Original Articles

Value of Exercise Treadmill Testing in the Risk

Stratification of Patients With Pulmonary Hypertension
Sanjiv J. Shah, MD; Thenappan Thenappan, MD; Stuart Rich, MD; James Sur, MD;
Stephen L. Archer, MD; Mardi Gomberg-Maitland, MD, MSc
BackgroundThe ability of the Naughton-Balke exercise treadmill test, an objective indicator of exercise capacity, to
predict abnormal hemodynamics and mortality in pulmonary hypertension is unknown.
Methods and ResultsWe performed a cohort study of 603 patients with pulmonary hypertension from 1982 to 2006, and
studied the utility of exercise treadmill test as a predictor of abnormal hemodynamics and death. We used multivariable
linear regression to determine whether exercise capacity, measured in metabolic equivalents, was associated with
abnormal hemodynamics, and we used a Cox proportional hazards model to determine whether decreased exercise
capacity predicted death. Mean age was 5015 years, 76% were women, 63% had World Health Organization category
I pulmonary arterial hypertension, and 23% were World Health Organization functional classes I and II. Mean exercise
capacity was 3.72.2 metabolic equivalents. Decreased exercise capacity was independently associated with elevated
right atrial and mean pulmonary artery pressure, decreased cardiac index, and increased pulmonary vascular resistance.
During median follow-up of 4.6 years, 36% of the patients died. Decreased exercise capacity was associated with
mortality (multivariable hazard ratio, 1.18; 95% CI, 1.01 to 1.37 for each 1-metabolic equivalent decrease in exercise
capacity; P0.031; P0.052 after adjusting for invasive hemodynamic variables). Decreased exercise capacity also
predicted mortality in functional classes III patients, 24% of whom died (hazard ratio, 1.53; 95% CI, 1.04 to 2.26 for
each 1-metabolic equivalent decrease in exercise capacity; P0.032), although this association did not persist after
adjusting for invasive hemodynamic variables (P0.63).
ConclusionsReduced exercise capacity on exercise treadmill test is associated with worse hemodynamics and is a
predictor of mortality in patients with pulmonary hypertension. (Circ Heart Fail. 2009;2:278-286.)
Key Words: pulmonary hypertension exercise hemodynamics mortality

n the evaluation of patients with cardiorespiratory illness,

assessment of functional capacity is extremely valuable.
Reduced functional capacity in patients with heart failure,
whether evaluated subjectively by New York Heart Association
class, or objectively by 6-minute walk test (MWT) or cardiopulmonary exercise testing with determination of peak oxygen
consumption (VO2), is associated with a worse hemodynamic
profile and adverse outcomes.1 In patients with pulmonary
hypertension (PH), World Health Organization (WHO) functional class2 (analogous to New York Heart Association
functional class) and 6-MWT are tools that are used to define
clinical severity. Functional class determination is low cost
and offers prognostic assessment in a variety of diseases but
is a subjective measure on behalf of both patient and
physician. Although 6-MWT is thought to be a more objective measure of exercise capacity, it is too dependent on
patient and physician factors and is susceptible to motivational factors, especially in patients with WHO functional

classes III symptoms, who have the ability to respond to

motivation with increased exercise.37 Cardiopulmonary exercise testing is considered the reference standard for the assessment of functional capacity and VO2, but it is not available in
many centers, is complex, is more expensive than determination
of functional class and 6-MWT, and (compared with 6-MWT)
may only add marginally to risk prediction in PH.7,8

Clinical Perspective on p 286

We have shown that exercise treadmill testing (ETT) may
offer a more objective alternative to 6-MWT.9 Using the
Naughton-Balke exercise protocol,10 we determined that ETT
was a reliable measure of exercise capacity in pulmonary
arterial hypertension and correlated with 6-MWT. In addition, ETT seemed to be more sensitive than 6-MWT in
detecting changes in exercise capacity in less sick patients.
Furthermore, a recent meta-analysis of pulmonary arterial
hypertension clinical trials11 found no association between

Received July 16, 2008; accepted March 30, 2009.

From the Division of Cardiology (S.J.S.), Department of Medicine, Northwestern University Feinberg School of Medicine; and the Section of
Cardiology (T.T., S.R., J.S., S.L.A., M.G.), Department of Medicine, University of Chicago, Chicago, Ill.
The online-only Data Supplement is available at http://circheartfailure.ahajournals.org/cgi/content/full/CIRCHEARTFAILURE.108.807826.
Correspondence to Mardi Gomberg-Maitland, MD, MSc, Director of Pulmonary Hypertension, University of Chicago Medical Center, 5841 S
Maryland Ave, MC 2016, Chicago, IL 60637. E-mail mgomberg@medicine.bsd.uchicago.edu
2009 American Heart Association, Inc.
Circ Heart Fail is available at http://circheartfailure.ahajournals.org

DOI: 10.1161/CIRCHEARTFAILURE.108.807826

278
Downloaded from http://circheartfailure.ahajournals.org/
by guest on March 7, 2016

Shah et al
change in 6-MWT and survival, causing some to conclude
that 6-MWT is not an adequate surrogate end point in PH.11,12
Given the possible advantages of ETT in PH, we sought to
understand the value of ETT as a marker of abnormal
hemodynamics, and as a noninvasive predictor of survival in
patients with PH. We hypothesized that poor exercise capacity on ETT would be highly associated with abnormal
hemodynamics and would be a strong predictor of poor
outcomes. Furthermore, we hypothesized that ETT would be
a valuable test even in less sick patients (WHO functional
classes I and II). We tested these hypotheses in the Pulmonary
Hypertension Connection (PHC) registry.13

Methods

Exercise Treadmill Testing in PH

279

on a treadmill because of orthopedic issues. WHO functional class

IV was not an exclusion criterion for ETT if the patient could
ambulate and was not hospitalized at the time of initial evaluation.
All treadmill tests used the Naughton-Balke protocol,10 and treadmill
time (in seconds) was converted to exercise metabolic equivalents
(METs) as described previously,9 although it should be noted that the
conversion formula for exercise time to METs has not been validated
in PH. Sex-specific nomograms for METs have been published
previously.15,16

Patient Characteristics and

Laboratory Measurements
We analyzed the following baseline variables at the time of referral
for characterization of clinical phenotype: demographic data including age and sex, comorbidities, WHO functional class, medications,
albumin, creatinine, antinuclear antibody, and pulmonary function
testing, including diffusing capacity of carbon monoxide.

