Obese Japanese Patients with Stroke Have Higher Functional

Recovery in Convalescent Rehabilitation Wards:

A Retrospective Cohort Study

Shinta Nishioka, RD,* Hidetaka Wakabayashi, MD,† Tomomi Yoshida, RD,‡

Natsumi Mori, RD,‡ Riko Watanabe, RD,‡ and Emi Nishioka, RD‡

Background: A protective effect of excessive body mass index (BMI) on mortality

or functional outcome in patients with stroke is not well established in the Asian
population. This study aimed to explore whether obese patients with stroke have
advantages for functional improvement in Japanese rehabilitation wards. Method:
This retrospective cohort study included consecutive patients with stroke admit-
ted and discharged from convalescent rehabilitation wards between 2011 and 2015.
Demographic data, BMI, Functional Independence Measure (FIM) score, and nu-
tritional status were analyzed. Participants were classified into 4 groups according
to BMI (underweight <18.5 kg/m2, standard 18.5-<23 kg/m2, overweight 23-
<27.5 kg/m2, obese ≥27.5 kg/m2). The primary outcome was the FIM gain, and
the secondary outcome was the FIM score at discharge. Multiple regression anal-
ysis was performed to analyze the relationship between BMI and functional recovery.
Results: In total, 897 participants (males 484, females 413; mean age 71.6 years)
were analyzed and classified as underweight (134), standard (432), overweight
(277), and obese (54). The median FIM gain and the FIM score at discharge were
30 and 114, respectively. The FIM gain in the obese group was significantly higher
than those in the other groups. Multiple regression analysis revealed that
obesity was independently correlated with the FIM gain, and those at discharge
after adjusting for confounders such as age, gender, and FIM score on admission.
Conclusions: Obese Japanese convalescent patients with stroke may have some
advantages for functional recovery in rehabilitation wards. Key Words: Obesity
paradox—stroke—convalescent rehabilitation wards—functional recovery—Functional
Independence Measure.
© 2015 National Stroke Association. Published by Elsevier Inc. All rights reserved.

From the *Department of Clinical Nutrition and Food Services, Nagasaki Rehabilitation Hospital; †Department of Rehabilitation Medicine,
Yokohama City University Medical Center; and ‡Department of Clinical Services, Nagasaki Rehabilitation Hospital.
Received May 30, 2015; revision received August 8, 2015; accepted August 24, 2015.
Authors’ contributions: Shinta Nishioka: conception and design of the study, collection, analysis and interpretation of data, writing of the
paper, drafting of the manuscript, and approval of the final version of the manuscript. Hidetaka Wakabayashi: conception and design of the
study, analysis and interpretation of data, revision of the manuscript, and approval of the final version of the manuscript. Tomomi Yoshida,
Natsumi Mori, Riko Watanabe, and Emi Nishioka: conception and design of the study, collection of data, revision of the manuscript, and
approval of the final version of the manuscript.
Grant support: This study was not sponsored or funded by any industry, government, or institution.
Address correspondence to Shinta Nishioka, Department of Clinical Nutrition and Food Services, Nagasaki Rehabilitation Hospital, 4-11,
Gin-ya machi, Nagasaki 850-0854, Japan. E-mail: shintacks@yahoo.co.jp.
1052-3057/$ - see front matter
© 2015 National Stroke Association. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2015.08.029

