Full text of DOI:10.1177/0148607101025004174

Communications
Original
in
Most
Fed
Patients
Trauma
Infectious
Complications
Due
Parenterally
Are Not
to
Elevated Blood Glucose Levels
A.
Kenneth
Anne
and M. Keith
MD
Kudsk, MD;
Laulederkind, RN, BSN;
Hanna,
From the
Tennessee
ofTennessee, Memphis,
University
ABSTRACT.
between
nutritional studies of trauma
To determine the
and infectious
below values considered to increase the risk for infection
Objective:
hyperglycemia
relationship
in
>
In
to
be
220
trial
levels tended
noninfected
statistical
(SG
2,
mg/dL). glucose
in in
infected than
complications
Methods:
Retrospec-
slightly higher
patients
patients.
the first
review ofserum
valuesintwo
random-
within
5
tive
daysreaching
had similar SGvalues but
significance by
glucose
studies of
published
day
either enteral
isocaloric
5. Unfed control
more
ized,
or
prospective
patients receiving
patients
infectious
signifi-
or
Conclusions:
(trial 1)
Trial
parenteral feeding
isonitrogenous,
cantly
major
complications.
diets
2
Patients
infections had
in
SG
levels
but these SG
enteral
tively
also included
(trial 2).
developing
slightly higher
patients prospec-
noninfected
followed who received little or no enteral
than
feeding.
patients early
admission,
risk
far
for
Results: Patients randomized to
or
values were
below levels considered
a
infective
explain
enteral
parenteral feeding
in trial
1
exhibited no
differences in the
until thefourth orfifth
does not
highest
complications. Significant hyperglycemia
significant
in
differences in infectious
ill
trauma
recordedserum
after
(SG)
glucose
day
complications
randomizedto various routes and
types
critically
SG
to be
4
in infected than non-
but SGwasfar
ofnutrition.
tended
protocol entry.
infected
patients
(Journal
higher
and
EnteralNutrition
in
the first
2001)
of Parenteral
25:174-179,
patients
hospital days,
and clinical studies demonstrate the
in the enteral-fed
the same time
evolved that
differ-
Laboratory
advantages
mg/dL
group during
ofenteral
ofnutrients in
frame. 13 As
a
result,
speculation 11
delivery
with
reducing
1-10
in
ences
infectious outcome between enteral- and
septic morbiditycompared
parenteralfeeding.
par-
and
mechanisms for this increase in with
are
likely
parenteral
due to
enteral-fed
Proposed
sepsis
changes
and
patients
with
overfeeding
feeding.
total
nutrition
bacterial
include
in
(TPN)
parenteral
gut
hyperglycemia
To
the incidence of
in
translocation,
permeability,
impaired
investigate
hyperglycemia
mucosal
due to our two
studies4°5 of trauma
or
immunity.
Recently, hyperglycemia
previously published
to
randomized to
leading
enteral,
parenteral,
overfeeding
immunosuppression
through patients
spe-
and
as
enteral
we
examined the rela-
(PMN)
has
polymorphonuclear leukocyte
diets,
dysfunction
cialty
specifically
been
between
and
mor-
of immunoglobulins
glycosylation
proposed
tionship
early hyperglycemia
septic
a
causeofincreased
with
and concluded that
fails to
septic
potential
complications
ofthis
argument
bidity
hyperglycemia
The
nutrition. 11
is
in
infectious
differences
significant
parenteral
support
explain
complica-
are
threefold.
diabetic
tion rates seen when route and
varied.