Study Sample
We studied patients in the PHC registry, which was initiated in
March 2004, and which has been described in detail previously.13,14
All patients evaluated at a single United States practice over time at
3 different university hospitals (University of Illinois at Chicago,
Rush University Medical Center, and University of Chicago Medical
Center) between 1982 and 2006 were entered into the database. Over
the study period, 4 physicians acquired all the clinical data. Data
were collected by a chart review and entered using an internet-based
electronic data capture system. Patients were entered retrospectively
from 1982 to February 2004 and prospectively from March 2004
through 2006.
Since the registry was initiated, only 2 investigators with expertise
in data management and understanding of clinical care of patients
with PH entered the data. Neither of these investigators participated
in the care of the patients in the registry. Hemodynamics were
entered separately without the knowledge of any of the other clinical
data, and outcomes (mortality data) were collected by the investigators after all other data had already been entered. Therefore, clinical
data entry was performed blinded to hemodynamics and eventual
outcome. Informed consent for participation in the registry was
obtained during initial evaluation for new patients, or during routine
office visits for patients who were already being followed before
initiation of the registry (1% refused entry into the study). Data
entry occurred after the complete initial evaluation. For all variables,
outliers were verified by a chart review to minimize data entry errors.
The PHC registry was approved by the respective institutional
review boards based on the location of the practice, and all actively
seen patients gave informed consent to be entered into the registry.
Greater than 95% of the patients in the registry were seen as
outpatients during initial evaluation (the remaining were inpatients
transferred from the referring institution). We collected baseline
demographic, clinical, medication, exercise testing, and cardiac
catheterization data on all patients seen by our practice.

ETT
Of the total PHC registry cohort (N1360), we excluded patients
who were 18 years at the time of referral (N43) and also
excluded patients if we could not determine whether or not they had
undergone ETT by a chart review (N114). Of the remaining 1203
patients, 309 with missing ETT data and 291 with missing ETT and
functional class data were excluded. The remaining 603 patients
comprised our study cohort. Supplementary table S1 demonstrates
the differences between the study cohort and those in the PHC
registry who had missing exercise capacity and functional class data.
Of the 603 patients in our study, 458 (76%) were entered retrospectively and the remaining 145 (24%) were entered prospectively.
Reasons for not undergoing ETT at baseline included the following:
(1) the patient had already undergone evaluation of exercise capacity
using another modality (6-MWT or cardiopulmonary exercise testing) or if the patient had undergone exercise testing at the referring
facility; (2) the patient was hospitalized during initial evaluation
(only outpatients underwent ETT); and (3) the patient could not walk

Invasive Hemodynamics
Of the 603 subjects included in our analysis, 521 (86%) underwent
baseline hemodynamic testing by right heart catheterization, the
majority (95%) of which were performed at our institution by PH
specialists. All hemodynamic testing performed at our institution
was completed within 1 month of initial referral, and all patients
were hemodynamically stable at the time of catheterization. Pulmonary vascular resistance (PVR) was calculated as follows:
PVR(mean pulmonary artery pressurepulmonary capillary
wedge pressure)/cardiac output.

Mortality
Vital statistics were collected for all patients by a chart review and
by a query of the Social Security Death Index. For each death, the
date of death was documented. Social Security Death Index data
were available on all patients. In all patients who were not identified
as deceased by the Social Security Death Index, we were able to
confirm vital status by a chart review.

Statistical Analysis
All continuous variables are expressed as meansSD unless otherwise noted, and P values 0.05 were considered statistically
significant. We first compared groups of patients depending on
baseline WHO functional class, dividing the cohort into 3 groups for
descriptive purposes: WHO functional classes III, class III, and
class IV. Among the WHO functional class groups, we compared
demographics, clinical characteristics, laboratory tests, and hemodynamics with analysis of variance and Kruskal-Wallis tests for
continuous variables, and 2 and Fisher exact tests for categorical
variables.
To better understand the relationship between ETT exercise
capacity and the various demographic, clinical, and laboratory
variables, we performed univariate linear regression with ETT
exercise capacity (METs) as the dependent variable. We then used
univariate and backward-selection multivariable linear regression to
determine whether ETT exercise capacity was associated with
abnormal hemodynamics after adjusting for other baseline risk
factors (linear regression assumptions were checked for all models).
We studied the effect of each 1-MET decrease in exercise capacity
on 4 hemodynamic variables, which have been shown to be important prognostic parameters in patients with PH17: right atrial pressure, mean pulmonary artery pressure, cardiac index, and PVR. We
repeated these analyses in the subset of patients with WHO functional classes III PH and in the subset of patients with WHO
category I PH, studying the effect of each 1-MET decrease in
exercise capacity on hemodynamics.
Next, we used a locally weighted smoothed scatterplot curve to
graphically depict the relationship between observed METs and
mortality. We used the Kaplan-Meier method to estimate survival
rates for patients with exercise capacity 3, 3 to 6, and 6 METs,
and Kaplan-Meier survival curves were compared by log-rank test.
To determine the univariate and multivariable risk of death by

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

280
Table 1.

Circ Heart Fail

July 2009

Baseline Demographic, Clinical, Laboratory, Echocardiographic, and Hemodynamic Characteristics by Functional Class
WHO Functional Class

Characteristics
Age, y
Female, n (%)

III (N138)

III (N286)

IV (N179)

4614

5115

5215

0.001

103 (75)

214 (76)

137 (77)

