsistencies may be the result of individ-
ual circadian rhythm variations which suggest that
optimal testing times are individual specific. Be-
63
cause of these discrepancies. many investigators
schedule repeat tests at the same time of the day,
whenever possible (Froelicher et al. 1974; Harrison
et al. 1980; McArdle et al. 1973).
5. Runner's Pretest Instructions
and Preparation
Even though Stamford et al. (1978) showed that
V0 2 max values were not altered by exhaustion and
elevated blood lactate levels. extremes of climate
and exercise should be avoided at least 24 hours
preceding the test (McArdle et al. 1973). The run-
ner should also be well rested and caffeine con-
taining beverages, drugs, alcohol. cigarette smoking
or nicotine in other forms such as smokeless to-
bacco and large meals should be avoided for at least
2 to 4 hours before the test (Falls & Humphrey
1973; Froelicher et al. 1974; Jones & Haddon 1973;
Hammond & Froelicher 1984).
A period of familiarisation with testing proce-
dures is important for allowing the runner to be-
come accustomed to the equipment. understanding
the objectives of the test. reducing anxiety and al-
lowing consistency of efficiency between tests (She-
phard et al. 1968). This is particularly important if
reliable steady-state O 2 requirements at each in-
crement is desired (Falls et al. 1973; Hammond &
Froelicher 1984; Stamford et al. 1976).
6. Test Termination
To obtain reliable V0 2 max values it essential that
test termination be based on some standard cri-
teria. The most commonly used criterion for de-
termining whether a maximal effort has been ob-
tained is the documentation of a 'levelling off or
plateauing' of VO~ with an increase in exercise in-
tensity. The plateau documents that V0 2 max has
been achieved and that the final work level was
accomplished largely through the contribution of
anaerobic energy sources (Stamford 1978).
Taylor et al. (1955) established their test ter-
mination point when VO~ increased by less than
150 ml/min (2.1 ml/kg/min) with an increase in
elevation of 2.5% at a constant speed. Only 7 of
V02 mall Testing in Runners
II S subjects failed to reach this criterion. Stamford
(1976) used changes in V0 2 of less than 112 mil
min between successive stages to be indicative of
a 'levelling off' of V0 2. Others have either sug-
gested the use of similar values or no increase or
a decline in V0 2 with an increase in exercise in-
tensity (Davies et al. 1984; McMiken & Daniels
1976; Shephard et al. 1968). Of 80 tests performed,
Katch et al. (1982) demonstrated a plateauing in
61 % of the tests. Also, there were no differences in
the mean V0 2 max values for the tests that dem-
onstrated a plateauing and those that did not in
the same subjects. Davies et al. (1984) showed that
74% of his trained runners were able to obtain a
plateauing of V02 and a plateau was not obtained
in only 6% of the total tests performed. Others have
also found that a plateauing of V02 can be ob-
tained in a high percentage of well trained indi-
viduals (Stamford 1976; Pollock 1977). Stamford
et al. (1978) demonstrated that tests preceded by
severe exhaustion did not demonstrate an ade-
quate plateau, even though the V0 2 max values were
not different than those obtained when a plateau-
ing was seen.
Even though the plateauing of V0 2 is felt to be
the most valid indicator of a maximal effort, a
number of the studies mentioned demonstrated that
V02 max values were not different between the same
subjects when a plateau was obtained and when it
was not. Other criterion have also been established
for indicating when a maximal effort has been or
is being approached during a test. These include:
I. Subjective fatigue, exhaustion, and an ina-
bility to continue (Gibson et al. 1979).
2. The estimated aerobic requirement of the last
stage is at least 10% greater than that measured.
This indicates a possible slower acceleration in
V0 2 (McMiken & Daniels 1976).
3. Respiratory exchange ratio of at least 1.0 or
1.10 (Davies et al. 1974; McMiken & Daniels 1976).
4. Blood lactate values approaching or exceed-
ing 10.0 mmol/L (Shephard et al. 1968).