of nutrition
First,
patients undergoing major
type
or abdominal
on
cardiovascular
hyperglycemia
surgery, sustaining
postoperative day 1,
(>220
mg/dL)
METHODS
had an infection rate 2.7 times that observed in dia-
betic
with serum
values <220
(VA)
patients
glucose
in the Veterans Affairs
A
review
on the
charts
was
retrospective
performed
Second,
of
mg/dL.11
Cooper-
of 149
trauma
severely injured
patients requiring
ative
nutrition of
Study
perioperative parenteral
who had
entered
feed-
been
parenteral
emergent celiotomy
previously
(>300
general surgical patients, hyperglycemia
g/dL)
1
2
into
of trials
enteral vs
or
comparing
patients)4
occurred in
more
receiving par-
patients
significantly
1: 98
to
randomized
(trial
ing
patients
5
enteral
in the
in the
nutrition
Third,
perioperative period.l2
isocaloric diets
2: 51
for
entry
(trial
isonitrogenous,
patients).5
of randomized trials
meta-analysis
comparing
In
latter
the
into the
study,
patients eligible
and enteral nutrition in
or
ill
parenteral
critically
surgi-
but without enteral access were
protocol
prospectively
cal
the mean
patients
critically injured patients,
not
followed
for the
randomized)
of
(but
development
in
value was >200
the
fed
with 132
glucose
group
mg/dL
parenterally
infectious
nutrient
Details
complications.
regarding specific
on
7
to
9
of the
days
study compared
diet
of infectious
criteria for
and
entry
formulas,
protocols,
infec-
diagnosis
tious
status are
demographics, specific
and measurements of nutritional
complications,
detailed in the
index
The
values of each
extensively
and are not
publica-
present
in this
reiterated
tions
paper.
Kenneth A.
requests:
956
Kudsk, MD,
mailmay
and
reprint
Correspondence
review evaluates all serum
Court
Suite
Memphis, TN38163. Electronic
E228,
glucose
Avenue,
be sent to kkudsk@utmem.edu.
the
5
first
of
for
to deter-
patient
174
days
hospitalization
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015

175
TABLE
I
values
between blood
mine the
relationship
subsequent
glucose
Maximum
values
of
the first five
glucose
(mgldl) during
patients receil’ing
days
The
the
of infection.
and
development
irz
erzteral
or
(EST)
hospitalization
TPwfeeditags
of
Review Board
The
of Ten-
Institutional
nessee
University
the
Health Science Center
study
approved
design.
1
Trial
<
.05 us ENT
4.
~‘p
tip
Day
Day
<
.02
ENT
us
5.
A
total of 98
Index
enteral
hours
received
with an Abdominal Trauma
were randomizedto either
patients
or
enteral
diet.
ENT,
diet; TPN,
parenteral
of 15
(ENT)
(ATI)
greater
or
24
diets within
(TPN)
parenteral
of
Patients
the enteral diet
injury.
receiving
t test.
of variance was used
priate
to
Two-way analysis
11.6 3.1
of
in
the formula
glucose
kcal/kg/d
significantly
differences for
and noninfected
andISS and
assess
infected
sustained
a
lower
in
and
incidence
septic
for
patients,
glucose perkilogram perday,
<
15.7% us TPN:
trauma.
after
fed
(ENT:
.02)
complications
blunt and
40%, p
ATI scores. Arepeated measures
of variance
analysis
values over time.
penetrating
Parenterally
compared glucose
received 14.1
2.4
of
in the
glucose
com-
kcal/kg/d
patients
In
TPNformula.
addition to enteral us
parenteral
into
1
RESULTS: TRIAL
values over
we divided trial
1
4
groups
parisons,
patients
based on route of nutrition and infectious status:
time in
Maximum blood
TPN
ENTand
glucose
infected
enteral noninfected
enteral
(ENT-NON),
patients
this
noninfected
and
had no
(ENT-INF),
parenteral
glucose
study
TPN,
(TPN-NON),
parenteral
In
in
than ENT
TPN resulted
more
infected
Nine
trial,
significantly
(TPN-INF).
patients
diedwithin
infectious
[39.5%]
imumblood
(17/43
The max-
levels recorded
4
of
(2
complications
feeding
days
patients
TPNus 8/46
[17.4%]
.02).
