If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Respiratory muscle weakness (RMW) was observed in 39% of patients with HFpEF.
•
RMW was associated with poor prognosis in both patients with HFrEF and HFpEF.
•
RMW showed the additive effect for risk prediction only in patients with HFpEF.
•
Respiratory muscle assessment is useful for risk prediction in HFpEF patients.
Abstract
Background
Although respiratory muscle weakness (RMW) is known to predict prognosis in patients with heart failure with reduced ejection fraction (HFrEF), RMW prevalence and its prognosis in those with preserved ejection fraction (HFpEF) remain unknown. We aimed to investigate whether the RMW predicted mortality in HFpEF patients.
Methods
We conducted a single-centre observational study with consecutive 1023 heart failure patients (445 in HFrEF and 578 in HFpEF). Maximal inspiratory pressure (PImax) was measured to assess respiratory muscle strength at hospital discharge, and RMW was defined as PImax <70% of predicted value. Endpoint was all-cause mortality after hospital discharge, and we examined the influence of RMW on the endpoint.
Results
Over a median follow-up of 1.8 years, 134 patients (13.1%) died; of these 53 (11.9%) were in HFrEF and 81 (14.0%) in HFpEF. RMW was evident in 190 (42.7%) HFrEF and 226 (39.1%) HFpEF patients and was independently associated with all-cause mortality in both HFrEF (adjusted hazard ratio [HR]: 2.13, 95% confidence interval [CI]: 1.17–3.88) and HFpEF (adjusted HR: 2.85, 95% CI: 1.74–4.67) patients. Adding RMW to the multivariate logistic regression model significantly increased area under the receiver-operating characteristic curve (AUC) for all-cause mortality in HFpEF (AUC including RMW: 0.78, not including RMW: 0.74, P = 0.026) but not in HFrEF (AUC including RMW: 0.84, not including RMW: 0.82, P = 0.132).
Conclusions
RMW was observed in 39% of HFpEF patients, which was independently associated with poor prognosis. The additive effect of RMW on prognosis was detected only in HFpEF but not in HFrEF.
Heart failure with preserved left ventricular ejection fraction (HFpEF), which is highly observed in elderly patients, comprises approximately 50% of the overall heart failure patients [
Mode of death in patients with heart failure and reduced vs. preserved ejection fraction: report from the registry of hospitalized heart failure patients.
]. Additionally, strategies that can improve prognosis in HFpEF patients remain unclear, although some pharmacological or nonpharmacological therapies are recognised to improve prognosis in HFrEF patients [
ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC.
]. Therefore, to establish the proper treatment strategy that can improve prognosis in HFpEF patients, it is crucial to identify clinically meaningful predictors.
Respiratory muscle weakness (RMW) is frequently observed in patients with chronic heart failure [
Validity, prognostic value and optimal cutoff of respiratory muscle strength in patients with chronic heart failure changes with beta-blocker treatment.
Eur. J. Cardiovasc. Prev. Rehabil.2009; 16: 424-429
], and several studies have reported that reduced respiratory muscle strength is caused by the atrophy of these muscles and/or decreased actin-myosin cross-bridges, resulting from the activation of neurohumoral factors due to heart failure [
]. Importantly, RMW, estimated using maximal inspiratory pressure (PImax), is a known predictor of exercise intolerance and ventilatory inefficiency, leading to decreased quality of life and lower survival in HFrEF patients [
Validity, prognostic value and optimal cutoff of respiratory muscle strength in patients with chronic heart failure changes with beta-blocker treatment.
Eur. J. Cardiovasc. Prev. Rehabil.2009; 16: 424-429
Respiratory muscle weakness increases dead-space ventilation ratio aggravating ventilation-perfusion mismatch during exercise in patients with chronic heart failure.
]. Conversely, a previous study by Habedank and colleagues has shown that PImax was not a significant predictor of mortality as it varied according to gender, body mass index, and cachexia in severe HFrEF patients [
]. Several statements have recommended the use of PImax, relative to a reference value (%PImax), for assessing weakness and dysfunction of respiratory muscles [
Inspiratory muscle weakness is associated with exercise intolerance in patients with heart failure with preserved ejection fraction: a preliminary study.