Etiology of pulmonary hypertension, n (%)

WHO category I (pulmonary arterial hypertension)

86 (62)

178 (62)

116 (65)

WHO category II (pulmonary venous hypertension)

14 (10)

44 (15)

23 (13)

WHO category III (pulmonary hypertension associated with hypoxemia)

0.89
0.13

2 (2)

13 (5)

10 (5)

WHO category IV (pulmonary hypertension due to chronic

thromboembolic disease)

10 (7)

16 (6)

14 (8)

WHO category V (miscellaneous)

26 (19)

35 (12)

16 (9)

Medications, n (%)
Diuretics

42 (31)

137 (49)

113 (64)

0.001

Calcium channel blockers

44 (32)

89 (32)

60 (34)

0.87

Digoxin

20 (15)

49 (17)

31 (18)

0.73

Warfarin

50 (37)

99 (35)

55 (31)

0.58

Prostacyclins

4 (3)

5 (2)

2 (1)

0.45

Endothelin blockers

3 (2)

11 (4)

7 (4)

0.68

Phosphodiesterase inhibitors

3 (2)

6 (2)

1 (0.5)

0.42

Comorbidities, n (%)
Systemic hypertension

36 (26)

101 (36)

66 (37)

0.082

Diabetes mellitus

10 (7)

20 (7)

19 (11)

0.36

Obesity

17 (12)

51 (18)

34 (19)

0.23

Coronary artery disease

4 (3)

22 (8)

12 (7)

0.15

Interstitial lung disease

11 (8)

37 (13)

25 (14)

0.23

Pulmonary embolism

34 (25)

60 (21)

41 (23)

0.69

Laboratory tests
Serum creatinine, mg/dL (N553)

0.90.3

1.10.3

1.21.2

0.019

Albumin, g/dL (N528)

4.00.4

3.90.5

3.80.5

0.0002

Positive antinuclear antibody test, n (%) (N429)

42 (42)

67 (54)

0.18

105 (51)

Pulmonary function tests

FEV1, % predicted (N504)

8022

7418

6920

0.0001

Forced vital capacity, % predicted (N503)

8219

7618

7220

0.0002

FEV1/FVC, % (N365)

9013

9213

8714

0.006

Total lung capacity, % predicted (N460)

9117

8717

8322

0.003

DLCO, % predicted (N442)

6522

5824

4920

0.0001

Treadmill exercise capacity, METs

6.22.4

3.51.5

2.10.8

0.0001

Hemodynamics (N521)
Mean right atrial pressure, mm Hg*

7 (410)

9 (514)

12 (816)

0.0001

Mean pulmonary artery pressure, mm Hg

4216

4815

5213

0.0001

Cardiac index, L/min/m2

2.70.8

2.40.9

2.20.8

0.0001

Pulmonary vascular resistance, WU

7.45.2

10.47

11.76.3

0.0001
0.001

Pulmonary capillary wedge pressure, mm Hg

105

116

137

Mean aortic pressure, mm Hg

9415

9616

9516

0.73

Pulmonary artery oxygen saturation, %

688

6011

5511

0.0001

Acute vasodilator response, % (N409)

9 (11)

8 (4)

2 (1.4)

0.006

*Values represent median (interquartile range) and P value was calculated using the Kruskal-Wallis statistic because right atrial pressure was right skewed.

METs, we used a Cox proportional hazards analysis (with backward

selection of covariates), and the proportionality assumption was
tested and confirmed for all models. From our univariate analysis of
variables associated with METs at P0.10 and from the National

Institutes of Health study of mortality in pulmonary arterial hypertension,18 we determined that the following covariates should be
entered into our multivariable models: age, etiology of PH, systemic
hypertension, diabetes mellitus, obesity, interstitial lung disease,

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

Shah et al

Exercise Treadmill Testing in PH

281

Table 2. Variables Associated With Exercise Capacity

(Metabolic Equivalents) on Univariate Analysis
Metabolic Equivalents
Variables

-Coefficient (95% CI)

Age*

0.36 (0.47 to 0.24)

0.0001

1.0 (1.9 to 0.15)

0.022

Etiology of pulmonary
hypertension
WHO category III (pulmonary
hypertension associated with
hypoxemia)
WHO category
V (miscellaneous)

0.9 (0.4 to 1.5)

0.0001

Comorbidities

Figure 1. Differences in exercise capacity among various etiologies of pulmonary hypertension. PAH indicates pulmonary arterial hypertension; PVH, pulmonary venous hypertension; Hypoxemia, pulmonary hypertension due to hypoxemia; CTEPH,
pulmonary hypertension due to chronic thromboembolic disease; Miscellaneous, pulmonary hypertension due to miscellaneous causes.

Systemic hypertension

0.82 (1.18 to 0.45)

Diabetes mellitus

0.75 (1.38 to 0.11)

0.021

Obesity

0.42 (0.88 to 0.04)

0.077

Interstitial lung disease

0.73 (1.26 to 0.20)

0.007

1.15 (1.15 to 0.81)

0.0001

Medications
Diuretics
Phosphodiesterase inhibitors

Results
Mean age on entry into the study was 5015 years and 76%
were women. Of the total cohort, 63% had pulmonary arterial
hypertension (WHO category I), 3.5% were taking endothelin
antagonists, 1.8% were taking prostacyclins, 1.7% were on
phosphodiesterase inhibitors, and 13% were on calcium
channel blockers for PH (19% of the cohort were on calcium
channel blockers for other reasonsie, systemic hypertension or Raynauds phenomenon). As is characteristic of PH,
the study patients had significantly reduced exercise capacity.
Mean exercise time was 5.44.3 minutes, and mean exercise
capacity was 3.72.2 METs (median, 3.1 METs; range, 1.1
to 11.1 METs). Even patients with WHO functional classes
III symptoms had decreased exercise capacity (mean,
6.22.4 METs; median, 6.1 METs; 29% with poor exercise
capacity 5 METs). Importantly, no patients had adverse
complications or died during or shortly after the NaughtonBalke ETT protocol. Table 1 lists the demographic, clinical,
laboratory, pulmonary function test, treadmill exercise capacity, and hemodynamic data stratified by WHO functional
class. Functional classes III patients were younger, less

