Journal of Cardiac Failure Vol. 14 No. 1 2008

Methods

Percent Predicted Value for the 6-Minute Walk Test:Using Norm-Referenced Equations to Characterize

Severity in Persons With CHF

KONSTANTIN BALASHOV, BSc,1 DEBBIE EHRMANN FELDMAN, PhD,1,2 SYLVIE SAVARD, MPs,3

STEFANIE HOUDE, MD, MSc,8 MARC FRENETTE, MD,4 ANIQUE DUCHARME, MD, MSc,5 NADIA GIANNETTI, MD,6

CAROLINE MICHEL, MD,7 AND LOUISE PILOTE, MD, PhD3

Montreal, Canada; British Columbia, Canada

ABSTRACT

Background: The 6-minute walk test (6MWT) is a widely used measure of functional capacity in patientswith chronic heart failure (CHF). Norm-referenced equations that predict the 6-minute walk distance(6MWD) according to age, height, weight, and gender have been proposed for healthy patients. Weexplored whether these equations apply to CHF patients.Methods and Results: The sample consisted of 213 patients newly admitted to specialized CHF clinics inMontreal, Canada. Percent predicted value (PPV) for 6MWD was calculated using norm-referenced equa-tions. We explored correlations between different measures: PPV, 6MWD, Minnesota Living with Heart Fail-ure Quality of Life score (MLHF-QOL). We compared severity among different age, gender, and BMI (bodymass index) subgroups and assessed consistency using different measures of severity. Mean age was 65.5years and 77.5% were men. Compared with the 6MWD, PPV had a slightly better correlation withMLHF-QOL score (�0.26 versus�0.20), and slightly more predictive power in linear regressions (adjustedr2 5 6.5% versus 4.2%). When PPV was used to differentiate severity between different age, gender, and BMIsubgroups, it consistently led to similar conclusions as the MLHF-QOL score, unlike 6MWD.Conclusion: The 6MWD in meters may give misleading results when used as an indicator of severity ofCHF condition to compare groups with different sex, age, and BMI distributions. It may be necessary tostandardize it using norm-referenced equations. (J Cardiac Fail 2008;14:75e81)Key Words: The 6-minute walk test, functional capacity, norms.

From the 1University of Montreal, School of Rehabilitation; 2Centre deRecherche Interdisciplinaire en Readaptation; 3Division of Clinical Epide-miology, McGill University Health Centre; 4Hopital Sacre-Cœur de Mon-treal; 5Montreal Heart Institute, University of Montreal; 6Royal VictoriaHospital, McGill University Health Centre; 7SMBD Jewish General Hospi-tal, Montreal, Canada; and 8Kootenay Boundary Regional Hospital, Trail,British Columbia, Canada.

Manuscript received April 13, 2007; revised manuscript received July15, 2007; revised manuscript accepted September 10, 2007.

Reprint requests: Dr. Debbie Ehrmann Feldman, Universite de Montreal,Ecole de readaptation, C.P. 6128, Succ. Centre-Ville, Montreal, Quebec,Canada H3C 3J7.

Supported by a grant from the Canadian Institutes of Health Research(CIHR).

Mr. Balashov was the recipient of a research bursary from University ofMontreal. Dr. Feldman holds a new investigator career award from The Ar-thritis Society. Dr. Pilote holds an investigator career award from CIHR.Dr. Ducharme is supported by ‘‘Fond de la Recherche en Sante duQuebec’’ (FRSQ).

1071-9164/$ - see front matter� 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.cardfail.2007.09.005

75

It is necessary to use functional testing for patients withchronic heart failure (CHF) to estimate their prognosis andplan effective treatment strategies. Therefore a test that is ac-curate, reproducible, safe, and inexpensive is required. Themaximal peak oxygen consumption test meets some of thesecriteria, but has several important limitations.1,2 First, mostpatients with CHF lead a sedentary lifestyle and thereforemeasurement of peak oxygen consumption may not reflecttheir true functional limitation during daily activities. Sec-ond, a significant proportion of these patients cannotperform the test because of comorbidities or technical prob-lems. For example, some patients are unable to tolerate themouthpiece of the gas analyzer. Finally, this test requiresspecialized equipment and expertise and thus is not availablefor the assessment of a vast proportion of CHF patients.2

