ESC HEART FAILURE ORIGINAL RESEARCH ARTICLE

ESC Heart Failure (2017)

Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/ehf2.12158

Relationship between skeletal muscle mass and

cardiac function during exercise in community-
dwelling older adults
Masamitsu Sugie1,3*, Kazumasa Harada1, Tetsuya Takahashi3,5, Marina Nara4,6, Joji Ishikawa1, Teruyuki
Koyama2,3, Hunkyung Kim3, Jun Tanaka1, Hajime Fujimoto1, Shuichi Obuchi3, Stephan von Haehling7,
Syunei Kyo8 and Hideki Ito9
1
Department of Cardiology, Tokyo Metropolitan Geriatric Hospital and Institute of Gerontology, Tokyo, Japan; 2Department of Rehabilitation, Tokyo Metropolitan Geriatric
Hospital and Institute of Gerontology, Tokyo, Japan; 3Department of Institute of Gerontology, Tokyo Metropolitan Geriatric Hospital and Institute of Gerontology, Tokyo,
Japan; 4Ochanomizu University, Tokyo, Japan; 5Tokyo University of Technology, Tokyo, Japan; 6Health Management Services Inc., Tokyo, Japan; 7Department of Cardiology
and Pneumology, Institute of Innovative Clinical Trials, University Medical Center Göttingen, Göttingen, Germany; 8Department of Cardiac Surgery, Tokyo Metropolitan
Geriatric Hospital and Institute of Gerontology, Tokyo, Japan; 9Department of Diabetes, Metabolism and Endocrinology, Tokyo Metropolitan Geriatric Hospital and Institute
of Gerontology, Tokyo, Japan

Abstract
Aims This study aimed to investigate the relationship between skeletal muscle mass and cardiac functional parameters in
older adults during cardiopulmonary exercise testing (CPET).
Methods and results Sixty-three Japanese community-dwelling older adults were enrolled (20 men and 43 women; mean
age 80 years, range 65–97 years). Cardiac functional parameters during exercise were assessed using CPET. Skeletal muscle
mass index (SMI) was calculated by dividing the appendicular lean mass (measured using dual-energy X-ray absorptiometry)
by height in metres squared. Subjects were divided into two groups: men with SMI ≥ 7.0 kg/m2 and women with
SMI ≥ 5.4 kg/m2 (non-sarcopenic group); or men with SMI < 7.0 kg/m2 and women with SMI < 5.4 kg/m2 (sarcopenic group).
There were significant positive correlations between SMI and peak oxygen uptake (VO2) (r = 0.631, P < 0.001), and between
SMI and peak VO2/heart rate (HR) (r = 0.683, P < 0.001). However, only peak VO2/HR significantly differed between groups in
both sexes. Multiple linear regression analyses with peak VO2/HR as a dependent variable showed that SMI was the only
independent determinant after adjusting for potential confounders. After 4 month follow-up of 47 participants, there was still
a significant positive correlation between SMI and peak VO2/HR (r = 0.567, P < 0.001), and between percent change of SMI
and percent change of peak VO2/HR (r = 0.305, P < 0.05).
Conclusions Peak VO2/HR, an index of stroke volume at peak exercise, was associated with SMI. This indicates that skeletal
muscle mass might affect cardiac function during exercise.

Keywords Community-dwelling older adults; Skeletal muscle mass index; Cardiopulmonary exercise testing; Peak oxygen pulse;
Sarcopenia
Received: 1 December 2016; Revised: 21 February 2017; Accepted: 25 February 2017
*Correspondence to: Masamitsu Sugie, Tokyo Metropolitan Geriatric Hospital and Institute of Gerontology, 35-2 Sakae-cho, Itabashi-ku, Tokyo 173-0015 Japan. Tel: 03-3964-1141;
Fax: 03-3964-1982. Email: masamitsu_sugie@tmghig.jp
This work was performed at the Tokyo Metropolitan Geriatric Hospital and Institute of Gerontology.

Introduction because of its association with significant morbidity and

mortality. Indeed, sarcopenia has recently been shown to be
Sarcopenia, that is, age-associated loss of muscle mass and strongly associated with increased mortality because of
strength, is highly prevalent in many ageing societies.1 It has cardiovascular disease in community-dwelling older adults,2
received much clinical and research attention in recent years and with an unfavourable prognosis in patients with chronic

© 2017 The Authors ESC Heart Failure published by John Wiley & Sons Ltd on behalf of European Society of Cardiology
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
2 M. Sugie et al.

