Observational Study Medicine ®

Neuromuscular electrical stimulation in the

intensive care unit prevents muscle atrophy in
critically ill older patients
A retrospective cohort study
Tadayoshi Nonoyama, PTa,b,*  , Hiroko Shigemi, MD, PhDc, Masafumi Kubota, PT, PhDb,d,
Akihiko Matsumine, MD, PhDe, Kenji Shigemi, MD, PhDf, Tamotsu Ishizuka, MD, PhDa

Abstract
Critically ill patients in the intensive care unit (ICU) develop muscle atrophy and decreased physical function. Though neuromuscular
electrical stimulation (NMES) therapy has been shown to be effective in preventing this, but its effect on older patients is unknown.
To examine the course of critically ill older patients treated with NMES in the ICU and to define the impact of its use.
A retrospective cohort study was conducted using older ICU patients (≥65 years) categorized into a control group (n = 20) and
an NMES group (n = 22). For subgroup analysis, each group was further classified into pre-old age (65–74 years) and old age
(≥75 years).
The control group showed significant decrease in muscle thickness during ICU and hospital stay. The NMES group showed
lower reduction in muscle thickness and showed decrease in muscle echo intensity during hospital stay, compared to the control
group. NMES inhibited decrease in muscle thickness in the pre-old age group versus the old age group. The decreasing effect of
NMES on echo intensity during hospital stay manifested only in the pre-old age group. We did not find much difference in physical
functioning between the NMES and control groups.
Lower limb muscle atrophy reduces in critically ill older patients (≥65 years) with NMES and is pronounced in patients aged < 75
years. The impact of NMES on the physical functioning of older patients in ICU needs to be further investigated.
Abbreviations: ADL = activities of daily living, CFS = clinical frailty scale, CRP = C-reactive protein, FSS-ICU = functional
status score for the intensive care unit, ICU = intensive care unit, ICU-AW = ICU-acquired weakness, MRC sum score = Medical
Research Council sum score, NMES = neuromuscular electrical stimulation, SOFA = sequential organ failure assessment, SPPB
= short physical performance battery.
Keywords: intensive care unit, lower limb muscle atrophy, muscle thickness, neuromuscular electrical stimulation, older patients,
physical function

1. Introduction (ADL).[3] In addition, motor dysfunction persists for months

to years[4,5] and continues in 32% of patients.[6] Preventing
Critically ill patients admitted in an intensive care unit (ICU) muscle atrophy and ICU-AW is essential for avoiding long-
suffer from severe muscle weakness and atrophy referred to as term motor dysfunction. In older patients, muscle atrophy,
ICU-acquired weakness (ICU-AW). The incidence of ICU-AW muscle weakness, and physical functioning decline progres-
is reportedly 46% among critically ill patients,[1] and decreased sively with age, even before ICU admission. The prevalence of
muscle thickness is associated with the onset of ICU-AW and sarcopenia increases with age,[7] and skeletal muscle mass loss
physical functioning at discharge.[2] Most critically ill patients is associated with a greater decline in ADL.[8] In particular,
have new onset of frailty, which is independently associated age is a risk factor for ICU-AW,[9] and compared to younger
with mortality and impairment of activities of daily living patients, older patients are at a higher risk of muscle mass

This work was partly supported by a Grant-in-Aid from the Ministry of Education, of Fukui, Eiheiji, Fukui, Japan, and f Department of Anesthesiology, Faculty of
Culture, Sports, Science, and Technology (MEXT) of Japan KAKENHI, https:// Medical Sciences, University of Fukui, Eiheiji, Fukui, Japan.
www.jsps.go.jp/english/index.html, grant number 17K17439. * Correspondence: Tadayoshi Nonoyama, PT, 23-3 Matsuoka-Shimoaizuki, Eiheiji-
The datasets generated during and/or analyzed during the current study are not cho, Yoshida-gun, Fukui 910-1193, Japan (e-mail: [email protected]).
publicly available, but are available from the corresponding author on reasonable Copyright © 2022 the Author(s). Published by Wolters Kluwer Health, Inc.
request. This is an open access article distributed under the Creative Commons
Supplemental Digital Content is available for this article. Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
a
Third Department of Internal Medicine, Faculty of Medical Sciences, University
of Fukui, Eiheiji, Fukui, Japan, b Department of Rehabilitation, University of Fukui How to cite this article: Nonoyama T, Hiroko Shigemi P, Kubota M, Matsumine A,
Hospital, Eiheiji, Fukui, Japan, c Division of Infection Control and Prevention, Shigemi K, Ishizuka T. Neuromuscular electrical stimulation in the intensive care
Faculty of Medical Sciences, Kyoto Prefectural University of Medicine, Kyoto, unit prevents muscle atrophy in critically ill older patients: A retrospective cohort
Japan, d Department of Physical Therapy, Graduate Course of Rehabilitation study. Medicine 2022;101:31(e29451).
Science, School of Health Sciences, College of Medical, Pharmaceutical, and Received: 27 August 2021 / Received in final form: 21 April 2022 / Accepted:
Health Sciences, Kanazawa University, Kanazawa, Japan, e Department of 22 April 2022
Orthopedics and Rehabilitation Medicine, Faculty of Medical Sciences, University
http://dx.doi.org/10.1097/MD.0000000000029451