The 6-minute walk test (6MWT) can be performed bymost patients, without the need for specialized equipment.2

The 6MWT was first used by Guyatt et al in 19853 and

76 Journal of Cardiac Failure Vol. 14 No. 1 February 2008

subsequently has gained widespread acceptance.2 The pa-tient is instructed to walk for 6 minutes at his or her ownpace while attempting to cover as much ground as possibleand being allowed to stop and resume as needed. The test iscarried out in a corridor at least 20 meters long, free fromobstacles, under supervision, and with standardized encour-agements every minute, according to American ThoracicSociety guidelines.4 After 6 minutes, the total distance cov-ered is measured. This test evaluates submaximal exercisecapacity and is usually well tolerated and accepted byCHF patients.3,5,6

Several studies have established a good correlation be-tween the 6-minute walk distance (6MWD) and maximaloxygen uptake.5,7e9 Other studies have suggested even bet-ter correlations between the 6MWD and the New YorkHeart Association (NYHA) functional class, the left ven-tricular ejection fraction (LVEF) and health-related qualityof life score measured by the Minnesota Living with HeartFailure (MLHF) questionnaire.6,8,10,11 All these parameterscould be used to evaluate the severity of a cardiorespiratorycondition such as CHF. Though widely employed, the6MWD may not be the best indicator of disease severity be-cause it is largely influenced by age, sex, height, andweight.12,13 Enright et al14,15 and Troosters et al13 have pro-posed reference equations that predict the 6MWD as a func-tion of the latter variables. By using these norm-referencedvalues, a percentage of the predicted value (PPV) for the6MWD is calculated as the quotient of the measured6MWD and the norm means.

The objective of this study was to verify the usefulnessof 6MWD as an indicator of CHF severity by determiningwhether norm-referenced equations apply to CHF patientsand if they could be used to construct an indicator ofseveritydPPVdthat would be superior to the 6MWDmeasured in meters.

Methods

Study Population

This research was carried out as a subanalysis of a large longi-tudinal cohort study (Access-Clinic) of patients newly referred tospecialized CHF clinics. The aim of the study was to determinewhether there were gender differences in access, management,and outcomes for patients newly enrolled in specialized CHFclinics. Patients were recruited by research nurses for the longitu-dinal cohort study in 5 different centers in Montreal, Canada, be-tween April 2004 and March 2007. Each of these clinics usesa common software program to record clinical information onpatients. We included patients older than age 40 who were newlyadmitted to the multidisciplinary CHF clinic and who could speakand understand English or French. All patients provided informedconsent. The study was approved by the ethics committees of thevarious institutions involved. The present analysis refers to thoseenrolled in the first year of the study.

Data Collection

Clinical characteristics of 213 patients were extracted from thecommon software program/patient database entitled (‘‘Vison

Cardiologie,’’ Montreal, Canada) used by the 5 centers that partic-ipated in the study. These data included LVEF, NYHA functionalclass, comorbidities (such as diabetes, pulmonary disease), smok-ing history, number of hospital admissions and emergency roomvisits over the study period, and medications prescribed. Patientsunderwent the 6MWT supervised by a trained nurse accordingto a standardized protocol. Patients were also asked to completethe MLHF quality of life (QOL) questionnaire.

Measures

Disease severity at admission into a CHF clinic is evaluated byseveral generally accepted parameters. These include (1) theLVEF; (2) the NYHA functional classification, which is a validmeasurement of functional status, but its reproducibility has notbeen established16e18; (3) the MLHF-QOL19e21; and (4) the6MWT.11,22,23

Using the reference equation developed by Troosters13 for the6MWT, we derived a new measured called the PPV. PPV is com-puted by dividing the actual 6MWD by the expected value of6MWD and then multiplying by 100. Troosters’ equation is as fol-lows: Predicted 6MWD 5 218 þ 5.14 * height (cm) � 5.32 * age(years) �1.8 * weight (kg) þ 51.31 * sex (1emale, 0efemale).We used this equation instead of Enrights’ because of similaritywith our population, age-wise.14,15

In addition, we collected data on the number of comorbidities,body mass index (BMI, computed using height and weight data inthe patient’s file), because these measures could strongly influencethe patient’s functional capacity. The number of comorbidities isthe list of associated conditions used in the ‘‘Vision Cardiologie’’database, which includes: diabetes I, diabetes IIdnoneinsulin-dependent, diabetes IIdinsulin-dependent, dyslipidemia, gout,hypothyroidism, hyperthyroidism, chronic renal failure, highblood pressure, anemia, chronic obstructive pulmonary disease,asthma, smoking, vascular disease (peripheral vascular disease),cardiovascular accident-transient ischemic attack, alcoholismand obesity, and other.

