Intensive nutritional supplements can improve outcomes in stroke ...

ARTICLES
M.H. Rabadi, MD,
MRCPI
P.L. Coar, RD, CDN
M. Lukin, MS
M. Lesser, PhD
J.P. Blass, MD, PhD
Address correspondence and
reprint requests to Dr. M.H.
Rabadi, VA Medical Center, 921
NE 13th Street, Oklahoma City,
OK 73104
mhrabadi@gmail.com
Editorial, page 1852
Intensive nutritional supplements can
improve outcomes in stroke rehabilitation
ABSTRACT
Objective: Poor nutrition is a common complication of strokes severe enough to require inpatient
rehabilitation. We therefore tested whether intensive nutritional supplements given to undernourished
patients from the time of their admission to a specialized stroke rehabilitation service would
improve patient outcomes.
Methods: Randomized, prospective, double-blind, single center study comparing intensive nutritional
supplementation to routine nutritional supplementation in 116 undernourished patients
admitted to a stroke service. The analysis included the 90% of patients who were not lost to
follow-up due to acute or subacute hospitalization (n 102; 51 in each group). The nutritional
supplements are commercially available and Food and Drug Administration approved. The primary
outcome variable was change in total score on the Functional Independence Measure (FIM).
The secondary outcome measurements included the FIM motor and cognitive subscores, length
of stay (taken from day of admission), 2-minute and 6-minute timed walk tests measured at admission
and on discharge, and discharge disposition (home/not home).
Results: Patients receiving intensive nutritional supplementation improved more than those on
standard nutritional supplements on measures of motor function (total FIM, FIM motor subscore,
2-minute and 6-minute timed walk tests, all significant at p 0.002). They did not, however,
improve on measures of cognition (FIM cognition score). A higher proportion of patients who received
the intensive nutritional supplementation went home compared to those on standard supplementation
(p 0.05).
Conclusion: Intensive nutritional supplementation, using readily available commercial preparations,
improves motor recovery in previously undernourished patients receiving intensive inpatient
rehabilitation after stroke. Neurology ® 2008;71:1856–1861
GLOSSARY
F-M Fugl-Myer; FIM Functional Independence Measure; FOOD Feed or Ordinary Diet; LOS length of stay; NIHSS
NIH Stroke Scale.
Half of the patients admitted to inpatient rehabilitation stroke services are reportedly
undernourished. 1-3 Many factors contribute to inadequate nutrition after moderate to severe
stroke: difficulty swallowing, depression, patients’ difficulty with feeding themselves (dependency),
increased energy demand during recovery, and often a poor nutritional status before
the stroke due to such factors as age, motor, and cognitive impairments from previous strokes
and depression.
Poor nutrition has been documented to adversely affect functional outcomes after strokes. 4-6
Several studies have shown that the nutritional support of stroke survivors improves their functional
outcome. 7,8 However, a recent Cochrane Report concluded that protein and energy supplementa-
e-Pub ahead of print on October 22, 2008, at www.neurology.org.
From the Stroke Service (M.H.R., P.L.C.), Burke Rehabilitation Hospital, an affiliate of Weill Medical College of Cornell University; Biostatistics
Unit at the Feinstein Institute for Medical Research (M. Lukin, M. Lesser), North Shore–Long Island Jewish Health System; Departments of Public
Health (M. Lesser) and Neurology and Neuroscience (J.P.B.), Weill Medical College of Cornell University; and Burke Medical Research Institute
(J.P.B.), New York, NY.
Disclosure: Neither Novartis pharmaceuticals nor the study funding source (Burke Medical Research Foundation) had any role in the design, conduct,
analysis, or presentation of the results of this study. None of the authors have any commercial interests related to this study.
1856 Copyright © 2008 by AAN Enterprises, Inc.

tion in elderly malnourished patients produces
small but consistent weight gain and decreased
mortality, but had no effect on functional outcomes
or length of hospital stay. 9
In contrast, a large multicenter international
study (n 4,023), the Feed or Ordinary
Diet (FOOD) study, concluded that
nutritional supplementation after stroke had
no effect on death or poor outcome (modified
Rankin score of 3). 10 That study, though
relatively large, has several important limitations.