S. Heart rate within 10 beats of predicted max-
imum (Gibson et al. 1979; Shephard et al. 1968).
6. Rating of perceived exertion of 19 or 20
(Hammond & Froelicher 1984).
64
7. Reproducibility Slid Accuracy
0/ ~02.wc
A number of investigators have discussed or di-
rectly investigated the variability of V0 2 max val-
ues. Katch et al. (1982), using repeat testing on the
same subjects, panitioned the variability ofV02 max
for an individual into several sources: technologi-
cal error and biovariation (biological fluctuation).
The technological error was computed for each
piece of equipment as the standard deviation of
multiple trials and includes the variable error of
the instruments used to measure and calculate
V0 2 max, uncontrolled environmental factors,
reading errors, and other unidentified errors. The
biological variation is reflective of the inherent
biological variation of the organism and is com-
puted as the standard deviation for the V0 2 max
values for each individual, after subtracting net
technological error. The combined (technological
and biological) sources of variance averaged ±
S.6%. Biological variability accounted for 90% or
more of the combined variability while technolog-
ical error accounted for less than 10%. They sug-
gested that in light of the magnitude of biological
variation multiple tests may be necessary to secure
reliable control values when using measures of
V0 2 max to document training effects.
When Taylor et al. (19SS) analysed the variance
in 5 duplicate tests on 12 subjects, in 28 duplicate
determinations of V0 2 max they found that the dif-
ference between individuals was highly significant
while the trend over time was not. The coefficient
of variation for the means over time was 2.4%. In
their group it was estimated that 18.7% of the var-
iability was random, 80.S% was accounted for by
differences between individuals, and less than 1%
was accounted for by day-to-day variability or
technological error. In an earlier investigation.
Taylor (1944) reponed that 30% of the total vari-
ability could be accounted for by biological vari-
ation and less than I% was due to technological
error.
The reproducibility of V0 2 max in individuals
can be measured to within a coefficient of vari-
ation of ± 3% (Gibson et al. 1979; Taylor et al.
VO! rna. Testing in Runners
1955) with the day-to-day variability ranging from
4 to 6% (Boileau et al. 1977). 10% ofthis variability
can be attributed to technical difficulties (Shephard
1984) and at least a 10% day-to-day variation in
the level of physical conditioning of subjects
(Wright et al. 1978).
Investigators must address the following con-
siderations to help minimise the variability of
V0 2 max·
I. Since test termination points are subjectively
determined by the runner, the reasons for test ter-
mination must be consistent for all tests. The in-
vestigator should use consistent motivational tech-
niques for encouraging the runners to continue as
long as possible. Suggestions include intensive ver-
bal motivation and informing the subject of the
amount of time left to complete a stage or allow
for data acquisition.
2. It is advisable to perform tests under con-
sistent conditions with regards to time of day. pos-
tabsorptive state. ambient conditions. and length
of time since previous exercise.
3. Problems with gas analysis. systematic errors
and technical personnel must also be given con-
stant attention. Staff must be given frequent in-
service training to ensure consistent procedures.
Calibration gases must be accurate. The gas ana-
Iysers must be calibrated frequently particularly
immediately preceding and following each test.
Respiratory and sampling tubing must be checked
for leaks daily.
The inception of automated metabolic units
have eliminated some of the technical problems
such as contaminated samples due to erroneously
turned valves or bags with undetected leaks. They
also allow metabolic responses to be continually
monitored and updated throughout the test. This
affords the investigator the opportunity to contin-
uously monitor values and may prevent the need
for retesting because of technical errors that were
previously unable to be detected until the test was
completed and all collected samples were analysed.
On the other hand, automated systems can be ex-
pensi ve. faulty assumptions on the part of the user
may result in inaccuracies and software limitations
in terms of timing sequences and lack of flexibility
65
may limit applicability. Users must also ensure the
accuracy and correctness of the computational and
correction equations within the program by check-
ing a list of software program steps.