26 Journal of Stroke and Cerebrovascular Diseases, Vol. 25, No. 1 (January), 2016: pp 26–33
OBESE PATIENTS WITH STROKE HAVE HIGHER FUNCTIONAL RECOVERY
Introduction
Obesity is a well-known risk factor for stroke inci-
dence, and is generally defined as a body mass index
(BMI) of 30 kg/m2 or more in Western populations1 or
27.5 kg/m2 or greater in Asian populations.2 The rela-
tive risks for stroke incidence by obesity (BMI ≥ 30 kg/m2
and ≥27.5 kg/m2 for Asian and Western populations, re-
spectively) and overweight (BMI 25-29.9 kg/m2 and 23-
27.4 kg/m2 for Asian and Western populations, respectively)
are 1.64 and 1.22, respectively, according to a large-
scale meta-analysis.3 Obesity raises the risk for stroke via
several pathways, including diabetes mellitus, hyperten-
sion, atherosclerosis, atrial fibrillation, and obstructive sleep
apnea.4
Interestingly, patients with stroke who are obese or over-
weight have a tendency toward better outcomes after
stroke. A prospective study showed that patients with
stroke with BMI 25-29.9 kg/m2, 30-34.9 kg/m2, or 35 kg/m2
or more had lower all-cause mortality than normal-
weight patients.5 Other studies reported a protective effect
of obesity for all-cause mortality and cardiovascular mor-
tality in patients with stroke aged 80 years or older6 and
a good survival rate in obese patients with stroke re-
gardless of age.7 The paradoxical effects of obesity are
also seen in Asian populations.8,9 This finding is called
the “obesity paradox” and has been reported in some other
chronic diseases, including chronic heart failure,10 chronic
kidney disease,11 chronic obstructive pulmonary disease,12
and rheumatoid arthritis.13
The existence of the obesity paradox in patients with
stroke is still controversial. Several studies on the rela-
tionship between obesity and stroke outcome have reported
the adverse effects of obesity/overweight on the surviv-
al rate14,15 or no correlation with BMI and stroke outcome.16
A few studies have been published regarding the rela-
tionship between obesity and stroke rehabilitation outcome,
with some observational studies reporting that obese or
overweight patients with stroke have higher rates of re-
covery for activities of daily living (ADLs) after 30 months17
or 3 months18 from stroke onset, as well as at the end
of the in-hospital rehabilitation period.19 Conversely, another
study indicated that obese patients with stroke with mod-
erate functional impairment have lower ADL improvement
than normal-weight patients,20 whereas another study re-
ported that the initial neurological symptom is a more
powerful predictor for functional recovery than BMI.21
However, whether obesity stroke inpatients have better
recovery of ADLs during the convalescent stage in Asian
populations remain unclear. Only one study investi-
gated the stroke recovery in Asian patients18 using the
BMI ranges recommended for Asian populations. Addi-
tionally, no data exist regarding the relationship between
in-hospital rehabilitation outcome and obesity in pa-
tients with stroke in Asian populations. Do Japanese stroke
inpatients with obesity have an advantage for ADL
27
improvement? What is the most beneficial BMI range for
ADL improvement? To answer these questions, we con-
ducted a retrospective cohort study on patients with stroke
in convalescent rehabilitation wards that were covered
by public health insurance22 in Japan.
Materials and Methods
Participants
The present study investigated 1281 consecutive sub-
jects who were admitted and discharged from our
convalescent rehabilitation hospital for poststroke reha-
bilitation between June 2011 and March 2015. Exclusion
criteria included being transferred to an acute care hos-
pital because of exacerbation of the clinical condition or
surgical operation (such as recurrent stroke, severe pneu-
monia, gastrointestinal bleeding, and gastrostomy), dying
in our hospital, being transferred to another rehabilita-
tion hospital in accordance with the patients’ or their
family’s wishes, readmission to our hospital after finish-
ing treatment for their comorbidities (except after recurrent
stroke), and lack of recorded body weight. Subjects with
missing covariate data were not excluded from the analysis.
In Japan, patients with stroke who require intensive
rehabilitation to improve ADLs and return to their own
home are usually admitted to convalescent rehabilita-
tion wards and were covered by public health insurance.22
These patients receive comprehensive rehabilitation support
from a multidisciplinary rehabilitation team including a
rehabilitation doctor, nurse, physical therapist, occupa-
tional therapist, speech language therapist, registered
dietitian, care worker, social worker, dental hygienist, and
pharmacist. All the participants in the present study
received intensive rehabilitation from a therapist for
approximately 3 hours/day between admission and
discharge.
Information regarding the participants’ characteristics
including age, gender, primary diagnosis (cerebral in-
farction, intracranial hemorrhage, and subarachnoid
hemorrhage), length of hospital stay (LOS), days from
stroke onset to admission to the rehabilitation hospital,
and nutritional status using the Mini Nutritional Assess-
ment Short Form (MNA-SF)23 was collected via medical
records. Prestroke certification for public long-term care
insurance (LTCI) was also investigated to assess the pa-
tients’ prestroke functional dependence. LTCI is the only
public care insurance that provides various social care
services in Japan for people aged 65 years or older who
are functionally dependent or those aged 40 years or older
who suffer from diseases such as stroke.24
The BMI was defined as the weight in kilograms divided
by the square of the height in meters. In the present study,
the participants were classified into 4 groups following