1
Nonoyama et al.  •  Medicine (2022) 101:31Medicine

loss and ICU-AW due to pre-existing sarcopenia and anabolic appropriate analgesia and sedation management, bedside exer-
resistance.[10] cises, walking practice during ventilatory control, and walking
Neuromuscular electrical stimulation (NMES) is used to treat in the hospital using a wheelchair were conducted. After dis-
skeletal muscle disorders in critically ill patients. Reportedly, charge from the ICU, rehabilitation was continued until dis-
NMES reduces muscle atrophy in critically ill patients,[11–13] charge from the hospital.
decreases the incidence of ICU-AW,[14] improves muscle strength
early, and reduces the duration of mechanical ventilation.[15] The
effects of NMES on older people have been reported, including 2.3. Neuromuscular electrical stimulation
an increase in cross-sectional areas of muscle tissue and muscle For NMES, belt electrode skeletal muscle electrical stimulation
strength[16] and improvements promoting postural control.[17,18] (G-TES; Homer Ion Corp., Osaka, Japan) was applied. Belt elec-
However, there are no reports on its effects on older patients trodes were attached to the patient’s proximal and distal thighs
admitted to the ICU, and such effects are unclear. In addition to and ankles (Fig. 1). The NMES settings were set at a frequency
quantitative assessment, qualitative assessment of muscle mass of 20 Hz, a pulse width of 250 μs, and a duty cycle of 5 seconds
is also attracting attention. Decreased muscle quality in criti- of stimulation followed by 2 seconds of pause. The intensity of
cally ill patients is assessed by enhanced muscle echo intensity electrical stimulation was set at maximum intensity without
on ultrasound images.[19,20] However, no studies have examined pain, and muscle contraction was confirmed by inspection and
the effect of NMES on muscle quality, and the recovery rate palpation. NMES was started within 48 hours of ICU admission
of physical functioning is worse in older persons than in the and applied once a day for 30 minutes, 5 days/wk, and only
young.[5] Although frailty is a major concern in older adults fol- during ICU stay. The inclusion criteria for NMES were patients
lowing ICU discharge, the preventive effect of NMES has not who were expected to stay in the ICU for >48 hours. The var-
been reported. ious contraindications were as follows: wearing a pacemaker,
We hypothesized that NMES in critically ill older ICU use of intra-aortic balloon pumping or support such as percu-
patients effectively reduces lower limb muscle atrophy, pro- taneous cardiopulmonary support, skin damage at the electrode
motes muscle quality and muscle strength, and improves phys- site, administration of muscle relaxants, and presence of deep
ical functioning at the time of hospital discharge. Therefore, vein thrombosis.
as a preliminary study, we retrospectively examined the course
of critically ill older patients treated with NMES in the ICU
to define the effects of NMES on muscle thickness (muscle 3. Outcome Measures
volume), echo intensity of muscle (muscle quality), muscle
strength, and physical function and evaluated its impact on The primary outcome was muscle thickness, and secondary
older patients multilaterally. outcomes were echo intensity of muscle, grip strength, Medical
Research Council (MRC) sum score, functional status score
for the ICU (FSS-ICU), and short physical performance battery
2. Material and Methods (SPPB).

2.1. Study design

This is a single-center retrospective cohort study, and data were 3.1. Basic characteristics
extracted from medical records and databases. The subjects The data collected as basic characteristics included age, gen-
were critically ill older ICU patients (≥65 years) in a Japanese der, height, weight (at the time of ICU admission), clinical
tertiary hospital between June 1, 2018, and August 31, 2019. frailty scale (CFS) before admission, sequential organ fail-
Echo measurements and physical function assessments were ure assessment (SOFA) score (value on ICU admission, peak
started with all patients on June 1, 2018, and NMES was intro- value), C-reactive protein (CRP) peak (value on ICU admis-
duced in the ICU on June 1, 2018. Therefore, patients admitted sion, peak value), emergency or plans, main diagnosis, num-
to the ICU between June 1, 2018, and January 31, 2019, before ber of days in the ICU and hospital, and days of mechanical
the introduction of NMES, formed the control group. Patients ventilation.
admitted to the ICU between February 1, 2019, and August 31,
2019, after NMES was introduced, were included in the NMES
group.
The exclusion criteria were as follows: ICU stay <72 hours,
mechanical ventilation <72 hours, not eligible for early mobi-
lization and rehabilitation protocols, motor paralysis before
hospitalization or ICU admission, case of bone fracture, case
of death, patients with missing data, patients who failed to
start NMES within 48 hours, those who had fewer than 3
NMES sessions, and those with swelling, wounds, or skin
disorders at the measurement site of the ultrasound. As this
was a retrospective study, patients with missing data were
excluded.

2.2. Early mobilization

Early rehabilitation was initiated within 48 hours and performed
for all the subjects by a team comprising physicians, nurses, and
clinical engineering technicians, led by a physical therapist ded-
icated to the ICU. This therapy was performed according to the
early mobilization and rehabilitation protocols of the hospital
ICU (Supplemental Digital Content 1, http://links.lww.com/ Figure 1.  Neuromuscular electrical stimulation. Belt electrodes were attached
MD/G890) specifically to ensure standardized rehabilitation. A to the proximal and distal thighs and ankles.
patient’s mobilization target was set at daily conferences. Under