Under carefully controlled and standardised
conditions. the variation in V0 2 max should reflect
the biological error. Current state of the art rapid
responding gas analysers have capabilities of ±
0.0 I % of full scale when properly calibrated. An-
alyser specifications should be obtained to ensure
that they meet the needs of the laboratory, espe-
cially if a rapid response time is desired. The in-
vestigator must also ensure the accuracy of their
calibration gases by using manual gas analysis or
crosscalibration with an already verified gas con-
centration. Calibration gas concentration should be
known to within ± 0.03% (Weber & Janicki 1986).
The automated metabolic systems also neces-
sitate the integration of a volume flow device that
has the capability of continuously reporting accu-
rate flow volumes ranging from the resting state to
peak exercise (5 to 150 L/min). The flow-volume
meter selected cannot be sensitive to the phasic na-
ture of respiration. turbulent flow at higher vol-
umes. moisture content, changes in flow rate,
changes in temperature. and must offer minimal
resistance, linearity. and the desired frequency re-
sponse (Froelicher 1983; Wasserman et al. 1987;
Weber & Janicki 1986).
Automated systems commonly use dry gas me-
ters, turbine flow meters, or pneumotachs with a
pressure ditTerentiating transducer. The major dif-
ference between systems is whether they directly
measure volume or flow. If flow is measured, it
must be integrated over time. The accuracy of ven-
tilation is ± 4% by measuring volume over time.
There is a 2% deviation from linearity when flow
is integrated over time (Harrison et al. 1980). For
information concerning flow-volume meters and
gas analysers see Weber and Janicki (1986) and
Wasserman et al. (1987). Recommendations for
selecting metabolic equipment include:
I. Investigators should obtain the specifications
for all components of a metabolic system from the
manufacturer.
2. Peers who are using the system being con-
V02 mu Testina in Runnen
sidered should be contacted and consulted con-
cerning the acquisition of reliable data, the fre-
quency and extent of malfunctions, and
maintenance support.
3. If possible, obtain equipment on consign-
ment for a period of time to allow for hands on
experience before final purchase decisions are made.
4. Laboratories must also be aware of possible
biases on the part of manufacturen, distributing
firms, and even peen who may have lost their ob-
jectivity concerning the systems they have pur-
chased.
S. Discuss prospective purchase with biomedi-
cal engineering personnel who may offer qualified
insight into other options that may be less expen-
sive or more desirable.
6. Discuss the costs, stipulations, and length of
service contracts.
7. Some automated systems offer many 'extras'
in terms of software capabilities and other options
that do not improve the quality of the data ob-
tained, give erroneous impressions, and result in
increased costs.
8. COIIC/",io,.
Even though V02 max is very reproducible and
resilient to minor changes in testing procedures, in-
vestigaton need to address the following to ensure
reliable values are obtained when assessing trained
runnen: exercise modality, protocol selection (speed
and grade), testing schedule, test termination points,
and reproducibility and accuracy of the measure.
The protocol selected should offer maximal effi-
ciency and comfort for the runner while fulfilling
data acquisition needs and meeting the overall ob-
jectives of the test.
American Heart Association: Subcommittee on Rehabilitation.
Tartet Activity Group. The exercise standards book. American
Heart AlIIOCiation. Dallas. 1979
Astrand p.o. Measurement of maximal aerobic capacity. Cana·
dian Medical Association Journal 96: 732·73S. 1967
Astrand p.o. Rodahl K. Textbook of work physioiOlY: physio-
lOIical basis of exercise. 2nd ed.• McGraw Hill Company. New
York. 1977
Astrand p.o. Saltin B. Maximal oxyaen uptake and hean rate in
66
various types of muscular activity. Journal of Applied Physi·
010lY 16: 977-981. 1961
Bernard RJ. Gardner GW, Diaco NV, MacAlpin RN, Kattus AA.