89 for this
n
ENT,
p <
(ENT,
analysis
46;
entry), leaving
values in
first 5
the
n =
mg/dL
glucose
during
43).
is shownin Table I. Therewere
days of hospitalization
differences in
no
blood
between
significant
glucoses
Trial
2
ENTandTPNpatients until the fourth andfifth
of
day
complicate
Trauma
admitted into this
required
patients
study
when infections
to
hospitalization
postinjury recovery. Despite
began
an ATI :2:: 25 or an
and had
emergency laparotomy
the
significantly higher
Score
Enteral
was
access
20.
(lSS) :2::
Injury
Severity
and
infectious rate
levels with
either the TPNor
higher glucose
TPN,
obtainedin 33 patients whowererandomizedto either
in
average maximumglucose values
n
the
diet
or an isoca-
16)
17).
(IED;
immune-enhancing
ENT
were far below levels considered
a
risk
group
diet
n
Patients who
of
(ISO;
loric, isonitrogenous
for
factor for increased
(>220
septic
complications
mg/dL).
met criteria
into the
entry
study by severity
or
but who had no enteral access to
(ISS
ATI),
injury
Mrzximum blood
values in
vs
infected
noninfected
glucose
route
start
enteral
served as controls
no
(n
special-
18)
feedings,
early
of
patients regardless
of feeding
in
to evaluate outcome
ized nutrition
patients receiving
route
are
infected
over
When
patients
of nutrition is
and
Patients
the immune-
ignored
support.
given
with
noninfected
diet received 14.4
1.3
of
compared
patients
no
enhancing
kcal/kg/d
glucose
patients
the first
5
of
therewere
in the enteral formula while the control diet
days
hospitalization,
signif-
2)
icant differences in maximal
serum
or mean
1.2 kcal/d. Sixteen of18
in
received 14.3
the
the
before
the
1)
day
(Fig.
(Fig.
patients
the
fifth
of
values until
but
studied arm of
nutrition
hospitaliza-
glucose
unrandomized,
prospectively
when
the
tion
stress
infectious
and
elicited
recov-
trial received no
complications
specialized
support
to
the
of infectious
postinjury
response began
complicate
complications;
development
Not until
5 did infected
have
in
IV
was
amount of
glucose
the
fluids
unknown.
ery.
day
signifi-
patients
Sig-
<
higher maximum
andmean
fewer
infectious
the
were
.02)
.007)
levels than noninfected
cantly
(p
(p
nificantly
major
complications
serum
There
notedin
diet
patients.
glucose
immune-enhancing
patients receiving
differences between infected and
were no
with
other2
the
(6%)
significant
(Isonitrogenous
compared
groups
unfedcontrol
in
5.9 vs
2.8 vs
noninfected
NON
NON
ATI
or ISS
40.5
(INF =
diet: 41%
patients
~- F.9,
z- 3.9,
nonrandomized,
.02;
p
group:
infec-
42.6
29.7
22.9
<
(INF
58%
tious
The
incidence of
ns)
ns).
.002).
p
major
highest
in the &dquo;unfed&dquo;
occurred
complications
patient
for all
Serumglucose levels were evaluable
population.
in
this trial.
patients
route
nutrition ar2d
infectious complications
time
values over
of
glucose
Influences of
on maximum
Statistical
analysis
for discrete
nutrition and the
route of
When stratified
both
by
mean
variables was
or absence of infectious
Significance
presence
or Fisher’s maximum
complications,
(categorical)
test of
in
determined
levels
infected
were
TPN
sta-
withthe
X2
homogeneity
glucose
patients
exact
all
in
bothENTand
continuous variables were tested with
thannoninfected
test;
higher
patients
4
either t tests or
each
U
tests. Before
of
the first
on
Mann-Whitney
during
significance
were well
days,
hospital reaching
making groups
we
tested the
or
tistical
2
(Table II ).
these
for
at risk
t test,
assumption
equal
day
However,
F
variances with an
test and used the
values
below levels considered
unequal
appro-
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015

176
TABLE II
Maximumglucose values
the
first
four
(mg/dL) during
hospitalization
days
of
<
<
us
c
dvs
.0002;
.007.