Choosing the CHF Severity Benchmark

By using PPV, we assumed that the variables for which we wereadjusting (sex, height, weight, age) had similar influence on the6MWD in our CHF population as they did in the general popula-tion. However, this may not be the case. Thus, to determinewhether the reference equations applied to the CHF population,we sought to compare the 6MWD results in our CHF study pop-ulation among subgroups with equivalent disease severity. Weused several severity indicators including 6MWD, PPV, MLHF-QOL score, NYHA functional class, and LVEF. The NYHA clas-sification may be less sensitive than the MLHF-QOL score.11 Inaddition, because of the small proportion of patients in Classes I(11.7%) and IV (1.8%), grouping Classes I and II and III andIV was necessary for statistical analysis, further decreasing its sen-sitivity. In addition, LVEF is not a reliable indicator of functionallimitation, exercise capacity, or QOL score in CHF patients.24,25

Consequently, the MLHF-QOL score, with proven reliability andvalidity,19,20 was selected as the best approximation of severityfor the purposes of this study and served as our main severitybenchmark.

Statistical Analysis

To help decide whether PPV is a better indicator of severity than6MWD, we performed the following analyses: (1) we conducted

Percent Predicted Value for the 6-Minute Walk Test � Balashov et al 77

correlations with other severity measures to determine whetherPPV correlated better with these measures than 6MWD, (2) we as-sessed explanatory power in regression models to determinewhether PPV explained more of the variance for QOL, and (3)we evaluated consistency of results when testing for differencesbetween sex, age, gender, and BMI subgroups. Given that thereare differences in female and male CHF populations26e28 andthe low number of women in our sample, we also conducted strat-ified analyses for gender. These analyses are described in furtherdetail in the following sections.

Correlations. We explored relationships between differentmeasures using correlation coefficients. Pearson’s correlationswere calculated for quantitative continuous variables: PPV,6MWD, MLHF-QOL score, LVEF, the number of comorbidities,and Spearman’s for categorical and discrete variables: NYHAand COPD. COPD was selected among the comorbidities because6MWD is often used as a measure of functional severity for thiscondition as well.29 Correlations were computed for the sampleas a whole and separately for men and women. We considered cor-relation coefficients of 0.6 to 0.8 as marked, 0.4 to 0.6 as moder-ate, and 0.2 to 0.4 as low.30

Regressions. Regression models were employed to determinewhich variables (PPV, 6MWD, or others such as height, weight,age, NYHA class, LVEF, number of comorbidities, presence ofCOPD) best explain differences in severity of CHF. The MLHF-QOL score was chosen as the dependent variable being the bestapproximation of severity for the purposes of the study. Again,we looked at men and women together and separately.

Mean comparison tests. We attempted to detect differencesin severity between different age, gender and BMI subgroups,using t-tests for continuous variables and chi-square tests forcategorical variables. We used various cutoffs to dichotomizeage and BMI. For age, we used 55, 65, and 75 years; for BMI,we used 25, 30, and 35. Again, MLHF-QOL score was used asthe main benchmark and we looked for consistency in conclusionsin tests with MLHF-QOL, PPV, 6MWD, and NYHA. Because oflow number of patients in NYHA functional Classes I and IV, wedichotomized the measure grouping Class I and Class II togetherversus Class III and Class IV, calling this dichotomized measure‘‘NYHAbin.’’

Data Quality

The parameters used for severity assessment were not recordedsimultaneously, and time differences could affect the correlationsperformed. For the majority, the spreads in terms of time of datacollection were acceptable except for the LVEF (Table 1).

Results

Of the 213 patients included in the present study, 77.5%were males and the average age was 65 years (SD 5 11).The majority (85%) displayed left ventricular systolic dys-function. There were no significant differences betweengenders with the exception that the 6MWD was lower inwomen (�42.3 meters, P 5 .01) (Table 2).