], its prevalence and prognostic potential in these patients remains unclear. Therefore, we conducted an observational study to clarify the relationships of RMW assessed by %PImax with mortality in heart failure patients.
This study aimed to investigate whether RMW predicted mortality in patients with HFrEF or HFpEF.
2. Methods
The study protocol was approved by the Kitasato Institute Clinical Research Review Board (KMEO B18-075) and was performed according to the ethical guidelines of the Declaration of Helsinki.
2.1 Study population
This study had a retrospective longitudinal observational design. We included consecutive patients with HFrEF or HFpEF who were admitted to the Kitasato University Hospital for heart failure treatment between May 2009 and December 2017. HFrEF was defined as a left ventricular ejection fraction (LVEF) < 40% on an echocardiogram, and HFpEF was diagnosed based on clinical guidelines and LVEF ≥ 50% [
ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC.
ACCF/AHA guideline for the management of heart failure: a report of the American college of cardiology foundation/American heart association task force on practice guidelines.
]. Patients who had undergone thoracic surgery within the last three months or had chronic respiratory diseases were excluded from the study.
2.2 Study protocol
Patients underwent haematological analysis, echocardiograms, and assessment of pulmonary function and respiratory muscle function at hospital discharge. The primary endpoint of this study was defined as all-cause mortality after hospital discharge. We collected the data on all variables from electronic database.
2.3 Patient characteristics
Information on age, gender, body mass index (BMI), smoking history, aetiology of heart failure, severity of heart failure based on the New York Heart Association functional classification (NYHA class), medications, and comorbidities such as hypertension, diabetes mellitus, dyslipidaemia, chronic kidney disease or atrial fibrillation, was obtained from medical records at study entry. Routine laboratory analysis included haemoglobin, serum albumin and C-reactive protein, and plasma brain natriuretic peptide (BNP). The estimated glomerular filtration rate (eGFR) was determined based on serum creatinine levels. Additionally, echocardiographic variables including LVEF, left atrial diameter (LAD), mitral early diastolic inflow velocity (E), mitral late diastolic inflow velocity (A), mitral annular early diastolic velocity (e′), and deceleration time of mitral early diastolic inflow (DCT) were measured. The AHEAD score was used for risk assessment and was calculated by assigning one point to each of the following factors: A: atrial fibrillation, H: haemoglobin <13 g/dL for men and 12 g/dL for women, Elderly (age > 70 years), A: abnormal renal parameters (creatinine > 130 μmol/dL) and D: diabetes mellitus [
2.4 Pulmonary function and respiratory muscle function
Pulmonary function was assessed by measuring forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) using a spirometer (Autospiro AS-507, Minato Medical Science, Osaka, Japan) and calculated their percentage based on predictive values issued by the Japanese Respiratory Society [
]. To assess respiratory muscle function, maximal inspiratory pressure (PImax) was measured using a pressure transducer (Autospiro AAM-377, Minato Medical Science, Osaka, Japan) according to the joint statement of the American Thoracic Society and European Respiratory Society [
]. To measure PImax, patients in a sitting position were asked to hold a 25–mm diameter mouthpiece in their mouth and perform a 3-s forced inspiration from the maximal expiratory level. PImax was determined based on the average value of the maximum inspiratory pressure over a 1-s period during the 3-s forced inspiration. PImax was expressed as its absolute value in the present study, although it has a negative value compared to atmospheric pressure. Respiratory pressure measurement was performed three times, and the maximum value in PImax was used for analysis. Subsequently, we calculated percentage PImax (%PImax) based on predictive values that were estimated using age, gender, height, and body weight [
The primary endpoint of this study was all-cause death identified through medical chart review. The time period for this event was calculated as the number of days from the date of the respiratory muscle strength measurement to event date. We also investigated whether the cause of death was cardiovascular (CV) death or non-CV death, including respiratory diseases.