.01;
d’: p
.003;
a vs a’: p
.0001;busb’:p
c’: p
<
<
us
e
f
vs
.02;
e’ : p
f’ : p
enteral
TPN-
enteral
ENT-NON,
NON,
noninfected; ENT-INF,
noninfected; TPN-INF,
infected;
infected.
parenteral
parenteral
the
them
occurred on the
and
of
instances,
majority
fourth or fifth
were
as infectious
hospital day
complications
and treated.
diagnosed
FIG. 1. Maximum
levels in nVdL over the first five
of
glucose
days
Cumulative incidence
over time
in infected and noninfected
There were no
patients had
of
episodes
hyperglycemic
hospitalization
significant
patients.
differences
until the fifth
wheninfected
day
Table III stratifies
with
or absence of
or
enteral
patients
paren-
noninfected
than
blood
higher
significantly
sugars
patients.
teral
miaandthe
stratification was carried out at
the
onwhich
feedingby
presence
hyperglyce-
occurred. The
day
hyperglycemia
2
blood
levels:
glucose
deleterious
infection
and
of
effects on
neutrophils
mg/dL).
development
blood
>220
or blood
values >_200
>
sugars
g/dL
glucose
220
SG
(ie,
Ofthe 25 ENT or TPN
who became
mg/dL.
patients
3
demonstrates the outcome of
with
Figure
patients
level above or all blood
20
never had
of
a
blood
>220
(80%)
infected,
glucose
at least
a
blood
single
glucose
the first
4
and 14
(56%)
during
days
hospitalization,
values below 220
Ofthe
who
a
glucose
mg/dL.
patients
never had
a
blood
value above 200
mg/dL.
of
glucose
an
68%
never had
(17
25)
infection,
developed
2
of 46
ofENTand
3
of43
TPN
(4.3%)
(6.9%)
complications preceded
at time the
15
Only
blood
TPN)
>220
5 ENT
Only
(17%;
single
glucose
mg/dL.
infectious
>220
developed
glucose
ofthe 89
entered
at
patients
and 10
into the studies
time
patients
a
blood
by
mg/dL
days.
any
during
a
220
glucose >
mg/dL
any
experienced
during
4
Of the
6
TPN
4
first
infected
the fifth
of
the first
5
and
8
6
hospitalization
ENTand
(2
days
hospitalization,
a
value >220
of
with
by
day,
ofthese
blood
patients
glucose
15 becameinfected. Ofthe
with
(23%)
TPN)
patients
had
a
or
and
3
of
as
of
single episode
day
hyperglycemia,
all
below 220
17 of 74
glucoses
infected,
mg/dL,
on the
these
episodes
infectious
infected or noninfected
occurred
fifth
hospital day
became
for two-thirds ofall infec-
accounting
The
maintained
complications developed.
majority
tions.
the incidence of infection
was
Although
signifi-
blood
in
patients
sug-
the small
of
that
cantly higher
group
patients
inbothENTand
ars below200
TPNgroups and
became
mg/dL
had blood
>220
of 15
(8
p
were
glucose
mg/dL
hyperglycemic
were
controlled when
<
rapidly
they
hyperglyce-
us 17 of 74
of
the
.03),
nonhyperglycemic,
majority
isolated
mic.
episodes
sporadic,
hyperglycemic
If blood
cations of
role in
infectious
after
Three of
glucose plays any
compli-
injury,
trauma
hyperglycemic
patients
5
(60%)
enteral
ENT
feeding appeared protective.
with
of
values >220
during
patients
glucose
mg/dL
the first
tious
5
no infec-
patients.
days
hospitalization developed
us
4
of 10
of TPN
>200
(40%)
complications
in
ENT
with
a
6 of
Likewise,
(75%)
remained
glucose
mg/dL,
patients
patients
8
18
TPN
(38.9%)
us
7
of
patients
free ofinfectious
complications.
