Publication
Article
We assessed the prognostic value of the 6-minute walk test (6-MWT) among 1592 subjects with differing degrees of left ventricular systolic dysfunction (LVSD). We found that the 6-MWT was an independent predictor of mortality, particularly among patients with more than mild LVSD. The 6-MWT provides less prognostic utility in patients with mild or lesser LVSD, however.
For patients with chronic heart failure (CHF), determining functional status is an important factor in identifying risk. The 6-minute walk test (6-MWT) is a submaximal,1 self-paced test that is tolerable for many patients with CHF because it imitates activities of daily living. The 6-MWT is commonly used to assess patients’ functional status in cardiology clinics around the world. Among a representative sample of CHF subjects, we evaluated the prognostic value of the 6-MWT and performed a subgroup analysis according to severity of left ventricular systolic dysfunction (LVSD).
Subjects and methods
The study was approved by the Hull and East Riding Ethics Committee, and all subjects provided informed consent prior to participation. The sample included patients referred to our clinic with the symptom of breathlessness. Heart failure was defined as having current symptoms of heart failure or a history of symptoms controlled by medical therapy. The symptoms had to result from cardiac dysfunction and not from another cause. Left ventricular systolic dysfunction was evaluated with 2-dimensional echocardiography, and subjects were considered to have LVSD if they had a left ventricular ejection fraction (LVEF) ≤ 45%. The baseline characteristics of the subjects and self-perceived symptom severity in surviving and deceased subjects are shown in Tables 1 and 2. The 6-MWT was performed according to a standardized procedure.2 Subjects walked as far as possible within 6 minutes in a 15-meter flat, obstacle-free corridor, turning 180° each time they reached the end of the corridor.
Base clinical characteristics in surviving and deceased subjects.
Table 1.
Variable (media ± IQR)
All patients
Alive
Dead
value
P
N (%)
1592 (100)
1380 (86.7)
212 (13.3)
-
Males, n (%)
955 (60)
826 (60)
129 (61)
.28
Median age, years (range)
74 (67-80)
73 (66-80)
76 (70-82)
.001*
Median BMI, kg/m-2 (range)
27.5 (24.9-30.9)
27.5 (25-31)
27.5 (25-31)
.165
Media LVEF, % (range)
48 (35-56)
49 (37-57)
40 (31-50)
.001*
LVI > mild, %
28.2
25.2
51.1
.001*
Median 6-MWT, m (range)
345 (240-420)
360 (270-430)
240 (101-330)
.001*
Median log-NT-proBNP, pg/mL-1 (range)
1.5 (1.1-2.1)
1.4 (1.0-2.0)
2.2 (1.7-2.7)
.001*
Median sodium, mmol/L-1 (range)
140 (138-141)
140 (138-141)
137 (134-140)
.001*
Potassium, mmol/L-1 (range)
4.4 (4.1-4.7)
4.4 (4.1-4.7)
4.4 (4.1-4.7)
.717
Median urea, mmol/L-1 (range)
5.8 (4.7-7.6)
5.7 (4.6-7.2)
7.6 (5.9-10.8)
.001*
Median creatinine, μmol/l-1 (range)
93 (80-114)
91 (78-109)
115 (93-144)
.001*
Median hemoglobin, g/dL-1 (range)
13.8 (12.8-14.8)
13.9 (12.9-14.8)
13.0 (11.8-14.3)
.001*
Diabetes mellitus, %
14.6
13.3
21.4
.001*
Atrial fibrillation, %
17.6
14.7
30.1
.001*
QRS duration, ms-1 (range)
96 (86-112)
96 (86-110)
107 (92-136)
.001*
Angina, %
16.2
14.8
22.3
.001*
Resting HR, beats/min-1 (range)
70 (60-80)
70 (60-80)
75 (63-88)
.001*
Resting systolic BP, mm Hg (range)
143 (126-161)
144 (128-161)
136 (117-155)
.001*
Loop diuretic, %
62.5
60.6
74.0
.001*
ACE-I, %
48.2
40.5
62.2
.001*
β-blocker, %
42.2
38.6
48.5
.001*
* Difference between living and deceased patients (P < .05).
IQR indicates interquartile range; BMI, body mass index; LVEF, left ventricular ejection fraction; LVI, left ventricular impairment; 6-MWT, 6-minute walk test; NT-proBNP, N-terminal B-type natriuretic peptide; HR, heart rate; BP, blood pressure; ACE-I, angiotensin-converting enzyme inhibitor. (Reprinted with permission from Ingle L, Rigby AS, Carroll S, et al. Prognostic value of the 6-min walk test and self-perceived symptom severity in older patients with chronic heart failure. Eur Heart J. 2007;28[5]:560-568.)