1.66 (0.30 to 3.02)

0.017

1.78 (1.92 to 1.64)

0.0001

Forced expiratory volume in

1 second

0.029 (0.020 to 0.039)

0.0001

Forced vital capacity

0.031 (0.021 to 0.041)

0.0001

Total lung capacity

0.028 (0.018 to 0.039)

0.0001

Diffusing capacity of carbon

monoxide

0.034 (0.026 to 0.043)

0.0001

WHO functional class

diuretic and phosphodiesterase inhibitor use, WHO functional class, albumin, creatinine, antinuclear antibody test, forced vital capacity,
diffusing capacity of carbon monoxide, and hemodynamic variables
(right atrial pressure, mean pulmonary artery pressure, and cardiac
index). To avoid multicollinearity, forced expiratory volume in 1
second and total lung capacity were not included in our multivariable
analyses, because they were highly correlated with forced vital
capacity (r0.89, P0.0001 and r0.63, P0.0001, respectively).
Transformation of hemodynamic variables for normality did not alter
the results of any of our multivariable analyses. Finally, we repeated
our Cox regression analyses in the subset of WHO functional classes
III patients and in the subset of patients who had WHO category
I PH (pulmonary arterial hypertension). All statistical analyses were
performed using Stata (version 9, StataCorp LP, College Station,
Tex). The authors had full access to the data and take responsibility
for its integrity. All authors have read and agree to the manuscript as
written.

0.0001

Pulmonary function tests

Laboratory data
Serum creatinine

0.034 (0.60 to 0.08)

0.009

Serum albumin

0.85 (0.46 to 1.23)

0.0001

Positive anti-nuclear antibody

test

0.67 (1.09 to 0.25)

0.002

*Per decade increase.

Referent groupWHO category I (pulmonary arterial hypertension).
Per 1-unit increase in WHO functional class.
Per 1-unit increase in % predicted value.
Per 0.1 mg/dL increase in serum creatinine.
Per 1 g/dL increase.

likely to be on diuretic therapy, and had lower serum

creatinine, higher serum albumin, and better results on
pulmonary function testing. There was a graded decrease in
exercise capacity and worsening hemodynamics with increasing functional class. In addition, functional classes III
patients were more likely to have a positive vasodilator
response. Figure 1 displays the variability in exercise capacity
by etiology of PH. WHO category III patients (PH associated
with hypoxemia) had the worst exercise capacity, and WHO
category V patients (PH due to miscellaneous causes) had the
best exercise capacity.
Table 2 lists the clinical variables that were associated with
exercise capacity on univariate linear regression analysis at a
significance of P0.10. Advanced age, systemic hypertension, diabetes mellitus, and diuretic use were all associated
with decreased exercise capacity. Lung disease, as indicated

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

WHO indicates World Health Organization; RAP, right atrial pressure; MPAP, mean pulmonary artery pressure; CI, cardiac index; PVR, pulmonary vascular resistance.
*Per 1-MET decrease in exercise capacity on univariate analysis.
Per 1-MET decrease in exercise capacity on multivariable analysis using backward selection with the following covariates: age, etiology of pulmonary hypertension, systemic hypertension, diabetes mellitus, obesity, interstitial
lung disease, diuretic and phosphodiesterase inhibitor use, WHO functional class, albumin, creatinine, antinuclear antibody test, forced vital capacity, and diffusing capacity of carbon monoxide.

0.013

0.0001
1.1 (0.5 to 1.7)
0.0001
0.9 (0.6 to 1.2)
0.003
1.0 (0.3 to 1.6)
0.005
0.6 (0.2 to 1.0)
0.0001
0.9 (0.6 to 1.2)
0.0001
0.9 (0.6 to 1.1)
PVR, WU

0.006
1.8 (0.5 to 3.1)

0.11 (0.20 to 0.02)

0.0001

0.0001
1.3 (0.6 to 1.9)

0.11 (0.20 to 0.02)

0.019

0.0001
2.9 (1.3 to 4.5)

0.12 (0.21 to 0.02)

0.031

0.001
2.0 (0.8 to 3.2)

0.07 (0.13 to 0.007)

0.0001

0.0001
1.9 (1.1 to 2.6)

0.10 (0.15 to 0.05)

0.0001

0.0001
2.1 (1.5 to 2.7)

0.12 (0.16 to 0.09)

MPAP, mm Hg

CI, L min1 m2

0.8 (0.2 to 1.4)

-Coefficient (95% CI)

P

0.0001
1.1 (0.8 to 1.4)
0.0001
0.9 (0.4 to 1.4)
0.0001
0.7 (0.3 to 1.0)
0.0001
0.6 (0.3 to 1.0)
0.0001
1.1 (0.9 to 1.3)
RAP, mm Hg

Pulmonary Arterial Hypertension, N380

-Coefficient* (95% CI)

P

-Coefficient (95% CI)

P

WHO Functional Classes III, N138

-Coefficient* (95% CI)

P

-Coefficient (95% CI)

Total Cohort, N603

-Coefficient* (95% CI)