the BMI categories for Asian populations recommended
by the World Health Organization 2 (<18.5 kg/m 2
28
2
[underweight group], 18.5-<23 kg/m [standard group],
23-<27.5 kg/m2 [overweight group], and ≥27.5 kg/m2 [obese
group]). The participants’ height and body weight were
measured by a nurse, care worker, or a registered dieti-
tian on admission.
The present study was conducted in accordance with
the Declaration of Helsinki. The Ethics Committee of the
Nagasaki Rehabilitation Hospital approved the present
study and exempted us from obtaining informed consent
from the participants because the study analyzed only
anonymous retrospective data.
Main Outcome Measurement
This study used the Functional Independence Measure
(FIM) gain as the main outcome measurement. The FIM
is a reliable, validated indicator of ADLs25,26 that has 13
items regarding motor function and 5 items regarding
cognitive function. Each item is given 1-7 points, from
“total assistance” to “complete independence.” The FIM
score was assessed by the multidisciplinary team on ad-
mission and at discharge. The primary outcome was the
FIM gain, which was calculated as changes in the FIM
score between admission and discharge, and the second-
ary outcome was the FIM score at discharge.
Sample Size Calculation
The study size was calculated using Power and
Sample Size Calculation software version 3.0 (William D.
Dupont and Walton D. Plummer, Vanderbilt University
School of Medicine, Nashville, TN) (available from
http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/
PowerSampleSize). A previous study in Japan found a
mean FIM gain in rehabilitative stroke patients of
19.7 ± 18.5.22 We inferred that there was an overall increase
of 5 in the FIM gain from underweight to obese. To detect
a mean difference of 5 in the FIM gain (with a standard
deviation of 18.5) among the 4 groups when the ratio is
1:1, 216 patients would be required in each group for a
power (1 – β) of .8 and an α of .05. Thus, we included
approximately 4 years of data in the present study to ensure
enough patients for sufficient statistical power.
Statistical Analysis
Statistical analysis was performed using the IBM Sta-
tistical Package for the Social Sciences version 21 software
(IBM Corporation, Armonk, NY). Parametric data were
reported as the mean ± standard deviation, while non-
parametric data were expressed as the median and
interquartile range. One-way analysis of variance, chi-
square, and Kruskal–Wallis tests were used to analyze
the differences among the 4 groups. After performing 1-way
analysis of variance and the Kruskal–Wallis test, the
Dunnett test and the Dunn test for multiple compari-
sons were performed, respectively. Nominal data that
S. NISHIOKA ET AL.
showed differences by chi-square test were examined by
residual analysis. A multiple regression analysis was used
to elucidate whether BMI independently affects the FIM
gain and the FIM score at discharge. A P value less than
.05 was considered statistically significant.
Results
Of the 1281 participants, 254 were excluded because
they transferred to other hospitals (acute care/rehabilitation)
or died in our hospital, 127 were excluded because of
readmission, and three were excluded because their weight
was not recorded. The proportion of underweight pa-
tients in the excluded participants was significantly higher
than those in the included participants, whereas the pro-
portion of overweight and obese patients in the excluded
participants was significantly lower than those in the in-
cluded participants. A total of 897 participants (484 males,
413 females) with a mean age of 71.6 ± 13.3 years were
investigated in the present study. Five hundred ninety
patients suffered from cerebral infarction (66%), 247 from
intracranial hemorrhage (28%), and 60 from subarach-
noid hemorrhage (7%). Sixty-eight participants had missing
data regarding prestroke certification for LTCI. There were
no missing data for any other covariates.
Table 1 shows the demographic characteristics of the
study participants. According to the BMI on admission,
134 were classified as the underweight group, 432 as the
standard group, 277 as the overweight group, and 54 as
the obese group. Age, gender, LOS, stroke subtype,
prestroke certification for LTCI, and nutritional status were
significantly different among the 4 groups. According to
multiple comparison tests, the underweight group was
older than the other groups, while those in the stan-
dard group were older than those in the obese groups.
The LOS of the underweight group was significantly longer
than those of the standard and overweight groups. Total,
motor, and cognitive FIM score on admission were sig-
nificantly lower in the underweight group than those in
the other 3 groups as shown by the Dunn test.
Table 2 shows the differences between each BMI group
in the ADLs and discharge outcomes. Total, motor, and
cognitive FIM score at discharge were significantly lower
in the underweight group than in the other BMI catego-
ries as shown by the Dunn test. Meanwhile, the FIM gain
and the FIM efficiency were significantly higher in the
obese group as compared with those in the other groups
by multiple comparison. According to residual analysis,
the home discharge rate in the underweight group was
significantly lower than those in the other groups, while
the rate of discharging to long-term care facilities was
significantly higher in the underweight group.
Age, gender, LOS, days from stroke onset to admis-
sion to rehabilitation hospital, stroke subtype, FIM score
on admission (motor and cognitive), presence of prestroke
certification for LTCI, nutritional status, and BMI
 OBESE PATIENTS WITH STROKE HAVE HIGHER FUNCTIONAL RECOVERY
Table 1. Characteristics of the 897 patients with stroke admitted to convalescent rehabilitation wards

