Optimizing the exercise protocol for cardiopulmonary ...

Optimizing the exercise protocol for cardiopulmonary assessment MARK J. BUCHFUHRER, JAMES E. HANSEN, TERRY E. ROBINSON, DARRYL Y. SUE, KARLMAN WASSERMAN, AND BRIAN ,J. WHIPP Division of Respiratory Physiology and Medicine, Department of Medicine, Harbor-UCLA School of Medicine, Torrance, California 90509 BUCHFUHRER, MARK J., JAMES E. HANSEN, TERRY E. ROBINSON, DARRYL Y. SUE, KARLMAN WASSERMAN, AND BRIAN J. WHIPP. Optimizing the exercise protocol for cardio­

pulmonary assessment. J. Appl. Physio!.: Respirat. Environ. Exercise Physio!. 55(5): 1558-1564, 1983.-Twelve normal men performed I-min incremental exercise tests to exhaustion in approximately 10 min on both treadmill and cycle ergometer. The maximal O2 uptake (V0 2 Illa.) and anaerobic threshold (AT) were higher (6 and 13%, respectively) on the treadmill than the cycle; the AT was reached at about 50% of V0 2 max on both ergometers. Maximal CO 2 output, heart rate, and O2 pulse were also slightly, but significantly higher on the treadmilL Maximal ventilation, gas exchange ratio, and ventilatory equivalents for O2 and CO 2 for both forms of exercise were not significantly different. To determine the optimum exercise test for both treadmill and cycle, we exercised five of the subjects at various work rate increments on both ergometers in a randomized design. The treadmill increments were 0.8, 1.7, 2.5, and 4.2%1 min at a constant speed of 3,4 mph, and 1.7 and 4.2%/min at 4.5 mph. Cycle increments were 15, 30, and 60 WImin. The V0 2 max was significantly higher on tests where the increment magnitude was large enough to induce test durations of 8-17 " min, but the AT was independent of test duration. Thus, for evaluating cardiopulmonary function with incremental exercise testing by either treadmill or cycle, we suggest selecting a work rate increment to bring the subject to the limit of his tolerance in about 10 min. maximal oxygen uptake; anaerobic threshold; cycle ergometer; treadmill; exercise test duration

EXERCISE TESTS are commonly used to evaluate cardi­ 0respiratory function, the maximal O2 uptake (Y02 max) having become a widely utilized means of assessing its upper limit. The test, first standardized by Taylor et a1. in 1955 (22), utilized a high initial work rate performed for 3 min on a treadmill at a speed of 7 mph. On a series of subsequent consecutive days the treadmill grade was increased by 2.5% until the increment in O2 uptake (Y0 2 ) was less than 150 ml/min (Le., the "plateauing" reflecting Y0 2 max). A continuous test with 2-min increments of work rate, beginning at 90-100% of predicted Y0 2 max, was recommended in 1968 by Shepard et a1. (20), and was shown to have the same Y0 2 max as the discontinuous test; the criterion for Y0 2 max was an increase in Y0 2 of less than 2 mI· kg-I. min- I with a further increase in the work rate. The continuous and discontinuous tests described 1558

above are not appropriate for patients with heart or lung disease as they typically cannot sustain the high work rates required by those protocols for 2-3 min. Further­ more, other parameters such as the anaerobic threshold (AT) cannot be determined by these protocols. Conse­ quently, several different continuous incremental tests have been employed (1, 2, 9, 18, 26). In these tests, the plateau criterion for V0 2 max is infrequently achieved (9, 10,19). Furthermore, Y0 2 max with the "standard" discon­ tinuous test of Taylor et al. (22) is the same as with incremental running tests (9, 10, 14, 16, 17,19). Walking tests, which are the most useful clinically, typically uti· lize small increment~ of work rate (1, 2). However, they may result in lower V0 2 max than either the Taylor or the continuous incremental running protocols (9, 10, 14, 16, 17, 19). Tests using a cycle ergometer have been shown to result in lower VO z max than for treadmill tests (8, 9, 11, 13, 28); thus the manner of testing is also important. To determine the efficacy of treadmill and cycle for cardiorespiratory assessment during exercise and the effect of varying incremental work rate patterns on the estimate of both the AT and Y0 2 max, we studied normal subjects on both ergometers with work rate increments that were constant for each particular test but different among tests. The objective was to determine any system­ atic differences in pattern of cardiorespiratory response, using either the cycle or the treadmill, to determine the optimum work rate protocol for standardization of incre­ mental exercise testing. METHODS