CardiovucuJar responses to sudden strenuous exercise: heart
rate. blood pressure. and ECG. Journal of Applied PhysiolOlY
34: 833-831, 1913
Boileau RA, Bonan A, Heyward VH, Massey BH. Maximal aero-
bic capacity on the treadmill and the bicycle efiometer of boys
11-14 yean of qe. Journal of Sports Medicine and Physical
Fitness 11: IS3-172, 1977
Clark BA. McConneU TR. Comparison of treadmill testina pro-
tocols to UICSI V02 .... in well trained runnen. Abstract.
Medicine and Science in Sports and Exercise 18 (Suppl.): S36,
1986
Conley DL. Krahenbuhl GS. Runnina economy and distance run-
nina performance of hiahly trained athletes. Medicine and Sci-
ence in Sports and Exercise 12: 3S1-360, 1980
Corry I. Powen N. Maximal aerobic power measurement in run-
nm and swimmers. British Journal of Sports Medicine 16:
IS4-160, 1982
Costill DL. Fink WJ, Pollock ML Muscle fiber composition and
enzyme activities of elite distance runners. Medicine and Sci-
ence in Sports 8: 96-100. 1916
Costill DL. Fox EL. Eneraetics of marathon run nina. Medicine
and Science in Sports I: 81-86, 1969
Costill DL. Thomason H, Roberts E. Fractional utilization of the
aerobic capacity durina distance runnina. Medicine and Sci-
ence in Sports S: 248-2S2. 1913
Daniels JT. A physiolOBist's view of runnina economy. Medicine
and Science in Sports and Exercise 11: 332-338, 1985
Daniels JT, Yarborouah RA, Foster C. Chanaes in V02 mao and
runnina performance with trainina. European Journal of Ap-
plied PhysioiotY and Occupational Physiology 39: 249-2S4, 1978
Davies B. Daaet A, Jakeman p, Mulhall J. Maximum oxyaen
uptake utilizina different treadmill protocols. British Journal
of Sports Medicine 18: 14-19, 1984
Davies CTM. Thompson MW. Aerobic performance of female
marathon and male ultramarathon athletes. European Journal
of Applied PhysiolOlY 41: 233-24S. 1919
deVries HA. PhysiolOlY of exercise, 3rd ed., W.e. Brown, Iowa.
1980
Diaz FJ, H&pn RD, Wright JE, Horvath SM. Maximal and sub-
maximal exercise in different positions. Medicine and Science
in Sports 10: 214-217. 1918
Dimri GP, Malhotra MS, Sen Gupta J. Samprath-Kumar T, Arora
BS. Alterations in aerobic/anaerobic proportions of metabol-
ism durina work in the heat. European Journal of Applied
PhysiolOlY 4S: 43-SO, 1980
Fainhter RD. Wallen J, Salness K. Fox M, Minh V-D. Wilson
AF. A comparison of incremental exercise tests during cycle
and treadmill ergometry. Medicine and Science in Sports and
Exercise IS: S49-SS4, 1983
Falls HB. Humphrey LD. A comparison of methods for elicitina
maximum oxyaen uptake from college women during tread-
mill walkina. Medicine and Science in Sports S: 239-241, 1913
Faria IE, Drummond BJ. Circadian changes in resting hean rate
and body temperature, maximal oxygen consumption and per-
ceived exertion. ErJonomics 25: 381-386, 1982
Farrell PA. Wilmore J. Coyle E, Billina J, Costill DL Plasma
lactate accumulation and distance running performance. Med-
icine and Science in Sports II: 338-344, 1919
Faulkner JA. Roberts DE, Elk RL Cardiovascular responses to
subrnaximum and maximum effort cycling and runnina. Jour-
nal of Applied PhysiolOlY 30: 4S1-46I, 1911
Foster C, Anholm JD, Hellman CK. Carpenter J, Pollock ML. et
al. Left ventricular function during sudden strenuous exercise.
Circulation 63: S92-S96, 1981
Foster C, Dymond OS, Carpenter J, Schmidt P. Effect of warm-
VO" rna. Testing in Runners
up on left ventricular response to sudden strenuous exercise.