Underweight Standard Overweight Obese

Characteristics Overall (<18.5 kg/m2) (18.5-<23 kg/m2) (23-<27.5 kg/m2) (≥27.5 kg/m2) P value

Age, years
Mean ± SD 71.6 ± 13.3 77.5 ± 13.3§ 71.9 ± 12.6‖ 69.7 ± 13.1 64.9 ± 14.7 <.001*
Gender, n (%) <.001†
Male 484 (54.0) 44 (32.8)** 242 (56.0) 172 (62.1) 26 (48.1)
Female 413 (46.0) 90 (67.2)** 190 (44.0) 105 (37.9) 28 (51.9)
LOS .004‡
Median (25th-75th percentiles) 122 (72.5-166) 141 (96.5-173)¶ 121 (76-163) 114 (60.5-165) 124.5 (68-150)
Days from stroke onset to admission .072‡
Median (25th-75th percentiles) 21 (17-31) 23 (18-32) 22 (17-32) 20 (15-29) 21 (17-27)
Stroke subtype, n (%) .038†
Cerebral infarction 590 (65.8) 77(57.5) 288 (66.7) 189 (68.2) 36 (66.7)
Intracerebral hemorrhage 247 (27.5) 40 (29.9) 114 (26.4) 77 (27.8) 16 (29.6)
Subarachnoid hemorrhage 60 (6.7) 17(12.7)†† 30 (6.9) 11 (4.0) 2 (3.7)
Prestroke certification for LTCI, n (%) <.001†
Yes 235 (26.2) 58 (43.2)‡‡ 109 (25.2) 59 (21.3) 9 (16.7)
No 570 (63.5) 64 (47.8)‡‡ 283 (65.5) 183 (66.1) 40 (74.1)
Under examination 24 (2.7) 5 (3.7) 12 (2.8) 7 (2.5) 0 (0)
Uncertain 68 (7.6) 7 (5.2) 28 (6.5) 28 (10.1) 5 (9.3)
FIM on admission, points
Median (25th-75th percentiles)
Total FIM 74 (46-94) 54.5 (30.8-78.3)# 74 (48-95.8) 80 (51-98.5) 78.5 (58-92.3) <.001‡
Motor FIM 50 (28-67) 35 (18-56)# 51 (30-68) 55 (31.5-71) 49.5 (35.8-61) <.001‡
Cognitive FIM 24 (15-30) 18 (11-25)# 23 (15-30) 26 (16-31) 27.5 (22-31) <.001‡
BMI
Median (25th-75th percentiles) 21.7 (19.8-24.1) 17.2 (15.9-17.9)# 20.9 (19.9-21.8)# 24.4 (23.7-25.8)# 29.0 (28.2-30.7)# <.001‡
Nutritional status, n (%)§§ <.001†
Malnourished 556 (62.0) 128 (95.5)‡‡ 291 (67.4)‡‡ 115 (41.5) 22 (40.7)
At risk of malnutrition 329 (36.7) 6 (4.5)‡‡ 141 (32.6)‡‡ 150 (54.2) 32 (59.3)
Normal 12 (1.3) 0 (0) 0 (0) 12 (4.3) 0 (0)

Abbreviations: ANOVA, analysis of variance; BMI, body mass index; FIM, Functional Independence Measure; LOS, length of hospital stay; LTCI, long-term care insurance; SD, standard deviation.
*One-way ANOVA.
†Chi-square test.
‡Kruskal–Wallis test.
§Significantly different from other groups according to the Dunnett test.
‖Significantly different from the obese group according to the Dunnett test.
¶Significantly different from standard and overweight groups according to the Dunn test.
#Significantly different from other groups according to the Dunn test.
**Significantly different from standard and overweight groups according to the residual test.
††Significantly different from overweight groups according to the residual test.
‡‡Significantly different from other groups according to the residual test.
§§Assessed by Mini Nutritional Assessment Short Form (malnourished: 0-7 points, at risk of malnutrition: 8-11 points, normal: 12-14 points).

