:/KC Vol. 22, blo I
:uly 1993: 175-82
175
--._
--
wi exercise testing has progressed, many
protocols have b n developed to assess certain patient
populations.
Rapidly paced protocols
screening subjects who arc:younger or active, or both (i.e.,
Bruce, Ellestad), whereaz more moderatd: protoco9s are
adapted to subjects who are eider or deconditione
tes Air Force
(i.e., Naughton, Ba
) (I). The main
School of Aerospace
disadvaniage of this plethora of techniques has been determining equivalent work loads among them (Le., what does
As experience
I-lifeactivities, such as hiking or
tions of maximal ventilator oxygen uptake during isotonic exercise have been
and validated as an excellent common language for
from
From the Cardiology Sections, Long Beach Veterans Affairs Medical
Center, Long Beach, Palo Alto Veterans Affairs MedicalCenter, Palo Alto
and Stanford University, Stanford, California. This study was supported in
part by the Veterans AffairsCooperativeStudy ProgramNo. 16. Department
of Veterans Affairs, kshington, D.C.
Manuscript received July 30, 1992;revised manuscript received December 7, 1992,accepted December9, 1992.
Address for correspondence: Victor F. Froelicher, MD. Cardiology
Section (lllc), PaKoAlto Veterans Alfairs Medical Center, 3801 Miranda
Avenue, Palo Alto, California94304.
0591993bv the Americsn Colleee of Cardioloav
ifiSpiied
air
E?Lkr
0
cause maxamaloxygen uptake IS
der, activity status and disease state, tables that
take these
0751097!93/$4.00
146
JACC Vd. 22, No. ’
July 1993:175-82
MORRIS ET AL.
NOMOGRAM USING METS TQ ASSESS EXERCISE CAPACITY
factors into account must be referred to SO that a certain
MET v&e as can be accurately categorized either normal or
abnormal.
We thus decided to fashion a nomogram that would make
it convenient for physicians to translate a MET level into a
percent of normal exercise capacity for men on the basis of
age and activity status, similar to that published by Bruce et
al. (3), except that we utilized METS instead of time on the
Bruce protocol. Because exercise capacity has been shown
to be an independent predictor of cardiovascular mortality
(4), use of METS and this nomogram can facilitate discussions between physicians and their patients with regard to
prognosis as well as disability.
stages, starting at 2.0 mph/O%grade, increasing to 3.3
and followed thereafter by grade increases of S%i’stage
Patients were not allowed to
handrails. The completion of
associated with an estimated additional 2
If a patient completed ~50%
highest stage, he was assigned
associated with the speed and grade of the previous stage.
various populations (5,6) an
time limits were
Studypatients. We retrospectively reviewed the exercise
test results of 3,583 male patients referred to our laboratory
during the period April 16, 1984to December 29,1989 for the
evaluation of possible or probable coronary artery disease.
Excluded were those who had a submaximal exercise test
(e.g., limited by chest pain, target heart rate or illness),
history of a previous myocardial infarction or more than one
Q wave on the rest electrocardiogram (EC@. history of
congestive heart failure, use of beta adrenergic blocking
agents or digitalis, previous coronary artery bypass surgery
or coronary angioplasty, valvular heart disease, chronic
obstructive pulmonary disease or claudication. Because
~2% of our population were women, they were excluded as
well. This left us with a male referral population of 1,388
with a mean age of 57 years (range 21 to 89). Their relatively
normal status was affirmed by a subsequent 3-year follow-up
(annual mortality rate < I%). Those who could be so classified were further classified as sedentary (n = 253) or active
(n = 346)patients, and an additional grouping was made of
those under age 54 years, again with similar subgroupings.
A separate nomogram was developed from 244 normal
men who volunteered for maximal exercise testing with
ventilatory gas exchange analysis. These subjects differed
from the former in that they were not referred for any clinical
reason and were a healthy, younger (mean age 45 r 14years,
range 18 to 72)$ free-living nonveteran population who
responded to advertisements for healthy subjects. Exercise
testing was performed for research purposes in these subjects and not for clinical reasons. They were also classified
into sedentary (n = 74) and active (n = 122)groups.