heart failure (HF).3 Just like sarcopenia, chronic HF is highly 80 years (range 65–97 years). None of the subjects were
prevalent and a major cause of death in ageing populations.4 currently hospitalized, but all were being treated on an
Previous reports have suggested a relationship between outpatient basis at the Tokyo Metropolitan Geriatric Hospital
skeletal muscle mass and HF, particularly among patients with and Institute of Gerontology.
HF with preserved ejection fraction (HFpEF).5 However, the Exclusion criteria were as follows: unable to walk
clinical interrelationship between skeletal muscle mass and independently and required nursing care, impaired vision,
cardiac function remains to be insufficiently defined. This is impaired hearing, musculoskeletal impairments that might
particularly true in association with exercise. Therefore, it is interfere with the ability to perform the symptom-limited
of an importance in super-ageing societies to better define exercise test, a clinically unstable condition, significant
the relationship between reduction in muscle mass and cognitive disorders and less than 64 years old. Potential
strength associated with sarcopenia and changes in cardiac participants that performed habitual exercise training were
function that are prevalent in patients with chronic HF. The also excluded from the study. The clinical characteristics of
aim of this study was to investigate the relationship between the subjects are summarized in Table 1A. A follow-up
skeletal muscle mass and parameters of cardiac function in assessment was conducted with 47 participants 4 months
community-dwelling older subjects. after the baseline evaluation by using the methods and
procedures similar to those used at the baseline.

Methods
Skeletal muscle mass index and body mass index
Participants
Appendicular skeletal muscle mass (ASM) was measured
Sixty-three consecutive community-dwelling older adults (20 using total body dual-energy X-ray absorptiometry (DEXA,
men and 43 women) living in the Tokyo metropolitan area Lunar iDXA, GE Healthcare, Tokyo, Japan). Participants were
participated in this study. The mean age of subjects was positioned for whole-body scans in accordance with the

Table 1A Clinical characteristics of the 63 chronically ill participants

Participant characteristics
Male [n(%)] 20(32%)
Age [years; mean(range)] 79(65–97)
Male: 82(68–97)
Female: 79(65–93)
2
Physiological Assessment Body mass index, kg/m 22.2±3.5
Male: 21.8±3.5
Female: 22.4±3.5
Brachial-ankle pulse wave velocity, cm/min 1881±445
2
Skeletal muscle mass index, kg/m 5.98±0.9
Male: 6.3±1.1
Female: 5.8±0.7
Cardiopulmonary exercise test Peak VO2, mL/min 811±301
Peak VO2/weight, mL/min/weight 15.5±4.6
Peak VO2/heart rate, mL/beat 6.9±2.3
Peak heart rate, bpm 117±22
Peak watt 63±25
Anaerobic threshold VO2, mL/min 581±173
Anaerobic threshold VO2/weight, mL/min/weight 11.1±2.8
Anaerobic threshold VO2 /heart rate, mL/beat 5.9±1.7
Anaerobic threshold heart rate, bpm 99±13
Anaerobic threshold watt 38.1±14
⊿VO2/⊿LOAD, mL/watt 8.0±2.0
VE vs. VCO2 slope 35.7±11.1
Type of illness [n(%)] Hypertension 36(57%)
Dyslipidemia 28(44%)
Diabetes mellitus 19(30%)
Coronary artery disease 17(27%)
Chronic heart failure 9(14%)
Atrial fibrillation 8(13%)
Drug [n(%)] Calcium channel blocker 25(39%)
Beta-blocker 20(30%)
Angiotensin-converting enzyme inhibitor 14(22%)

VE vs. VCO2 slope, minute ventilation vs. carbon dioxide output slope; VO2, oxygen uptake.

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
Relationship between skeletal muscle mass and cardiac function 3

Table 1B Univariate correlations between a skeletal muscle mass previous study.9 The mean of the right and left baPWV values
index and age, body mass index, and the results of
cardiopulmonary exercise testing were used for analysis.