Journal of Applied Physiology 53: 380-383. 1982
Foster VL. Hume JE. Dickinson AL. Chatfield SJ. Byrnes WC
The reproducibility of VOl max. ventilation. and lactate thresh-
old in elderly women. Medicine and Science in Sports and
Exercise 18: 425-430. 1986
Froelicher Jr VF. Exercise testing and training: clinical applica-
tions. Journal of the American College of Cardiology I: 114-
125. 1983
Froelicher Jr VF. Brammel H. Davis G. Noguera I. Stewart A.
Lancaster MC A comparison of three maximal treadmill ex-
ercise protocols. Journal of Applied Physiology 36: 720-725.
1974
Gass GC'. Rogers S. Mitchell R. Blood lactate concentration fol-
lowing maximal exercise in trained subjects. British Journal of
Sports Medicine 15: 172-176. 1981
Gibson TM . Harrison MH. Wellicome RM. An evaluation of a
treadmill work test. British Journal of Sports Medicine 13: 6-
II. 1979
Glassford RG. Baycroft GHY. Sedgwick AW. Comparison of
maximal oxygen uptake values determined by predicted and
actual methods. Journal of Applied Physiology 20: 509-513.
1965
Hammond HK. Froelicher VF. Exercise testing for cardiorespir-
atory fitness. Sports Medicine I: 234-239. 1984
Harrison MH. Brown GA. Cochran LA. Maximal oxygen uptake:
its measurement. application. and limitations. Aviation and
Space Environmental Medicine 51 : 1123-1127. 1980
Hermansen L. Saltin B. Oxygen uptake during maximal treadmill
and bicycle exercise. Journal of Applied Physiology 26: 31-37.
1969
Hermansen L. Ekblom B. Saltin B. Cardiac output during sub-
maximal and maximal treadmill and bicycle exercise. Journal
of Applied Physiology 29: 82-86. 1970
Ishiko T . Merits of various standard test protocols: a comparison
between I.CP.F.R .. W.H.O .. I.B.P. and other groups. In Shep-
hard & Lavellee (Eds) Physical fitness assessment - principles.
practice and application. pp. 7-17. CC Thomas. lIIinois. 1978
Jacobs I, Sjodin B. Relationships of ergometric specific VOl max
and muscle enzymes to blood lactate during submaximal ex-
ercise. British Journal of Sports Medicine 19: 77-80, 1985
Jones NL. Haddon RWT. Effect of a meal on cardiopulmonary
and metabolic changes during exercise. Canadian Journal of
Physiology and Pharmacology 51: 445-450. 1973
Kamon E. Pandolf KB. Maximal aerobic power during laddermill
climbing. uphill running. and cycling. Journal of Applied Phys-
iology 32: 467-473. 1972
Katch VL. Sady S. Freedson P. Biological variability in maxi-
mum aerobic power. Medicine and Science in Sports and Ex-
ercise 14: 21-25 . 1982
Kenney WL. Hodgson JL. Variables predictive of performance
in elite middle-distance runners. British Journal of Sports
Medicine 19: 207-209. 1985
Kumagi S. Tanaba K. Matsuara y, Matsuzaka A. Hirakoba K.
Relationship of the anaerobic threshold with the 5km. 10km.
and 10 mile races. European Journal of Applied Physiology
49: 13-23. 1982
Leger L. Mercier D. Gross energy cost of horizontal treadmill and
track running. Sports Medicine 1:270-277. 1984
Londeree BR . The use oflaboratory test results with long distance
runners. Sports Medicine 3: 201-213. 1986
Magel JR. Foglia GF. McArdle WD. Gutin B. Pelchar GS. et al.
Specificity of swim training on maximum oxygen uptake. Jour-
nal of Applied Physiology 38: lSI-ISS. 1975
Maksud MG. Coutts KD. Comparison of a continuous and dis-
continuous graded treadmill test for maximal oxygen uptake .