2.6 Statistical analyses
Differences in clinical parameters between patients with and without RMW were compared using the unpaired Student's t-test or the Mann-Whitney U test for continuous variables and by the Chi-square test or Fisher's exact test for categorical variables, as appropriate. We examined the influence of respiratory muscle weakness on survival using the Kaplan-Meier method with the log-rank test. To identify predictors of all-cause mortality, we used a multivariate Cox proportional hazard model that incorporated clinical characteristics and RMW as covariates. The subgroup analysis of RMW in various subgroups relevant to the heart failure prognosis was analysed to assess any potential effect modification of the association between RMW and all-cause mortality. To assess the additive effect of RMW on the predictive capability of all-cause death, the area under the receiver-operating characteristic curves (AUC) of multivariate logistic regression models for all-cause death were compared between models with and without RMW. The predictive accuracy and model fit of the logistic regression analyses were examined using Hosmer-Lemeshow statistics. To confirm whether the sample size was adequate, we calculated the sample power with an alpha value of 0.05, using mortality and RMW rates, hazard ratio, accrual time during which patients were recruited, and follow-up time. Continuous variables are reported as mean ± standard deviation or median with interquartile range and categorical variables are expressed as patient numbers and their percentages. A two-tailed P value of <0.05 was considered statistically significant. All statistical analyses were performed using SPSS, ver. 25.0 (IBM, Armonk, NY), JMP, ver. 14.1.0 (SAS Institute, Cary, NC), and R, ver. 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1 Patient characteristics
The potential study population consisted of 2335 consecutive patients with HFrEF or HFpEF. Patients who had undergone thoracic surgery within the last three months (n = 527) or had chronic respiratory diseases (n = 175) were excluded, along with those who could not perform a respiratory muscle function test at hospital discharge (n = 610). Consequently, 1023 heart failure patients were included for analysis; of these 445 had HFrEF and 578 had HFpEF.
Table 1 shows demographic and clinical characteristics of both HFrEF and HFpEF patients. RMW was observed in 190 (42.7%) HFrEF patients and 226 (39.1%) HFpEF patients. Among HFrEF patients, compared to those without RMW, those with RMW were significantly older, and had higher values for AHEAD score, pack-years, and BNP levels. HFpEF patients with RMW showed significantly lower BMI values compared to those without RMW. Both HFrEF and HFpEF patients with RMW showed significantly lower albumin levels. Moreover, a higher percentage of patients with RMW were NYHA class III at hospital discharge, and patients with RMW also showed significantly lower FVC, %FVC, FEV1, and %FEV1 values than those without RMW.