Twelve male volunteers (Table 1), determined to be normal by history, physical exam, electrocardiogram, and pulmonary function testing, gave their informed consent for this study. The subjects did not vary the level of their exercise activity during the testing period. All studies were performed in a comfortable laboratory environment with a mean temperature of 21'C and a range of 18­ 23"C. Each subject performed a I-min incremental test on both a calibrated, electronically braked cycle (Lanooy, Godart, Sweden) and a treadmill (Warren E. Collins, Braintree, MA). The increment was chosen in each case to yield an exercise duration of about 10 min. During the tests, the subjects breathed room air through a low-resistance breathing valve (Koegel, San Antonio, TX) with a dead space of 50 ml. The valve

0161-7567/83 $1.50 Copyright © 1983 the American Physiological Society

OPTIMIZING THE EXERCISE PROTOCOL

1559

TABLE 1. Physical characteristics and maximal treadmill aerobic power of study group Subset of 5 Men

12 Men Means ±SD

Range

Means ±SD

Range

A

-­ -- .="

­

B

+ N

0

U

Age, yr 39 ± 8.2 28-54 36 ± 9.7 28-54 Wt, kg 72 ± 11.7 44-87 75 ± 4.7 68-81 Ht,cm 173 ± 7.4 160-183 179 ± 3.4 175-183 MVV,I/min 164 ± 27.3 128-204 176 ± 19.1 152-197 1 V02 m... ml·min- • 43 ± 10.2 29-59 52 8.4 39-59 kg- 1 MVV, maximal voluntary ventilation determined from 15-s proce­ dure; V0 2 rna.. maximal O2 uptake.

.>..

directed the cooled expirate through a room temperature pneumotachograph (Fleisch no. 3) that was connected to a transducer (Hewlett-Packard 47304A) for the deter­ mination of expiratory flow. Expired air was sampled from the mouthpiece at a rate of 1 ml/s for determination of partial pressures of O2 and CO 2 (P0 2 and PC02) by mass spectrometry (Perkin-Elmer MGA-llOO). The heart rate (HR) was determined by a cardiotachometer from the electrocardiograph (Hewlett-Packard 78332A). These signals underwent analog-to-digital conversion and were sent on line to a calculator (Hewlett-Packard 9825A). Determinations of VE, V0 2 , VC02, gas exchange ratio (R), HR, end-tidal P0 2 (PETo), end-tidal PC02 (PETeo.), and ventilatory equivalents (calculated with the dead space of the valve excluded) for O 2 and CO 2 CVE(V02 and VE/VC02) were made as previously de­ scribed by Sue et al. (21) on a breath-by-breath basis. Calculations were made from data collected over 20-s periods twice each minute and displayed graphically as shown in Fig. 1. The AT was discerned from these graphs by four of the authors independently, using the criteria described by Wasserman, Whipp, Davis, and colleagues (4, 6, 7, 23, 24, 25): a systematic increase in VE/V0 2 without an increase in VE/VC02, and a systematic in­ crease in PETo2 without a decrease in PETeo2. A subset of five subjects was used to test the effect of different work rate increments (thus altering test dura­ tion) on the reproducibility of important indices of car­ diorespiratory and metabolic function, such as V0 2 rna., AT, maximal ventilation (VEmaJ, maximal heart rate (HRmax ), maximal CO 2 production (VC02 max), maximal gas exchange ratio (Rmax ), and maximal O 2 pulse (0 2 pulse max = V0 2 max/HRmax) on the cycle and treadmill. Each of these five subjects, who were generally more fit than the other subjects (Table 1), performed the tread­ mill and cycle tests in a randomized order with at least 24 h separating the tests. The treadmill test consisted of walking for 2 min at a constant speed on the level; the work rate was then incremented equally each minute at the same constant speed by increasing the grade to the limit of tolerance. The grade increases on the different tests were 0.8, 1.7, 2.5, and 4.2%/min at 3.4 mph, and 1.7 and 4.2%/min at 4.5 mph. (Two of the least fit subjects were unable to perform the 4.2%/min at 4.5 mph test for sufficient time). Each subject also performed 5-min steady-state tests on the treadmill at work rates corresponding to

-+

60

U

50

-> .......

w .>.