Patient-perceived physical symptom severity in surviving and deceased patients.
Table 2.
Variables
All patients
% (n)
Alive
% (n)*
Dead
% (n)†
value
P
Swelling of ankles
.05‡
None, very little, a little
70.6 (1084)
71.3 (969)
65.3 (115)
Some
13.8 (212)
13.4 (182)
17.0 (30)
A lot, very much
15.6 (213)
15.3 (182)
17.7 (31)
Signs of breathlessness at rest
.001‡
None, very little, a little
78.3 (1202)
79.7 (1083)
67.6 (119)
Some
13.0 (200)
12.4 (169)
17.6 (31)
A lot, very much
8.7 (133)
7.9 (17)
14.8 (26)
Signs of breathlessness at night
.004‡
None, very little, a little
82.7 (1269)
84.3 (1146)
69.9 (123)
Some
9.2 (141)
8.1 (110)
17.6 (31)
A lot, very much
8.1 (125)
7.6 (103)
12.5 (22)
Signs of breathlessness
during normal activity
< .001‡
None, very little, a little
57.0 (875)
77.6 (813)
35.2 (62)
Some
13.1 (201)
12.4 (168)
18.8 (33)
A lot, very much
25.0 (383)
11.4 (304)
44.9 (79)
Fatigue at rest
.005‡
None, very little, a little
72.6 (1114)
74.1 (1007)
60.8 (107)
Some
13.1 (201)
12.4 (168)
18.8 (33)
A lot, very much
14.3 (220)
13.5 (184)
20.5 (36)
Fatigue limiting daily activity
< .001‡
None, very little, a little
62.0 (951)
64.3 (874)
43.8 (77)
Some
17.4 (267)
17.0 (231)
20.5 (36)
A lot, very much
10.6 (317)
18.7 (254)
35.8 (63)
*Note that in the alive group, 21 patients had missing data.
†Note that in the dead group, 36 patients had missing data.
‡Differences between alive and dead patients (P < .05). P values calculated by the chi-square test for trend.
(Reprinted with permission from Ingle L, Rigby AS, Carroll S, et al. Prognostic value of the 6-min walk test and self-perceived symptom severity in older patients with chronic heart failure. Eur Heart J. 2007;28[5]:560-568.)
Analyses were performed using SPSS software (version 13.0). A level of 5% statistical significance was used throughout (2-tailed). The 6-MWT data were divided into quartile ranges (≤ 240 m, 241-345 m, 346-420 m, > 420 m), and Kaplan-Meier curves were constructed to show mortality data. A Cox proportional hazards model was developed to predict all-cause death rates based on the variables shown in Tables 1 and 2.
Figure. Kaplan-Meier curve showing survival based on the 6-minute walk test (6-MWT)
distance (quartiles).
Results
P
The death rate for the course of the study was 13.3% (212 of 1592 subjects). For surviving subjects, the mean duration of follow-up was 37 months (28-45 months). The Kaplan-Meier curve shown in the Figure displays cumulative survival in relation to the 6-MWT. The prognosis was worse for subjects with poorer 6-MWT results ( < .001). Worsening 6-MWT distance, signs of breathlessness at night, increased N-terminal B-type natriuretic peptide (NT-proBNP) levels, decreased hemoglobin concentration, and use of beta blocking agents were all shown to independently predict all-cause mortality.
When subjects were stratified according to whether they had more than mild LVSD or mild or less than mild LVSD, 6-MWT results, sodium levels, and QRS duration were shown to predict mortality among subjects with more than mild LVSD, as shown in Table 3. Sex, hemoglobin concentration, signs of breathlessness at night, log-NT-proBNP, and use of beta blocking agents were shown to predict all-cause mortality among subjects with mild or less than mild LVSD.
Table 3. Cox regression models in patients with greater than mild LVSD and mild or less than mild LVSD.