First, the vast majority of these patients
(92%) were well nourished to begin with and
therefore unlikely to benefit from nutritional
supplementation. The low percentage of undernourished
patients (8%) reflected a flexible
definition of the nutrition status, and enrolled
only patients with normal swallowing. The effects
of nutritional supplementation on this
low percentage of undernourished patients
could easily have been lost in the statistical
analyses by the majority of well-nourished patients.
Second, the study did not measure
stroke severity by any of the standardized, validated
clinical scales such as NIH Stroke Scale
(NIHSS) or Fugl-Myer (F-M) scale. Accurate,
conventional measurement of stroke severity
is critical in a study of effects on outcome,
since stroke severity is itself a major predictor
of functional outcome. 11 Third, the study reported
that patients in the gastrostomy (PEG)
group did not improve. This finding disagrees
with a previous study, which reported the rate
of functional recovery and home discharge
were similar to case-matched controls in PEG
patients who survived. 12 Finally, the FOOD
study functional outcome measures were assessed
6 months after enrollment. Medical
events during this interim period may well influence
the results (for instance, the reported
rate of infection was 13%).
Because of the lack of consistency in reports
on the effects of nutritional supplementation
in stroke survivors, we tested the
hypothesis that modern, commercially available
intensive nutritional supplements might
improve the functional outcomes of undernourished
stroke survivors, decrease the
length of stay, and result in home discharge
Table 1 Inclusion and exclusion criteria
The inclusion criteria were:
1. First acute stroke event within 4 weeks of admission
to an inpatient rehabilitation facility
2. Hemorrhagic or ischemic stroke documented clinically
and by neuroimaging
3. Significant weight loss as indicated by unintentional
weight loss of at least 2.5% within 2 weeks following
stroke onset
4. Medically stable from a cardiorespiratory standpoint
that they could participate in their daily therapies
5. Ability to ingest food including supplements either
orally or via the PEG tube
6. Informed consent, if possible from the patient; where
it was not possible, proxy consent was obtained from
the next of kin according to institutional IRB
standards
The exclusion criteria were:
1. Patients with prior documented history of alcohol
abuse, renal and liver diseases, and malabsorption
2. Patients medically unstable or demented
3. Patients terminally ill (e.g., patients with stroke as a
complication of a terminal cancer)
4. Patients participating in any other structured
therapeutic trial in the acute care hospital or at Burke
rather than institutionalization when admitted
on a designated stroke rehabilitation unit.
METHODS Patients. Successive patients admitted to the
stroke service of the Burke Rehabilitation Hospital with either ischemic
or hemorrhagic strokes (n 313) were offered the chance to
participate in this study, if they met preset inclusion and exclusion
criteria (table 1). Some patients (n 197) refused to participate
from fear of regaining weight and had been trying unsuccessfully to
lose weight before their strokes (figure). There were no differences
between participants and refusers in age, gender, stroke type or severity,
and FIM scores. None of the patients who fulfilled the study criteria
were on nutritional supplementation at the acute hospital prior to their
transfer to our facility. The study was approved and monitored by the
Burke Rehabilitation Hospital I.R.B., and registered at ClinicalTrials.
gov, #NCT 00332800. Those who gave written informed consent were
randomized prospectively within 72 hours of admission using sealed,
opaque envelope block randomization of 10 patients (5 patients in each
group) at a time. Envelopes were identical for the two groups of patients.
Participants were assigned to either the standard or the intensive
nutritional supplement treatment group by a designated dietician not
associated with the study (P.L.C.).
The nutritional status of the patients was evaluated within
24 hours of their admission. Patients were weighed the same way
on admission and at discharge, i.e., in a hospital gown and on a
bed scale. The stroke unit designated dietician compared admission
weight at the acute care hospital to the weight on admission
to our rehabilitation facility. Nutritional deficiency was defined
conventionally according to standards of the American Dietetic
Association 13 and the Omnibus Budget Reconciliation Act regulations
of 1987, 14 as loss of 2.5% or more of body weight within
2 weeks after the acute stroke.