All data were collected prospectively and entered into a
personal computer data base (R:Base, Microrim). in approximately 50%of the men tested for clinical reasons and 80%of
volunteers tested with ventilatory gas exchange, activity
status was classified as either sedentary or active on the
basis of two questions: I) Do you perform 20 min or more of
brisk walking three times a week, and 2) do you regularly
participate in an aerobic sport?
Exe&e testing. All referred patients underwent exercise
testing using the USAFSAM treadmill protocol with 2-min
oms. Standard criteria for
art rate or
were followed (?), b
as encourosed, and a m~~irnal
Among volunteers who pe
ventilatory gas exchange, an
fatigue, leg fatigue or shortness of breath. A previous study
demonstrated that similar results are obtained by these two
protocols (6).
S~~~~~les.Simple univariate linear regression was performed with age as the independent variable and the various
hemodynamic responses including
rate, maximal systolic blood pressure and maximai ratepressure product as the dependent variables. Statgraphics
(STSC) was used to calculate the regression lines, SE
p values and confidence limits. These relations were studi
for all patients together and for the subgroups classified as
sedentary and active. Age limits were chosen for sciectcd
relations for comparison with other populations. .4nalysis of
variance (ANOVA) was also used to create box plots of
METSfor selected age categories and to test for differences
among age categories.
Cakulationof exerciseca city. The predicted
age were calculated by the formulas obtained from the
regression analysis using age as the independent variable.
This was done for the entire group as well as for the
sedentary and active groups separately. All equations are
numbered in sequence.
The observed metabolic equivalent level was obtained
from final treadmill speed and grade, as explained previously. A percent exercise capacity was then obtained from
the following equation:
Exercisecapacity =
ObservedMETLevel
-- x 100
PredictedMETS
111
Exercise capacity represents the actual percent of r;ai~.~ity for a given age on the basis of METs performed, with
100% the average for age. Values for percent exercise
capacity (i.e., 50%, 80%, lOO%,120%, ISO%)were calc.ulated for various ages using the equation 1. A nomogram was
JACC Vol. 22, No. P
JEII!’ 1993: 175-82
0
1
25
30
2
20
35
;
a
3
40
4
45
5
50
6
56
7
60
6
65
9
:
5
65
70
75
of
percenl
exercise
(based on
ca
lit equivak
for
age in
xercise capacity i
ETS = ~eta~~~~c
E-b]).
by plotting specific ages
regressed against
age
(as shown in Fig.
axirnd heart rate =
196 -
&9(Age).
s, sedentary (~1= 253) ad active
following equations.
Predicted
ts for the
METS
=
18.0 -
O.IS(Age).
PI
tive subgroup (n = 346; S E = 3.0; r = -0.49;
ETs = 18.7 - 0.15(Age).
I31
For the sedentary subgroup (n = 253; §EE = 3.2; r = -0.43;
p < WMl),
Predicted
The
METS
= 16.6 -
0.16(Agej.
I41
2 is shown in Figure 1; t
gure 2, and a
is shown
n in Figure 3.
ation 2 is
For the entire group of 1,388referral ~~t~e~tsjthe
maximal Borg scale was 16, which is consistent with a
maximal effort. For a population with a mean age of 57
years, the mean maximal heart rate of 144 beatslmin is
somewhat lower than expected. aximal heart rate was
mean
JACC Vol. 22. No. 1
July 1993.175-82
MORRIS ET AL.
NOMOGRAM USING METS TO ASSESS EXERCISE CAPACITY
178
following equations: For all
SEE = 2.5; r = --0.53: p < 0.
250
Predicted METS
rmal subjects (n = 24
= 14.7 - O.II(Age).
[6!
For the active subgroup (n = 122; SEE = 2.5; r = -0.5
p -CUOi),
PredictedMETS= !0.4 - O.l3(Age).
I7
For the sedentary subgroup (n = 74; SE
p < O.OOl),
ETs = Il.9 - ~.~7~Ag~~.
.._. ..”
70
_ 1:.......... .....
.
.
I
:..:..I
t
00
40
20
_...+
.._...._!
i
I
.