Table 1Demographic and clinical characteristics in HFrEF and HFpEF patients with (%PImax < 70%) and without (%PImax ≥ 70%) RMW.
HFrEF
HFpEF
Variable
Overall
%PImax ≥ 70%
%PImax < 70%
Overall
%PImax ≥ 70%
%PImax < 70%
n (%)
445
255 (57.3)
190 (42.7)
P value
578
352 (60.9)
226 (39.1)
P value
Age, yrs
63.4 ± 15.2
61.5 ± 14.8
65.9 ± 15.5
0.002
71.4 ± 10.8
70.9 ± 10.6
72.2 ± 11.1
0.154
Gender female, n (%)
123 (27.6)
72 (28.2)
51 (26.8)
0.830
244 (42.2)
155 (44.0)
89 (39.4)
0.301
BMI, kg/cm2
22.4 ± 4.4
22.5 ± 4.4
22.2 ± 4.4
0.582
22.5 ± 3.7
22.7 ± 3.7
22.0 ± 3.6
0.029
SBP, mm Hg
116 ± 30
116 ± 30
116 ± 30
0.829
124 ± 29
124 ± 28
125 ± 29
0.581
DBP, mm Hg
70 ± 20
70 ± 20
69 ± 20
0.422
69 ± 17
69 ± 16
68 ± 18
0.745
HR, beats/min
84 ± 21
85 ± 22
83 ± 20
0.464
79 ± 20
80 ± 20
77 ± 19
0.064
NYHA at hospital discharge, n (%)
II
342 (72.9)
216 (84.7)
126 (66.3)
<0.001
337 (58.3)
223 (63.4)
114 (50.4)
0.003
III
103 (23.1)
39 (15.3)
64 (33.7)
241 (41.7)
129 (36.6)
112 (49.6)
AHEAD score
1.63 ± 1.23
1.47 ± 1.16
1.84 ± 1.30
0.002
2.09 ± 1.15
2.02 ± 1.14
2.20 ± 1.17
0.064
Prior history of HF, n (%)
157 (35.3)
83 (32.5)
74 (38.9)
0.192
157 (27.2)
92 (26.1)
65 (28.8)
0.503
Smoking history, n (%)
282 (65.3)
161 (65.2)
121 (65.4)
1.000
305 (53.8)
186 (54.1)
119 (53.4)
0.931
Pack-years
37.8 ± 32.9
34.3 ± 29.1
42.2 ± 36.7
0.047
37.3 ± 31.0
36.4 ± 28.4
38.7 ± 34.7
0.535
Medical history
IHD, n (%)
191 (42.9)
109 (42.7)
82 (43.2)
1.000
257 (44.5)
156 (44.3)
101 (44.7)
0.932
Hypertension, n (%)
261 (58.7)
146 (57.3)
115 (60.5)
0.497
397 (68.7)
249 (70.7)
148 (65.5)
0.199
Diabetes mellitus, n (%)
179 (40.2)
101 (39.6)
78 (41.1)
0.770
235 (40.7)
148 (42.0)
87 (38.5)
0.435
Dyslipidaemia, n (%)
212 (47.6)
125 (49.0)
87 (45.8)
0.504
257 (44.5)
162 (46.0)
95 (42.0)
0.391
CKD, n (%)
273 (61.3)
153 (60.0)
120 (63.2)
0.555
404 (70.0)
243 (69.2)
161 (71.2)
0.642
Atrial fibrillation, n (%)
98 (22.0)
49 (19.2)
49 (25.8)
0.106
151 (26.1)
88 (25.0)
63 (27.9)
0.440
Medications
ACE-I, n (%)
261 (58.7)
152 (59.6)
109 (57.4)
0.697
191 (33.0)
118 (33.5)
73 (32.3)
0.786
ARB, n (%)
157 (35.3)
89 (34.9)
68 (35.8)
0.920
260 (45.0)
158 (44.9)
102 (45.1)
1.000
Beta-blockers, n (%)
410 (92.1)
233 (91.4)
177 (93.2)
0.594
388 (67.1)
238 (67.6)
150 (66.4)
0.786
Diuretics, n (%)
401 (90.1)
227 (89.0)
174 (91.6)
0.424
364 (63.0)
214 (60.8)
150 (66.4)
0.186
Blood examination
Haemoglobin, g/dL