RESULTS: TRIAL
2
There were no
in ATI or
differences
ISS between
or between infected and noninfected
groups
patients.
differences in rates of infectious
Despite significant
the
them
among
3
serum
lev-
among
groups,
complications
glucose
were no
els were similar
There
4).
(Fig.
in
differences
mean or maximum
4)
(Fig.
significant
serum
values between infected and nonin-
glucose
the first
5
fected
tion.
of
patients during
infected
days
hospitaliza-
FIG. 2. Mean
the first five
differences
tions
level in infected us noninfected
over
glucose
of
patients
tendedto have
higher
Although
glucose
reach statistical
patients
were
There
no
days
hospitalization.
until the fifth
significant
complica-
levels than noninfected
this did not
values
as
patients,
infectious
day
hospital
and the
developed.
significance,
glucose
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015

177
of
Theincidence
or enteral
infected and
feeding.
3.
FIG.
noninfected
in normal
and
randomizedto
nutrition
patients
glycemic
hyperglycemicpatients
parenteral
(ENT)
(TPN)
far
were
below 220
became
Of the 27
5
who levels in this select
of
above 220
was
mg/dL.
group
patients
complications
patients
(Table IV)
patients
mg/dL,
the incidence of infectious
eventually
infected,
only
signifi-
hadanySGlevels above 220
immune-enhanc-
isocaloric control
(1
to
reduced.
affect outcomein
mg/dL
cantly
Although hyperglycemia appears
3
and1
diet,
diets,
a
diabetic
wedonot
agree
ing
isonitrogenous,
patient).
infection and
population,
unfed control
There was no association
in
and
the clinical
tion studies oftrauma
that
levels
elevated
overfeeding
postoperative glucose
between
this
values
glucose
DISCUSSION
study.
in
outcome differences
as
nutri-
these
in 2
explain
patients
of
suggested by
In
authors.
our
values obtained
analysis
glucose
studies of
torso
The
elevated serum randomized,
mechanisms
which
sustaining
the aver-
levels recorded over the
prospective
patients
proposed
by
severe blunt and
maximum serum
is
increase infectious
penetrating
trauma,
glucose
through
of
complications
of PMNfunction and the
Thelevel
increased risks
age
glucose
glycosylation
impairment
first
165
5
to
associ-
diabetic
from 120
randomizedto enteral or
immunoglobulins.11
of hyperglycemia
days of hospitalization ranged
mg/dL
mg/dLin
patients
in
trauma
with
of infection
to be >220
ated
and
These values are well below the
has been
thereis some
patients
mg/dL,
parenteral feeding.
reported
with
to interfere
necessary
evidencewith
220
considered
regard
gran-
and
never
although
conflicting
mg/dL
in
have
In
of TPN
70%
infected
whobecame
to PMN function
function.
diabetes,13-17
investigators
adherence,18
ulocyte
addition,
patients
in
ofthe
noted
che- 75%
enteral
granulocyte
patients
220
improvement
the
and
and
withcontrol ofblood
had
infectious
a
blood
of
motaxis,l
glu-
glucose >
mg/dL,
not be
phagocytosis2°
majority
cose below 220
could
mg/dL.
complications
explained by
even
and
confirmed
when
lowervalueof 200
the
Clinically, Pomposelli
colleagues 10
hyperglycemia
mg/dL
the
of
control in the first
2
was chosen. Aneven
can be
in
made the
case
importance
glucose
diabetic
postop-
stronger
in
erative
second
of
where
days
By
avoiding glucose
patients.
study
very severely injured patients
TABLE III
Cumulative incidence
over the
firstfive
ofinfectious
of
days hospitalization
complications
a
remain in the
total even if the
was an
*Once
are listed on
diet.
cumulative
isolated,
event.