80
..j
The nomogram for equation 6 is illustrated
the nomogram for equations 7 and 8 is illustr
For normal suL-jects,the values obser
heart rate and maximal perceived exertio
and 19.0 + 1.2, respectively, consistent
effort. Maximal heart rate regressed with age yie
following equation (SEE = 15.3; r = -0.
i
100
AGE
Figure 4. Graphic representation of regression equation of maximal
heart rate versus age for referral patients. Inner dashed lines
represent 95% confidence limits for the mean; outer dashed lines
represent 95% prediction limits (r = -0.43; p < 0.001).
h&XirtIdl heart rate =
An ANOVA was performed for ME’Psby the age grou,ring ~40, 40 to 49, 50 to 59, 60 to 69. 70 to 79 and 80 to 89
years (p c O.OOl),and box plots are illustrated in Figure 5.
The relations were then reanalyzed, excluding referral
patients 54 years of age or older to match a previous study
population (9). Examination of this new grouping (mean age
43 years; n = 442) led to regression equations not appreciably different from equations 2 to 5 for referral patients of all
ages.
Healthy volunteers
tested with vent&tory gas exchange
analysis (normal subjects). Regression analysis of METS
(measured maximal oxygen uptake, in milliliters per kilogram per minute, divided by 3.5) versus age for all nnrmal
subjects and for each of the two subgroups yielded the
24
_,.l
;
;
/
j
~
1
I
20
_..
:
W
s
12
e
4
0
AGE
I
.
.
.
.
,...
_“.
I
.i
.+..
/
,
.I...
“i.
.,.
..,
“..
i
~
/
,
,.....
,I
..?‘..
191
.,.,
...”.._..................
.. ,.,,
!
..,..
..,..
i
..!........ ..~
1
1
...”
L...
!
/
13
.
0.72(Age).
Many physicians find exercise capacity relative to the
peers in an age group to be a useful means of assessing a
pati:l&t’scardiovascular status. The term liiEI’s is a more
meaningful and useful expression of exercise capacity than
are the various protocol times an stages often used. Use of
the term facilitates comparisons of data using different
protocols and tailoring of protocols for particuiar patients.
Metabolic equivalent levels can be used for exercise prescription and for estimating levels of disability by utilizing
tables listing the MET demands of most common activities
(Table I). Thus, it seems clear that use of METs can improve
comm~tmication3mn~ 0 nbk-imc.
r”, I__.-__. , whereas expressing exer-
; ,
.j
._
i
1;
it(~I
4j~
:/
200 -
.
1
1
i ,.,...
i
I ............
..,...,...,,...,,
._,i,,.
,..... .;
.
40
40’s
JO’S
80’s
70’9
80+
n=134
n=201
n=367
n=537
n=l3i
n-g
Figure5. Box plots of metabolicequivalents(MET<!b)
age decades for referral patients. Each borjzo~ta~ line
represents a quartile, and the extremes of the age decade
are represented by the ends of the vertical lines.
JACC Vol. 22, No. I
July 1993:135-82
30
35
35
4c
40
45
45
50
60
55
:
el
55
60
g
4
65
60
55
70
75
a0
85
90
16
Figure 6. Nomogram of percent norrnr;! exercise capaci!y among
healthy normal subjects tested using ventilatory oxygen uptake.
Elfs = metabolic equivalents.
Long Beach Veterans
Affa
dical Center
dilters
from
mall oxygen uptake regressed against age and obtained the
following equation for 710 asymptomatic
men of all activity
levels (age range 20 to 53 years):
Predicted METS -z 13.1
O.Og(Age)
(n z-. 710: SEE = 1.7; r = -0.32).
Comparing this with that obtained
patients -44 years old:
Predicted METS
(n
ilQ1
from our
= 18.8 - O.iS(Age)
; 442; SEE -. 3.3; p K O.OOl:r -
0.33).
IIrl
The slope and intercept are obviously different. consistent
with differences in the populations (see later)
derived his nomogram for functional exercise capacity from
Figwe ?. Nnmgram
cf percLnt normai sxerck
capacity among
active and sedentary healthy normal subjects teste
tory oxygen uptake. METS = met&olic equivalenls.