Correlation
Related factors coefficient P value
Cardiopulmonary exercise testing
Age
0.127 n.s.
Body mass index 0.770 P < 0.001
Brachial-ankle pulse wave velocity
0.278 P < 0.05 All patients underwent symptom-limited bicycle ergometer
Peak VO2 0.631 P < 0.001 cardiopulmonary exercise testing (CPET) using an upright,
Peak VO2/weight 0.274 P < 0.05
Peak VO2/heart rate 0.683 P < 0.001 electromagnetically braked, cycle ergometer (Aerobike
Peak heart rate
0.079 n.s. Strength Ergo-8, Mitsubishi Electronic, Tokyo, Japan), a
Peak watts 0.540 P < 0.001 metabolic analyser (Aeromonitor AE-310S, Minato Medical
Anaerobic threshold VO2 0.584 P < 0.001
Anaerobic threshold VO2/weight 0.150 n.s. Science, Osaka, Japan), and an electrocardiogram and heart
Anaerobic threshold VO2/heart rate 0.626 P < 0.001 rate (HR) (Stress test system ML-9000, Fukuda denshi, Tokyo,
Anaerobic threshold heart rate
0.017 n.s. Japan). The exercise test began with a 3 min rest on the
Anaerobic threshold watts 0.386 P < 0.01
ΔVO2/Δwork load 0.297 P < 0.05 ergometer followed by a 4 min warm-up at 0 W at 60 rpm.
VE vs. VCO2 slope
0.166 n.s. The load was then increased incrementally by 15 W/min
during the exercise test. All CPET parameters were measured
n.s., not significant; peak VO2/HR, peak oxygen uptake/heart rate;
VE vs. VCO2 slope, minute ventilation vs. carbon dioxide output from the beginning of the initial resting period on the cycle
slope; VO2, oxygen uptake. ergometer until the end of the exercise session.
P values were calculated using Student’s t-test. The CPET was terminated upon the patient’s request or if
abnormal physiologic responses occurred.10 The CPET was
also ceased if a patient was unable to continue to perform
manufacturer’s protocol. Participants lay in a supine position the pedalling exercise correctly. Oxygen uptake (VO2), carbon
on the DEXA table with limbs close to the body. The dioxide output (VCO2), minute ventilation (VE), tidal volume,
whole-body lean soft tissue mass was divided into several and frequency of respiration were smoothed with an
regions, that is, arms, legs, and the trunk. The sum of the 8-breath moving average. Peak VO2 was defined as the
muscle mass (lean soft tissue) of the four limbs was highest value of VO2 obtained during the last minute of the
considered as ASM, and the skeletal muscle mass index CPET. Peak watt was defined as the power at measured peak
(SMI) was calculated as ASM divided by the height in metres VO2 with CPET. VO2/HR, known as oxygen pulse, was
squared (kg/m2). Subjects were then divided into two groups calculated by dividing the moving averaged VO2 by the HR.
based on their SMI: men with an SMI ≥ 7.0 kg/m2 and women When respiratory exchange ratio (VCO2/VO2, RER) was less
with an SMI ≥ 5.4 kg/m2 (non-sarcopenic group), or men with than 1.0 at peak exercise, the test was considered insufficient
an SMI < 7.0 kg/m2 and women with an SMI < 5.4 kg/m2 because of the participant’s poor effort and the data at peak
(sarcopenic group). The threshold levels for group exercise were not used in the statistics. The anaerobic
assignment were based on the criteria of the Asian Working threshold was determined synthetically by gas exchange
Group for sarcopenia.6 Body mass index (BMI) was calculated criteria at the point of non-linear increase in the ventilatory
as bodyweight/height2 (kg/m2). equivalent for oxygen and the V-slope analysis (VCO2–VO2
plot). The slope of the VE–VCO2 relationship was calculated
by linear regression analysis using the values from the
beginning of ramp exercise to the respiratory compensation
Brachial-ankle pulse wave velocity measurement point during the CPET and was used as an index of the
ventilatory efficiency.
Participants were observed under quiet resting conditions in
the supine position. The brachial-ankle pulse wave velocity
(baPWV) and blood pressure were measured with a vascular Statistical analysis
testing device (form PWV/ABI device; BP-203PREIII, Omron
Colin, Kyoto, Japan), according to the method previously Pearson’s correlation analyses were performed to evaluate
described.7 Bilateral brachial and ankle arterial pressure the relationship between SMI and age, BMI, and cardiac
waveforms were stored for 10 s by the extremity cuffs function parameters during exercise, including peak VO2,
connected to a plethysmographic sensor and an oscillometric peak VO2/HR, peak watts, ΔVO2/Δwork load, and the VE vs.
pressure sensor wrapped around the participant’s arms and VCO2 slope. Comparisons of the clinical characteristics of
ankles. The baPWV was calculated from the distance patients in the non-sarcopenic and sarcopenic groups,
between the two arterial recording sites divided by the including BMI, and cardiac functional parameters during
transit time.8 The reproducibility of baPWV was shown in a exercise, were performed using unpaired Student’s t-test. In

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
4 M. Sugie et al.

addition, to examine the independent associations between Results

peak VO2/HR and SMI, we applied serial multiple linear
regression models with peak VO2/HR as dependent variable. We enrolled 63 patients, 68% were women, and their
All statistical analyses were performed using the Statistical mean age was 80 years. Patients’ baseline demographics
Package for the Social Sciences (SPSS version 22.0, IBM and medication are shown in Table 1A. A total of 24
Japan, Tokyo, Japan) and a two-tailed significance level was (38%) subjects fulfilled the criteria of the sarcopenic group
set at P < 0.05 for all tests. and 39 (62%) those of the non-sarcopenic group. The
sarcopenic group had a significantly lower mean peak
VO2 (mL/min) (692 ± 241 vs. 884 ± 313, P < 0.05) and
Ethical considerations
peak VO2/HR (mL/beat) (6.1 ± 1.8 vs. 7.4 ± 2.4,
This study was approved by the Ethics Committee of the Tokyo P < 0.05) than the non-sarcopenic group (Table 2). Of
Metropolitan Geriatric Hospital and Institute of Gerontology the 20 male subjects, 13 (65%) were in the sarcopenic
(Authorization Number: 240301) and conforms with the group and 7 (35%) were in the non-sarcopenic group. Male
principles outlined in the Declaration of Helsinki. All participants patients in the sarcopenic group had significantly lower
gave their written informed consent prior to data collection. mean values for peak VO2 (mL/min) (683 ± 254 vs.