--

.40

0

30

0

N

E

.....

0

.>

N

0

0 0 0

~

N

0

.>

20

"

w

(>

10

-

+ -

1.40

U tw

100 'b:>cnR>

N

0

..

-~

0

N

0

120

1

6b~

~~oo

()

0.0

o

I

0

:t: 80

E , E 60'

t-

1

.40 +

w

~

20'

++++1""+/ ""++1++++-1+++ ,.. ++ +-I++tI:t+t­

... 0

5

10

15

0

5

INCREMENT Al TIME (min) Determination of anaerobic threshold (AT) and maximal O2 uptake (V0 21M• represented by MAX) in response to treadmill tests utilizing different increments of power in 1 subject. Grade was incre­ mented by 1.7%/min (A) and 4.2%/min (B) at a constant treadmill speed of 3.4 mph. Note that AT occurs at same V0 2 (1.95 l/min) in both tests, whereas V0 2 max (MAX) is clearly greater in A (4.79 l/min). See text for details in determining AT from ventilatory equivalents for O2 and CO2 (VE!VOz and VE!VC02 and end"tidal P0 2 and Pcoz (PETo2 and PETeo,). f'IG. 1.

approximately 40, 70, and 90% of their treadmill AT. On the cycle, each of the five subjects exercised for 2 min at 0 W (unloaded pedaling); then the work rate was increased in equal increments of 15, 30, or 60 W /min to the limit of tolerance. Five-minute steady-state tests on

BUCHFUHRER ET AL.

1560 the cycle were also performed at 0 Wand at approxi­ mately 40, 70, and 90% of their cycle AT. Where multiple comparisons were made, the repeated

CYCLE

TREADMILL 4.0­

RESULTS

*

c:

+

3.0­

E

'"--.J

N

o

measure analysis of variance test was applied with sub­ sequent post hoc Scheffe test (3). Where two, and only two, variables were compared, t tests were employed (3). Where multiple comparisons of unequal groups were made, analysis of variance was employed (3). All values are reported significantly different at the P < 0.05 level.

2.0­

:>

15

30

45

60

watts/min

% grade /min at 3.4 mph

2. Effects of different increments of power on anaerobic thresh­ old (hatched) and maximal O 2 uptake (V02) (open bars) for incremental treadmill at 3A mph (left) and cycle ergometry (right) on 5 subjects. Large SE reflect wide range of fitness levels of subjects. *t:j:Pairs of tests significantly different at P < 0.05. FIG.

I

SHORT TREADMill

91.0:± 2.2

CYCLE

89.0:± 3.4

INTERMEDIATE

On both the cycle ergometer and as shown in Fig. 1, on the treadmill, V0 2 increased as a linear function of time (i.e., work rate) with all test protocols, except for the short lag phase, which occurs when work rate is rapidly increased and the plateauing as V0 2 max is ap­ proached. The V0 2 at the AT was found to be indepen­ dent of the work rate increment on a given ergometer (Figs. 1 and 2). In contrast, the highest V0 2 which could be obtained on a given test was influenced by the test protocol (Figs. 1 and 2). As shown in Figs. 2 and 3, the V0 2 max did not increase monotonically with the increase in either work rate in­ crement or test duration; rather it was highest for test durations between 8-17 min on both ergometers (Fig. 3). Tests lasting less than 8 min resulted in a mean reduction in V0 2 max of about lO% (P < 0.05). Tests lasting longer than 17 min also showed a decrease in V0 2 max with the

LONG

97.3:± 2.4

94.7± 3.3

± 1.3

9S.3± 1.6

99.2

~P, ........

100

I

I I

I I

/" ,,

, V

!

I I

I

~

,

88 I

I

!

0

I I

I

/:" 84

o

5

10

15

20

25

30

35

TEST DURATION (min) 3. Effect of test duration on 5 subject's maximal O2 uptake (V0 2 m . . ), expressed as percent of each individual's highest V0 2IDa, on each ergometer for incremental treadmill (closed symbols) and cycle ergometry (open symbols). Note V0 2 max is typically highest for test durations of 8-17 min (intermediate duration). Top of figure gives FIG.

percent means ± SD of highest V0 2 rna. for all tests in each test duration group. Short duration group (less than 8 min) was significantly lower than either intermediate or long-duration group (greater than 17 min). *One anomolous datum that reached highest V0 2 rna, at 22 min.