180
MORRISST AL.
NOMOGRAM USING METS TO ASSESS EXERCISE
TaBlo 1. Estimated
Energy Requirements
Activity
Mild
Baking
Billiards
Bookbinding
Canoeing (leisurely)
Conducting an orchestra
Dancing, ballroom (slow)
Golf (with cart)
Horseback riding (walking)
Playing a musical instrument
Volleyball (noncompetitivel
Walking (2mph)
Writing
Moderate
Calisthenics (no weights)
Croquet
Cycling (i?isUrely)
Gardening (no lifting)
Golf (without cart)
Mowing lawn (power mower)
Playing drums
Sailing
Swimming (slowly)
Walking (3 mph)
Walking (4 mph)
Vigorous
Badminton
Chopping wood
Climbing hills
Cycling (moderate)
Dancing
Field hockey
Ice skating
Jogging(IO-min mile)
Karate or judo
Roller skating
Rope skipping
Skiing (water or downhill)
Squash
Surfmg
Swimming (fast)
Tennis (doubles)
of
CAPACITY
SelectedActivities*
METS,
2.0
2.4
2.2
2.5
2.2
2.9
2.5
2.3
2.0
2.9
2.5
1.7
4.0
3.0
3.5
4.4
4.9
3.0
3.8
3.1
4.j
3.3
4.5
5.5
4.9
7.0
5.7
6.0
7.1
5.5
10
6.5
6.5
I2
6.8
12
6.0
7.0
6.0
JACC Voi. 22, No. 1
July 1993: 175-82
der. The latter is not a question in this con
stud& dealt with men. There are significant di
the age ranges and means of the studies q
population being the oldest. For this reaso
separate analysis of patients <54 years of age
slope was still obtained. The decline in maxi
with age is also steeper in our referral group, par
maximal oxygen uptake slope. Thus, maximal
decreased with age at a greater rate than in previ
(1I), an observation that
effort or complicating illnesses in older pat
ulation was classified by a
The study
ing” examination in
exclusion criteria in
having to fulfillcriteria to
Air Force study. Failure to adequately exclude sicker cardiac patients could cause variations in the results. Activity
status was not clas%ed in the study by
was classified by a similar met
activity st
irs and U.S. Air Force studies
Veterans
varying levels of past or current co~ditiQ~i~g
factor in the divergence of the regression equations. Differences in methodology must also be considered when examining divergent results. Only our healthy volu
subjects in the U.S. Air .?orce study were
measured maximal oxygen uptake values,
treadmill protocols were different, numerous
demonstrated similar maximal values with Ed
investigations (Table 3).
It is difficult to determine which study provides the most
universal regression equations because all studies have
weaknesses in either population selection or methodology.
One solution would be for each center to develop its own
regression equations, but this is neither economical nor
likely. Our equations derived from the referral patients (Fig.
1 and 2) are specific to male veterans referred for exercise
testing for evaluation of possible heart disease, excluding
those with obvious medical problems that might compromise
. *
..
*These actrvmes can often be performed at variabie iniensrttes, assmmmg
t’“;:. exercise cap&%yy.Hov~vcr, they vKX!!dcertainly be
tha: the intensity is not excessive and that the courses are flat (no hills) unless
applicable
to hospital- or office-based settings that care for
so specified. Adapted, with permission, from Fletcher et al. (1).
similar patients. In contrast, the earlier studies by Bruce et
al. (3) and Froeiicher et al. (9), which consisted primarily of
apparently healthy normal subjects and volunteers, are not
Predicted METS = 15.2 - O.II(Age).
1141
as applicable to the patients seen by practicing physicians.
Ail of these equations are listed in Table 2 for comparison.