13.2 ± 2.4
13.4 ± 2.5
13.0 ± 2.3
0.060
11.9 ± 2.1
11.9 ± 2.0
11.8 ± 2.2
0.635
Albumin, g/dL
3.66 ± 0.51
3.71 ± 0.51
3.59 ± 0.49
0.009
3.53 ± 0.53
3.57 ± 0.49
3.46 ± 0.58
0.009
CRP, mg/dL
0.65 ± 1.16
0.69 ± 1.32
0.61 ± 0.90
0.472
0.92 ± 1.43
0.97 ± 1.49
0.84 ± 1.32
0.276
eGFR, mL/min/1.73m2
52.9 ± 23.1
53.9 ± 20.9
51.6 ± 25.8
0.295
49.3 ± 22.9
49.5 ± 20.1
49.0 ± 26.8
0.807
BNP, pg/mL
390.2 [196.7– 789.5]
348.4 [181.9 –720.0]
478.5 [215.3 –898.4]
0.014
236.5 [114.6–443.6]
226.2 [115.0 –384.1]
254.8 [115.6 –509.6]
0.090
Echocardiographic measurements
LVEF, %
28.6 ± 7.0
28.7 ± 6.9
28.4 ± 7.2
0.652
61.5 ± 8.1
61.1 ± 7.7
62.2 ± 8.8
0.103
LAD, mm
44.4 ± 8.6
43.9 ± 8.4
45.1 ± 8.9
0.154
43.3 ± 10.4
42.9 ± 9.5
43.9 ± 11.6
0.267
E/A
1.6 ± 1.1
1.6 ± 1.1
1.7 ± 1.1
0.566
1.2 ± 1.0
1.2 ± 1.1
1.3 ± 0.9
0.383
E′, cm/s
5.8 ± 2.7
5.8 ± 2.7
5.7 ± 2.7
0.754
6.8 ± 3.2
6.7 ± 3.2
6.8 ± 3.2
0.858
E/E′
16.3 ± 7.9
15.9 ± 7.8
16.9 ± 8.0
0.280
14.8 ± 7.5
14.7 ± 7.9
14.8 ± 6.9
0.873
DCT, ms
176.3 ± 67.0
179.4 ± 69.5
172.2 ± 63.4
0.317
214.6 ± 78.4
219.4 ± 79.3
207.2 ± 76.6
0.113
Respiratory function
FVC, L
2.58 ± 0.87
2.79 ± 0.84
2.28 ± 0.83
<0.001
2.18 ± 0.81
2.28 ± 0.78
2.02 ± 0.83
<0.001
FEV1, L
2.02 ± 0.77
2.2 ± 0.73
1.77 ± 0.75
<0.001
1.67 ± 0.67
1.77 ± 0.64
1.51 ± 0.69
<0.001
FEV1/FVC, %
78.0 ± 10.1
78.8 ± 8.8
76.9 ± 11.5
0.052
76.1 ± 9.9
77.6 ± 8.2
73.9 ± 11.7
<0.001
%FVC, %
76.3 ± 17.0
81.4 ± 15.1
69.4 ± 17.2
<0.001
74.4 ± 19.0
78.4 ± 17.5
68.0 ± 19.6
<0.001
%FEV1, %
73.3 ± 18.6
78.5 ± 15.8
66.4 ± 19.7
<0.001
71.6 ± 20.6
76.4 ± 18.7
63.9 ± 21.1
<0.001
PImax, cmH2O
56.8 ± 28.6
73.2 ± 25.9
34.8 ± 15.7
<0.001
50.1 ± 25.9
63.4 ± 22.9
29.3 ± 14.0
<0.001
%PImax, %
76.2 ± 31.6
97.5 ± 23.0
47.8 ± 14.7
<0.001
79.3 ± 35.4
101.1 ± 25.7
45.3 ± 16.5
<0.001
All-cause mortality, n (%)
53 (11.9)
17 (6.7)
36 (18.9)
<0.001
81 (14.0)
26 (7.4)
55 (24.3)
<0.001
Values are mean ± SD or median [interquartile range].
A, mitral late diastolic inflow velocity; ACE-I, angiotensin convertor enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BNP, brain natriuretic peptide; CKD, chronic kidney disease; CRP, C-reactive protein; DBP, diastolic blood pressure; DCT, deceleration time of mitral early diastolic inflow; E, mitral early diastolic inflow velocity; E′, mitral annular early diastolic velocity; eGFR, estimated glomerular filtration rate; FEV1, forced expiratory volume in 1-s; FVC, forced vital capacity; HFpEF, heart failure preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; HR, heart rate; IHD, ischemic heart disease; LAD, left atrial diameter; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; RMW, respiratory muscle weakness; PImax, maximal inspiratory pressure; SBP, systolic blood pressure.