Nutritional supplements. We compared the effects of adding
one of two commercially available supplements as additions
to normal hospital diet. Both are made by Novartis Pharmaceu-
Neurology 71 December 2, 2008 1857

Figure Study design
ticals; the dose of each was 120 mL every 8 hours by mouth,
accompanied by multivitamins with minerals. The “standard”
nutritional supplement was Resource Standard, Appendix 1
(127 calories, 5gofprotein). This nutritionally balanced formula
is routinely prescribed for poorly nourished patients as part
of clinical care on our stroke rehabilitation service. The “intensive”
nutritional supplement was Novasource 2.0 Appendix 2
(240 calories, 11 g of proteins). Novasource 2.0 is a calorically
dense, high protein, nutritionally complete formula often used
to treat patients with relatively high nutritional requirements,
such as in healing wounds. The intensive supplement contained
almost twice as many calories per unit volume, because it contained
about twice as much protein and four times as much fat.
The only micronutrient to differ significantly between the two
preparations was vitamin C, which was 90 mg/dose in the intensive
and 36 mg/dose in the routine supplement. Previous study
in this unit has demonstrated that vitamin C supplementation of
1,000 mg a day had no effect on outcome of stroke rehabilitation.
15 The costs of these two supplements are comparable.
The supplements were always given within 72 hours after
arriving at our rehabilitation facility. This routinely coincided
with the start of rehabilitation therapies (physical, occupational,
speech). The supplements were continued throughout their inpatient
stay but not prescribed after discharge.
The staff nurses gave all the supplements as medication with
compliance documented in the nurse’s notes. One of the investigators
(M.H.R.) monitored compliance. Occasionally a dose was
missed. When that happened, the treating physician would give
the patient a make-up dose of the appropriate (but still blinded)
1858 Neurology 71 December 2, 2008
preparation. This process led to excellent compliance. During
the course of the study neither the patient nor the therapists
providing therapy and assessing outcomes were aware of the
groups to which their patients were assigned.
Outcome measures. The primary outcome variable was
change in total score on the Functional Independence Measure
(FIM), measured at admission and on discharge. This is a standard,
clinically relevant 16 scale for measuring the outcome of rehabilitation.
10 Secondary outcome measurements included the
FIM motor and cognitive subscores, which are also wellstandardized
clinical measures. 17 FIM subscores were measured
to assess whether change in FIM was due to an improvement in
the motor or cognitive scores. Additional secondary outcome
measures were length of stay ([LOS] taken from day of admission),
discharge disposition (home/not home), and 2-minute
and 6-minute timed walk tests.
Both the 2- and 6-minute timed walk tests are valid, reliable,
and sensitive measures that are easy to administer. 18 These clinical
measures were obtained within 72 hours of admission to the
acute rehabilitation unit and on the day of discharge, by trained
and certified therapists. Each patient was instructed by the physical
therapist to cover as much ground as possible in 6 minutes at
a comfortable walking speed, and that speed and endurance were
being evaluated. Rest periods were allowed during the evaluation.
Patients ambulated with or without an orthotic device
(ankle-foot or knee-ankle-foot orthosis), and with or without a
walker (hemi or rolling) or a cane, if it was determined that gait
quality or safety were enhanced by the aids. If patients needed
assistance to ambulate during the timed walking test, the assigned
physical therapist could help advance the patient’s weaker
leg or provide assistance in weight-shifting. The physical therapist
measured elapsed time with a stopwatch once the patient
stood, and the distance covered was measured using Trumeter
Mini-Measure Distance-Measuring Wheel, a device that accurately
measures up to 10,000 feet. The distances covered in feet,
at both a 2-minute and then at an additional 4-minute time
interval (for a total of 6 minutes), were noted separately. The
ambulation speed was solely based on the 2-minute portion of
the total 6-minute timed walk test, as it is impractical to carry
out two separate timed walk tests on two separate occasions for
every patient on admission.