The regression lines in our referral group may differ from
Several factors account for the differences among the
those in “free-living populations” or volunteers owing to
regression equations obtained from the referred and volunvarying levels of subclinical disease, activity status and
teer groups in the present study, the U.S. Air Force study (9)
obvious differences in attitude toward the test by both test
and the study of Bruce et al. (3). The steeper slope in the
administrators and subjects.
referral group is consistent with a faster decline in maximal
The combination of these factors explains the upward
oxygen uptake with age than that found in these previous
shift in the slope of the nomogram scale among volunteers in
studies. Regression analyses can vary owing to population
whom oxygen uptake was determined directly from ventiiadifferences in such areas as age, activity status, state of
tory gas-exchange analysis in the present study. Also, estiheah
definition of normal or healthy individuals and genmating metabolic equivalent levels from treadmill work
Peesent study
Refeerral patieai:
1,388
PieMETS =
18. i - 0.16(hge)
57 (28-89)
Simple
ofCAPG, CHF, B
Drg. COPD.ciaudisation,
No hismy
quesimnawe
UK
Est
or rl Q wave on ECG
Patients 64
yr old
Normalsubjects
Pr METS =
18.8 - O.l?(Age)
479
Pr METs=
.2@
Apparently healthy
710
NormaPexamination,
normal
415(21-53)
Simple
questionnaire
No history of CA%.
oe Dig, C@PD,, ci
angina, previous
~by~~rnias or >
on EC6
CHE, 5
5%
eas
14.7 - O.II(Age)
Fraelicher et al. (9)
Pr METS =
13. I - 0.0X(&e)
Rruce et al. (3)
Pr METS =
13.7 - ~.O~(A~e)
Wolthuis ct al. (II)
PK METS =
13 -- O.OS(Age)
resting and exercise ECG, no
by~e~e~si~~
2,092
44.4 (NA)
37 (2S-54)
70.4
None
Cardiac screening examination
ES4
Questionnaire/
interview
Normal history and physical
MWeFlS
exercise ECC, and Halter
monitoring; no arrhythmias.
hypertension or medicaiions
Dehn and Bruce (IO)
7H?
Pr METS =
16.2 - O.ll(Rge)
BB = beta-blocker;
CA5G
= coronary
artery
52.2 (40-72)
bypass graft surgery; CHF
Mixed
-
= congestive heart failure;
COPD
= chronic obstrucbve
CXR = chest X-ray film; Dig = digitalis; ECG = electrocardiogram;
Est = estimaled; MTN = hypertension:
NA = not available; Pr MET’s = predicted metabolic equivalents;
ref. = reference; USAFSAM
= United States Air Force School
results in an ovemiiiiiatioii of exercise capacity (6,?5,?6).
The approximately 1.0 to 1.5
equivalent values for any given age a
in whom exemse capaciry w
work load are not surprisi
IPreported previously (6,15,16).Because
imal heart rate versus age ~e~a~~Q~
was
steeper in the referra! grou
hat lower maximal
Table
3. MetLbolic Equivalent (MET) Norms fbr tiera, From
Availablle Studies of Large Populations
MetabolicEquivalent Norms
_-
4ge (yr)
20 to 29
30 to 39
40 to 49
50 to 59
60 to 69
70to 79
80 to89
Froelicher
et al. (9)
Hossack
et al. (18)
11 +2
13 * I
BOk 2
IO + 2
-
12 * 2
11 zk2
102 2
822
Sk1
-
Pollock and
Wilmore
(19)*
12 + 2
I2 2 2
II k2
IO + 2
822
8k2
-
Morris
(Present
Study0
II +4
9t4
8?3
7%3
523
*Cooper’sclinic. tP.eferralpatients. Values presented are mean values +
I SD.
182
JACC Vol. 22, No. I
MORRIS ET AL.
be used for sirbjects tested for screening or preexercise
program evaluations.
Conciusion;;.Neariy 20 years ago, Bruce et aI. (3) deveioped a nomogram for assessing a patient’s “functional
aerobic impairment” based on age and time exercised in the
Bruce protocol. With the proliferation of different treadmill
and ergometer protocols, the term metabolic equivalent, or
MET, has come into use as the common denominator to
assess exercise capacity. For these reasons, a new nomogram of percent normal exercise capacity based on age and
METSis very much needed. Expression of exercise capacity
relative to a patient’s peers of similar age is a usefili means
of assessing cardiovascular status. Our referral population
provides reference values that can be applied by physicians
who see similar patients in office- or hospital-based practices. The nomogram and box plots CMIbe used to facilitate
the description of relative exercise capacity for patients,
physicians and employers. They can be used to estimate
functionality and disability in easily understood terms.