3.2 Association between RMW and all-cause mortality
During the median follow-up period of 1.8 years, 134 patients died; of these, 53 had HFrEF and 81 had HFpEF, and the mortality rate was 72.8/1000 person-years. The cumulative all-cause mortality rate was 6.7% and 18.9% in HFrEF patients without and with RMW, respectively, and 7.4% and 24.3% in HFpEF patients without and with RMW, respectively. The Kaplan-Meier survival curves for the two groups are shown in Fig. 1. Patients with RMW had a significantly lower survival rate than those without RMW in both HFrEF (log-rank: 16.429, P < 0.001) and HFpEF (log-rank: 38.295, P < 0.001) groups.
Fig. 1Kaplan-Meier survival curves for the association between respiratory muscle weakness and all-cause death in HFrEF and HFpEF. (A), patients with HFrEF and (B), patients with HFpEF. Solid line, patients without RMW; dotted line, patients with RMW. HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; PImax, maximal inspiratory pressure; RMW, respiratory muscle weakness.
The CV deaths occurred in 93 patients (40 in HFrEF and 53 in HFpEF) and non-CV deaths occurred in 41 patients (13 in HFrEF and 28 in HFpEF). There were no statistical differences in rates of CV and non-CV death between HFrEF and HFpEF (Supplementary file). Patients with RMW showed significantly higher rates of CV death (P = 0.026 for HFrEF and P = 0.027 for HFpEF) and non-CV death (P = 0.012 for HFrEF and P = 0.005 for HFpEF) compared to those without RMW, both in HFrEF and HFpEF (Fig. 2).
Fig. 2Rates of cardiovascular death and non-cardiovascular death between presence or absence of RMW in HFrEF and HFpEF. White bar, patients without RMW; black bar, patients with RMW. CV, cardiovascular; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; RMW, respiratory muscle weakness.
3.4 Cox proportional hazard models for RMW and all-cause death
Table 2 shows the results of the Cox proportional hazard models for RMW and all-cause mortality. In the univariate Cox proportional hazard model, RMW, defined as %PImax < 70%, was a significant predictor of all-cause mortality in HFrEF and HFpEF patients. The multivariate model identified RMW as a significant independent predictor for all-cause mortality even after adjustment for clinical confounding factors, both in HFrEF and in HFpEF patients. No significant interactions were observed in the association between RMW and poor prognosis across the various subgroups in both HFrEF and HFpEF patients, even after adjustment for confounding factors used in the multivariate Cox proportional hazard model (Fig. 3). The sample size in this study was sufficient, as reflected by a sample power of HFrEF and HFpEF of 0.995 and 0.998, respectively.
Table 2Cox proportional hazard models of respiratory muscle weakness for all-cause mortality in HFrEF and HFpEF.
%PImax ≥ 70%
%PImax < 70%
HR
95% CI
HR
95% CI
P value
HFrEF
Univariate analysis
1
Reference
3.09
1.74–5.48
<0.001
Multivariate analysis
1
Reference
2.13
1.17–3.88
0.011
HFpEF
Univariate analysis
1
Reference
3.93
2.46–6.27
<0.001
Multivariate analysis
1
Reference
2.85
1.74–4.67
<0.001
Multivariate analyses were adjusted for age, gender, BMI, AHEAD score, NYHA class at hospital discharge, IHD, LVEF, and BNP. CI, confidence interval; HFrEF, heart failure with reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; HR, hazard ratio; PImax, maximal inspiratory pressure.
Fig. 3Forest plots of hazard ratios for the association between respiratory muscle weakness and all-cause mortality according to major subgroups. Hazard ratios were adjusted for age, gender, BMI, AHEAD score, NYHA class at hospital discharge, ischemic heart disease, LVEF, and BNP. ACE-I, angiotensin convertor enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BNP, brain natriuretic peptide; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association functional classification.