29
 30
Table 2. Kruskal–Wallis test and chi-square test of differences among different BMI groups

Underweight Standard Overweight Obese

Variables Overall (<18.5 kg/m2) (18.5–<23 kg/m2) (23–<27.5 kg/m2) (≥27.5 kg/m2) P value

FIM at discharge, points

Median (25th-75th percentiles)
Total FIM 114 (87-123) 94 (65.8-114)‡ 114 (87.3-122) 118 (96-124) 119.5 (109-124) <.001*
Motor FIM 84 (63-89) 69 (43-85)‡ 84 (64-89) 87 (70.5.-90) 87 (77.8-89) <.001*
Cognitive FIM 31 (23-34) 25 (18-31)‡ 30 (23-34)§ 32 (25-35) 34 (30.8-35) <.001*
FIM gain
Median (25th-75th percentiles) 30 (17-42) 30 (17.5-44) 29 (17.3-39) 29 (15-42.5) 37.5 (27.8-46)‖ .015*
FIM efficiency
Median (25th-75th percentiles) .27 (.17-.38) .24 (.15-.34) .26 (.17-.38) .27 (.19-.39) .34 (.25-.44)¶ .001*
Discharge outcome, n (%) <.001†
Home 738 (82.3) 90 (67.2)# 355 (82.2) 243 (87.7) 50 (92.6)
Long-term care facility 129 (14.4) 37 (27.6)# 61 (14.1) 27 (9.7) 4 (7.4)
Geriatric health services facility 19 (2.1) 3 (2.2) 10 (2.3) 6 (2.2) 0
Long-term care hospital 11 (1.2) 4 (3.0) 6 (1.4) 1 (.4) 0

Abbreviations: BMI, body mass index; FIM, the Functional Independence Measure.
*Kruskal–Wallis test.
†Chi-square test.
‡Significantly different from other groups according to the Dunn test.
§Significantly different from the obese group according to the Dunn test.
‖Significantly different from standard and overweight groups according to the Dunn test.
¶Significantly different from underweight and standard groups according to the Dunn test.
#Significantly different from other groups according to the residual analysis.

S. NISHIOKA ET AL.
OBESE PATIENTS WITH STROKE HAVE HIGHER FUNCTIONAL RECOVERY 31
Table 3. Multiple regression analysis of FIM gain

Unstandardized coefficient

Standard 95% confidence Standardized

Variables β error interval of β coefficient P value

LOS .070 .017 .037 .103 .182 <.001

Female −4.607 1.192 −6.947 −2.267 −.118 <.001
Motor FIM on admission −.293 .039 −.370 −.217 −.344 <.001
Prestroke certification for LTCI −5.463 1.486 −8.381 −2.546 −.124 <.001
Obesity 5.581 2.419 .834 10.328 .068 .021
Age −.153 .051 −.253 −.052 −.104 .003
Days from stroke onset to admission −.128 .046 −.218 −.039 −.086 .005
Constant 58.774 5.844 47.303 70.245 <.001

Abbreviations: FIM, Functional Independence Measure; LOS, length of hospital stay; LTCI, long-term care insurance.