Hemodynamic
Parameter

by a diagnosis of WHO category III PH, presence of

interstitial lung disease, or low lung volumes on pulmonary
function testing, was strongly associated with decreased
exercise capacity.
Table 3 summarizes the association of exercise capacity
and abnormal hemodynamics. On univariate and multivariable analyses, reduced exercise capacity was associated with
increased right atrial and mean pulmonary artery pressure,
decreased cardiac index, and increased PVR. These associations persisted in the subgroup of patients who were WHO
functional classes III and in the subgroup of patients with
pulmonary arterial hypertension. Figure 2 graphically depicts
the differences in hemodynamic variables in subjects with
exercise capacity 3, 3 to 6, and 6 METs.
Mean follow-up time was 5.54.1 years (minimum
follow-up time 1 day, maximum follow-up time 14.6 years).
Of the total cohort, 217 (36%) died during the follow-up.
Figure 3 shows the relationship between exercise capacity
and mortality over the range of observed METs. Figure 4
displays the Kaplan-Meier survival curves for patients with
exercise capacity 3, 3 to 6, and 6 METs. Table 4
summarizes the results of our Cox proportional hazards
analysis. On univariate analysis, each 1-MET decrease in
exercise capacity was associated with a 1.24-fold increased
hazard ratio (HR) of death (95% CI, 1.14 to 1.35; P0.0001).
On multivariable analysis, each 1-MET decrease in exercise
capacity continued to independently predict death (HR, 1.18;
95% CI, 1.01 to 1.37; P0.031). Similar results were
obtained in the subset of patients with WHO functional
classes III PH (Table 4). When invasive hemodynamic data
were taken into consideration, reduced exercise capacity was
associated with increased death, although at borderline significance (P0.052) in the total cohort, but not in the
functional classes III patients (P0.63). In the subgroup of
patients with pulmonary arterial hypertension (N380), decreased exercise capacity was an independent predictor of
death on univariate and multivariable analyses, even when
adjusted for hemodynamic variables (Table 4).
We performed the following additional statistical analyses
to further explore the association between ETT and death: (1)
the effect of adding PVR to our multivariable models; (2)
analysis of the subgroup of patients who did not have WHO
category I pulmonary arterial hypertension (ie, WHO categories II to V PH); and (3) analysis of the subgroup of patients
who were not on pulmonary arterial hypertension therapies
(calcium channel blocker, endothelin antagonist, phosphodiesterase inhibitor, or prostacyclin) at the time of referral.
Although PVR was highly associated with exercise capacity, we did not include it in our original multivariable Model
2 (Table 4) because PVR is a direct function of other
variables in the multivariable model (namely mean PA
pressure and cardiac output). Including PVR with these other
variables in our multivariable models, would have resulted in
multicollinearity. However, we did analyze the effect of
replacing mean pulmonary artery pressure and cardiac output
with PVR in our Model 2, and the results of our multivariable
models did not change.
Although WHO category I patients comprised the majority
(63%) of our cohort, we found that the association between

0.006

July 2009

Association of Treadmill Exercise Capacity and Abnormal Hemodynamics on Univariate and Multivariable Analysis

Circ Heart Fail

Table 3.

282

Shah et al

Exercise Treadmill Testing in PH

283

Figure 2. Differences in hemodynamic parameters in patients with exercise capacity 3, 3 to 6, and 6 metabolic equivalents.

exercise capacity applied to the other categories of PH (WHO

categories II to V). In this group of patients, each 1-MET
decrease in exercise capacity resulted in a univariate
HR1.51 (95% CI, 1.12 to 2.04; P0.007) and multivariable
HR1.86 (95% CI, 1.07 to 3.24; P0.027). However, this
association did not persist after adjusting for invasive hemodynamic variables (P0.12) due to the decreased number of
patients in this subgroup who had full hemodynamic data
(N156).
When we repeated our Cox proportional hazards analyzes
after excluding patients who were on treatment for PH at the

Figure 3. Locally weighted smoothed scatterplot of the relationship between exercise capacity and mortality rate.

time of referral, exercise capacity continued to predict mortality. On univariate analysis, each 1-MET decrease in exercise capacity was associated with HR1.23 (95% CI, 1.11 to
1.35; P0.0001) of death. On multivariable analysis, this
association persisted (HR, 1.24; 95% CI, 1.08 to 1.44;
P0.003). Adding invasive hemodynamic variables to the
multivariable model did not erase the association between
exercise capacity and mortality in this subgroup (HR, 1.18;
95% CI, 1.01 to 1.38; P0.04).

Discussion
In our large series of patients diagnosed with PH due to a
variety of etiologies, we found that reduced exercise capacity,
determined objectively by ETT, is associated with abnormal
hemodynamics and is a predictor death. These results applied
to our overall cohort and those with less severe symptoms
who were WHO functional classes III. Reduced exercise
capacity was also an independent predictor of death in the
more homogenous subgroup of patients with pulmonary
arterial hypertension (WHO category I PH) who are often
studied in clinical trials of PH-specific therapies. Excluding
patients on treatment for PH with calcium channel blockers,
endothelin antagonists, prostacyclins, or phosphodiesterase
inhibitors did not remove the association between exercise
capacity and death.
Importantly, adding invasive hemodynamic variables to
our multivariable models attenuated the association between
ETT exercise capacity and death in most of our analyses

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

284

Circ Heart Fail

July 2009

Figure 4. Kaplan-Meier survival curves for

patients with exercise capacity 3, 3 to 6, and
6 metabolic equivalents.

except in the subgroup of patients with WHO category I PH

(pulmonary arterial hypertension). These results show that
ETT exercise capacity is an especially powerful predictor in
WHO category I patients who have a more homogenous
underlying pathophysiology. We hypothesize that in the other
subgroups, including functional classes III patients and
non-WHO category I patients, adding invasive hemodynamics to the multivariable models most likely attenuated the
association between exercise capacity and death because of
the decreased statistical power of these subgroups (due to
smaller sample size), and because both ETT exercise capacity
and invasive hemodynamic testing evaluate the same underlying phenomenon: pulmonary vascular disease and its effect
on right ventricular function. Regardless of the effect of
adding invasive hemodynamics to our multivariable models,
it is clear that ETT exercise capacity is a strong noninvasive
predictor of abnormal hemodynamics and mortality.
Table 4.

The Naughton-Balke ETT protocol has several advantages.