3.5 Additive effect of RMW on predictive capability for all-cause death
The AUCs of the multivariate logistic regression models for all-cause mortality are shown in Fig. 4. The AUC of a model that used clinical characteristics as covariates in HFrEF patients was 0.82 (95% CI: 0.75–0.87), and the addition of RMW to the model did not increase the AUC (AUC: 0.84, 95% CI: 0.78–0.88, P = 0.132). In HFpEF patients, the AUC of the model without RMW was 0.74 (95% CI: 0.68–0.80), which significantly increased to 0.78 (95% CI: 0.72–0.83, P = 0.026) when RMW was included in the model. In the Hosmer-Lemeshow statistics of logistic regression models, both the HFrEF and HFpEF models reached statistical significance for predicting all-cause mortality (HFrEF: chi-squared = 7.47, predictive value = 88%, P = 0.487; HFpEF: chi-squared = 11.19, predictive value = 86%, P = 0.191).
Fig. 4Receiver-operating characteristic curves of logistic regression models to predict all-cause mortality in HFrEF and HFpEF. (A), patients with HFrEF and (B), patients with HFpEF. Independent variables of the logistic regression model (model 1) were age, gender, BMI, AHEAD score, NYHA class at hospital discharge, ischemic heart disease, LVEF, and BNP. Solid line, model 1 with RMW; Dotted line, model 1 without RMW. AUC, area under the receiver-operating characteristic curve; BMI, body mass index; BNP, brain natriuretic peptide; CI: confidence interval; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association functional classification; RMW, respiratory muscle weakness.
The novel findings of this study are as follows. First, RMW was observed in 39% of the patients with HFpEF. Second, both HFrEF and HFpEF patients with RMW had a significantly higher mortality rate compared to those without RMW, and RMW increased the risk of all-cause mortality in HFrEF and HFpEF patients by two- and three-fold, respectively. However, the additive effect of RMW on predictive capability for poor prognosis was observed only in patients with HFpEF but not in those with HFrEF.
4.1 Previous studies
To the best of our knowledge, this is the first study to demonstrate that 39% of HFpEF patients have RMW and that it is a significant indicator of poor prognosis in these patients. Previous studies have reported RMW in 30%–50% of HFrEF patients [
], which is similar to that seen in the present study. Thus, we show that the prevalence of RMW is comparable among HFpEF and HFrEF patients. Conversely, mean absolute value of PImax was significantly lower in HFpEF patients (50.1 cmH2O) than in HFrEF patients (56.8 cmH2O), suggesting that the value of PImax as a respiratory muscle strength in heart failure differs between HFpEF and HFrEF patients (Supplementary file). Generally, respiratory muscle strength is lower in older patients, females, and in patients with severe heart failure [
]. Further, our study population tended to include older patients and a higher proportion of females in the HFpEF group compared to the HFrEF group, which is consistent with the reported higher prevalence of HFpEF in the elderly and in females [
]. Additionally, it is notable that %PImax levels were comparable between HFrEF and HFpEF patients in the present study. Thus, we believe that the use of %PImax values would be both useful and important in assessing RMW for heart failure patients.
4.2 Interpretations of findings
Although there were no differences in age, gender, smoking habits, comorbidities, or medications between HFpEF patients with and without RMW, patients with RMW showed significantly lower pulmonary function, as evidenced by the lower FVC and FEV1 values, and a higher mortality rate compared to those without RMW. This observation is partially different from that seen in HFrEF patients, i.e., RMW in HFrEF was associated with older age, presence of comorbid conditions, smoking habits, and higher BNP. Bowen et al. have demonstrated that activated levels of ROS and the ubiquitin proteasome system in respiratory muscles, which are known to cause muscle atrophy in HFrEF [
Heart failure with preserved ejection fraction induces molecular, mitochondrial, histological, and functional alterations in rat respiratory and limb skeletal muscle.
]. These molecular and histological alterations could have contributed to the observed differences between HFrEF and HFpEF patients with or without RMW. Furthermore, although RMW was identified as an independent predictor of all-cause mortality in both HFrEF and HFpEF in this study, additional effect of RMW to traditional prediction model of heart failure that included NYHA and comorbid conditions was observed only in HFpEF patients but not in HFrEF patients. Recent studies on the cause of death in heart failure patients have reported that while the main causes of death in HFrEF patients were heart failure exacerbation and sudden death, non-cardiovascular death, including due to respiratory failure or infections of the respiratory system, were the main causes of death in HFpEF patients, apart from cardiovascular causes of death [
Mode of death in patients with heart failure and reduced vs. preserved ejection fraction: report from the registry of hospitalized heart failure patients.