Table 2 Comparison of clinical characteristics between the non-sarcopenic and sarcopenic groups

All Male Female

Non- Non- Non-
sarcopenic Sarcopenic sarcopenic Sarcopenic sarcopenic Sarcopenic
group group group group group group
(n = 39) (n = 24) P value (n = 7) (n = 13) P value (n = 32) (n = 11) P value
Age, years 80±7 80±7 n.s. 82±7 81±8 n.s. 79±7 78±7 n.s.
2
Body mass index, kg/m 23.8±3.0 19.7±2.5 <0.001 25.0±2.7 20.2±2.6 <0.01 23.6±3.1 19.0±2.4 <0.001
Brachial-ankle pulse wave 1846±437 1938±461 n.s. 1673±669 1939±331 n.s. 1884±372 1937±597 n.s.
velocity, cm/min
Peak VO2, mL/min 884±313 692±241 <0.05 1140±350 683±254 <0.01 828±280 704±236 n.s.
Peak VO2/weight, mL/min/weight 16.1±4.6 14.4±4.5 n.s. 17.8±5.3 12.8±4.3 <0.05 15.8±4.4 16.2±4.2 n.s.
Peak VO2/heart rate, mL/beat 7.4±2.4 6.1±1.8 <0.05 10.2±3.4 6.5±2.0 <0.01 6.8±1.6 5.6±1.6 <0.05
Peak heart rate, bpm 119±21 114±23 n.s. 114±23 104±22 n.s. 120±21 124±19 n.s.
Peak watts 68±26 53±21 <0.05 87±26 51±21 <0.01 64±25 56±20 n.s.
Anaerobic threshold VO2, mL/min 619±179 514±141 <0.05 700±114 513±125 <0.01 599±187 514±163 n.s.
Anaerobic threshold VO2/weight, 11.3±2.8 10.6±2.9 n.s. 11.5±2.7 9.4±2.7 n.s. 11.3±2.9 11.8±2.6 n.s.
mL/min/weight
Anaerobic threshold VO2/heart 6.2±1.7 5.2±1.5 <0.05 7.8±1.8 5.4±1.5 <0.01 5.9±1.5 5.1±1.7 n.s.
rate, mL/beat
Anaerobic threshold heart rate, 100±13 100±14 n.s. 92±16 98±16 n.s. 102±12 102±11 n.s.
bpm
Anaerobic threshold watt 39.7±14 35±12 n.s. 46.4±15 33.1±14 n.s. 38.1±14 37.2±10 n.s.
ΔVO2/Δwork load, mL/watt 8.1±2.2 8.0±1.7 n.s. 9.7±1.3 7.5±1.7 <0.01 7.7±2.2 8.5±1.7 n.s.
VE vs. VCO2 slope 35.7±10.4 35.8±12.6 n.s. 34.8±6.4 37.6±11.2 n.s. 35.9±11.1 34.0±14.1 n.s.

Hypertension (+) 26 10 n.s. 5 6 n.s. 21 4 n.s.

(
) 13 14 2 7 11 7
Dyslipidemia (+) 21 7 n.s. 3 4 n.s. 18 3 n.s.
(
) 18 17 4 9 14 8
Diabetes mellitus (+) 11 8 n.s. 1 5 n.s. 10 3 n.s.
(
) 28 16 6 8 22 8
Coronary artery disease (+) 9 8 n.s. 3 6 n.s. 6 2 n.s.
(
) 30 16 4 7 26 9
Chronic heart failure (+) 4 5 n.s. 1 4 n.s. 3 1 n.s.
(
) 35 19 6 9 29 10
Atrial fibrillation (+) 6 2 n.s. 0 2 n.s. 6 0 n.s.
(
) 33 22 7 11 26 11
Beta-blocker treatment (+) 11 9 n.s. 2 7 n.s. 9 2 n.s.
(
) 28 15 5 6 23 9

n.s., not significant; peak VO2/HR, peak oxygen uptake/heart rate; VE vs. VCO2 slope, minute ventilation vs. carbon dioxide output slope;
VO2, oxygen uptake.
Numerical data are expressed as mean ± SD.
P values were calculated using Student’s t-test.
Participants were classified as being in the non-sarcopenic group and sarcopenic group based on the Asian sarcopenia cut-off values for
2 2
muscle mass measurements (7.0 kg/m for men and 5.4 kg/m for women as measured by dual X-ray absorptiometry).6