patients
particular day, they
single
episode
TPN.
parenteral
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015

178
intra-abdominal
the sixth
made
abscess maybe
thatinfectious
CATscan on
cer-
by
hospital day,
complication
been
the
before
of stress.
had
tainly
present
diagnosis, inducing
the metabolic
almost
there is
changes
an
Similarly,
active infective
process occurring
certainly
before more obvious
within
pulmonary parenchyma
infiltrates and
manifestation of
which
purulent
pulmonary
of
leads to the
diagnosis
pneu-
sputum,
readily
It
seems
monia and the institution of
therapy.
hyperglycemia
logical,
noted
in
was
but
that the
unprovable,
infected
after
in the first few
days
injury
patients
metabolic
the
to
a
infection;
thus,
developing
response
that
it is no
who
become
ultimately
surprise
infected aremore
patients
in
their
below
early
hospital
hyperglycemic
most
far
the
220
remain
stay.
this,
Despite
whichhasbeenassociated with
mg/dL,
function.
in
neutrophil dys-
hyperglycemia
increase in
the
Unfortunately,
who
tool because
not
as
a
infections is
patients
develop
helpful
FIG. 4. Meanglucose values for
an
an immune-
or anunfed
differences in blood
randomized to
patients
a
15
difference in
isocaloric control
diagnostic
glu-
diet,
isonitrogenous,
diet,
mg/dL
enhancing
There were no
cose between infected and noninfected
population.
patient
glucose
significant
(shown
groups
of
anyof the groups during
days
values between
thefirst five
on
is too subtle to be
relevant
and
Table I)
too low to
clinically
hospitalization.
immune defenses.
probably
This
some
impair
is not to
that
not
say
hyperglycemia may
play
but
diet
with
values >220
glucose
were influenced
role;
patients
mg/dL
complica-
major septic episodes
almost no
by
had
tions
below 220
a
rate of infectious
significantly
higher
developed hyperglycemic episodes.
patients
than
unless one
mechanisms for the
different
of infectious
whose blood
sugars
were
(55%)
patients
Therefore,
totally
implicates
development
and
infectious
(23%),
mg/dL
thereby, hyperglycemia
compli-
rela-
ofabscess
the
is
But this
in
the
cations in these
2
a
studies,
may aggravate
complications.
plays
hyperglycemia
in
effect
minor as shown
trauma
the
probably
relatively
unimportant
tively
part
development
second trial where
and
after severe trauma. Weconclude that
of nutrition
determinant of
severely injured
patients
pneumonia
>
>
ATI
with an
25 or an ISS
20 had
a
of
the
most
route and
to be
range
6%to 58%
from
major
type
appear
infectious
in
infectious
septic
complications
nutrition but
depend-
important
complications
blood
rather than level
these studies
ofserum
ing upon
therapy,
in
average
groups.
sugars
glucose.
In
remained almost identical between
at odds
These results seem to be
of
with
fact,
Pomposelli’s
had
few
this
data from the VA
study
significant hyper-
interpretation
very
patients
at
published
and the
coop-
andpatients whoreceived
erative
of
little,
time,
any
study&dquo;
meta-analysis
enterally glycemia
if any,
nutrition
support
had the
and could not
the increase
rate of
and
fed
and
trauma
highest
major
parenterally
general surgical
3
have been
In the VA
possibly
patients.3
cooperative study, however,
septic complications
in
infectious com-
overfed. Taken in
levels were
in
the
and
toto,
although glucose
significantly higher
2
cannot be
the
within
anygroup
studies
fed
it
in
is unknown when
plications
explained
parenterally
patients,
occurred
the clinical course
by hyperglycemia.
havebeen
Anumberofmechanisms
during
hyperglycemia
for the
whetherit occurred
or after the onset of
before,
proposed
during,
complications
associatedwith
infectious or noninfectious
selves
which them- increasedinfectious
par-
occur within the
complications
enteral
release of
changes
Histologic
feedings.