Statistical analyses. The primary objective addressed by the
statistical analysis was to determine whether intensive treatment
was more efficacious than standard treatment. Comparability of
the two randomized treatment groups with respect to demographic
and clinical baseline variables was established using the
Mann-Whitney test for continuous variables (age, weight on admission,
Mini-Mental State Examination, Upper and Lower extremity
Motricity Index). A 2 test or Fisher exact test (as
appropriate) was used to compare the two groups on categorical
variables (sex, stroke type, comorbidities, visual fields, discharge
disposition). Mann-Whitney compared outcome measures that
were continuous (Total, Motor, and Cognitive FIM subscores,
LOS) and Fisher exact tests compared the categorical measure
(discharge disposition). The p values were two-sided, and adjusted
for multiple variables.
The study accrued 50 subjects in each treatment arm. This
enrollment objective was based on a published study, 7 in which
the observed pre-to-post change in Barthel score was 45 in the
supplemental group vs a change of 40 in the control group, for a
difference in change of 5 (or, 12%). The observed SD for the
change was 6.0; thus, the effect size was ES 5/6 0.83. In
order to achieve 80% power to detect an effect size of 0.83, a

Table 2 Demographics of the study sample
Standard group
(n 58)
Intensive group
(n 58) p Value
Age, y 75.00 10.58 73.58 13.02 0.52
Sex, M/F 33/25 35/23 0.70
Onset to admission, d 16.36 15.70 14.10 11.23 0.37
Weight on admission, lb 145.36 32.01 146.87 29.72 0.79
Stroke type, n (%) 0.36
Hemorrhagic, ICH/SAH/SDH 6/1/0 (12.0) 12/0/1 (22.4)
Ischemic
Thrombotic 38 (65.5) 34 (58.6)
Embolic 8 (13.7) 10 (17.2)
Carotid occlusion with stroke 3 (5.1) 1 (1.7)
Stroke severity
MMSE 15.53 10.09 13.59 10.50 0.31
Visual field, n (%) 0.11
Normal 33 (56.8) 44 (75.8)
Left H hemianopsia 11 (18.9) 3 (5.1)
Right H hemianopsia 9 (15.5) 8 (13.7)
Left visual neglect 4 (6.8) 2 (3.4)
Right visual neglect 1 (1.7) 0 (0)
Undetermined 0 (0) 1 (1.7)
Upper extremity MI 38.53 39.24 47.50 39.96 0.22
Lower extremity MI 44.41 37.02 55.91 34.50 0.08
Limb placement task 6.89 3.74 5.77 3.79 0.11
Admission UDS
Total FIM Score 45.79 18.55 45.87 17.40 0.97
Motor FIM subscore 25.93 12.41 25.93 11.78 1.00
Cognitive FIM subscore 18.44 7.12 18.53 6.78 0.94
Admission
2-Minute walk test, ft 40.12 64.39 45.58 49.44 0.61
6-Minute walk test, ft 87.58 160.00 95.69 126.51 0.76
Dysphagia 36 (62%) 32 (55.1%) 0.41
Depressed 27 (46.5%) 27 (46.5%) 1.00
Pressure sores 12 (20.6%) 5 (8.6%) 0.06
Admission % significant weight loss 9.75 6.56 9.92 5.66 0.88
Admission % IBW 102.17 18.75 102.63 24.72 0.90
Admission albumin level 3.64 0.49 3.75 0.37 0.17
Admission prealbumin level 19.26 5.74 21.58 6.70 0.07
Admission transferrin level 190.60 48.93 210.47 40.85 0.02
Blood urea nitrogen 24.12 11.44 23.93 10.49 0.92
Each mean and SD was determined for the 51 patients in each of the relevant groups.
ICH intracranial hemorrhage; SAH subarachnoid hemorrhage; SDH subdural hemorrhage;
MMSE Mini-Mental State Examination; MI Motoricity Index; UDS Unified Data
System; FIM Functional Independence Measure; % IBW percent ideal body weight.
sample size of 26 per arm was required (alpha 0.05, two-tailed
t test). Prior to implementing the trial the investigative team
decided to increase the sample size from 26 to 50 patients per
arm, in order to detect smaller effect sizes for the several FIM
variables and to account for multiple endpoints.
RESULTS A total of 116 patients were randomized,
58 to each group. Seven patients in each group
(12%) were transferred to acute or other hospitals
during the study and therefore lost to follow-up (figure).