The statistical correlations were low for both PPV and6MWD and the following: NYHA class, MLHF-QOLscore, the number of comorbidities, and presence ofCOPD (Fig. 1, 2). The correlation of PPV with our bench-mark, the MLHF-QOL score, was slightly greater than that

of 6MWD (�0.26 versus �0.20 for the sample as a wholeand �0.27 versus �0.23 for men). Among males, theMLHF-QOL score had significant correlations with allother severity measures (except LVEF): NYHA (0.36,P ! .001), NYHAbin (0.33, P ! .001), comorbidities(0.20, P 5 .016), and COPD (0.24, P 5 .003). For menand women taken together, the results were similar to these,but for women alone, none of the correlations was statisti-cally significant. The LVEF was not significantly correlatedwith any measure of severity in our sample.

Several regression models were examined to determinewhich measure, PPV or 6MWD, explained more of the var-iance of the MLHF-QOL score (the dependent variable).For the sample as a whole, PPV alone explained 6.5% ofthe variance, slightly more than 6MWD alone (4.2%) andvery close to 6MWD with gender, height, age, and weightin the model (6.3%). We also conducted separate regressionanalyses for men and women. For men, PPV explained7.3% of variance, more than 6MWD alone (5.5%) or withheight, age, and weight in the model (5.9%) (Table 3). Ina stepwise multiple regression, PPV was selected over6MWD in both cases. The results were quite different forwomen: only age and BMI showed significant correlationswith MLHF-QOL score, together explaining 19.9% of thevariance.

When testing for differences in severity between gendersusing the various indicators: PPV, 6MWD, MLHF-QOL,and NYHAbin, only 6MWD showed a significant differ-ence (42 meters less for women, P 5 .01), whereas therewere no differences for the other severity indicators.

Regarding age, CHF functional limitation was more se-vere in elderly patients (O65 years of age) using 6MWD(mean difference of 46 meters, P 5 .001) and NYHAbin(P 5 .012); there were no significant differences betweenthe 2 age groups for either the MLHF-QOL score or PPV.When stratifying by gender, there was a significant differ-ence in severity using the 6MWD only, with older males be-ing more limited than younger males (mean difference of49 meters, P 5 .002). The latter results held true forPPV, 6MWD, and MLHF-QOL using age break points of55 and 75 years. There were no significant differences for

Table 1. Difference in Timing (in Days) Between DifferentMeasures and MLHF-QOL (Top Lines) and 6MWT

(Bottom Lines)

Difference in Days Between Mean MedianStandardDeviation

MLHF-QOL 6-minute walk test 3.4 0 17.6Left ventricular

ejection fraction�78.7 �28.0 189.5

New York Heart Associationfunctional class

3.9 0 12.4

6MWT Left ventricularejection fraction

�81.6 �30.0 189.5

New York Heart Associationfunctional class

2.8 0 9.5

MLHF-QOL, Minnesota Living with Heart Failure Quality of LifeScore; 6MWT, 6-minute walk test.

78 Journal of Cardiac Failure Vol. 14 No. 1 February 2008

Table 2. Characteristics of the Study Population

All Men Women P Value

Participants 213 (100%) 165 (77.5%) 48 (22.5%) ! .0001Mean age (SD) years 65.5 (11.2) 64.9 (11.1) 67.7 (10.8) .06New York Heart Association functional class .61

Class I 25 (11.7%) 17 (10.3%) 8 (16.6%)Class II 125 (58.7%) 100 (60.6%) 25 (52.1%)Class III 59 (27.7%) 45 (27.3%) 14 (29.2%)Class IV 4 (1.8%) 3 (1.8%) 1 (1.2%)

Mean LVEF (SD) 27.7% (12.7) 26.8% (12.4) 30.3% (13.4) .06LVEF !40% 182 (85.4%) 142 (86.1%) 40 (83.3%) .19Mean number of comorbidities (SD) 3.4 (2.0) 3.5 (2.0) 3.1 (1.8) .09COPD 31(14.5%) 24 (14.6%) 7 (14.5%) .99Smoker 65 (30.5%) 54 (32.7%) 11 (22.2%) .19Mean BMI (SD) 27.13 (5.1) 27.4 (4.7) 26.3 (5.9) .14Mean dM-MLHF-QOL (SD) 48.1 (24.6) 48.5 (25.0) 46.8 (23.5) .34Mean 6MWD in meters (SD) 295.7 (99.5) 305.0 (98.9) 262.7 (95.8) .007Mean percentage

predicted value of 6MWDdPPV (SD)46.4% (14.7) 45.8% (14.1) 48.5% (16.7) .17

LVEF, left ventricular ejection fraction; COPD, chronic obstructive pulmonary disease; BMI, body mass index; MLHF-QOL, Minnesota Living with HeartFailure Quality of Life Score; 6MWD, 6-minute walk distance; PPV, percent predicted value.