It is well-tolerated (no patients had adverse events during or
shortly after testing in our study, indicating that the
Naughton-Balke protocol is safe, even in WHO functional
class IV patients), is more standardized than the 6-MWT, and
less expensive and more widely available than cardiopulmonary exercise testing. In one of the original descriptions of the
Naughton-Balke protocol, investigators found that exercise
capacity determined by this test held prognostic significance
regardless of functional class in patients with cardiac disease.10 We have extended this finding to patients with PH by
showing that ETT exercise capacity predicted abnormal
hemodynamics and death even in the subgroup of PH patients
who were less sick (WHO functional classes III). Although
reproducible, inexpensive, and easily performed, the 6-MWT
has several limitations in that it is effort-dependent and
susceptible to motivational factors. The use of the 6-MWT as

Hazard Ratios for Treadmill Exercise Capacity as a Predictor of Mortality

WHO Functional Classes III*
(N138)

Total Cohort (N603)

Predictor

WHO Category I PAH* (N380)

Hazard Ratio (95% CI)

Hazard Ratio (95% CI)

Hazard Ratio (95% CI)

Unadjusted

1.24 (1.14 to 1.35)

0.0001

1.22 (1.04 to 1.44)

0.015

1.22 (1.12 to 1.34)

0.0001

Multivariable Model 1

1.18 (1.01 to 1.37)

0.031

1.53 (1.04 to 2.26)

0.032

1.18 (1.01 to 1.38)

0.035

Multivariable Model 2

1.17 (1.00 to 1.37)

0.052

1.12 (0.71 to 1.75)

0.63

1.20 (1.02 to 1.41)

0.032

Treadmill exercise capacity

(per each 1-MET decrease)

WHO indicates World Health Organization; MET, metabolic equivalent; PAH, pulmonary arterial hypertension.
*In the WHO functional classes III subgroup, 33 deaths occurred during the follow-up; in the WHO category I PAH subgroup, 182 deaths occurred during the
follow-up.
Model 1: backward selection multivariable analysis using the following covariates: age, etiology of pulmonary hypertension, systemic hypertension, diabetes
mellitus, obesity, interstitial lung disease, diuretic and phosphodiesterase inhibitor use, WHO functional class, albumin, creatinine, antinuclear antibody test, forced
vital capacity, and diffusing capacity of carbon monoxide.
Model 2: backward selection multivariable analysis using all covariates listed for Model 1 plus hemodynamic variables (right atrial pressure, mean pulmonary artery
pressure, and cardiac index).
For Model 2, N566 for the total cohort, N122 for WHO functional classes III, and N366 for WHO category I PAH because hemodynamic data was not
available in all study patients.

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

Shah et al
a measure of exercise capacity for left heart failure is no
longer considered adequate. A recent systematic review of
randomized controlled trials that used 6-MWT between 1988
and 2004 in heart failure demonstrated that the 6-MWT
poorly discriminated pharmacological treatment effects, and
had little prognostic and discriminative value in patients
without advanced disease.6 In addition, the most recent
American Heart Association guidelines on exercise stress
testing did not advocate time-based walk tests (such as the
6-MWT) and instead advocated ETT and cardiopulmonary
exercise stress testing.3 As more therapies become available
for PH, and patients get treated earlier in their disease course,
the ETT may be better at risk stratification in patients with
higher functional capacity. Because one of the difficulties in
managing patients with PH is the challenge in identifying
at-risk patients early in the course of disease, it would be very
helpful to have an objective test, such as low-intensity ETT,
which could noninvasively predict abnormal hemodynamics
and increased risk for death in patients with mild symptoms.
Since its inception, the Naughton-Balke ETT protocol has
served as a reliable, objective, noninvasive test, which has
been validated in the cardiac catheterization laboratory as a
predictor of oxygen consumption during exercise. In addition
to a possible role in the evaluation of patients with PH, ETT
could be used after the diagnosis of PH has been confirmed
by invasive hemodynamic testing to evaluate response to
therapies and for monitoring of disease progression and/or
remission, although the use of ETT in this manner requires
further study. With the continued development of PH-specific
therapies, the need for a noninvasive objective assessment to
follow an individual patients response to therapy has become
more critical, and the Naughton-Balke ETT may be just the
right tool.
There are several limitations to consider when interpreting
our results. Our data collection started in early 2004, so most
patients were studied retrospectively. Patients did not systematically undergo 6-MWT, so we are unable to directly
compare 6-MWT to Naughton-Balke ETT. We do not have
data on serial measurements of exercise capacity of ETT, so
we cannot determine whether changes in ETT predict outcome. In addition, we did not have data on blood pressure,
heart rate, and oxygen saturation response to exercise, and
therefore cannot determine the prognostic implications of
these variables. Other limitations include a study population
comprised of patients cared for by a single tertiary referral
practice, and the fact that ETT is not a direct measure of VO2
and thus only an approximation of METs. However, the
practical nature of ETT and its ability to closely correlate with
direct measurement of VO2 allows the test to be easily
administered with good reliability and reproducibility.9,10
Finally, the conversion formula for exercise time to METs
used in our study has not been validated in PH. However,
because METs in our study were derived directly from
exercise time, we have effectively shown that decreased
exercise time alone is an important clinical predictor of
adverse outcomes.
In summary, we found that reduced exercise capacity,
determined objectively by Naughton-Balke ETT and measured in METs, is a powerful predictor of abnormal hemo-

Exercise Treadmill Testing in PH

285

dynamics and death in PH. Importantly, ETT was able to

noninvasively risk stratify patients with PH who are less sick
(WHO functional classes III), the very patients who are
more likely to respond to vasodilator challenge, and in whom
early treatment may be beneficial. Although 6-MWT has
played an important role in the development of therapeutics
for pulmonary arterial hypertension, alternative measures of
therapeutic response (especially those already validated in
other disease settings) appropriate to the study population
should improve investigators ability to detect early signals of
efficacy and better define the magnitude of benefit specific to
cardiopulmonary function. ETT is a simple, widely available,
and inexpensive modality for the objective determination of
exercise capacity. Future clinical trials in PH should
include the ETT as a surrogate end point to prospectively
validate its use.