OPTIMIZING THE EXERCISE PROTOCOL

1561

2. Effect of speed and grade variations on cardiopulmonary and metabolic responses to incremental treadmill testing in five subjects

TABLE

Test No. 2

Time of incremental exercise, min HR.n"", beats/min VErn... l/min VEm../MVV, % Ve02 rnax.l/min

3

4

5

Signif

0.8%, 3.4 mph

1.7%, 3.4 mph

2.5%, 3.4 mph

1. 7%, 4.5 mph

4.2%, 3.4 mph

26.4 190.6 143.:3 82.1 4.47

15.1 186.2 144.3 82.5 4.8:3

11.4 ± 188.0 ± 152.7 ± 87.5 ± 4.98 ±

10.6 187.8 144.7 82.6 4.82

7.0 ± 185.0 ± 152.0 ± 87.2 ± 5.16 ±

± ± ± ± ±

1.6 3.9 17.8 9.8 0,44

± ± ± ± ±

0.9 3.8 17.1 9.3 0.45

1.0 2.4 15.0 8.5 0.41

± ± ± ± ±

1.0 3.5 15.3 7.8 0,48

0.5 2.0 15.8 9.1 0.50

R.n..

1.19 ± 0.02

1.26 ± 0.04

1.29 ± 0.05

1.25 ± 0.02

1.41 ± 0.05

O2 pulse mru" mlfbeat V02 rna., l/min AT,l/min

19.8 ± 2.0 3.77 ± 0.39 2.07 ± 0.29

20.7 ± 2.0 3.88 ± 0,44 2.02 ± 0.28

20.9 ± 2.1 3.94 ± 0,43 2.05 ± 0.27

20.6 ± 2.1 3.89 ± 0.44 2.16 ± 0.27

19.8 ± 2.0 3.68 ± 0,41 2.02 ± 0.27

NS* NS* NS* 1 vs. 2, 1 vs. 1 vs. 4, 1 vs. 2 VS. 5, 4 VS. 1 VS. 5, 2 vs. 3 VB. 5. 4 VS. 1 VS. 3 3 VS. 5 NS*

Values are means ± SE. Test format: grade increment/min utilized at constant treadmill speed. HRrnax, maximal heart rate; ventilation; MVV, maximal voluntary ventilation; VC02 rna" maximal CO2 production; R.n." maximal gas exchange ratio; O2 pulse O2 pulse; V0 2 ma.. maximal O2 uptake; AT, anaerobic threshold. * Not significant at P < 0.05.

longest of these tests having the greatest reduction. The mean reduction in V0 2 max of tests lasting longer than 17 min was about 5%. This was not statistically different from the V0 2 max in "optimum" duration tests (8-17 min) owing to the behavior of a single subject who reached his highest V0 2mall at 22 min (the * in Fig. 3). This was the only one of 10 occasions when the longest duration test on each ergometer had a higher V02 max than the test of next shorter duration. The HRmax , VEmax , and O2 pulse max, however, did not show this systematic variation with test duration on either the treadmill (Table 2) or the cycle (Table 3). The VEmax approximated 85% of the 15-s maximal voluntary ventilation (MVV) on all tests. There was a tendency for VC02 max and Rmax to increase systematically with tests of shorter duration. The relationship of VE to VC02 below the AT was linear with a positive VE intercept. The results were strikingly similar whether on the cycle or treadmill dur­ ing incremental and steady-state exercise. All five subjects complained of increasing low back pain with the longer treadmill studies, whereas seat discomfort commonly occurred with longer cycle exer­ cise. Consequently, the subjects described the longer tests as more uncomfortable and thus more "stressful." There­ fore the work rate increment was selected for each of the 12 subjects in the treadmill-cycle comparison studies so that the test duration would be the same on the cycle and the treadmill (Table 5). Subjectively, tests of similar duration were similarly stressful on both ergometers. Metabolic measurements on the treadmill were no more difficult than on the cycle and allowed clear discernment ofthe AT and V0 2 max (Fig. 1). Although no significant differences between treadmill and cycle were found for VEmax , Rma.. VEmax /V0 2 max, and VEmax/VC02 max, the V0 2 max, VC02 max, HRmax , O2 pulse max, and AT were all significantly higher on the treadmill. The AT, when expressed as a percent ofV0 2 max, was not significantly different between the treadmill and the cycle and approximated 50%. Similarly, the LlV0 2 (V 02 max AT) was not significantly different between

In. . .

max.