the female subgroup regarding age, but the numbers aresmall. Thus, using the 6MWD to compare severity, impliedhigher limitation in older men, whereas PPV and MLHF-QOL showed no differences between older and youngermen or women. We also compared disease severity amongoverweight (BMI O25), obese (BMI O30), severely obese(BMI O35), and nonobese patients. Regarding obesity,there was no significant difference in severity for any ofthe indicators for the sample as a whole and for maleswhen stratifying by gender. On the other hand, obesewomen had a higher MLHF-QOL score (indicative of lowerperceived health-related QOL) than nonobese women. Se-verely obese women had more severe CHF limitationthan non-severely obese, according to MLHF-QOL (P 5

Correlations (men and women)

-0.4-0.35

-0.3-0.25

-0.2-0.15

-0.1-0.05

0

PPV -0.255 -0.345 -0.3 -0.25 -0.225

6MWD -0.203 -0.362 -0.355 -0.233 -0.248

QOL NYHA NYHA bin Comorb COPD

Fig. 1. Shown, for both men and women, are the correlations be-tween both percent predicted value (PPV) (white boxes) and 6-minute-walk-distance (6MWD) (darkened boxes) with each ofthe following: Minnesota Living with Heart Failure Quality ofLife Score (MLHF-QOL), New York Heart Association functionalclass (NYHA), NYHA binary measure: NYHAbin, which is thebinary classification of NYHA Classes I and II versus ClassesIII and IV, number of comorbidities, and presence of chronic ob-structive pulmonary disease.

.002), PPV (P 5 .029), and 6MWD (P 5 .008), but notaccording to NYHAbin.

Discussion

We found significant but low correlations between eitherPPV or 6MWD with other known measures of severity, butPPV had a slightly better correlation with the MLHF-QOLscore. In general, correlations with the 6MWD in our sam-ple were somewhat lower compared with Demers’ et al,11

which could be explained by the lower proportion ofwomen in their sample (17%) versus ours (22.5%). Further-more, their study was conducted in the context of

Correlations (men)

-0.45-0.4

-0.35-0.3

-0.25-0.2

-0.15-0.1

-0.050

PPV -0.271 -0.399 -0.367 -0.297 -0.273

6MWD -0.228 -0.413 -0.395 -0.302 -0.29

QOL NYHA NYHA bin Comorb COPD

Fig. 2. Shown, for men only, are the correlations between bothpercent predicted value (PPV) (white boxes) and 6-minute-walk-distance (6MWD) (darkened boxes) with each of the following:Minnesota Living with Heart Failure Quality of Life Score(MLHF-QOL), New York Heart Association functional class(NYHA), NYHA binary measure: NYHAbin, which is the binaryclassification of Classes I and II versus Classes III and IV, numberof comorbidities, and presence of chronic obstructive pulmonarydisease.

Percent Predicted Value for the 6-Minute Walk Test � Balashov et al 79

Table 3. Comparing PPV and 6MWT Using Regression Models in Men

Regression Model Independent

Variable(s)

Estimate (95 CI ) P

QOL = PPV PPV −0.48 (−0.77, −0.19) 0.001 0.073

QOL = 6MWD 6MWD −0.06 (−0.10, −0.02) 0.006 0.055

QOL = 6MWD Height Age

Weight

6MWD

Height

Age

Weight

−0.08 (−0.12, −0.03)

0.48 (−0.22, 1.18)

−0.27 (−0.68, 0.14)

−0.13 (−0.41, 0.15)

0.001

0.184

0.192

0.368

0.059

QOL = PPV 6MWD Height Age

Weight BMI NYHAbin LVEF

Comorb COPD *

NYHAbin

PPV

12.97 (3.08, 22.7)