Sources of Funding
This work was supported by a Heart Failure Society of America
Research Fellowship Award, an Actelion Entilligence Young
Investigator Award, a Northwestern Memorial Foundation Dixon
Translational Award, an American Heart Association Scientist Development grant 0835488N (all to S.J.S.), a Doris Duke Clinical
Scientist Development Award (to M.G.M.), the Canadian Institutes
for Health Research, and National Institutes of Health grant
HL071115 (to S.L.A.).

Disclosures
Dr Gomberg-Maitland has received research grant support from
Actelion, Co-Therix, Encysive, Gilead, Lilly/Icos, Pfizer, and United
Therapeutics and has served as a consultant and/or on advisory
boards for Alkermes, Biomarin, Encysive, Gilead, Glaxo SmithKline, Medtronic, Pfizer, and United Therapeutics. She has a
pending patent entitled, Compositions and Methods for Treating
Pulmonary Hypertension, Gomberg-Maitland et al, WO/2007/
087575. Dr Rich previously served as a part-time salaried employee
at United Therapeutics, which concluded in January 2007. Dr Archer
holds a provisional patent for the use of mitochondrial modulators
for the treatment of cancer.

References
1. Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds LH Jr, Wilson JR.
Value of peak exercise oxygen consumption for optimal timing of cardiac
transplantation in ambulatory patients with heart failure. Circulation.
1991;83:778 786.
2. Rubin LJ. Diagnosis and management of pulmonary arterial hypertension:
ACCP evidence-based clinical practice guidelines. Chest. 2004;126:
7S10S.
3. Arena R, Myers J, Williams MA, Gulati M, Kligfield P, Balady GJ,
Collins E, Fletcher G. Assessment of functional capacity in clinical and
research settings: a scientific statement from the American Heart Association Committee on Exercise, Rehabilitation, and Prevention of the
Council on Clinical Cardiology and the Council on Cardiovascular
Nursing. Circulation. 2007;116:329 343.
4. Frost AE, Langleben D, Oudiz R, Hill N, Horn E, McLaughlin V,
Robbins IM, Shapiro S, Tapson VF, Zwicke D, DeMarco T, Schilz R,
Rubenfire M, Barst RJ. The 6-min walk test (6MW) as an efficacy
endpoint in pulmonary arterial hypertension clinical trials: demonstration
of a ceiling effect. Vascul Pharmacol. 2005;43:36 39.
5. Hoeper MM, Oudiz RJ, Peacock A, Tapson VF, Haworth SG, Frost AE,
Torbicki A. End points and clinical trial designs in pulmonary arterial
hypertension: clinical and regulatory perspectives. J Am Coll Cardiol.
2004;43:48S55S.
6. Olsson LG, Swedberg K, Clark AL, Witte KK, Cleland JG. Six minute
corridor walk test as an outcome measure for the assessment of treatment
in randomized, blinded intervention trials of chronic heart failure: a
systematic review. Eur Heart J. 2005;26:778 793.

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

286

Circ Heart Fail

July 2009

7. Oudiz RJ, Barst RJ, Hansen JE, Sun XG, Garofano R, Wu X, Wasserman
K. Cardiopulmonary exercise testing and six-minute walk correlations in
pulmonary arterial hypertension. Am J Cardiol. 2006;97:123126.
8. Groepenhoff H, Vonk-Noordegraaf A, Boonstra A, Spreeuwenberg MD,
Postmus PE, Bogaard HJ. Exercise testing to estimate survival in pulmonary hypertension. Med Sci Sports Exerc. 2008;40:17251732.
9. Gomberg-Maitland M, Huo D, Benza RL, McLaughlin VV, Tapson VF,
Barst RJ. Creation of a model comparing 6-minute walk test to metabolic
equivalent in evaluating treatment effects in pulmonary arterial hypertension. J Heart Lung Transplant. 2007;26:732738.
10. Patterson JA, Naughton J, Pietras RJ, Gunnar RM. Treadmill exercise in
assessment of the functional capacity of patients with cardiac disease.
Am J Cardiol. 1972;30:757762.
11. Macchia A, Marchioli R, Marfisi R, Scarano M, Levantesi G, Tavazzi L,
Tognoni G. A meta-analysis of trials of pulmonary hypertension: a
clinical condition looking for drugs and research methodology. Am
Heart J. 2007;153:10371047.
12. Farber HW. The status of pulmonary arterial hypertension in 2008.
Circulation. 2008;117:2966 2968.

13. Thenappan T, Shah SJ, Rich S, Gomberg-Maitland M. A USA-based

registry for pulmonary arterial hypertension: 19822006. Eur Respir J.
2007;30:11031110.
14. Shah SJ, Thenappan T, Rich S, Tian L, Archer SL, Gomberg-Maitland M.
Association of serum creatinine with abnormal hemodynamics and mortality
in pulmonary arterial hypertension. Circulation. 2008;117:24752483.
15. Gulati M, Black HR, Shaw LJ, Arnsdorf MF, Merz CN, Lauer MS,
Marwick TH, Pandey DK, Wicklund RH, Thisted RA. The prognostic
value of a nomogram for exercise capacity in women. N Engl J Med.
2005;353:468 475.
16. Morris CK, Myers J, Froelicher VF, Kawaguchi T, Ueshima K, Hideg A.
Nomogram based on metabolic equivalents and age for assessing aerobic
exercise capacity in men. J Am Coll Cardiol. 1993;22:175182.
17. McLaughlin VV, Presberg KW, Doyle RL, Abman SH, McCrory DC, Fortin
T, Ahearn G. Prognosis of pulmonary arterial hypertension: ACCP
evidence-based clinical practice guidelines. Chest. 2004;126:78S92S.
18. DAlonzo GE, Barst RJ, Ayres SM, Bergofsky EH, Brundage BH, Detre
KM, Fishman AP, Goldring RM, Groves BM, Kernis JT. Survival in
patients with primary pulmonary hypertension. Results from a national
prospective registry. Ann Intern Med. 1991;115:343349.