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
Relationship between skeletal muscle mass and cardiac function 5

1140 ± 350, P < 0.01), peak VO2/HR (mL/beat) (6.5 ± 2.0 Figure 1 (A) Statistically significant positive correlation between skeletal
vs. 10.2 ± 3.4, P < 0.01), peak watts (W) (51 ± 21 vs. muscle mass index and peak oxygen pulse (r = 0.683, P < 0.001) in a
population of 63 chronically ill older adults. (B) Statistically significant
87 ± 26, P < 0.01), and ⊿VO2/⊿work load (mL/W) positive correlation between per cent change of skeletal muscle mass
(7.5 ± 1.7 vs. 9.7 ± 1.3, P < 0.01) than male patients in index and per cent change of peak oxygen pulse (r = 0.305, P < 0.05)
the non-sarcopenic group (Table 2). In contrast, 11 (26%) in a population of 47 chronically ill older adults after 4 months of exercise
of the 43 female subjects were in the sarcopenic group training.
and 32 (74%) were in the non-sarcopenic group. Female
patients in the sarcopenic group had lower peak VO2/HR
(mL/beat) values (5.6 ± 1.6 vs. 6.8 ± 1.6, P < 0.05) than
female patients in the non-sarcopenic group (Table 2).
Only peak VO2/HR and BMI significantly differed between
the two groups in both sexes. There were significant
positive correlations between SMI and peak VO2
(r = 0.631, P < 0.001), as well as between SMI and peak
VO2/HR (r = 0.683, P < 0.001) (Figure 1A). Moreover, there
were significant positive correlations between SMI and
peak watts (r = 0.540, P < 0.001), SMI and ⊿VO2/⊿work
load (r = 0.297, P < 0.05), and SMI and BMI (r = 0.770,
P < 0.001). Results of univariate correlation analyses are
shown in Table 1B.
Table 3 shows the results of multiple linear regression
analyses with peak VO2/HR as dependent variable. The linear
regression models show that SMI is an independent
determinant of peak VO2/HR after adjustment for potential
confounders (age, sex, baPWV, hypertension, dyslipidemia,
diabetes mellitus, coronary artery disease, chronic HF, atrial
fibrillation, and treatment with beta-blockers; B = 1.561;
P < 0.001).

Exercise capacity after exclusion of potential

confounders

Moreover, after excluding participants who had atrial

fibrillation and those being treated with beta-blockers
from the analyses, we performed parallel statistical
analyses of the data from the remaining 42 participants
(mean age 78 years, 26% men). Of these 42 participants,
15 (36%) were in the sarcopenic group, and 27 (64%)
were in the non-sarcopenic group. The results of this
subpopulation were similar to those obtained from all
63 subjects; there were significant positive correlations
between SMI and peak VO2/HR (r = 0.697, P < 0.001),
and patients in the sarcopenic group had significantly
lower peak VO2/HR (mL/beat) values compared with
patients in the non-sarcopenic group (6.2 ± 1.7 vs.
7.7 ± 2.2, P < 0.05).

VO2/HR (r = 0.567, P < 0.001). Twenty-seven subjects

Follow-up assessment were in the SMI-increasing group, and 20 subjects were
in the SMI-decreasing group. There was significant
The assessment after 4 months of follow-up using data positive correlations between percent change of SMI
from 47 participants, we found that there was still a and percent change of peak VO2/HR (r = 0.305,
significant positive correlations between SMI and peak P < 0.05) (Figure 1B).

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
6 M. Sugie et al.