3 5, 5 5 5

maximal maximal

TABLE 3. Effect of different increments of work rate on cardiopulmonary and metabolic responses to incremental cycle ergometry in five subjects Test No.

15W

2

3

30W

60W

.~-.--"--.--.--"--.--.--

Time, min HRma.. beat/min VEmax.l/min VEmax/MVV, % Ve02m.» l/min Rtnax

O2 pulse max, mlfbeat V0 2 max, l/min AT, l/min

18 ± 1.6 183.6 ±:3.1 147.6 ±17.4 84.9 ±10.5 4.37 ±0.42 1.21 ±0.04 19.6 ±1.9 3.62 ±0,40 1.86 ±0.21

10.6 ±1.0 18:3.4 ±3.7 155.5 ±17.2 89.1 ±9.7 4.87 ±0,47 1.:31 ±0.04 20.5 ±2.1 3.77 ±0.43 1.80 ±0.24

5.8 ±0.5 177.8 ±:3.9 150.7 ±14.6 86.:3 ±8.1 4.71 ±0.43 1.43 ±0.04 18.8 ±1.9 3.35 ±0.38

Signif .. .. ~

NS* NS* NS* NS* 1 VS. 2, 1 vs. 3 2 vs. 3 2 VS. 3 1 VS. 3, 2 VS. 3 NS*

Values are means ± SE. Test format: work rate increments/min in Watts. See Table 1 for definitions. * Not significant at P < 0.05 level.

the treadmill and the cycle. It was not possible to deter­ mine the AT rigorously from the 60-W incremental work tests owing to. the curvilinearity of the pattern of re­ sponse of VE/V0 2 as shown in Fig. 4. DISCUSSION

We have shown that the V0 2 max with a I-min incre­ mental test varies with increment size (Fig. 2, Tables 2 and 3). The highest V0 2 max is generally achieved with tests of intermediate increments (i.e., 2.5%/min at 3.4 mph on treadmill and 30 W/min on cycle) for subjects who are reasonably fit. In subjects who are less fit, the intermediate increments mentioned above may be rela­ tively large and thus yield a lower V0 2 max' Subjects of

BUCHFUHRER ET AL.

1562

+

B

A 60

'-'

N

0

50

U

.> .......

.>,

.40

o

loU

-

00

30

o + ':i' + + ++

0

'-'

N

o

20

0

.> .......

10

loU

.>

0 0

2

3

5

.4

o

V02 (Llmin) FIG. 4. Example of response of ventilatory equivalents for O2 and CO 2 (VE/VO z and VE!VC02) to 60 W/min (A) and 30 W fmin (B) incremental cycle ergometer test. Note that continuously curvilinear

response pattern of VE/V0 2 (A) makes clear recognition of anaerobic threshold (AT) impossible. unlike the response to smaller power incre­ ments (E) where the AT is clearly discernible at vertical arrow.

different fitness seem to have similar "stress" from tests of similar duration rather than of similar incremental size. This concept is best exemplified by Table 2, which shows no significant difference in any cardiorespiratory index on the three treadmill tests in the intermediate test duration range (i.e., "optimum" range of 8-17 min) that utilize different grade increments and different treadmill speeds. The V0 2 max was highest on tests of intermediate incremental size (Fig. 2) and consequently intermediate duration (8-17 min on Fig. 3). Short tests (less than 8 min) with large increments had significantly lower '\;02 max probably because of problems with mus­ cular force limitation. Long tests (greater than 17 min) with small increments had lower '\;02 max in all but one case. Possible factors for this reduction in '\;0 2 max on these prolonged tests include: higher body temperature, greater dehydration, different substrate utilization, temperature, subject discomfort, or ventilatory muscle fatigue. Bore­ dom or lack of sustained committment were, we believe, unlikely causes in these five well-motivated subjects. We did not measure rectal temperature, but skin and respi­ ratory water loss and diversion of blood flow to the skin could result in reduced muscle blood flow in the longer tests. The lower R at maximum exercise in longer tests is explainable by the lower rate of induction of metabolic acidosis in such tests; however, a greater proportional contribution of fatty acids cannot be excluded. Ventila­ tory muscle fatigue is a plausible factor. In our experi­ ence, nonathletic subjects who exercised to their symp­ tom-limited maximum in 6-12 min uncommonly have a '\;Emax/MVV exceeding 80%. In well-trained athletes, Grimby and colleagues (12) found that flow rates im­ pacted on the outer envelope of the flow-volume curve,