Respiratory muscle weakness increases dead-space ventilation ratio aggravating ventilation-perfusion mismatch during exercise in patients with chronic heart failure.
]. Therefore, in the present study, RMW worsened HFpEF patient prognosis because it can decrease pulmonary function that leads to respiratory complications and the incidence of cardiovascular events, even in the subgroup analysis stratified into the previously reported indicator of HFpEF.
4.3 Clinical implications
The results presented here have clinical implications that RMW has been identified as a new predictor of prognosis in HFpEF patients. As respiratory muscle strength is easily measured in clinical practice, RMW might be a useful marker for risk classification not only in HFrEF but also in HFpEF patients. Furthermore, an increase in respiratory muscle strength due to inspiratory muscle training has been reported to improve exercise tolerance and quality of life in HFrEF patients [
], and our results imply similar potential benefits due to greater respiratory muscle strength in HFpEF patients as well.
4.4 Potential limitations
There are some limitations in the present study. As this was a single-centre study that only included Japanese patients, it is unclear whether these results can apply to patients in other hospitals or in other populations. Nevertheless, the sample size used here is larger than that of previous studies on respiratory muscle strength in HFpEF patients [
Inspiratory muscle weakness is associated with exercise intolerance in patients with heart failure with preserved ejection fraction: a preliminary study.
], which statistically satisfied the power analysis for estimating sample size. However, given half of potential study population was excluded from the analysis, external validity could have been reduced. We also performed the multivariate analyses using multiple confounders. Such multiple test might increase the rate of false positives (type I error). Therefore, future multi-centre prospective studies are required to investigate the validity and reliability of RMW, assessed by %PImax, as a predictor of prognosis in these patients.
5. Conclusions
RMW, defined as %PImax < 70%, was independently associated with poor prognosis in both HFrEF and HFpEF patients. However, the additive effect of RMW on risk prediction was observed only in HFpEF and not in HFrEF patients.
Funding
This work was supported by Japan Society for the Promotion of Science Grant-in-Aid [JSPS KAKENHI Grant Number JP16K16442].
Author contributions
NH, KK, and TM contributed to the conception and design of the study. NH and TM wrote the manuscript. NH, KK, RM, KN, TI, ST, TN, and MY contributed to data collection. NH, TM, KK, RM, KN, ST, EM, CN, MT, AM, and JA contributed to interpretation. NH and KK contributed to the statistical analysis. JA contributed to supervision and mentorship. All authors have critically revised and assisted in the preparation of the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.
Research ethics and patient consent
Ethical approval for the study was given by Kitasato Institute Clinical Research Review Board (KMEO B18-075).
Declaration of competing interest
None.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Mode of death in patients with heart failure and reduced vs. preserved ejection fraction: report from the registry of hospitalized heart failure patients.
ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC.
Validity, prognostic value and optimal cutoff of respiratory muscle strength in patients with chronic heart failure changes with beta-blocker treatment.
Eur. J. Cardiovasc. Prev. Rehabil.2009; 16: 424-429
Respiratory muscle weakness increases dead-space ventilation ratio aggravating ventilation-perfusion mismatch during exercise in patients with chronic heart failure.
Inspiratory muscle weakness is associated with exercise intolerance in patients with heart failure with preserved ejection fraction: a preliminary study.
ACCF/AHA guideline for the management of heart failure: a report of the American college of cardiology foundation/American heart association task force on practice guidelines.
Heart failure with preserved ejection fraction induces molecular, mitochondrial, histological, and functional alterations in rat respiratory and limb skeletal muscle.
30348-8&id=fx1.jpg){kind=link}
30348-8&id=gr1.jpg){kind=link}
30348-8&id=gr2.jpg){kind=link}
30348-8&id=gr3.jpg){kind=link}
30348-8&id=gr4.jpg){kind=link}