Table 3 Multiple linear regression analysis with peak VO2/HR as the Moreover, it was previously reported that peak AVO2diff did
dependent variable
not change after exercise training in either the young or in
B β P value LCI UCI older adults.14 Based on the Fick principle and this previous
Skeletal muscle 1.561 0.625 <0.001 1.031 2.091 report, our results suggest that a reduction in SMI is one of
mass index the most important factors affecting the deterioration of
Age
0.103
0.304 <0.05
0.174
0.033
Sex
1.054
0.207 n.s.
2.158 0.051
peak SV.
2
R = 0.615 Recently, the relationship between sarcopenia and
cardiovascular disease has been recognized to be of a great
B, regression coefficient; LCI, lower 95% confidence interval; peak
VO2/HR, peak oxygen uptake/heart rate; UCI, upper 95% importance in super-ageing societies. Both sarcopenia and
confidence interval. chronic HF are highly prevalent in advanced ageing
Adjusted for conventional risk factors (age, sex, brachial-ankle societies.1,4 In particular, HFpEF has received much attention
pulse wave velocity, hypertension, dyslipidemia, diabetes mellitus,
coronary artery disease, chronic heart failure, atrial fibrillation, in recent years because of its high prevalence among older
and treatment with beta-blockers) in addition to peak VO2/HR. adults.15 It has been reported that HFpEF is associated with
P values were calculated using Student’s t-test. reduced lean body mass,5 and exercise intolerance is a
hallmark of both sarcopenia and HFpEF.16,17 The association
Discussion between exercise intolerance and a lower peak VO2 is
explained by the Fick principle. However, our finding that
Our study using data from unselected outpatients a geriatric peak VO2/HR, an index of peak SV correlated with SMI,
outpatient clinic in Japan shows that 37% of subjects suggests, for the first time, a relationship between cardiac
presented with reduced skeletal muscle mass that fulfills the functional reserve and muscle wasting. This may be the case
criteria of sarcopenia. Similarly, previous study reported that with the exercise intolerance in patients with HFpEF, which
the prevalence rate of sarcopenia for community-dwelling Phan et al. attributed to deterioration in peak SV.18 On the
Japanese women was less than 7% for ages 60–69 years, and other hand, Dhakal et al. reported that a reduction in peak
24% for ages 70–80 years, and the prevalence rate of AVO2diff was the cause of the exercise intolerance in
sarcopenia for community-dwelling Japanese men was less than HFpEF,19 although previous studies showed no changes in
33% for ages 60–69 years, and 47% for ages 70–85 years.11 peak AVO2diff with ageing.13
We also showed that skeletal muscle mass assessed using Both muscle wasting and HFpEF are associated with
DEXA scanning was a major determinant of exercise capacity exercise intolerance. Muscle wasting is associated with a
in elderly subjects, and this fact remained true after reduction in peak SV, whereas HFpEF is associated with a
restricting the analysis to those without beta-blocker use reduction in peak SV and/or peak AVO2diff. Thus, muscle
and those without atrial fibrillation. wasting in community-dwelling older adults might be one of
Skeletal muscle mass remained a major predictor of several possible phenotypes of ageing, which may
exercise capacity in both groups and determines the level of subsequently develop to HFpEF.
exercise that can be achieved in either group, even though There are several potential mechanisms that may underlie
sarcopenic subjects had overall lower peak VO2 values than the relationship between muscle wasting and deterioration
non-sarcopenic subjects. of cardiac function. It was known that the most evident
The loss of muscle mass that occurs with ageing is clinically metabolic explanation for muscle wasting is an imbalance
important because it leads to diminished muscle strength, between protein catabolism (e.g. members of the ubiquitin–
reduced exercise tolerance, and a decreased quality of life.12 proteasome system, myostatin, and apoptosis inducing
In the present study, SMI was positively correlated with VO2. factors) and anabolism (e.g. members of the ubiquitin–
This suggests that there is a relationship between muscle proteasome system, myostatin, and apoptosis inducing
wasting and exercise intolerance. However, ageing-related factors).20,21 Even more, it was known that the muscle
muscle wasting is thought to be sex-dependent. In the wasting in HF is also an imbalance between protein
present study, the only CPET parameter that was significantly catabolism and anabolism.22 Recently, Mangner and
correlated with SMI in both sexes was peak VO2/HR. colleagues show an animal model in that the antioxidative
In general, peak VO2/HR is calculated using the Fick and metabolic capacities are heterogeneous in their response
principle: to chronic HF between the diaphragm and quadriceps, but
similar activation of protein degradation pathways (e.g. the
Peak VO2 =HR ¼ Stroke Volume ðSVÞ ubiquitin–proteasome system) was evident in both
muscles.23 Ubiquitin–proteasome system is known as the
arterial-venous oxygen difference ðAVO2 diffÞ:
system that induces degradation of sarcomeric proteins
including cTnI,24 myosin heavy chain,25 and myosin-binding
Peak VO2/HR strongly correlates with peak stroke protein.26 These changes occur in both skeletal muscle and
volume,13 and therefore, it is considered an index of SV. cardiomyocytes. In addition, MuRF-1 affects fatty acid and