2
Nonoyama et al.  •  Medicine (2022) 101:31www.md-journal.com

3.2. Muscle thickness and muscle quality a dedicated ICU physiotherapist who was technically trained
An ultrasound imaging system (Vscan Extend, GE Healthcare) in the use of ultrasound equipment and had >2 years of expe-
was used to measure muscle thickness and echo intensity. The rience performing measurements.
measurement positions were supine, mid-hip, and knee exten-
sion. The measurement site was the midpoint of the line con- 3.3. Muscle strength, basic motor function, and physical
necting the superior anterior iliac spine and the upper patella
(Fig. 2A). A 7.5-MHz linear probe was used to record the trans- performance
verse ultrasound layer images when the deep probe was in per- At the time of discharge from the ICU and hospital, MRC sum
pendicular contact with the skin surface. Excess gel was used score, grip strength (measured twice; and maximum value was
to minimize image distortion due to skin indentation caused used), and FSS-ICU were assessed for muscle strength and basic
by the pressure of the ultrasound probe.[19,20] Muscle thickness movement ability, respectively. SPPB was used to evaluate phys-
was measured from the subcutaneous fat just above the femur ical performance at the time of discharge from the hospital.
(Fig. 2B), and echo intensity was used to evaluate muscle quality. The ICU-AW was defined as an MRC sum score of < 48 points,
Ultrasound transverse layer images of the rectus femoris muscle frailty was defined as an SPPB score of <9 points,[23] and the
were measured with the Image J software (ver 1.53, National onset of ICU-AW and frailty was evaluated.
Institutes of Health, the United States) using the same images
as those used for measuring muscle thickness. The measurement
area was the region of interest (excluding fascia) of the rectus 3.4. Sample size
femoris muscle (Fig.  2C), which was evaluated using an 8-bit The study sample size was calculated by the method described
gray-scale in the range of 0 to 255, and the average value was by Gerovasili et al,[11] who reported that the rate of change in
used. In this connection, the echoes are attenuated in the deeper rectus femoris muscle thickness was −8.0 ± 3.9% in the NEMS
layers of tissue.[21] It is difficult to use the echo intensity of deep group and −13.9 ± 6.4% in the control group. GPower software
muscle as a qualitative evaluation. Therefore, the echo intensity (version 3.1) was used to determine the minimum sample size
of the vastus medialis muscle was not measured in this study. required for the study. An alpha level of 5% with a statistical
For subgroup analysis, to investigate the effect of age on power of 90% was included in the power analysis. Thus, a min-
muscle atrophy and muscle quality, the control and NMES imum of 18 patients were required per group.
groups were further classified into 2 groups, consistent with
the definition proposed by the Joint Committee of Japan
Gerontological Society and the Japan Geriatrics Society on the 3.5. Statistical analysis
definition and classification of older people[22]: a pre-old age
group (65–74 years old) and an old age group (≥75 years). Among the data collected, the Shapiro–Wilk test was used to
In each group, muscle thickness and echo intensity were mea- check continuous variables for normality. Normally distrib-
sured at the time of admission to the ICU, discharge from the uted indicators were expressed as mean ± standard deviation,
ICU, and discharge from the hospital, and the changes were and indicators that did not show normal distribution were
examined. In addition, the reduction rate in muscle thickness expressed as median (interquartile range). Category variables
during ICU stay (from ICU admission to ICU discharge) and were expressed as frequencies (%). The 95% confidence interval
hospital stay (from ICU admission to hospital discharge) were (95% CI) for the percentage reduction rate of muscle thickness
calculated and compared between the control and NMES was also shown. A one-way analysis of variance (ANOVA) and
groups. The formula for the reduction rate of muscle thick- the Tukey post hoc test were performed to compare the changes
ness was as follows: reduction rate of muscle thickness during in muscle thickness and echo intensity in each group at the time
ICU stay = (ICU discharge − ICU admission)/ICU admission × of ICU admission and discharge from the ICU and hospital. The
100. The ultrasound muscle measurements were performed by unpaired t-test or Mann–Whitney U test compared the reduction

Figure 2.  Measurement by ultrasound imagery. (A) Measurement site; (B) measurement of muscle thickness; (C) measurement area of echo intensity.

3
Nonoyama et al.  •  Medicine (2022) 101:31Medicine

rate of muscle thickness and echo intensity, muscle strength, and patients; in the NMES group, step 1 included 1 patient, step 2
physical function between the 2 groups. A chi-square test was included 3 patients, step 3 included 3 patients, step 4 included
used for the qualitative variables of patient characteristics. The 3 patients, and step 5 included 2 patients. The subjects were
IBM SPSS software (version 27.0) was used for performing all followed up until they were discharged from the hospital. The
measurements and calculations. A P-value <.05 was considered follow-up period was 39.4 ± 17.7 days for the control group
statistically significant. and 41.6 ± 20.5 days for the NMES group. Age, gender, height,
weight (at the time of ICU admission), CFS before admission,
SOFA score at ICU admission, peak SOFA score, peak CRP,
3.6. Data analysis admission type, surgical/internal medicine, number of days
in the ICU, number of days in the hospital, and number of
The Research Ethics Committee of University of Fukui approved
patients in each group were evaluated. There were no signifi-
the study protocol (Assurance No. 20190136).
cant differences in the number of days in the ICU, number of
days in the hospital, or number of ventilator days. The con-
trol group comprised 7 patients in the pre-old age group and
4. Results 13 patients in the old age group. The NMES group had 10
4.1. Patient characteristics patients in the pre-old age group and 12 patients in the old
age group. Data on clinical and demographic characteristics
A total of 273 subjects were considered for analysis during showed no significant differences for the pre-old age group in
the study period; 231 patients met the exclusion criteria, leav- the control and NMES groups and the old age group in the
ing 42 patients for analysis (Fig. 3). Table 1 presents the clin- control and NMES groups (Supplemental Digital Content 2,
ical and demographic characteristics of the patients. Twenty http://links.lww.com/MD/G890).
patients in the control group (78.0 [72.8 − 81.5] years, 13
men and 7 women) and 22 patients in the NMES group (78.0
[72.0 − 84.0] years, 22 men and 7 women) were included 4.2. Muscle thickness
in the analysis. The number of patients at each stage of the
early mobilization protocol at 48 hours after admission to 4.2.1. Comparison of muscle thickness.  Muscle thickness at
the ICU was as follows: in the control group, step 1 included ICU admission, discharge, and hospital discharge showed no
11 patients, step 3 included 2 patients, and step 4 included 7 significant difference between the control and NMES groups

Before the introduction of NMES After the introduction of NMES

(from June 1, 2018 to January 31, 2019) (from February 1, 2019 to August 31, 2019)
n=151 n=122