hyperglycemia through
and catecholamines.
aggravate
in
with increases
in
association
tract
gastrointestinal
cortisol,
Similarly,
cytokines,
within the
in
bacterial trans-
does
of humans with
mucosal
increases
and
blood
fed
were
meta-analysis,
in
sugars
patients
signifi-
permeability
in
location
occur within
animal models.21-25
Although atrophy
at baseline
cantly higher
parenterally
mucosa
or at the
ofthe
4
How-
associ-
to
the duodenal
the
(160
6)
ever,
midpoint
study (days
mg/dL)
of the
extent
dramatic as
whenglucose levels
178
histologic
in
parenteral nutrition,22
approximated
mg/dL.
data
are
noted
not
as
the severe
(>200
changes
mg/dL)
nearly
hyperglycemia
have
a
small
with
which
in
ated with increased
infectious rates
rats,
proximal
experiments
parenteral
mass
40% to
50%
bowel that reduces its mucosal
did not occur until the end of the
(on
by
patients
study
7
to
in
a
time frame characterized
the
9)
of
days
by
in
a
presence
trauma
We
septic complications
severely injured
TABLE IV
population.
ourresults
IED vs ISO vs No
Feeding
to meanthat
interpret
to be
hyperglycemia
rather
a
to
response
septic morbidity
appears
than the cause
infectious
The
of
of septic morbidity.
majority
were
between the
how-
complications
diagnosed
fourth and
It is
eighth hospital day.
to determine
impossible,
how
infectious
ever,
exactly
long
compli-
cations manifested themselves
before
immune
isocaloric control
metabolically
diet; ISO,
TED,
enhancing
isonitrogenous,
diet.
their
For
of an
diagnosis.
example, although diagnosis
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015

179
Li
KB:
Loss of
6. Kudsk
J.
tract
1996
nutrition
KA,
with
fed.23 Increases in
do
Renegar
upper respiratory
not
when
occur
enterally
permeability
to normal
returned
feeding. Ann
223:629-638.
Surg
immunity
parenteral
have
but
trauma
after
byday
to be
7. Li
on
Kudsk
Gocinski
et al: Effects of
B,
KA,
J,
parenteral
39:44-52,
of
in
do not
increases
and
appear
permeability
increases in infectious
cause is
7,
tissue.
et al:
KB,
J
Trauma
1995
gut-associated lymphoid
Li
with
more
8. Janu
associated
tions.~~°25
complica-
to atro-
P,
J,
Renegar
Recovery
gut-associated
after
immunity
tissue and
tract
The
lymphoid
upperrespiratory
par-
secondary
likely
enteral nutrition. Ann
1997
225:707-717,
Surg
Kudsk KA:
J,
tissue
asso-
(GALT)
ofthe
phy
lymphoid
gut-associated
parenteral
9.
Li
A
of TPN-induced
BK,
King
changes
Arch
temporal study
nutrition associated with
with
ciated
in
tissue and mucosal immu-
gut-associated lymphoid
intestinal and
of
and
levels
decreased
respiratory IgA
1997
132:1303-1309,
nity.
Surg
antiviral and antibacterial
established
al:
et
infection rate in
10.
Baxter
Babineau
JK, TJ,
of
loss
JJ,
Pomposelli
Early postopera-
tive
control
nosocomial
diabetic
it hasnot
maintenance ofmucosalimmu-
beentestedin
yet
glucose
predicts
1998
defenses.7-9,26-28
Although
JPEN
22:77-81,
patients.
clinical
the
population,
11. The
Affairs TPN
Veteran
in
Group:
325:525-532,
Cooperative Study
Periopera-
for differences in
a
cogent explanation
provides
nity
NEngl
tive TPN
J
Med
1991
surgical patients.
abscesses and
ofintra-abdominal
fed
the rates
in
12.
13.
14.