−0.34 (−0.64, -0.03)

<0.001

0.033

0.087*

0.118*

* Stepwise multiple regression at 0.1 significance. The r2 values are cumulative.

r2

(Adjusted)

PPV, percent predicted value; 6MWT, 6-minute walk test; QOL, quality of life; PPV, ; NYHAbin, dichotomized measure grouping New York Heart As-sociation Class I and Class II together versus Class III and Class IV; 6MWD, 6-minute walk distance; BMI, body mass index; LVEF, left ventricular ejectionfraction; COPD, chronic obstructive pulmonary disease; PPV, percent predicted value.

a randomized controlled trial (in which there is specific se-lection criteria including 6MWD !500 meters, LVEF!0.4), whereas our study was an observational cohortthat included all newly enrolled patients in the CHF clinic.

Regression analysis indicated that a slightly higher per-centage of the variance of MLHF-QOL score was explainedby PPV than by 6MWD or 6MWD and age, sex, height, andweight. This was especially evident when stratifying bygender. Nevertheless, the r2 values are low and the differ-ences are small; based on this criterion, we could concludethat PPV is at least just as good as 6MWD as a measure ofCHF severity. Confirming these results, regression modelsindicated that neither PPV nor 6MWD were strong predic-tors of MLHF-QOL score. Thus QOL and exercise capacityin CHF are only minimally related. This implies that instudies that measure CHF outcomes, both QOL and exer-cise capacity should be measured, because they refer to dif-ferent aspects of disease severity. Age and BMI hada significant effect on MLHF-QOL score in women, butnot in men, suggesting differences between males and fe-males with CHF on their perceived health-related QOL,as demonstrated in previous studies.27e29 Possibly, womenmay be more affected by both obesity and aging in terms ofhealth-related QOL31. It may be important to account fordifferences between men and women using an indicatorsuch as PPV that takes age, weight, and height into consid-eration.

The main advantage of using PPV rather than 6MWD isseen when comparing severity of CHF in specificsubgroups. It is consistent with other measures of severity,especially with MLHF-QOL score, when assessing differ-ences (or lack thereof) between different age or gender

subgroups. For example, when comparing severity ofCHF between women and men, the MLHF-QOL score,NYHA classification, and PPV show no differences,whereas the 6MWD is lower for women. Other authorsfound health-related QOL to be either lower in women32,33

or not different,34 depending on the indicator used. Whenstratifying by age between older ($55) and younger agegroups (!55), no difference was found in MLHF-QOLscore, NYHA functional class, and PPV, but, as expected,6MWD is significantly lower for the older group, andmay reflect musculoskeletal, pulmonary, and other limita-tions more prevalent in the elderly population. With the ex-ception of NYHA classification, the results are the samewhen comparing patients older and younger than 65 andolder and younger than 75. Thus PPV may be a better mea-sure of functional limitation and perceived health-relatedQOL in the aging population, being more consistent withthe other severity measures. Using 6MWD to compare se-verity among different age groups may lead to an erroneousconclusion.

We found higher severity of CHF in severely obesewomen (but not in severely obese men) based on PPV,6MWD, and MLHF-QOL. There were no differences inNYHA functional classifications (III and IV versus I andII) between any BMI groups, possibly because of lowersensitivity in measuring severity using the dichotomousclassification. In our study, BMI does not seem to affect6MWD; however, because of the small number of womenin our sample, we cannot be certain that this is true forthe entire CHF female population. Because previous find-ings confirm that obesity is a more important factor inwomen than in men with CHF,35 we would think that

80 Journal of Cardiac Failure Vol. 14 No. 1 February 2008

adjusting for height and weight by using PPV would bemore appropriate.

A limitation of the present study is its cross-sectional de-sign: we could not verify whether PPV is superior to the6MWD when evaluating severity of the same patient overtime. However, because PPV takes into account age andweight, which could potentially change over time, and be-cause we have found that these measures affect severity, itis conceivable to think that using PPV would give moremeaningful results as compared with the 6MWD. The smallsample size is a limitation. In particular, the relatively smallnumber of women in our sample may be the reason for thelack of statistically significant findings in women with re-spect to severity and age. Also, we had few patients thatwere either very mild or very severe (the majority couldbe deemed ‘‘moderate’’), making it difficult to generalizeour results to those with mild disease or end-stage disease.Another limitation is that, although we adjusted for thenumber of comorbid conditions, there are certain specificcomorbid conditions that may influence 6MWD (eg, arthri-tis, neurologic conditions). Because we had no indication ofseverity of these comorbid conditions, we opted to adjust forthe total number of conditions but also included COPD be-cause this condition is known to highly influence 6MWD.