categories were included in a multiple regression anal-
ysis of FIM gain and FIM score at discharge. No
multicollinearity was observed between the variables.
Obesity was independently associated with FIM gain in
addition to LOS, female gender, motor FIM, certifica-
tion for LTCI, age, and days from stroke onset to admission
[FIM gain = .070 × LOS + 5.581 × obesity (BMI ≥27.5 kg/
m 2 = 1, <27.5 kg/m 2 = 0) − 4.607 × gender (male = 1,
female = 2) − .293 × motor FIM on admission – 5.463 × pres-
ence of prestroke certification for LTCI (yes = 1,
no = 0) − .153 × age − .128 × days from stroke onset to ad-
mission + 58.774, R2 = .233, P < .001] (Table 3).
Obesity was also independently associated with FIM
score at discharge in addition to motor FIM, cognitive
FIM, certification for LTCI, LOS, female gender, age, and
days from stroke onset to admission [FIM at dis-
charge = .668 × motor FIM on admission + 1.149 × cognitive
FIM on admission + .071 × LOS + 5.159 × obesity (BMI
≥27.5 kg/m 2 = 1, <27.5 kg/m 2 = 0) − 5.420 × presence
of prestroke certification for LTCI (yes = 1,
no = 0) − 4.682 × gender (male = 1, female = 2) − .144 ×
age − .125 × days from stroke onset to admission + 56.647,
R2 = .663, P < .001] (Table 4).
Discussion
Two clinical observations were made in the present study.
First, obesity offers an advantage for ADL improve-
ment in Japanese patients with stroke during in-hospital
rehabilitation. Second, the most beneficial BMI range for
ADL improvement of Japanese patients with stroke in
convalescent rehabilitation wards is 27.5 kg/m2 or higher.
Obesity has an advantage for the recovery of ADLs in
Japanese patients with stroke during in-hospital rehabil-
itation. The present study indicates that being obese is
positively correlated with the FIM gain and the FIM score
at discharge from rehabilitation wards even after adjust-
ing for confounders. These results suggest the presence
of the obesity paradox in relation to functional recovery
of patients with stroke in convalescent rehabilitation wards
in Japan. A recent prospective cohort study in China in-
dicated that overweight was an independent explanatory

Table 4. Multiple regression analysis of total FIM at discharge

Unstandardized coefficient

Standard 95% confidence Standardized

Variables β error interval of β coefficient P value

Motor FIM on admission .668 .046 .577 .758 .520 <.001

Cognitive FIM on admission 1.149 .097 .958 1.339 .351 <.001
Prestroke certification for LTCI −5.420 1.486 −8.337 −2.503 −.081 <.001
Female −4.682 1.193 −7.023 −2.340 −.080 <.001
LOS .071 .017 .038 .104 .123 <.001
Obesity 5.159 2.435 .379 9.938 .042 .034
Age −.144 .052 −.245 −.042 −.065 .006
Days from stroke onset to admission −.125 .046 −.215 −.036 −.056 .006
Constant 56.647 6.038 44.796 68.497 <.001