In general, the standard and intensive groups
were comparable on demographic and clinical variables
at randomization (table 2). Comparisons
between the two groups found no significant differences
in age, lesion type, lesion location, extent of
disability (admission FIM score), other clinically relevant
variables, or combination of any demographic
or clinical variables at the beginning of the study.
These comparisons also included measures of nutritional
state (weight loss post stroke, albumin, transferrin,
BUN). The sole exception was a slightly
higher level of prealbumin in the intensive supplement
group. PEG tubes were in place in two patients
prior to randomization, both in the intensive supplement
group. None of our patients were on nasogastric
feeding tubes as a hospital wide policy.
The patients receiving the intensive nutritional
supplement improved significantly in the noncognitive
outcome variables (table 3): total FIM was 31.49
for intensive vs 22.94 for standard (p 0.001), FIM
motor subscore was 24.25 vs 16.71 (p 0.001),
change in 2-minute walk test was 101.60 vs 43.98
(p 0.001), and change in 6-minute walk test was
299.28 vs 170.59 (p 0.001). These differences
were perceived clinically important by the patients,
their families, and the physicians and other health
care providers caring for them. Performance on the
FIM cognitive subscore remained similar for the two
groups at 4.61 vs 4.37 (p 0.80).
A higher percentage of patients in the intensive
group returned home (63%) compared to those
given the standard nutritional supplement (43%)
(p 0.05). There were four patients in the intensive
group with pressure sores and nine in the standard
group. Three patients in each group experienced
healing of their pressure sores, and their rates of healing
were not significantly different.
Both groups of patients gained weight. On average,
the intensive treatment group gained more, but the
group difference was not significant (table 3). The insignificant
drop in the serum albumin level in both groups
is hard to explain. A possible explanation is inadequate
fluid intake by the patients before their transfer to the
rehabilitation setting. Serum albumin is reportedly an
unreliable marker of nutritional state in older people
with impaired activities of daily living, who constituted
the bulk of the patient population in this study. 19
Clinical side effects were not observed with either
of the nutritional supplements.
Seven patients in each group did not complete the
study as they were transferred to acute or subacute
Neurology 71 December 2, 2008 1859

Table 3 Functional outcome measures
Standard group (n 51) Intensive group (n 51)
p Value for
the difference
95% CI for the
difference
95% CI for the
difference Admission Discharge Difference
Admission Discharge Difference
Grouping variables
Total FIM score 46.53 18.92 69.47 22.78 22.94 11.79 (0.17, 46.05) 46.84 18.03 78.33 23.61 31.49 14.26 (3.54, 59.44) 0.001
Motor FIM subscore 26.63 12.65 43.33 16.63 16.71 9.64 (2.18, 35.60) 26.78 12.08 51.04 17.43 24.25 11.83 (1.06, 47.44) 0.001
Cognitive FIM subscore 18.43 7.17 22.80 6.77 4.37 3.53 (2.55, 11.29) 18.69 7.10 23.29 6.83 4.61 3.27 (1.80, 11.02) 0.80
2-Minute walk test, ft 41.00 65.22 84.98 80.08 43.98 62.46 (78.44, 166.40) 49.92 50.93 148.90 107.15 101.60 79.41 (54.04, 257.24) 0.001
1860 Neurology 71 December 2, 2008
6-Minute walk test, ft 92.51 168.77 263.10 284.90 170.59 198.61 (218.6, 559.87) 106.80 131.32 396.37 276.46 299.28 201.54 (95.74, 694.30) 0.001
Weight, lb 146.88 32.78 147.55 32.38 0.67 8.46 (15.91, 17.25) 149.16 28.98 151.47 27.87 2.31 7.35 (12.10, 16.72) 0.37
%IBW 104.08 18.76 105.92 18.72 1.84 7.18 (12.23, 15.91) 103.18 20.85 103.90 18.38 2.18 15.57 (28.34, 32.70) 0.46
Albumin 3.67 0.47 3.51 0.38 0.16 0.37 (0.89, 0.57) 3.78 0.38 3.61 0.32 0.17 0.35 (0.86, 0.52) 0.87
Pre-albumin 19.16 5.42 21.22 5.49 2.13 6.12 (9.87, 14.13) 21.58 6.73 22.75 5.40 1.17 5.71 (10.02, 12.36) 0.77
Transferrin 195.43 41.63 193.12 35.04 1.00 35.33 (68.25, 70.25) 209.69 42.83 199.69 42.05 10.00 37.29 (83.09, 63.09) 0.23
Discharge disposition
Home 22 (43.14%) 32 (62.75%) 0.05
Not home 29 (56.86%)* 19 (37.25%)†
Length of stay, d 25.44 7.32 25.98 10.12 0.77
Healed decubiti 3/9 (33.33%) 3/4 (75.00%) 0.26
*26 Skilled nursing facility, 3 acute hospital.