..““----.~
July 1993:175-82
NOMOGRAM USING METS TO ASSESS EXERC1SE CAPACITY
-
V-.-Y
We thank Karl Hammermeister. MD (Chief of Cardiology, Denver Veterans
Affairs Medical Center, Denver, Colorado) for suggesting the idea of a
nomogram and Dr. Myrvin Ellestad (Director, Long Beach Memorial Heart
Institute, Long Beach, California) for his helpful suggestions.
_----.-.
4.
5.
6.
7.
8.
9.
IO.
II
12.
13.
94.
15.
16.
Fletcher GF, Froelicher VF. Hartley LH. Haskell W. Pollock M Exercise sfandards: a slatement for health professionals from the American
Heart Association. Circulation 1990$2:2286-322.
Jette M, Sidney M, Blumchen G. Metabolic equivalenlk (METS) in
exercise testing. exercise prescription. and evaluation of functional capacity. Clin Cardiol 1990:13555-65.
Bruce RA. Kusumi F, Hosmer D. Ma_. intake and nomo.
*::k! oavqen
17.
18.
graphic assessment of functional aerobic impairment in cardiovascular
diseas<:.Am Heart J 1913:85:546-62.
Morris. CK: lreshime K. Kewraeuchi T; Hideg A. Froelicber VF. The
prognostic value of exercise capacity: a review of the literature. Am
J 1991;122:1423-31.
Wolthuis RA. Froelichcr VF, Fischer J, et al. New practical treadmill
protocol for clinical use. Am J Cardiol 1977;39:697-7QO.
Myers J, Buchanan N, Walsh D, et al. Comparison of the ramp versus
standard exercise protocols. J Am Coil Cardiol 1991;17:1334-42.
American College of Spo.-ts Medicine. Guidelines for Exercise Testing
and Prescription. Philadelphia:
Myers J, Buchanan N, Smith
t aI. Individualized ramp treadmill:
observations on a new protocol
st 1992;101:2363_41S.
Froelicher VF, Allen M, Lan
normal USAF aircrewmen. Aerosp Med 1974:45:310-5.
Dehn MM. Bruce RA. Lon~~tM~i~at variations in maximal oxvaen intake
with age and aclivily. J Apii Physiol 1972;33:805-7.
-*
Hammond HK, Fro&her VF. Normal and abnormal bean rate responses 10 exercise. Prog Cardiovasc Dis 9985;27:271-96.
Froelicher VF, Thompson AJ. Noguera 3, et al. Prediction of maximal
oxygen consumption: comparison of the Bruce and Balke treadmill
protocols. Chest 19?5;68:331-6.
Pollock ML, Bohannon RL, Cooper KM, et al. A comparative analysis of
four protocols for maximal treadmill stress testing. Am Heart J 1976;92:
39-46.
Froelicher VF, Brammeli 9-1, Davis G, et al. A comparison of the
reproducibility and physiologic response to three maximal treadmill
exercise protocols. Chest 1974;65:512-20.
Sullivan M. McKirnan D. Errors in predicting functional capacity for
post-myocardial infarclion patients using a modified Bruce protocol. Am
Heart J 1984;907:486-91.
Roberts JM. Sullivan M. Froelicher VF, Genter F,
oxygen uptake from treadmill testing in normal subjects and coronary
artery disease patients. Am Heart J 1984;108:l454-60.
Wolthuis R. Froelicher V. Fischer J. Triebwasser J. The response of
healthy men to treadmill exercise . Circulation 1977;55:153-7.
Hosseck KF. Bruce RA. Green 5. Kusumi F, Derouen TA, Trimble S.
Maximal cardiac output during upright exercise: approximate
normal
standards and variations with coronary heart disease. Am J Cardiol
1980:46:204-12.
19. Pollock M. Wilmore J. Exercise in Health and Disease. Philadelphia:
Saunders,
1990:326.
WB