Abbreviations: FIM, Functional Independence Measure; LOS, length of hospital stay; LTCI, long-term care insurance.
32
factor for better functional outcome 3 months after a
stroke.18 Additionally, a German cohort study showed that
overweight, obese, and severely obese patients tend to
have better survival and rate of functional independence.17
The findings from these studies may be supported by
our results. Conversely, a retrospective study in a reha-
bilitation unit showed that obese patients with stroke have
no advantages for FIM gain.20 These inconsistencies may
be explained by the number of participants. Relatively
few participants (<100) were included in the above ret-
rospective study, while the current study investigated
approximately 900 participants. The reasons for why obese
patients have advantages for functional recovery remain
unknown, but may partially be explained by a protec-
tive effect of obesity for the initial neurological severity.21
The most beneficial BMI range for ADL improvement
in Japanese patients with stroke in convalescent rehabil-
itation wards is 27.5 kg/m2 or higher. Previous studies
on BMI and functional outcome used different BMI
ranges.17-21 Although obesity is usually defined as a BMI
of 30 kg/m2 or more, different BMI categories have been
recommended for Asian people by the World Health Or-
ganization because of the health risks associated with
obesity.2 One Chinese study using BMI ranges for Asian
populations suggested that overweight (BMI 23-27.4 kg/m2)
patients have significantly better functional outcomes,
whereas obese (27.5-32.5 kg/m 2 ) or severely obese
(≥32.5 kg/m2) patients showed better functional recov-
ery, but this finding was not statistically significant.18 The
differences between the published literature and our study
may be explained by differences in the timing of mea-
suring BMI, functional outcome instruments, statistical
analysis, and sample size. According to a previous study,
stroke survivors usually lose approximately 3 kg (equiv-
alent to a BMI loss of 1.3 kg/m2) in acute care hospitals.27
Furthermore, weight loss in stroke survivors correlates
with lower functional recovery, even in those that are over-
weight (BMI 25-29.9 kg/m2) or obese (≥30 kg/m2).28 In fact,
about 40% of the obese participants in the present study
were defined as malnourished according to MNA-SF that
includes a weight loss component. Therefore, it is pos-
sible that the ADL improvement in obese patients who
experienced unintentional weight loss was less than
expected.
The exclusion criteria of the current study may affect
the FIM gain and FIM score at discharge of the under-
weight patients. More underweight participants existed
among the excluded participants than among the in-
cluded ones. A previous study suggested that patients
with stroke with lowest quantiles of BMI (<21.2) showed
relatively worse stroke severity and worse functional re-
covery compared with those with higher quantiles of BMI.21
Therefore, the FIM gain and FIM score at discharge of
the included underweight patients may be overestimated.
Consideration of body composition is important for in-
terpreting the relationship between obesity and functional
S. NISHIOKA ET AL.
outcome for patients with stroke, because skeletal muscle
mass plays a key role in physical function. Patients with
stroke usually experience muscle loss caused by dener-
vation, disuse atrophy, remodeling, and spasticity,29 which
may occur even in obese patients. Furthermore, obese pa-
tients with sarcopenia had lower instrumental ADLs than
those without sarcopenia.30 Therefore, maintaining muscle
mass and reducing fat mass via rehabilitation and nu-
tritional support may be important for patients with
stroke.31 According to the guidelines for secondary pre-
vention of stroke from the American Heart Association
and the American Stroke Association,4 it is uncertain
whether obese patients should reduce their weight as the
evidence is weak. To answer these questions, further re-
search on the correlation between body composition and
functional outcome is needed.
The current study has several limitations. First, the
retrospective study design means that we cannot render
a cause–effect relationship between BMI and ADL im-
provement. Second, there are no data for neurological
severity at the time of stroke onset, which may be a
confounder for the primary outcome. A previous study
reported that there was positive relationship between
the initial stroke severity and the FIM score on admission.32
Thus, multiple regression analysis with the FIM score
on admission was conducted in the present study. Third,
patients with milder and more severe stroke who did
not adapt for convalescent rehabilitation were not in-
cluded because the current study was performed in the
convalescent rehabilitation ward. They may be dis-
charged from acute care hospitals to their own homes
or may die.
In conclusion, we found 2 clinical observations. First,
obesity offers an advantage for ADL improvement in Asian
patients with stroke during in-hospital rehabilitation.
Second, the most beneficial BMI range for ADL improve-
ment in Asian patients with stroke in convalescent
rehabilitation wards is 27.5 kg/m2 or higher. These find-
ings are useful additions to the literature on weight
management of obese patients with stroke in rehabilita-
tion settings. As 40% of the obese patients with stroke
in the present study were malnourished, further study
is needed to clarify whether obese patients with stroke