†17 Skilled nursing facility, 2 acute hospital.
FIM Functional Independence Measure; % IBW percent ideal body weight.
hospital. When these seven patients in the standard
group were followed up, three patients went home,
two died, one was in long-term care, and one could
not be traced. Similarly, in the intensive nutrition
group, four patients went home, two were in longterm
care, and one could not be traced.
DISCUSSION The principal findings of this study
are that patients who are undernourished after their
acute strokes benefited from intensive nutritional
supplementation during inpatient stroke rehabilitation.
Those receiving more intensive nutrition
achieved a higher level of functional independence
and more of them were able to go home rather than
to institutional care. The findings of this study differ
from those of Gariballa et al. 7 and of the FOOD
trial, 10 because this study differs in several important
aspects. First, patients in our study were randomized
10–14 days poststroke, rather than within 7 days in
the other two studies. Second, 53% to 62% of our
patients were dysphagic, but none were in the other
two studies; undernutrition is more frequent in
dysphagic stroke patients than those with normal
swallowing. 20 Importantly, the present study included
only patients with evidence of weight loss obtained
from nutrition evaluation post acute stroke,
rather than examining the effect of nutritional supplementation
on both well nourished and malnourished
stroke survivors. These differences may explain
why this is the first study to show improvement in
functional outcomes in patients with stroke administered
nutritional supplements. Finally, prior studies
were in an acute hospital setting, while our patient
population was in a rehabilitation setting.
A key issue in this study was whether the group
that received the intensive supplement was in any
way “healthier” than the group ingesting the “standard”
supplement. We tested carefully for this possibility
and found no evidence for it. Table 3 shows
that none of the functional outcome measures on
admission were significantly different between the
two groups. The possibility that a combination of
factors may have been significant between groups
was examined in detail using several different models
of covariance (ANCOVA). Despite intensive investigation
of this critical point, we found no evidence for
significant differences in baseline health and disability
status between the two groups. LOS, both from
admission and from randomization, varied among
patients, raising the possibility that LOS was a confounding
variable. Accordingly, ANCOVA models
were used to adjust for LOS and once again the initial
study conclusions remained unchanged.
It is not clear whether the beneficial effects of the
supplement were primarily on the brain or on mus-

cle; the lack of significant effects on cognition argues
for the latter possibility.
The study limitations are several. First, this was a
single center investigation. Second, each patient’s estimated
protein and energy needs were not determined
separately on admission to our rehabilitation
service (the design was chosen to facilitate generalization
of the results to normal clinical practice). Third,
ethical considerations prevented our use of a “no nutritional
supplement” control, since previous studies
have established the clinical utility of nutritional supplements
in malnourished patients. 7-9 Fourth, it is
possible to have a weight change of 2.5% in a small
sized person, by using a different scale on admission
to an acute rehabilitation hospital from the one used
in the previous acute hospital. However, 2.5%
weight loss in 2 weeks is still considered significant
weight loss irrespective of the type of scale used to
measure weight. Finally, lack of documentation of
dietary intake (calorie count) may be considered a
weakness of this study. We considered calorie count
to be a rough subjective estimation of the nutritional
intake, which is difficult to obtain and is usually limited
to a 24- to 72-hour period. 14 Despite these limitations,
the strength of this study is the close watch
on compliance of nutritional supplementation intake
within the context of routine care. The data reflect
real life practice, the results of which can be incorporated
into routine clinical care of patients with stroke
admitted to rehabilitation units.
AUTHOR CONTRIBUTIONS
Statistical analysis was undertaken by Meredith Lukin, MS, and Martin
Lesser, PhD.