Medicine and Science in Sports 3: 63-65. 1971
Martin 81. Robinson S. Wiegman DL. Aulick LH. Effect of warm-
67
up on metabolic responses to strenuous exercise. Medicine and
Science in Sports 7: 146-149. 1975
McArdle WD. Katch FI. Pechar GS. Comparison of continuous
and discontinuous treadmill and bicycle tests for max V0 2•
Medicine and Science in Sports 5: 156-160. 1973
McArdle WD. Magel JR. Physical work capacity and maximum
oxygen uptake in treadmill and bicycle exercise. Medicine and
Science in Sports 2: 118-123. 1970
McArdle WD. Magel JR. Delio DJ. Toner M, Chase JM. Speci-
ficity of run training on VO e max and heart rate changes during
running and swimming. Medicine and Science in Sports 10:
16-20. 1978
McConnell TR. Swett DD, Jeresaty RM. Missri JC AI-Hani AJ.
The hemodynamic and physiologic differences between exer-
cise modalities. Journal of Sports Medicine and Physical Fit-
ness 24: 238-245. 1984
McKay GA, Banister EW. A comparison of maximum oxygen
uptake determination by bicycle ergometry at various pedall-
ing frequencies and by treadmill running at various speeds.
European Journal of Applied Physiology 35: 191-200. 1976
McMiken DF. Daniels JT. Aerobic requirements and maximum
aerobic power in treadmill and track running. Medicine and
Science in Sports 8: 14-17. 1976
Miles DS. Critz JB, Knowlton RG. Cardiovascular, metabolic.
and ventilatory responses of women to equivalent cycle er-
gometer and treadmill exercise. Medicine and Science in Sports
and Exercise 12: 14-19. 1980
Mitchell JH . Sproule BJ , Chapman C8. The physiological mean-
ing of the maximal oxygen intake test. Journal of Clinical In-
vestigation 37: 538-547. 1958
Moreira-Da-Costa M, Russo AK. Picarro IC Silva AC Leite-De-
Barros-Neto JK. et al. Maximal oxygen uptake during exercise
using trained or untrained muscles. Brazilian Journal of Med-
icine Biological Research 17: 197-202. 1984
Morgan WP. Psychogenic factors and exercise metabolism. Med-
icine and Science in Sports and Exercise 17: 309-316. 1985
Neiderberger M, Bruce RA. Kusumi F. Disparities in ventilatory
and circulatory responses to bicycle and treadmill exercise.
British Heart Journal 36: 377-382. 1974
Pelchar GS. McArdle WD. Katch Fl. Magel JR. DeLuca J. Spec-
ificity of cardiorespiratory adaptation to bicycle and treadmill
training. Journal of Applied Physiology 36: 753-756. 1974
Pollock ML. Submaximal and maximal working capacity of elite
distance runners. Part I: Cardiorespiratory aspects. Annals of
the New York Academy of Sciences 301 : 310-321. 1977
Reilly T. Brooks GA. Circadian constancy in work stress indices
and physiologic capacity. Ergonomics 25: 329-330. 1982a
Reilly T. Brooks GA. Investigation of circadian rhythms in meta-
bolic responses to exercise. Ergonomics 25: 1093-1107. 1982b
Rowell LB. Taylor HL. Wang Y. Limitations to prediction of
maximal oxygen intake. Journal of Applied Physiology 19: 919-
927. 1964
Saltin B. Astrand P-O. Maximum oxygen uptake in athletes. Jour-
nal of Applied Physiology 23: 353-358. 1967
Shephard RJ. The prediction of athletic performance by labora-
tory and field tests - an overview. In Shephard & Lavellee
(Eds) Physical fitness assessment - principles. practice, and ap-
plication. pp. 113-141. CC'. Thomas. lIIinois. 1978
Shephard RJ. Tests of maximum oxygen intake - a critical re-
view. Sports Medicine I: 99-124, 1984
Shephard RJ. Allen C Benade AJS. Davies C'TM. de Prampero
PE. et al. The maximal oxygen intake - an international ref-
erence standard of cardiorespiratory fitness. Bulletin of the
World Health Organization 38: 757-764. 1968
Sjodin B. Svedenhag J. Applied physiology of marathon running.