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
Relationship between skeletal muscle mass and cardiac function 7

glucose oxidation, as well as mitochondrial function in proteasome system. Secondly, we did not measure SV
cardiomyocytes.27 These mechanisms might underlie the directly.
relationship between muscle wasting and cardiac function, In conclusion, peak VO2/HR (an index of stroke volume
in both of deconditioning and reconditioning. at peak exercise) was strongly associated with skeletal
It was known that peripheral circulation significantly muscle mass. SMI was an independent determinant of peak
contribute to exercise intolerance in patients with chronic VO2/HR after adjustment for potential confounders. These
HF.28 Therefore, we evaluated baPWV as a parameter of results suggest that there is a bidirectional relationship
peripheral circulation in this study. We found that there is a between muscle wasting and cardiac function in
low negative correlation between SMI and baPWV, and there community-dwelling older adults. A large number of
is no difference of baPWV between sarcopenic group and longitudinal studies are needed to evaluate cardiac function
non-sarcopenic group (Table 3). Furthermore, in multiple over time and to prove a causal relationship between SMI
linear regression analyses, no relation was found between and peak VO2/HR.
baPWV and each of peak VO2/HR and SMI (Table 3). This
might be due to the difference of participant’s characteristics
between chronic HF patients in previous report28 and
community-dwelling older adults in this study. Conflict of interest
This study had several limitations. Firstly, we did not
measure the biomarker which related with ubiquitine– None declared.

References
1. Morley JE, Anker SD, von Haehling S. the Asian Working Group for 11. Abe T, Thiebaud RS, Loenneke JP, Loftin
Prevalence, incidence, and clinical Sarcopenia. J Am Med Dir Assoc 2014; M, Fukunaga T. Prevalence of site-
impact of sarcopenia: facts, numbers, 15: 95–101. specific thigh sarcopenia in Japanese
and epidemiology-update 2014. J 7. Yamashina A, Tomiyama H, Takeda K, men and women. Age (Dordr) 2014;
Cachexia Sarcopenia Muscle 2014; 5: Tsuda H, Arai T, Hirose K, Koji Y, Hori 36: 417–426.
253–259. S, Yamamoto Y. Validity, reproducibility, 12. Thomas DR. Loss of skeletal muscle
2. Atkins JL, Whincup PH, Morris RW, and clinical significance of noninvasive mass in aging: examining the
Lennon LT, Papacosta O, Wannamethee brachial-ankle pulse wave velocity relationship of starvation, sarcopenia
SG. Sarcopenic obesity and risk of measurement. Hypertens Res 2002; 25: and cachexia. Clin Nutr 2007; 26:
cardiovascular disease and mortality: a 359–364. 389–399.
population-based cohort study of older 8. Sugawara J, Hayashi K, Yokoi T, Cortez- 13. Ehsani AA, Spina RJ, Peterson LR,
men. J Am Geriatr Soc 2014; 62: 253–260. Cooper MY, DeVan AE, Anton MA, Rinder MR, Glover KL, Villareal DT,
3. Narumi T, Watanabe T, Kadowaki S, Tanaka H. Brachial-ankle pulse wave Binder EF, Holloszy JO. Attenuation of
Takahashi T, Yokoyama M, Kinoshita velocity: an index of central arterial cardiovascular adaptations to exercise
D, Honda Y, Funayama A, Nishiyama stiffness? J Hum Hypertens 2005; 19: in frail octogenarians. J Appl Physiol
S, Takahashi H, Arimoto T, Shishido T, 401–406. 2003; 95: 1781–1788.
Miyamoto T, Kubota I. Sarcopenia 9. Matsui Y, Kario K, Ishikawa J, Eguchi K, 14. Wang E, Næss MS, Hoff J, Albert TL,
evaluated by fat-free mass index is an Hoshide S, Shimada K. Reproducibility Pham Q, Richardson RS, Helgerud J.
important prognostic factor in patients of arterial stiffness indices (pulse wave Exercise-training-induced changes in
with chronic heart failure. Eur J Intern velocity and augmentation index) metabolic capacity with age: the role of
Med 2015; 26: 118–122. simultaneously assessed by automated central cardiovascular plasticity. Age
4. Hunt SA, Baker DW, Chin MH, pulse wave analysis and their associated (Dordr) 2014; 36: 665–676.
Cinquegrani MP, Feldman AM, Francis risk factors in essential hypertensive 15. Senni M, Tribouilloy CM, Rodeheffer RJ,
GS, Ganiats TG, Goldstein S, Gregoratos patients. Hypertens Res 2004; 27: Jacobsen SJ, Evans JM, Bailey KR,
G, Jessup ML, Noble RJ. ACC/AHA 851–857. Redfield MM. Congestive heart failure
guidelines for the evaluation and 10. Fletcher GF, Ades PA, Kligfield P, Arena in the community: a study of all incident
management of chronic heart failure in R, Balady GJ, Bittner VA, Coke LA, Fleg cases in Olmsted County, Minnesota, in
the adult: executive summary. J Am Coll JL, Forman DE, Gerber TC, Gulati M, 1991. Circulation 1998; 98: 2282–2289.
Cardiol 2001; 38: 2101–2113. Madan K, Rhodes J, Thompson PD, 16. Evans WJ, Campbell WW. Sarcopenia
5. Haykowsky MJ, Brubaker PH, Morgan Williams MA. American Heart Association and age-related changes in body
TM, Kritchevsky S, Eggebeen J, Kitzman Exercise, Cardiac Rehabilitation, and composition and functional capacity. J
DW. Impaired aerobic capacity and Prevention Committee of the Council on Nutr 1993; 123: 465–468.
physical functional performance in older Clinical Cardiology, Council on Nutrition, 17. Abudiab MM, Redfield MM, Melenovsky
heart failure patients with preserved Physical Activity and Metabolism, V, Olson TP, Kass DA, Johnson BD,
ejection fraction: role of lean body mass. Council on Cardiovascular and Stroke Borlaug BA. Cardiac output response to
J Gerontol A Biol Sci Med Sci 2013; 68: Nursing, and Council on Epidemiology exercise in relation to metabolic demand
968–975. and Prevention. Exercise standards in heart failure with preserved ejection
6. Chen LK, Liu LK, Woo J, Assantachai P, for testing and training: a scientific fraction. Eur J Heart Fail 2013; 15:
Auyeung TW, Bahyah KS, Chou MY, statement from the American Heart 776–785.
Chen LY, Hsu PS, Krairit O, Lee JS. Association. Circulation 2013; 128: 18. Phan TT, Abozguia K, Shivu GN,
Sarcopenia in Asia: consensus report of 873–934. Mahadevan G, Ahmed I, Williams L,