Received December 11, 2007. Accepted in final form May 21, 2008.
REFERENCES
1. d’Avalos A, Ricart W, Gonzales-Huix F, et al. Effect of
malnutrition after acute stroke on clinical outcome. Stroke
1996;27:1028–1032.
2. Finestone HM, Greene-Finestone LS, Wilson ES, et al.
Malnutrition in stroke patients on the rehabilitation service
and at follow-up: prevalence and predictors. Arch
Phys Med Rehabil 1995;76:310–316.
3. Newmark SR, Sublett D, Black J, et al. Nutritional assessment
in a rehabilitation unit. Arch Phys Med Rehabil
1981;62:279–282.
4. Newmark SR, Sublett D, Black J, et al. Nutritional support
in a rehabilitation unit. Arch Phys Med Rehabil 1981;
62:634–637.
5. Finestone HM, Greene-Finestone LS, Wilson ES, et al.
Prolonged length of stay and reduced functional improvement
rate in malnourished stroke rehabilitation patients.
Arch Phys Med Rehabil 1996;77:340–345.
6. Davis JP, Wong AA, Schluter PJ, Henderson RD,
O’Sullivan JD, Read SJ. Impact of premorbid undernutrition
on outcome in stroke patients. Stroke 2004;35
8:1930–1934.
7. Gariballa SE, Parker SG, Taub N, Castleden CM. A randomized,
controlled, a single-blind trial of nutritional supplementation
after acute stroke. J Parenter Enteral Nutr
1998;5:315–319.
8. Potter JM, Roberts MA, McColl JH, Reilly JJ. Protein
energy supplements in unwell elderly patients: a randomized
controlled trial. JPEN J Parenter Enteral Nutr 2001;
25:323–329.
9. Milne AC, Potter J, Avenell A. Protein and energy supplementation
in elderly people at risk from malnutrition. Cochrane
Database Syst Rev 2005;2:CD003288.
10. Dennis MS, Lewis SC, Warlow C, et al. Routine oral nutritional
supplementation for stroke patients in hospital
(FOOD): a multicentre randomized control trial. Lancet
2005;365:755–763.
11. Kelly PJ, Furie KL, Shafqat S, Rallis N, Chang Y, Stein J.
Functional recovery following rehabilitation after hemorrhagic
and ischemic stroke. Arch Phys Med Rehabil 2003;
847:968–972.
12. Iizuka M, Reding M. Use of percutaneous endoscopic gastrostomy
feeding tubes and functional recovery in stroke
rehabilitation: a case-matched controlled study. Arch Phys
Med Rehabil 2005;865:1049–1052.
13. American Dietetic Association. Manual of Clinical Dietetics,
6th edition. American Dietetic Association; 2000;15.
14. Blackburn GL, Bistrian BR, Maini BS, Schlamm HT,
Smith MF. Nutritional and metabolic assessment of the
hospitalized patient. JPEN J Parenter Enteral Nutr 1977;
1:11–22.
15. Rabadi Meheroz H, Kristal Bruce S. Effect of vitamin C
supplementation on stroke recovery: a case-control study.
Clin Interventions Aging 2007;2:147–151.
16. Stineman MG, Maislin G. Validity of functional independence
measure scores. Scand J Rehabil Med 2000;32:143–
144.
17. Stineman MG, Shea JA, Jette A, et al. The functional independence
measure: tests of scaling assumptions, structure,
and reliability across 20 diverse impairment categories.
Arch Phys Med Rehabil 1996;77:1101–1108.
18. Rabadi MH, Blau A. Admission ambulation velocity predicts
length of stay and discharge disposition following
stroke in an acute rehabilitation hospital. Neurorehabil
Neural Repair 2005;19:20–26.
19. Kuzuya M, Izawa S, Enoki H, Okada K, Iguchi A. Is serum
albumin a good marker for malnutrition in the physically
impaired elderly? Clin Nutr Epub 2006 Sep 21.
20. DePippo KL, Holas MA, Reding MJ, Mandel FS, Lesser
ML. Dysphagia therapy following stroke: a controlled trial.
Neurology 1994;44:1655–1660.
Neurology 71 December 2, 2008 1861