Conclusion

The reference equation developed by Troosters13 for pre-dicting the 6MWD in healthy elderly subjects is suitable fora CHF patient population. Compared with the 6MWD, PPVcomputed with the help of Troosters’ equation has a slightlybetter, albeit low correlation with the MLHF-QOL score,similar correlations with other indicators of severity, andslightly more predictive power in linear regressions. Mostimportant, the 6MWD in meters may give misleading re-sults when used as an indicator of severity of CHF condi-tion to compare groups with different sex, age, and BMIdistributions. Alternatively, using PPV consistently leadsto similar conclusions as employing the MLHF-QOL scorewhen comparing CHF severity levels. For this reason, webelieve that it is advisable to standardize 6MWD usingnorm-referenced equations.

References

1. Mancini D, LeJemtel T, Aaronson K. Peak VO(2): a simple yet endur-

ing standard. Circulation 2000;101:1080e2.

2. Sharma R, Anker SD. The 6-minute walk test and prognosis in chronic

heart failuredthe available evidence. Eur Heart J 2001;22:445e8.

3. Guyatt G, Sullivan M, Thompson P, Fallen E, Pugsley S, Taylor D, et al.

The 6-minute walk: a new measure of exercise capacity in patients with

chronic heart failure. Can Med Assoc J 1985;132:919e23.

4. ATS statement: guidelines for the six-minute walk test. Am J Respir

Crit Care Med 2002;166:111e7.

5. Morales FJ, Martinez A, Mendez M, Agarrado A, Ortega F, Fernandez-

Guerra J, et al. A shuttle walk test for assessment of functional capacity

in chronic heart failure. Am Heart J 1999;138:291e8.

6. Bittner V, Weiner DH, Yusuf S, Rogers WJ, McIntyre KM,

Bangdiwala SI, et al. Prediction of mortality and morbidity with

a 6-minute walk test in patients with left ventricular dysfunction.

SOLVD Investigators. JAMA 1993;270:1702e7.

7. Hendrican MC, McKelvie RS, Smith T, McCartney N, Pogue J,

Teo KK, et al. Functional capacity in patients with congestive heart

failure. J Card Failure 2000;6:214e9.

8. Miyamoto S, Nagaya N, Satoh T, Kyotani S, Sakamaki F, Fujita M,

et al. Clinical correlates and pronostic significance of six-minute

walk test in patients with pulmonary hypertension. Am J Respir Crit

Care Med 2000;161:487e92.

9. Cahalin LP, Mathier MA, Semigran MJ, Dec GW, DiSalvo TG. The

six-minute walk test predicts peak oxygen uptake and survival in

patients with advanced heart failure. Chest 1996;110:325e32.

10. Shah M, Hasselblad V, Gheorghiade M, Adams KF Jr, Swedberg K,

Califf RM, et al. Prognostic usefulness of the six-minute walk in

patients with advanced congestive heart failure secondary to ische-

mic or nonischemic cardiomyopathy. Am J Cardiol 2001;88:

987e93.

11. Demers C, McKelvie RS, Negassa A, Yusuf S. Reliability, validity,

and responsiveness of the six-minute walk test in patients with heart

failure. Am Heart J 2001;142:698e703.

12. Teramoto S, Ohga E, Ishii T, Yamaguchi Y, Yamamoto H, Mastuse T.

Reference value of six-minute walking distance in healthy middle-

aged and older subjects. Eur Respir J 2000;15:1132e3.

13. Troosters T, Gosselink R, Decramer M. Six minute walking distance in

healthy elderly subjects. Eur Respir J 1999;14:270e4.

14. Enright PL, Sherrill DL. Reference equations for the six-minute

walk in healthy adults. Am J Respir Crit Care Med 1998;158:

1384e7.