Sports Medicine 2: 83-99. 1985
Smodlaka VN . Treadmills \'S bicycle ergometers. Physician and
Sportsmedicine \0: 75-79. 1982
V02 mu Testing in Runners
Snell PG. Mitchell JH. The role of maximal oxyaen uptake in
exercise peri"onnanc:e. Clinics in Sports Medicine S: SI~2. 1984
Stamford BA. Maximal oxyaen uptake duri", tradmill walki",
and runnina at various speeds. Journal of Applied PhysiolOlY
39: 386-389. 1975
Stamford BA. Step increment versus constant load tests for de-
termination of maximal oxyaen consumption. European Jour-
nal of Applied PhysiolOlY and Occupational PhysiolOl)' 3S: 89-
93. 1976
Stamford BAI Rowland R. Moffatt RJ. Effect of severe prior ex-
ercise on assessment of maximal oxyaen uptake. Journal of
Applied PhysioJo&y 44: SS9-S63. 1978
Stremme SB. Inlier F. Meen HD. Assessment of maximal aerobic
power in specifically trained athletes. Journal of Applied Phys-
iolOlY 42: 833-837. 1977
Strydom NB. Environmental variables affectina fitness testina. In
Shephard" Lavellee (Eds) Physical fitness assessment - prin-
ciples. practice and application. pp. 94-101. C.c. Thomas.
Illinois. 1978
SvedenhqJ. Sjodin B. Maximal and submaximal oxyaen uptakes
and blood lactate levels in elite male middle and lona distance
runners. International Journal of Sports Medicine S: 2SS-261.
1984
Taylor CI. Some properties of maximal and submaximal exercise
with reference to physiolOlY variation and the measurement
of exercise tolerance. American Journal of PhysiolOlY 142: 200-
212. 1944
Taylor HL. Buskirk E, Henschel A. Maximal oxyaen intake as an
68
objective measure of cardiorespiratory performance. Journal
of Applied PhysiolOlY 8: 73-80. 19S5
Wasserman K. Hansen JE, Sut DY, Whipp BJ. Principles of ex-
ercise testina and interpretation, Lea " Febiaer, Philadelphia,
1987
Weber KT. Janicki JS. Cardiopulmonary exercise testing: physi-
oloaic principles and clinical applications, W.B. Saunders
Company, Philadelphia. 1986
Whipp BJ. Davis J. Torres F. Wasserman K. A test to determine
parameters of aerobic function durina ellercise. Journal of Ap-
plied PhysioJo&y SO: 217-221. 1981
Williams 00. Wyndam CH. Kok R. Rehden MJE. Effect oftrain-
ina on maximal oxyaen intake and on anaerobic metabolism
in man. Internationale Zeitschrift fUr Anaewandte Physiologie
einschlieshich Arbeitsphysiologie 24: 18-23. 1967
Wojtczak-Jaroszowa J. Banaskiewicz Z. Physical work capacity
durina the day and at niabt. Eraonomics 17: 193-198. 1974
Wriaht GR. Sidney KH. Shephard RJ. Variance of direct and
indirect measurements of aerobic power. Journal of Sports
Medicine and Physical Fitness 18: 33-42. 1978
Yamaji K. Sakamoto H. Nakaauchi M. Kitamura K. Shephard
RJ. BiolOlical rhythms of PWCI70 and maximal oxyaen in-
take. Journal of Human ErgolOlY 10: 213-219. 1981
Author's address: Dr Timothy R. McConn~II, Director. Cardiac
Rehabilitation. Department of CardiolOlY, Geisinaer Medical
Center, Danville, PA 17822 (USA).