Exclusion Exclusion
ICU stay < 72 hours n=49 ICU stay < 72 hours n=17

Mechanical ventilation < 72 hours n=48 Mechanical ventilation < 72 hours n=41

Not eligible for early mobilisation Not eligible for early mobilisation
and rehabilitation protocols n=13 and rehabilitation protocols n=16

Motor paralysis before Motor paralysis before

hospitalisation or ICU admission n=4 hospitalisation or ICU admission n=4

Bone fracture n=1 Bone fracture n=0

Death n=6 Contraindications of NMES n=8

Swelling, wounds, or skin disorders n=0 Swelling, wounds, or skin disorders n=0
Missing data n=10 Death n=4

NMES during ICU stay n=32

Exclusion
Failed to start NMES within 48 hours n=3
Fewer than 3 NMES sessions n=2
Death n=1
Missing data n= 4

Control group (n=20) NMES group (n=22)

Figure 3.  Flow diagram of the study. ICU = intensive care unit, NMES = neuromuscular electrical stimulation.

4
Nonoyama et al.  •  Medicine (2022) 101:31www.md-journal.com

Table 1
Clinical and demographic characteristics of patients

Control group, n = 20 NMES group, n = 22 P value

Age, y, median [IQR] 78.0 [72.8 − 81.5] 78.0 [72.0 − 84.0] .71
Male, n (%) 13 (65.0) 15 (68.2) .82
Height, cm, mean ± SD 157.5 ± 10.2 158.6 ± 11.1 .73
Weight, kg, mean ± SD 56.4 ± 12.5 55.9 ± 13.2 .90
CFS, point 2.7 ± 1.1 3.1 ± 3.1 .20
SOFA, mean ± SD
 ICU admission 9.0 ± 2.9 9.1 ± 3.6 .90
 Peak 10.1 ± 2.6 10.2 ± 3.2 .93
CRP peak, mL/dL, mean ± SD 20.4 ± 7.7 19.4 ± 10.2 .72
Type of ICU admission
 Emergency, n (%) 14 (70.0) 16 (72.7) .84
Main diagnosis, n (%)
 Sepsis 6 (30.0) 10 (45.5) .30
 Abdominal/pelvic surgery 8 (40.0) 7 (31.8) .58
 Cardiac surgery 5 (25.0) 3 (13.6) .35
 Thoracic surgery 3 (15.0) 5 (22.7) .52
 Respiratory failure 1 (5.0) 3 (13.6) .34
 Heart failure 0 1 (4.5) .33
 Others 0 1 (4.5) .33
ICU stay, d, mean ± SD 12.0 ± 5.5 12.6 ± 5.9 .70
Hospital stay, d, median [IQR] 33.5 [28.5 − 41.0] 38.0 [26.3 − 49.0] .71
Mechanical ventilation, d, median [IQR] 8.5 [4.0 − 10.0] 8.5 [5.3 − 15.8] .10
No significant differences were observed in the parameters between the control and NMES groups.
Normally distributed indicators were expressed as mean ± standard deviation, and indicators that did not show normal distribution were expressed as medians (interquartile range [IQR]).
CFS = clinical frailty scale, CRP = C-reactive protein, ICU = intensive care unit, SOFA = sequential organ failure assessment.

(control group: admission 24.5  ± 6.1 mm; discharge from ICU stay (−21.8% [95% CI −36.5 to −7.2] vs −4.7% [95% CI
ICU 19.5 ± 6.1 mm; discharge from hospital 17.8 ± 5.7 mm, −13.8 to −4.3], P < .05) and hospital stay (−29.3% [95% CI
NMES group: admission 21.3 ± 6.0 mm; discharge from ICU −42.2 to −16.4] vs −8.1% [95% CI −23.5 to −7.2], P < .05;
19.5 ± 5.9 mm; discharge from hospital 18.0 ± 6.1 mm). Fig. 5B, E). The significantly suppressed reduction rate in muscle
In the pre-old age group, muscle thickness at ICU admission thickness observed in the NMES group was not identified in the
was significantly lower in the NMES group compared to the old age group (Fig. 5C, F).
control group (control group 29.4 ± 4.1 mm vs NMES group
20.4 ± 5.9 mm, P < .05; Supplemental Digital Content 3, http:// 4.3. Echo intensity of muscle
links.lww.com/MD/G890).
In the control group, no major change in echo intensity was
4.2.2. Change in muscle thickness.  A significant decrease was observed (Fig. 6A–C). In the NMES group, the echo intensity
observed in muscle thickness at the time of discharge from the of muscle decreased significantly at the time of hospital dis-
ICU and hospital compared to ICU admission. charge compared to ICU admission (admission 69.3 ± 13.2 AU
vs discharge from hospital 58.0 ± 14.1 AU, P < .05; Fig. 6D).
1. Control group: admission 24.5 ± 6.1 mm versus discharge Notably, NMES reduced the echo intensity of muscle dras-
from ICU 19.5 ± 6.1 mm; P < .05 versus discharge from tically in the pre-old age group; however, the same was not
hospital 17.8 ± 5.7 mm, P < .01 (Fig. 4A). observed in the old age group (admission 69.2 ± 13.6 AU vs
2. Pre-old age control group: admission 29.4 ± 4.1 mm ver- discharge from hospital 51.8 ± 14.2 AU, P < .05; Fig. 6E, F).
sus discharge from ICU 22.5 ± 4.0 mm; P < .05 versus dis- The echo intensity of muscle at ICU admission, discharge, and
charge from hospital 20.7 ± 5.6 mm, P < .01 (Fig. 4B). hospital discharge showed no significant difference between
3. Old age control group: admission 21.9 ± 5.4 mm, dis- the control and NMES groups (control group: admission
charge from hospital 16.2 ± 5.2 mm, P < .05 (Fig. 4C). 66.7 ± 16.0 AU; discharge from ICU 65.1 ± 15.7 AU; dis-
As determined by the one-way ANOVA, no significant dif- charge from hospital 62.8 ± 15.7 AU, NMES group: admission
ferences were observed between the mean values of muscle 69.3 ± 13.2 AU; discharge from ICU 63.6 ± 15.1 AU; discharge
thickness at the 3 assessment time points in the NMES group from hospital 58.0 ± 14.1 AU). In the pre-old age group, echo
(Fig. 4D) or when the subjects were divided into pre-old and old intensity of muscle at ICU discharge and hospital discharge
age groups (Fig. 4E, F). was significantly lower in the NMES group compared to the
control group (Control group 73.5 ± 14.2 AU vs NMES group
4.2.3. Reduction rate in muscle thickness.  The reduction rate 51.8 ± 14.2 AU, P < .05; Supplemental Digital Content 3,
in muscle thickness during ICU and hospital stay was −19.8% http://links.lww.com/MD/G890).
(95% CI −27.7 to −11.8) and −26.8% (95% CI −34.6 to −19.0)
in the control group; −7.9% (95% CI −14.1 to −1.0) and −14.5%
(95% CI −23.7 to −5.3) in the NMES group, respectively. The 4.4. Physical function and onset of ICU-AW and frailty
decline in muscle thickness during ICU stay was significantly Table  2 presents details of physical functioning and onset of
lower in the NMES group, compared to the control group. The ICU-AW and frailty. There were no significant differences in the
reduction rate in muscle thickness during hospital stay was FSS-ICU, grip strength, or MRC sum score at the time of ICU
suppressed drastically in the NMES group (P < .05; Fig.  5A, discharge or hospital discharge or in SPPB score at hospital dis-
D). The pre-old age group showed a considerably suppressed charge. In addition, there was no significant difference observed
reduction rate in muscle thickness in the NMES group during in the onset of ICU-AW and frailty.