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158
8 M. Sugie et al.

Dwivedi G, Patel K, Steendijk P, Ashrafian Cachexia Sarcopenia Muscle 2014; 5: RING-finger 3 predisposes the heart to
H, Henning A. Heart failure with 193–198. cardiac rupture after myocardial
preserved ejection fraction is 22. von Haehling S, Steinbeck L, Doehner infarction. Proc Natl Acad Sci 2007; 104:
characterized by dynamic impairment of W, Springer J, Anker SD. Muscle 4377–4382.
active relaxation and contraction of the wasting in heart failure: an overview. 26. Mearini G, Gedicke C, Schlossarek S,
left ventricle on exercise and associated Int J Biochem Cell Biol 2013; 45: Witt CC, Krämer E, Cao P, Gomes MD,
with myocardial energy deficiency. J Am 2257–2265. Lecker SH, Labeit S, Willis MS,
Coll Cardiol 2009; 54: 402–409. 23. Mangner N, Weikert B, Bowen TS, Eschenhagen T. Atrogin-1 and MuRF1
19. Dhakal BP, Malhotra R, Murphy RM, Sandri M, Höllriegel R, Erbs S, regulate cardiac MyBP-C levels via
Pappagianopoulos PP, Baggish AL, Weiner Hambrecht R, Schuler G, Linke A, Gielen different mechanisms. Cardiovasc Res
RB, Houstis NE, Eisman AS, Hough SS, S, Adams V. Skeletal muscle alterations 2010; 85: 357–366.
Lewis GD. Mechanisms of exercise in chronic heart failure: differential 27. Patterson C, Willis MS, Portbury A. Rise
intolerance in heart failure with effects on quadriceps and diaphragm. J above: muscle ring-finger-1 (MURF1)
preserved ejection fraction: the role of Cachexia Sarcopenia Muscle 2015; 6: regulation of cardiomyocyte size and
abnormal peripheral oxygen extraction. 381–390. energy metabolism. Trans Am Clin
Circ Heart Fail 2014; DOI: 10.1161 24. Kedar V, McDonough H, Arya R, Li HH, Climatol Assoc 2011; 122: 70–81.
CIRCHEARTFAILURE.114.001825. Rockman HA, Patterson C. Muscle- 28. Manetos C, Dimopoulos S, Tzanis G,
20. Lenk K, Schuler G, Adams V. Skeletal specific RING finger 1 is a bona fide Vakrou S, Tasoulis A, Kapelios C,
muscle wasting in cachexia and ubiquitin ligase that degrades cardiac Agapitou V, Ntalianis A, Terrovitis J,
sarcopenia: molecular pathophysiology troponin I. Proc Natl Acad Sci USA Nanas S. Skeletal muscle
and impact of exercise training. J 2004; 101: 18135–18140. microcirculatory abnormalities are
Cachexia Sarcopenia Muscle 2010; 1: 25. Fielitz J, van Rooij E, Spencer JA, Shelton associated with exercise intolerance,
9–21. JM, Latif S, van der Nagel R, ventilatory inefficiency, and impaired
21. Palus S, von Haehling S, Springer S. Bezprozvannaya S, de Windt L, autonomic control in heart failure. J
Muscle wasting: an overview of recent Richardson JA, Bassel-Duby R, Heart Lung Transplant. 2011; 30:
developments in basic research. J Olson EN. Loss of muscle-specific 1403–1408.

ESC Heart Failure (2017)

DOI: 10.1002/ehf2.12158