15. Enright PL. The six-minute walk test. Respir Care 2003;48:783e5.

16. Bennett JA, Riegel B, Bittner V, Nichols J. Validity and reliability of

the NYHA classes for measuring research outcomes in patients with

cardiac disease. Heart Lung 2002;31:262e70.

17. Dunselman PH, Kuntze CE, van Bruggen A, Beekhuis H, Piers B,

Scaf AH, et al. Value of New York Heart Association classification,

radionuclide ventriculography, and cardiopulmonary exercise tests

for selection of patients for congestive heart failure studies. Am Heart

J 1988;116:1475e82.

18. Gibelin P, Poncelet P, Gallois H, Sebaoun A, Avierinos C. [Evaluation

of 3 functional classifications of cardiac insufficiency: a national mul-

ticenter study. National College of French Cardiologists]. Ann Cardiol

Angeiol (Paris) 1995;44:304e9.

19. Rector TS, Kubo SH, Cohn JN. Validity of the Minnesota Living with

Heart Failure questionnaire as a measure of therapeutic response to

enalapril or placebo. Am J Cardiol 1993;71:106e7.

20. Rector TS, Cohn JN. Assessment of patient outcome with the Minne-

sota Living with Heart Failure questionnaire: reliability and validity

during a randomized, double-blind, placebo-controlled trial of pimo-

bendan. Pimobendan Multicenter Research Group. Am Heart J

1992;124:1017e25.

21. Lewis EF, Lamas GA, O’Meara E, Granger CB, Dunlap ME,

McKelvie RS, et al. Characterization of health-related quality of life

in heart failure patients with preserved versus low ejection fraction

in HARM. Eur J Heart Failure 2007;9:83e91.

22. Kervio G, Carre F, Ville NS. Reliability and intensity of the six-minute

walk test in healthy elderly subjects. Med Sci Sports Exerc 2003;35:

169e74.

23. Domenico PG, Opasich C, Mazza A, Tangenti A, Maestri R,

Sanarico M. Reproducibility of the six-minute walking test in chronic

heart failure patients. Stat Med 2000;19:3087e94.

24. Franciosa JA, Park M, Levine TB. Lack of correlation between exer-

cise capacity and indexes of resting left ventricular performance in

heart failure. Am J Cardiol 1981;47:33e9.

25. Mitani H, Hashimoto H, Isshiki T, Kurokawa S, Ogawa K,

Matsumoto K, et al. Health-related quality of life of Japanese patients

with chronic heart failure: assessment using the Medical Outcome

Study Short Form 36. Circ J 2003;67:215e20.

Percent Predicted Value for the 6-Minute Walk Test � Balashov et al 81

26. Hussey LC, Hardin S. Comparison of characteristics of heart failure by

race and gender. Dimens Crit Care Nurs 2005;24:41e6.

27. Mitchell J, Taylor A. Congestive heart failure in women. J Fam Pract

2005;24(Suppl):6e7.

28. Altimir S, Lupon J, Gonzalez B, Prats M, Parajon T, Urrutia A, et al.

Sex and age differences in fragility in a heart failure population. Eur J

Heart Failure 2005;7:798e802.

29. Sciurba F, Criner GJ, Lee SM, Mohsenifar Z, Shade D, Slivak W, et al.

Six-minute walk distance in chronic obstructive pulmonary disease:

reproducibility and effect of walking course layout and length. Am J

Respir Crit Care Med 2003;167:1522e7.

30. Franzblau A. A primer of statistics for non-statisticians. New York:

Harcourt, Brace & World; 1958.

31. Muennig P, Lubetkin E, Jia H, Franks P. Gender and the burden of dis-

ease attributable to obesity. Am J Public Health 2006;96:1662e8.

32. Chin MH, Goldman L. Gender differences in 1-year survival and qual-

ity of life among patients admitted with congestive heart failure. Med

Care 1998;36:1033e46.

33. Cline CM, Willenheimer RB, Erhardt LR, Wiklund I, Israelsson BY.

Health-related quality of life in elderly patients with heart failure.

Scand Cardiovasc J 1999;33:278e85.

34. Burns RB, McCarthy EP, Moskowitz MA, Ash A, Kane RL, Finch M.

Outcomes for older men and women with congestive heart failure.

J Am Geriatr Soc 1997;45:276e80.

35. Lund LH, Mancini D. Heart failure in women. Med Clin North Am

2004;8:1321e45.