5
Nonoyama et al.  •  Medicine (2022) 101:31Medicine

A B C
** ** *
40 * 40 * 40
MT (mm)

30 30 30

20 20 20

10 10 10

0 0 0
I II III I II III I II III

D E F
40 40 40
MT(mm)

30 30 30

20 20 20

10 10 10

0 0 0
I II III I II III I II III
Figure 4.  Change in muscle thickness. A significant decrease in muscle thickness was observed in the control group at the time of ICU and hospital discharge,
compared to ICU admission (A). In the pre-old age group (B), a significant decrease in muscle thickness was observed at hospital discharge compared to ICU
admission (C). The NMES group suppressed the decrease in muscle thickness (D–F). (A) All subjects in the control group; (B) pre-old age group of the control
group; (C) old age group of the control group; (D) all subjects of the NMES group; (E) pre-old age group of the NMES group; (F) old age group of the NMES
group; I. ICU admission; II. ICU discharge; III. Hospital discharge. *P < .05, **P < .01. ICU = intensive care unit, MT = muscle thickness, NMES = neuromuscular
electrical stimulation.

A B C
20 20 20
10 10 10
0 0 0
RR of MT (%)

-10 -10 -10

-20 -20 -20
-30 -30 -30
-40 -40 -40
-50 * -50 * -50 N.S.

D E F
20 20 20
10 10 10
0 0 0
RR of MT (%)

-10 -10 -10

-20 -20 -20
-30 -30 -30
-40 -40 -40
-50 -50 -50
* * N.S.

Control group NMES group

Figure 5.  Reduction rate in muscle thickness. Application of NMES inhibited decrease in muscle thickness during ICU stay (A) and hospital stay (D). This inhi-
bition was observed in the pre-old age group (B, E) but not the old age group (C, F). (A) All subjects during ICU stay; (B) pre-old age group during ICU stay; (C)
old age group during ICU stay; (D) all subjects during hospital stay; (E) pre-old age group during hospital stay; (F) old age group during hospital stay. *P < .05,
MT = muscle thickness, NMES = neuromuscular electrical stimulation, N.S.= not significant, RR = reduction rate.

6
Nonoyama et al.  •  Medicine (2022) 101:31www.md-journal.com

A B C
120 120 120
EI (AU)

90 90 90
60 60 60
30 30 30
0 0 0
I II III I II III I II III

D E F
120 * 120 * 120
EI (AU)

90 90 90
60 60 60
30 30 30

0 0 0
I II III I II III I II III
Figure 6.  Change in echo intensity of muscle. Muscle echo intensity did not change in the control group (A, B, C) during the hospital stay. It decreased during the
hospital stay in the NMES group (D). A significant decrease was observed in the pre-old age group (E) but not the old age group (F). (A) All subjects in the control
group; (B) pre-old age group of the control group; (C) old age group in control group; (D) all subjects of the NMES group; (E) pre-old age group of the NMES group;
(F) old age group of the NMES group; I. ICU admission; ICU discharge; Hospital discharge. *P < .05. EI = echo intensity, NMES = neuromuscular electrical stimulation.

Table 2
Muscle strength, basic motor function, physical performance, and onset of ICU-AW and frailty.