while the '\;Emax averaged 83% of the MVV. On average, our 5 subjects stopped at similar levels of ventilation for all tests; in longer tests they necessarily sustained high levels of ventilation for longer times. Consequently, ven­ tilatory muscle fatigue may have occurred. Since the longer tests waste time, supply no additional information (26), and may not yield maximal values even in well­ motivated individuals, we recommend a test duration of 10 ± 2 min. The Balke protocol, using a 1% increment in grade each minute, was shown by Froelicher et al. (10) to have a lower '\;02 max than discontinuous tests or the Bruce test. Pollack et aI. (19) found the '\;02 max to be similar on Balke and Bruce protocols, probably because the test duration for both test formats were similar and of "op­ timum" duration. Some of the test protocols used by cardiologists (2, 13, 17, 19) are less suitable for clinical exercise testing as they have unnecessarily long and unequal work increments (changing both speed and grade), which may make determinations such as the AT less reliable. Whipp et al. (26,27) found '\;02 max to be the same on incremental tests of different durations. In these tests they varied the duration of the work increment rather than increment size which was kept constant at 15 W. In "ramp" exercise tests with different work rate slopes, Davis et al. (7) found no significant differences in '\;02 max with slopes of 20, 30, 50, and 100 W Imin. Together, all the above studies suggest that increments that are too large or too prolonged will result in a lower '\;02 max' AT has been found to be a reproducible parameter on "ramp" tests with different slopes (and thus different durations) (7, 26) and on incremental tests of different duration where the increment duration was varied (27).

OPTIMIZING THE EXERCISE PROTOCOL

1563

Relationship between VE and VC02 during incremental exercise below anaerobic threshold

TABLE 4.

Cycle

Treadmill Test

SS 1.7% 2.5% 4.2% Means ±SD*

Slope

21.9 21.7 20.9 20.7

± ± ± ±

0.8 1.2 0.9 0.9

21.8 ± 2.2

TABLE 5. Comparison of indices of cardiopulmonary and metabolic function during incremental and treadmill ergometry of equal duration to exhaustion

intercept

5.3 4.8 6.2 6.2

± ± ± ±

1.2 0.7 1.2 1.0

4.5 ± 2.4

Test

SS 15W 80W 60W

Slope

22.3 22.0 22.2 22.8

± ± ± ±

0.8 0.7 0.7 0.6

22.2 ± 1

intercept

8A ± 0.8

4.6 ± 0.8 4A ± 0.9 4A ± 0.6

3.7 ± 1.7

Values are means ± SE for 5 subjects except where noted. Test format: grade increment/min on treadmill at constant speed of 8A mph and W/min on cycle. VE, minute ventilation (BTPS); VC02, CO 2 output (STPD); SS, steady state. * Values are for 12 subjects.

We have found that the AT is a reproducible parameter on I-min incremental tests of different test durations or different incremental size (Figs. 1 and 2). We showed that each subject had a constant linear relationship of VE to Vcoz below the AT both in steady­ state tests and in tests of different increments on the cycle and the treadmill (Table 4). This reflects the link­ ing of VE to Vcoz; hence reliable and reproducible char­ acterization of ventilatory response to exercise may be appropriately determined regardless of the work-rate format as long as VC02 is used as the frame of reference. We found the V0 2 max to be higher (6%) on treadmill than cycle as previously described (5, 8, 11, 13, 28). The AT was higher (13.5%) on treadmill than on cycle. This is in agreement with Koyal et al. (15) and Withers et al. (28). Davis et aL (6) found no difference in the AT between the two erogmeters, but his subjects were stu­ dents who were cycling daily, whereas only one of our subjects cycled regularly. We found that the AT was not significantly different between the cycle (47%) and tread­ mill (50%) when expressed as a percent of V0 2 max (Table 5). Our values are similar to those of sedentary middle­ aged men before endurance training (4) but lower than after training (4) or those studies with more fit subjects (6,28). We also found that the additional aerobic power

Test duration, min VEmax. l/min VCOzmax, l/min HRmax. beats/min O2 pulse max> ml/beat Rmax VEmax!VOzmax VE""",/VCOzruax VO zmrun l/min AT, l/min AT!V02nw,,% i1V0 2 (V0 2max AT) Values are means ± tions. * Significance