should lose weight for functional improvement in reha-
bilitation settings.
References
1. WHO. Obesity: preventing and managing the global
epidemic. Report of a WHO consultation. World Health
Organ Tech Rep Ser 2000;894(i–xii):1-253.
2. WHO Expert Consultation. Appropriate body-mass index
for Asian populations and its implications for policy and
intervention strategies. Lancet 2004;363:157-163.
3. Strazzullo P, D’Elia L, Cairella G, et al. Excess body
weight and incidence of stroke: meta-analysis of
prospective studies with 2 million participants. Stroke
2010;41:e418-e426.
OBESE PATIENTS WITH STROKE HAVE HIGHER FUNCTIONAL RECOVERY
4. Kernan WN, Ovbiagele B, Black HR, et al. Guidelines
for the prevention of stroke in patients with stroke and
transient ischemic attack: a guideline for healthcare
professionals from the American Heart Association/
American Stroke Association. Stroke 2014;45:2160-2236.
5. Olsen TS, Dehlendorff C, Petersen HG, et al. Body mass
index and poststroke mortality. Neuroepidemiology
2008;30:93-100.
6. Towfighi A, Ovbiagele B. The impact of body mass index
on mortality after stroke. Stroke 2009;40:2704-2708.
7. Vemmos K, Ntaios G, Spengos K, et al. Association
between obesity and mortality after acute first-ever stroke:
the obesity-stroke paradox. Stroke 2011;42:30-36.
8. Kim BJ, Lee SH, Ryu WS, et al. Paradoxical longevity
in obese patients with intracerebral hemorrhage.
Neurology 2011;76:567-573.
9. Kim BJ, Lee SH, Jung KH, et al. Dynamics of obesity
paradox after stroke, related to time from onset, age, and
causes of death. Neurology 2012;79:856-863.
10. Curtis JP, Selter JG, Wang Y, et al. The obesity paradox:
body mass index and outcomes in patients with heart
failure. Arch Intern Med 2005;165:55-61.
11. Kalantar-Zadeh K, Abbott KC, Salahudeen AK, et al.
Survival advantages of obesity in dialysis patients. Am
J Clin Nutr 2005;81:543-554.
12. Landbo C, Prescott E, Lange P, et al. Prognostic value
of nutritional status in chronic obstructive pulmonary
disease. Am J Respir Crit Care Med 1999;160:1856-1861.
13. Escalante A, Haas RW, del Rincon I. Paradoxical effect
of body mass index on survival in rheumatoid arthritis:
role of comorbidity and systemic inflammation. Arch
Intern Med 2005;165:1624-1629.
14. Bazzano LA, Gu D, Whelton MR, et al. Body mass index
and risk of stroke among Chinese men and women. Ann
Neurol 2010;67:11-20.
15. Yi SW, Odongua N, Nam CM, et al. Body mass index
and stroke mortality by smoking and age at menopause
among Korean postmenopausal women. Stroke 2009;
40:3428-3435.
16. Dehlendorff C, Andersen KK, Olsen TS. Body mass index
and death by stroke: no obesity paradox. JAMA Neurol
2014;71:1-7.
17. Doehner W, Schenkel J, Anker SD, et al. Overweight and
obesity are associated with improved survival, functional
outcome, and stroke recurrence after acute stroke or
transient ischaemic attack: observations from the TEMPiS
trial. Eur Heart J 2013;34:268-277.
18. Zhao L, Du W, Zhao X, et al. Favorable functional
recovery in overweight ischemic stroke survivors: findings
33
from the China National Stroke Registry. J Stroke
Cerebrovasc Dis 2014;23:e201-e206.
19. Burke DT, Al-Adawi S, Bell RB, et al. Effect of body mass
index on stroke rehabilitation. Arch Phys Med Rehabil
2014;95:1055-1059.
20. Kalichman L, Rodrigues B, Gurvich D, et al. Impact of
patient’s weight on stroke rehabilitation results. Am J
Phys Med Rehabil 2007;86:650-655.
21. Kim Y, Kim CK, Jung S, et al. Obesity-stroke paradox
and initial neurological severity. J Neurol Neurosurg
Psychiatry 2015;86:743-747.
22. Miyai I, Sonoda S, Nagasi S, et al. Results of new policies
for inpatient rehabilitation coverage in Japan. Neurorehabil
Neural Repair 2011;25:540-547.
23. Rubenstein LZ, Harker JO, Salvà A, et al. Screening for
undernutrition in geriatric practice: developing the
short-form mini-nutritional assessment (MNA-SF). J
Gerontol A Biol Sci Med Sci 2001;56:M366-M372.
24. Tamiya N, Noguchi H, Nishi A, et al. Population ageing
and wellbeing: lessons from Japan’s long-term care
insurance policy. Lancet 2011;378:1183-1192.
25. Chumney D, Nollinger K, Shesko K, et al. Ability of
Functional Independence Measure to accurately predict
functional outcome of stroke-specific population:
systematic review. J Rehabil Res Dev 2010;47:17-29.
26. Ottenbacher KJ, Hsu Y, Granger CV, et al. The reliability
of the Functional Independence Measure: a quantitative
review. Arch Phys Med Rehabil 1996;77:1226-1232.
27. Paquereau J, Allart E, Romon M, et al. The long-term
nutritional status in stroke patients and its predictive
factors. J Stroke Cerebrovasc Dis 2014;23:1628-1633.
28. Jönsson AC, Lindgren I, Norrving B, et al. Weight loss
after stroke: a population-based study from the Lund
Stroke Register. Stroke 2008;39:918-923.
29. Scherbakov N, Anker SD, Dirnagl U, et al. Stroke induced
sarcopenia: muscle wasting and disability after stroke.
Int J Cardiol 2013;170:89-94.
30. Rolland Y, Lauwers-Cances V, Cristini C, et al. Difficulties
with physical function associated with obesity, sarcopenia,
and sarcopenic-obesity in community-dwelling elderly
women: the EPIDOS (EPIDemiologie de I’OSteoporose)
Study. Am J Clin Nutr 2009;89:1895-1900.
31. Wakabayashi H, Sakuma K. Rehabilitation nutrition for
sarcopenia with disability: a combination of both
rehabilitation and nutrition care management. J Cachexia
Sarcopenia Muscle 2014;5:269-277.
32. Roth EJ, Heinemann AW, Lovell LL, et al. Impairment
and disability: their relation during stroke rehabilitation.
Arch Phys Med Rehabil 1998;79:329-335.