Control group, n = 20 NMES group, n = 22 P value

MRC sum score, median [IQR]
 Discharge from ICU
   Upper extremity 25.0 [22.8−28.0] 24.0 [22.2−27.5] .54
   Lower extremity 23.5 [21.8−26.5] 26.0 [21.3−26.8] .89
  Total score 49.0 [44.8−56.0] 47.6 [42.0−54.0] .53
 Discharge from Hospital
   Upper extremity 27.0 [24.0−28.5] 28.0 [24.0−30.0] .40
   Lower extremity 26.0 [23.8−28.0] 27.5 [24.5−28.0] .63
  Total score 52.0 [48.0−56.3] 53.2 [49.3−58.0] .49
Grip strength, kg, mean ± SD
 Discharge from ICU 17.8 ± 6.4 16.2 ± 7.3 .45
 Discharge from Hospital 18.7 ± 6.1 18.5 ± 7.0 .93
FSS − ICU, median [IQR]
 Discharge from ICU
   Rolling 3.5 [3.0−6.0] 3.0 [2.0−5.0] .35
  Supine to sit transfer 4.0 [3.0−4.5] 3.0 [2.3−4.8] .33
  Sitting on the edge of bed 5.5 [5.0−7.0] 5.0 [5.0−7.0] .53
  Sit to stand transfer 6.0 [4.0−6.0] 4.5 [3.0−6.0] .33
  Walking 5.0 [3.0−6.0] 4.5 [3.0−5.0] .41
  Total score 22.9 [19.3−27.3] 21.0 [13.3−27.5] .45
 Discharge from Hospital
   Rolling 7.0 [6.0−7.0] 7.0 [6.0−7.0] .55
  Supine to sit transfer 7.0 [6.0−7.0] 6.0 [6.0−7.0] .23
  Sitting on the edge of bed 7.0 [7.0−7.0] 7.0 [7.0−7.0] .12
  Sit to stand transfer 7.0 [6.0−7.0] 6.0 [6.0−7.0] .22
  Walking 6.5 [6.0−7.0] 6.0 [6.0−7.0] .65
  Total score 24.0 [19.3−27.3] 20.5 [13.3−27.5] .42
SPPB, median [IQR]
   Balance 4.0 [1.8−4.0] 3.0 [2.0−4.0] .95
   Gait 3.5 [3.0−4.0] 3.0 [3.0−4.0] .83
   Stand up 3.0 [0−3.3] 1.0 [0−3] .42
  Total score 9.5 [5.0−11.0] 7.5 [6.0−9.8] .57
ICU-AW, n (%)
 Discharge from ICU 9 (45.0) 11 (50.0) .77
Discharge from hospital 4 (20.0) 3 (13.6) .69
Frailty, n (%) 10 (50.0) 13 (59.1) .55
No significant differences were observed between the 2 groups in any of the parameters.
Normally distributed indicators were expressed as mean ± standard deviation, and indicators that did not show normal distribution were expressed as medians (IQR).
FSS-ICU = functional status score for the intensive care unit, ICU = intensive care unit, IQR = interquartile range, ICU-AW = intensive care unit-acquired weakness, MRC sum score = Medical Research
Council sum score, SPPB = short physical performance battery.