Variable
Hazard ratio (95% CI)
Mild or less than
mild LVSD
Men
0.438 (0.233-0.824)
Hemoglobin, g/dL-1
0.751 (0.622-0.905)
Log-NT-proBNP, pg/mL-1
2.036 (1.609-2.576)
SOB at night
Some
0.671 (0.385-1.171)
> Some
2.519 (1.325-4.792)
β-blocker usage, %
0.390 (0.198-0.770)
Explain variation, %
72
More than mild LVSD
6-MWT, m
0.996 (0.994-0.998)
Sodium, mmol/L-1
0.881 (0.813-0.954)
QRS duration, ms-1
1.013 (1.005-1.021)
Explained variation, %
92
CI indicates confidence interval; SOB, shortness of breath; LVSD, left ventricular systolic dysfunction; NT-proBNP, N-terminal B-type natriuretic peptide; 6-MWT, 6-minute walk test. Variables selected based on 100 bootstrapped samples with an inclusion frequency |≥| 65%. (Reprinted with permission from Ingle L, Rigby AS, Carroll S, et al. Prognostic value of the 6-min walk test and self-perceived symptom severity in older patients with chronic heart failure. Eur Heart J. 2007;28[5]:560-568.)
Discussion
In our study, results of the 6-MWT were shown to be an independent predictor of mortality among subjects with CHF. The most powerful predictor of mortality, however, was an increased NT-proBNP level. When 6-MWT results were added to NT-proBNP levels, an additive prognostic effect was demonstrated. For subjects with more than mild LVSD, the 6-MWT remained an important predictor of mortality. This relationship was not shown for subjects with mild or lesser LVSD, however.
Most prior studies assessing the prognostic value of the 6-MWT have used sample sizes with fewer than 310 subjects. A study by Bittner and colleagues, however, which included 898 subjects, showed better survival rates than in our study, with approximately a 10% mortality rate for subjects with a 6-MWT distance < 350 m.3 Our subjects, however, had more severe LVSD, with a worse mean 6-MWT performance. In Bittner and colleagues' study, LVEF and 6-MWT distance were both shown to be independent predictors of mortality or hospitalization, whereas our results showed that LVEF did not remain predictive of all-cause mortality.
In another study, 121 subjects with LVSD and in New York Heart Association functional class II-III were followed for 18 months.4 Subjects were stratified into 2 groups: those who reached the combined end point of cardiovascular death or hospitalization and those who were free of events. A limitation of this study was that only 47 patients experienced an event over the course of follow-up, and this may be the reason the findings were nonsignificant. Study results also showed that subjects who walked < 300 m had a higher rate of cardiovascular-related events.
P
Rostagno and colleagues assessed the value of the 6-MWT as a predictor of event-free survival among 214 subjects with mild-to-moderate CHF who were followed for a mean of 34 months.5 Results of their study showed that the survival rate among subjects who walked < 300 m was 62% compared with 82% for subjects who walked > 300 m. Independent predictors of mortality were left ventricular fractional shortening ( = .009) and the distance walked in the 6-MWT. It is difficult to compare our results with those of this study because all subjects were included in our primary analysis, which was followed by a supplementary analysis stratified according to the severity of LVSD.
As far as we know, only 1 study has shown the prognostic significance of the 6-MWT among subjects with severe heart failure.6 In that study, peak oxygen uptake was compared with the 6-MWT in 307 subjects with advanced heart failure, and peak oxygen uptake was shown to be a better predictor than the 6-MWT. Therefore, the authors contended that the 6-MWT could not be used as a surrogate predictor for peak oxygen uptake in patients with advanced heart failure. However, we have shown that in the absence of peak oxygen uptake, the 6-MWT is an important predictor of mortality in patients with more than mild LVSD.
The results of our study showed that exercise capacity was significantly decreased in subjects with more than mild LVSD. Although a comprehensive discussion regarding the mechanisms responsible for the abnormal exercise response in CHF patients is beyond the scope of this article, it may be useful to briefly discuss these factors. During exercise, cardiac output is reduced,7 and muscle strength, mass, and endurance are decreased or lost.8,9 Ventilation-perfusion mismatches stem from hemodynamic dysfunction and from an altered control of ventilation, as shown by the augmentation of the chemoreflex. In patients with CHF, abnormal hemodynamic function correlates with a poor prognosis.10,11 Mechanisms of chemoreflex overactivity are associated with sympathetic activation and neurohormonal imbalance, both of which also affect survival.12,13 The chemoreflex may be directly affected by a reduced blood flow to the chemoreceptors, reflecting hemodynamic dysfunction. For an in-depth discussion of this topic, see the article by Clark.14
Conclusions
The results of our study showed that the 6-MWT was an important independent predictor of mortality in 1592 subjects with CHF, and this was especially evident in those with more than mild LVSD.