Treadmill

Cycle

9.3± 2A 125± 38.0 4.06± 1.11 185± 6.0 16.6± 5.0 1.34± 0.10 40.8± 6.6 30.7± 3.6 8.08± 1.00 1.55± 0.58 50.3± 9.4 1.54± 0.57

9A± 1.9 128± 43.0 3.87± 1.18 179± 9.0 16.0± 5.1 1.36± 0.09 44.4± 7.1 32.8± 4.7 2.90± 1.02 1.34± 0.53 47.2± 11.0 1.58± 0.72

-0.183 -0.681 2.628* 3.588' 3A33* -1.068 -2.120 -1.818 7A90* 2.824* 0.988 -OA07

SD for 12 subjects. See Table 2 for defini­ at P < 0.05 level.

that could be achieved after reaching the AT (Ll V0 2 = V0 2 max - AT) was the same on treadmill and cycle (Table 5). This suggests that the aerobic increment of nonathletes after reaching the AT may be independent of ergometer. In summary, we have determined that the AT is a reproducible parameter for different work increments in non highly trained adults and is reached at about 50% of V0 2 max on both cycle and treadmill in this group. Both V0 2 max and AT are greater on the treadmill, and thus, the type of ergometer should be taken into consideration when interpreting exercise tests. We further demon­ strated that the V0 2 max may be affected by incremental test duration. Therefore, to obtain the highest V0 2 max during incremental exercise on a given ergometer, we suggest selecting a work rate increment to bring the subject to his limit of tolerance in 10 ± 2 min. The authors thank Norma Harris and Barbara Young for secretarial assistance. Address for reprint requests: ,J. E. Hansen, Harbor-UCLA Medical Center, Div. of Respiratory Physiology and Medicine, Box 24, Torrance, CA 90509. Received 27 December 1982; accepted in final form 30 June 1983.

REFERENCES 1. BALKE, B., AND R. WARE. An experimental study of "physical fitness" of Air Force personnel. US Armed Forces Med. J. 10: 6: 675-688, 1959. 2. BRUCE, R. A. Exercise testing of patients with coronary artery disease. Ann. Clin. Res. 3: 328-332, 1971. 3. BRUNIG, J. L., AND B. L. KINTZ. Computational Handbook of Statistics (2nd ed). Glenview, IL: Scott, Foresman, 1977. 4. DAVIS, J. A., M. H. FRANK, B. J. WHIPP, AND K. WASSERMAN. Anaerobic threshold alterations caused by endurance training in middle-aged men. J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 46: 1039-1046, 1979. 5. DAVIS, J. A., AND F. W. KASCH. Aerobic and anaerobic differences between maximal running and cycling in middle aged males. Aust. J. Sports Med. 7: 81-84, 1975. 6. DAVIS, J. A., P. VODAK, .T. H. WILMORE, J. VODAK, AND P. KURTZ. Anaerobic threshold and maximal aerobic power for three modes of exercise. J. Appl. Physiol. 41: 544-550, 1975. 7. DAVIS, J. A., B. J. WHIPP, N. LAMARRA, D. J. HUNTSMAN, M. H. FRANK, AND K. WASSERMAN. Effect of ramp slope on measurement of aerobic parameters from the ramp exercise test. Med. Sci. Sports Exercise 14: 339--843, 1982.

8. FAULKNER, J. A., D. E. ROBERTS, R. L. ELK, AND J. CONWAY. Cardiovascular responses to submaximal and maximal effort cy­ cling and running. J. Appl. Physiol. 30: 457-461, 1971. 9. FROELICHER, V. F., H. BRAMMEL, G. D. DAVIS, I. NOGUERA, A. STEWART, AND M. C. LANCASTER. A comparison of three maximal treadmill exercise protocols. J. Appl. Physiol. 36: 720-725, 1974. 10. FROELICHER, V. F., H. BRAMMEL, G. D. DAVIS, I. NOGUERA, A. STEWART, AND M. C. LANCASTER. A comparison of the reprodu­ cibility and physiologic response to three maximal treadmill exer­ cise protocols. Chest 65: 512-517, 1974. 11. GLASSFORTH, R. G., G. H. Y. BAYCROFT, A. W. SEDGEWICK, AND R. B. J. MACNAB. Comaprison of maximal oxygen uptake values det.ermined by predicted and actual methods. J. Appl. Physiol. 20: 509-518, 1965. 12. GRIMBY, G., B. SALTIN, AND L. WILHELMSEN. Pulmonary flow­ volume and pressure-volume relationship during submaximal and maximal exercise in young well-trained men. Bull. Eur. Physio.­ Pathol. Respir. 7: 157-167,1971. 13. HERMANSEN, L., AND B. SAL'rIN. Oxygen uptake during maximal treadmill and bicycle exercise. J. Appl. Physiol. 26: 31-37, 1969. 14. KNOWLTON, R. G_, N. M. SAWKA, AND D. T. DEUTSCH. Varied