CLINICAL PERSPECTIVE
In patients with pulmonary hypertension (PH), World Health Organization functional class and the 6-minute walk test are
noninvasive tools that are used to define clinical severity. However, these tests may suffer from their subjective nature, and
the 6-minute walk test may not be able to discriminate functional capacity in World Health Organization classes III
patients. The ability of the Naughton-Balke exercise treadmill test (ETT), an objective indicator of exercise capacity, to
predict abnormal hemodynamics and mortality in PH is unknown. In our cohort study of 603 patients with PH, we studied
the use of ETT as a predictor of abnormal hemodynamics and death. Decreased exercise capacity was independently
associated with elevated right atrial and mean pulmonary artery pressure, decreased cardiac index, and increased
pulmonary vascular resistance. Decreased exercise capacity was associated with mortality in the study cohort overall, the
subgroup of patients who were functional classes III, and the subgroup of patients who had World Health Organization
category I PH (pulmonary arterial hypertension). Although the 6-minute walk test has played an important role in the
development of therapeutics for pulmonary arterial hypertension, ETT may improve the ability to detect early signals of
efficacy and better define the magnitude of benefit specific to cardiopulmonary function. ETT is a simple, widely available,
and inexpensive modality for the objective determination of exercise capacity. Future clinical trials in PH should include
the ETT as a surrogate end point to prospectively validate its use.

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

Value of Exercise Treadmill Testing in the Risk Stratification of Patients With Pulmonary
Hypertension
Sanjiv J. Shah, Thenappan Thenappan, Stuart Rich, James Sur, Stephen L. Archer and Mardi
Gomberg-Maitland
Circ Heart Fail. 2009;2:278-286; originally published online May 13, 2009;
doi: 10.1161/CIRCHEARTFAILURE.108.807826
Circulation: Heart Failure is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX
75231
Copyright 2009 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-3289. Online ISSN: 1941-3297

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circheartfailure.ahajournals.org/content/2/4/278

Data Supplement (unedited) at:

http://circheartfailure.ahajournals.org/content/suppl/2009/05/13/CIRCHEARTFAILURE.108.807826.DC1.html

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Heart Failure can be obtained via RightsLink, a service of the Copyright Clearance Center, not
the Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Heart Failure is online at:
http://circheartfailure.ahajournals.org//subscriptions/

Downloaded from http://circheartfailure.ahajournals.org/ by guest on March 7, 2016

Supplementary Material

Table S1: Baseline Demographics, Clinical, Laboratory, Echocardiography, Hemodynamic Characteristics by Type of Missing
Data

Characteristics

Missing Data
None

Age, years
Female, n (%)

Missing

Missing Exercise

Exercise

Capacity and

Capacity Only

Functional Class

(N = 603)

(N = 309)

(N = 291)

5015

5215

5616*

454 (76)

240 (78)

210 (72)

380 (63)

245 (79)

9 (3)

81 (13)

35 (11)

111 (38)

Etiology of pulmonary hypertension, n (%)

WHO Category I (pulmonary arterial

hypertension)

WHO Category II (pulmonary venous

hypertension)

WHO Category III (pulmonary hypertension

25 (4)

6 (2)

29 (10)

40 (7)

5 (2)

38 (13)

77 (13)

19 (6)

102 (35)

associated with hypoxemia)

WHO Category IV (pulmonary hypertension due

to chronic thromboembolic disease)

WHO Category V (miscellaneous)

Functional class, n (%)

45 (7)

12 (4)

II

93 (15)

35 (11)

III

286 (47)

175 (57)

IV

179 (30)

97 (28)

Medications, n (%)

Diuretics

292 (49)

167 (54)

166 (57)*

Calcium channel blockers

193 (32)

87 (28)

70 (24)*

Digoxin

100 (17)

50 (16)

67 (23)*

Warfarin

204 (34)

84 (27)*

90 (31)

Prostacyclins

11 (2)

6 (2)

2 (0.7)

Endothelin blockers

21 (4)

14 (5)

2 (0.7)*

Phosphodiesterase inhibitors

10 (2)

10 (3)

6 (2)

203 (34)

112 (36)

116 (40)

49 (8)

43 (14)

55 (19)*

102 (17)

70 (23)

58 (20)

Co-morbidities, n (%)

Systemic hypertension

Diabetes mellitus

Obesity

Coronary artery disease

38 (6)

21 (7)

47 (16)*

Interstitial lung disease

73 (12)

50 (16)

29 (10)

Pulmonary embolism

135 (22)

47 (15)*

55 (19)

Laboratory tests

Serum creatinine, mg/dL (N=1053)

1.10.7

1.21.0*

1.20.8*

Albumin, g/dL (N=979)

3.90.5

3.70.7*

3.70.6*

Positive anti-nuclear antibody test, n (%) (N=730)

215 (50)

79 (43)

78 (67)*

Pulmonary function tests

FEV1, % predicted (N=1062)

7420

7220

6824*

Forced vital capacity, % predicted (N=960)

7619

7419

7023*

FEV1/FVC, % (N=650)

9014

8816

9117

Total lung capacity, % predicted (N=873)

8719

8421

8325*

DLCO, % predicted (N=841)

5723

5123

5723

9 (6-14)

11 (6-17)

10 (7-16)

Hemodynamics (N=1055)

Mean right atrial pressure, mm Hg, median

(interquartile range)

Mean pulmonary artery pressure, mm Hg

4815

5015

4217*

Cardiac index, L/min/m2

2.40.9

2.31.0*

2.60.9*

Pulmonary vascular resistance, Wood units

107

128*

86*

Pulmonary capillary wedge pressure, mm Hg

116

116

158*

Mean aortic pressure, mm Hg

9516

9214*

9517

Pulmonary artery oxygen saturation, %

6011

5713*

6012

* p<0.05; p<0.05 across all groups