7
Nonoyama et al.  •  Medicine (2022) 101:31Medicine

5. Discussion In the NMES group, echo intensity of skeletal muscle

Our study shows that NMES inhibited reduction in muscle decreased at the time of hospital discharge rather than at
thickness during ICU stay and prevented muscle atrophy in crit- admission. An examination of the results by age group shows a
ically ill older patients. With the aging population and advances significant decrease in the pre-old age group but not in the old
in medical technology, the number of older patients receiving age group, similar to the trend in muscle thickness. Echo inten-
intensive care is increasing. Muscle atrophy in the ICU may sity in skeletal muscle reflects intramuscular adipose tissue and
accelerate the onset and progression of sarcopenia and frailty, interstitial fibrous tissue,[32,33] and echo intensity increases in
leading to an increased burden of care and medical costs. Thus, atrophied muscles.[19] In addition, echo intensity diminishes
this is a social problem and an urgent issue. The fact that NMES with strength training,[34] enabling the capture of qualitative
effectively prevents muscle atrophy in older ICU patients is of changes in muscles. In this study, changes were observed only
significance because it will help address these problems. As a in the NMES group, reflecting the qualitative improvement
mechanism by which NMES was effective in inhibiting mus- in muscle due to the treatment. However, it is unlikely that
cle atrophy, the studies on the effects of NMES with a large intramuscular fat would be reduced at the time of hospital
number of older subjects reported a decrease in branched-chain discharge, compared to at admission. It is hypothesized that
amino acids[13] and a decrease in urinary excretion of 3-meth- muscle quality had already decreased at the time of first ultra-
ylhistidine,[24] an indicator of protein catabolism; the same was sound imaging evaluation compared to before hospitalization.
expected to be true when only older subjects are included in the The subsequent improvement process may have been captured.
study. However, when subjects were categorized into the pre-old Future studies can verity the effects on qualitative changes in
and old age groups, the effect was observed only in the pre- muscles.
old age group and not in the old age group. It is speculated One limitation of this study is that the data were collected
that age-related decreases in growth hormone and insulin-like retrospectively and thus the methodological quality is not
growth factor-1 concentrations[25] may influence NMES-related high enough. Although there were no significant differences in
inhibition of muscle atrophy in older patients; however, the patient background between the 2 groups, the possibility of bias
factors are unclear, as differences in muscle protein catabolism due to unmeasured factors cannot be ruled out because this was
and responsiveness to NMES in different age groups of older not a randomized controlled trial. The study is based on data
patients are not known. This mechanism needs to be clarified collected at a single hospital, and the small number of subjects
in the future. and especially the small sample size for the subgroup analysis
There were no significant differences in muscle strength may make it unreliable, although significant differences were
and physical functioning at all of the endpoints between the found. The large number of septic and postoperative patients
NMES and control groups. Although some reports include introduced disease bias. Therefore, external validity is limited;
many older people in the target,[12] there are no detailed eval- further case data collection and prospective multicenter studies
uations of the impact of EMS on physical functioning. In pre- are needed to confirm the efficacy of NMES in elderly critically
vious reports that showed significant improvement in muscle ill patients admitted to the ICU. In addition, only 1 person was
strength by NMES, no active rehabilitation was performed in measured by ultrasound and was not blinded. Future prospec-
the control group.[15,26,27] In contrast, early rehabilitation was tive studies should be blinded by multiple evaluators to ensure
performed in the control group in a study that did not demon- higher quality. We used muscle thickness as an indicator of mus-
strate improvement in muscle strength after exercise intensifi- cle atrophy, and it may be underestimated because measuring
cation using NMES or in-bed ergometer.[28–30] In this study, all the cross-sectional muscle areas is superior to muscle thickness
subjects experienced active weaning and rehabilitation based in assessing muscle atrophy.[35] Although it was difficult to cap-
on the early weaning protocol; at the time of ICU discharge, ture the entire muscle cross-section due to the equipment used,
the FSS-ICU was approximately 21 to 22 points, and wean- we captured the changes in muscle thickness.
ing using the lower limbs, such as standing and walking, had
progressed. In our study, the effect of early mobilization may 6. Conclusion
have masked the impact of NMES on muscle strength and
physical functioning. There may be several reasons for NMES Critically ill older patients who underwent neuromuscular elec-
changing muscle thickness but having no effect on strength trical stimulation in the ICU showed less reduction in muscle
or performance. The MRC sum score assessed lower extrem- thickness. It is proposed that NMES during ICU stay may reduce
ity muscle strength because it could not capture changes in lower limb muscle atrophy in these patients. However, it did not
muscle strength as it was not evaluated in detail. In addition, seem to be effective in older patients (>75 years). In addition,
NMES suppressed atrophy of the quadriceps muscle; how- the effect of NMES on physical function at discharge from the
ever, as only the latter was assessed by ultrasound imaging, hospital is not clear and needs to be tested in future studies.
it is possible that atrophy in other muscles was not inhib-
ited. As a result, it may be deduced that NMES did not affect
Acknowledgments
physical functioning. As muscle strength and physical func-
tioning did not differ in the NMES group from those in the The authors are grateful to the ICU doctors and nurses, namely
control group, the incidence of ICU-AW and frailty did not Dr Michiko Kitamura, Dr Ritsuko Saito, Mr. Shingo Haneda,
vary between the 2 groups. As one of the major goals of reha- Mr. Takeshi Okamoto, Mr. Yuji Kuwabara, for extending their
bilitation in the ICU is minimizing the incidence of ICU-AW support and cooperation toward fulfillment of this study’s
and frailty, the exclusive addition of NMES may have little objectives. We appreciate Dr Takahiro Tokunaga of the Medical
impact on reducing the incidence of these conditions. In ICU Research Support Center, University of Fukui Hospital for his
patients, muscle atrophy occurs early and rapidly within a advice on statistical analysis.
few days of admission.[2] Muscle mass loss during ICU stay
is a factor in post-ICU functional impairment.[19] Increased
quadriceps thickness during hospitalization in ICU patients Author contributions
is associated with improved physical function 3 months after Conceptualization: T.N., H.S., M.K., A.M., K.S., T.I.
discharge.[31] Controlling muscle atrophy from the time of Data curation: T.N., H.S.
ICU admission may be of value for long-term physical func- Formal analysis: T.N., H.S., M.K.
tion. More studies on rehabilitation to reduce the incidence of Funding acquisition: T.N.
ICU-AW and frailty are needed in the future. Investigation: T.N., H.S., T.I.