1564 intensity treadmill protocols for the measurement of maximal oxygen uptake. J. Sports Med. 17: 263-268, 1977. 15. KOYAL, S. N., B. J. WHIPP, D. HUNTSMAN, G. A. BRAY, AND K WASSERMAN. Ventilatory response to the metabolic acidosis of treadmill and cycle ergometry. J. Appl. Physiol. 40: 864-868, 1976. 16. MASKUD, M. G., AND K D. COUTTS. Comparison of a continuous and discontinuous graded treadmill test for maximal oxygen up­ take. Med. Sci. Sports 3: 63-65, 1971. 17. McARDLE, W. D., F. J. KATCH, AND G. S. PECHAR. Comparison of continuous and discontinuous treadmill and bicycle tests for

V0 2 max' Med. Sci. Sports 5: 156--160, 1972. 18. NAGLE, I., B. BALKE, G. BAPTISTA, J. ALLEYIA, AND E. HOWELL. Compatibility of progressive treadmill, bicycle, and step test based on oxygen uptake responses. Med. Sci. Sports 3: 149-154, 1971­ 19. POLLOCK, M. L., R. L. BOHANNON, K H. COOPER, J. J. AYRES, A. WARD, S. R. WHlTE, AND N. C. LINNERUD. A comparative analysis of four protocols for maximal treadmill stress testing. Am. Heart J. 92: 39-46, 1976. 20. SHEPHARD, R. J., C. ALLEN, A. J. S. BENADE, C. T. M. DAVIES, P. E. DIPRIMPERO, R. HEDMAN, J. E. MERRIMAN, K MYHRE, AND R. SIMMONS. The maximum oxygen intake. W.H.O. BulL 38: 757-764,1968. 21. SUE, D. Y., ,J. E. HANSEN, M. BLAIS, AND K. WASSERMAN. Measurement and analysis of gas exchange during exercise using a programmable calculator. J. Appl. Physiol.: Respirat. Environ. Ex-

BUCHFUHRER ET AL.

ercise Physiol. 49: 456-461, 1980. 22. TAYLOR, H. L., E. BUSKIRK, AND A. HENSCHEL. Maximal oxygen intake as an objective measure of cardio-respiratory performance. J. Appl. Physiol. 8: 73-80, 1955. 23. WASSERMAN, K., AND B. J. WHIPP. Exercise physiology in health and disease. Am. Rev. Respir. Dis. 112: 219-249, 1975. 24. WASSERMAN, K, B. J. WHIPP, AND J. A. DAVIS. Respiratory physiology of exercise: metabolism, gas exchange, and ventilatory controL In: Respiratory Physiology Ill, edited by J. G. Widdicombe. Baltimore, MD: University Park, 1981, vol. 3, p. 149-211. (Int. Rev. PhysioL Ser.) 25. WASSERMAN, K, B. J. WHIPP, S. N. KOYAL, AND W. L. BEAVER. Anaerobic threshold and respiratory gas exchange during exercise. J. Appl. Physiol. 35: 236-243, 1973. 26. WHIPP, B. J., J. A. DAVIS, F. TORRES, AND K. WASSERMAN. A test to determine parameters of aerobic function during exercise. J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 50: 217-221, 1981. 27. WHIPP, B. J., S. N. KOYAI., AND K. WASSERMAN. AT and V0 2 kinetics for work increments of various durations (Abstract). Med. Sci. Sports 6: 67, 1974. 28. WITHERS, R. T., W. M. SHERMAN, J. M. MILLER, AND D. L. COSTILL. Specificity of the anaerobic threshold in endurance trained cyclists and runners. Eur. J. Appl. Physiol. 47: 93-104, 1981.