8
Nonoyama et al.  •  Medicine (2022) 101:31www.md-journal.com

Project administration: T.I. [17] Mignardot J-B, Deschamps T, Le Goff CG, et al. Neuromuscular electri-
Supervision: T.I. cal stimulation leads to physiological gains enhancing postural balance
Visualization: T.N., H.S., M.K. in the pre-frail elderly. Physiol Rep. 2015;3:e12471.
[18] Bondi D, Jandova T, Verratti V, et al. Static balance adaptations after-
Writing–original draft: T.N.
neuromuscular electrical stimulation on quadysps and lumbarparaspi-
Writing–review & editing: H.S., M.K., A.M., K.S., T.I. nal muscles in healthy older. Sport Sci Health. 2022;18:85–96.
The final manuscript has been read and all authors have agreed [19] Parry SM, El-Ansary D, Cartwright MS, et al. Ultrasonography in the
to the content of the manuscript and acknowledge that all intensive care setting can be used to detect changes in the quality and
those entitled to authorship are listed as authors. quantity of muscle and is related to muscle strength and function. J Crit
Care. 2015;30:1151.e9–1151.e14.
[20] Puthucheary ZA, Phadke R, Rawal J, et al. Qualitative ultrasound in
References acute critical illness muscle wasting. Crit Care Med. 2015;43:1603–11.
[1] Stevens RD, Dowdy DW, Michaels RK, et al. Neuromuscular dysfunc- [21] Haar G. The safe use of ultrasound in medical diagnosis [Internet].
tion acquired in critical illness: a systematic review. Intensive Care Med. London: British Institute of Radiology. 2012. Available at: http://www.
2007;33:1876–91. birpublications.org/pb/assets/raw/Books/SUoU_3rdEd/Safe_Use_of_
[2] Mayer KP, Thompson Bastin ML, Montgomery-Yates AA, et al. Acute Ultrasound.pdf. [access date February 18, 2022].
skeletal muscle wasting and dysfunction predict physical disability [22] Ouchi Y, Rakugi H, Arai H, et al. Redefining the elderly as aged 75 years
at hospital discharge in patients with critical illness. Critical Care. and older: proposal from the joint committee of Japan Gerontological
2020;24:637. Society and the Japan Geriatrics Society: redefining elderly: proposal
[3] Brummel NE, Girard TD, Pandharipande PP, et al. Prevalence and from Japan. Geriatr Gerontol Int. 2017;17:1045–7.
course of frailty in survivors of critical illness*. Crit Care Med. [23] Chen L-K, Woo J, Assantachai P, et al. Asian working group for sarco-
2020;48:1419–26. penia: 2019 consensus update on sarcopenia diagnosis and treatment. J
[4] Iwashyna TJ, Ely EW, Smith DM, et al. Long-term cognitive impair- Am Med Dir Assoc. 2020;21:300–307.e2.
ment and functional disability among survivors of severe sepsis. JAMA [24] Iwatsu K, Iida Y, Kono Y, et al. Neuromuscular electrical stimulation
2010;304:1787–94. may attenuate muscle proteolysis after cardiovascular surgery: a pre-
[5] Herridge MS, Tansey CM, Matté A, et al. Functional disability 5 years after liminary study. J Thorac Cardiovasc Surg. 2017;153:373–379.e1.
acute respiratory distress syndrome. N Engl J Med. 2011;364:1293–304. [25] Giovannini S, Marzetti E, Borst SE, et al. Modulation of GH/IGF-1
[6] Kawakami D. Prevalence of post-intensive care syndrome among axis: potential strategies to counteract sarcopenia in older adults. Mech
Japanese intensive care unit patients: a prospective, multicenter, obser- Ageing Dev. 2008;129:593–601.
vational J-PICS study. Crit Care. 2021;25:69. [26] Abdellaoui A, Préfaut C, Gouzi F, et al. Skeletal muscle effects of elec-
[7] Baumgartner RN, Koehler KM, Gallagher D, et al. Epidemiology trostimulation after COPD exacerbation: a pilot study. Eur Respir J.
of sarcopenia among the elderly in New Mexico. Am J Epidemiol. 2011;38:781–8.
1998;147:755–63. [27] Rodriguez PO, Setten M, Maskin LP, et al. Muscle weakness in septic
[8] Nakamura H, Fukushima H, Miwa Y, et al. A longitudinal study on the patients requiring mechanical ventilation: protective effect of transcuta-
nutritional state of elderly women at a nursing home in Japan. Intern neous neuromuscular electrical stimulation. J Crit Care. 2012;27:319.
Med. 2006;45:1113–20. e1–8.
[9] Druschky A, Herkert M, Radespiel-Tröger M, et al. Critical illness [28] Kho ME, Truong AD, Zanni JM, et al. Neuromuscular electrical stim-
polyneuropathy: clinical findings and cell culture assay of neurotoxicity ulation in mechanically ventilated patients: a randomized, sham-con-
assessed by a prospective study. Intensive Care Med. 2001;27:686–93. trolled pilot trial with blinded outcome assessment. J Crit Care.
[10] McKendry J, Thomas ACQ, Phillips SM. Muscle mass loss in the older 2015;30:32–9.
critically Ill population: potential therapeutic strategies. Nutr Clin [29] Fischer A, Spiegl M, Altmann K, et al. Muscle mass, strength and func-
Pract. 2020;35:607–16. tional outcomes in critically ill patients after cardiothoracic surgery:
[11] Gerovasili V, Stefanidis K, Vitzilaios K, et al. Electrical muscle stimula- does neuromuscular electrical stimulation help? The Catastim 2 ran-
tion preserves the muscle mass of critically ill patients: a randomized domized controlled trial. Critical Care. 2016;20:30.
study. Critical Care. 2009;13:R161. [30] Fossat G, Baudin F, Courtes L, et al. Effect of in-bed leg cycling and elec-
[12] Nakamura K, Kihata A, Naraba H, et al. Efficacy of belt electrode skel- trical stimulation of the quadriceps on global muscle strength in criti-
etal muscle electrical stimulation on reducing the rate of muscle volume cally Ill adults: a randomized clinical trial. JAMA 2018;320:368–78.
loss in critically ill patients: a randomized controlled trial. J Rehabil [31] Chapple L-AS, Deane AM, Williams LT, et al. Longitudinal changes in
Med. 2019;51:705–11. anthropometrics and impact on self-reported physical function after
[13] Nakanishi N, Oto J, Tsutsumi R, et al. Effect of electrical muscle stimula- traumatic brain injury. Crit Care Resusc. 2017;19:29–36.
tion on upper and lower limb muscles in critically Ill patients: a two-center [32] Reimers K, Reimers CD, Wagner S, et al. Skeletal muscle sonography: a
randomized controlled trial. Crit Care Med. 2020;48:e997–e1003. correlative study of echogenicity and morphology. J Ultrasound Med.
[14] Routsi C, Gerovasili V, Vasileiadis I, et al. Electrical muscle stimula- 1993;12:73–7.
tion prevents critical illness polyneuromyopathy: a randomized parallel [33] Pillen S, Tak RO, Zwarts MJ, et al. Skeletal muscle ultrasound: correla-
intervention trial. Critical Care 2010;14:R74. tion between fibrous tissue and echo intensity. Ultrasound Med Biol.
[15] Abu-Khaber HA, Abouelela AMZ, Abdelkarim EM. Effect of electrical 2009;35:443–6.
muscle stimulation on prevention of ICU acquired muscle weakness [34] Radaelli R, Botton CE, Wilhelm EN, et al. Low- and high-volume
and facilitating weaning from mechanical ventilation. Alexandria J strength training induces similar neuromuscular improvements in mus-
Med. 2013;49:309–15. cle quality in elderly women. Exp Gerontol. 2013;48:710–6.
[16] Mancinelli R, Toniolo L, Di Filippo ES, et al. Neuromuscular electrical [35] Puthucheary ZA, McNelly AS, Rawal J, et al. Rectus femoris cross-sec-
stimulation induces skeletal muscle fiber remodeling and specific gene tional area and muscle layer thickness: comparative markers of muscle
expression profile in healthy elderly. Front Physiol. 2019;10:1459. wasting and weakness. Am J Respir Crit Care Med. 2017;195:136–8.