Walters
Stevenson T: Oxidative metabolism of
pneumonia
LM,
leukocytes
from nondiabetic anddiabetic
J Lab Clin Med78:158&mdash;
patients.
patients.
parenterally
166, 1971
Nielson
RJ: Reversible abnormalities in
controlled diabetic patients. Am
JD,
KL,
Bulger
Bagdade
function in
J
CONCLUSIONS
phagocytic
poorly
MedSci
1972
263:451-456,
with
in many
reported
histologic
rates ofinfectious
Walters E:
adherence.
A
con-
paren-
clini-
complications
Higher
Bagdale JSM,
Impaired granulocyte
reversible defect in host defense in
with
have been demonstrated
teral nutrition
patients
poorly
Diabetes
1978
trolled diabetes.
27:677-681,
have
studies. These
work
been
both
cal
by
findings
IV. Measurementof
diabetes.
Immunityin
15. RichardsonR:
phago-
1942
leuko-
and
experimental
immunologic
demonstrating
in
Am
of
Sci
Med
diabetes mellitus.
J
204:29-35,
cytic activity
with
nutri-
serum
associated
changes
parenteral
16.
J: The
Bybee
cytes
Med
activity
polymorphonuclear
phagocytic
mild but
in
with diabetes mellitus.
J
Lab Clin
tion.
increases
obtained from
patients
Although
significant
1964
64:1-13,
be markers
in
infections
of
glucose may
subsequent
17. Jones
alterations in
Peterson CM:
diabetes
RL,
mellitus. Am
Hematologic
1981
levels
uncontrolled
critically
injured patients,
glucose
J
Med
70:339-352,
not
do
ated
surgical patients.
the increased rates of infection associ-
explain
with
18. Mowat BaumJ: Chemotaxis of
A,
patients
627, 1971
19. Kudsk
polymorphonuclearleukocytes
with diabetes mellitus. J Med 284:621&mdash;
at
in
least
trauma and
from
N
parenteral nutrition,
Engl
Stone
et al: Effects of enteral and
G,
Carpenter
KA,
JM,
of
malnourished rats on
1982
parenteral feeding
body composition.
J
Trauma
ACKNOWLEDGMENTS
22:904-906,
Offenbartl
et al:
Westrom
B,
20. Purandare
Increased
gut
S,
K,
The authors thank
Kazuhiko
theillustrations and
for his
Par-
Doris
to fluorescein
after TPN
Fukatsu,
MD,
permeability
isothiocyanate-dextran
in rat. Scand
J
Gastroenterol 1989
24:678-682,
assistancein
preparing
21. Deitch EA:
tial future
22. van der
failure:
and
Multiple organ
poten-
et al:
Pathophysiology
sons for secretarial
support.
Ann
von
1992
therapy.
Surg 216:117-134,
Hulst
van Kreel
BK,
RRWJ,
MF,
Meyenfeldt
Glutamine and the
of
Lancet 341:
preservation
gut integrity.
REFERENCES
1363-1365, 1993
et
intake in
Gastroenter-
et
23. Levine
Derin
maintenance of
al: Role of oral
1. Alexander
effects of
Ann
Macmillan
Stinnett
al: Beneficial
burnedchildren.
GM,
JJ,
E,
Steiger
JW,
JD,
BG,
massanddisaccharide
in
gut
activity.
aggressive protein feeding
severely
1974
1980
Jones
trauma&mdash;reduced
67:975-982,
ology
192:505-517,
Surg
Donovan
al: Increased
trauma.
2.
Moore
abdominal
1989
al: TENversus TPNfollowing
septic morbidity.
24.
Pate
blunt and
et
Moore
B,
TB,
LM,
FA,
EE,
TN,et
Langkamp-Henken
J
Trauma
intestinal
permeability following
penetrating
major
Crit Care Med
1995
et al:
23:660-664,
Minard
29:916-923,
al:
et
enteral feed-
Early
com-
25. Janu
Li
interleukin-6
Forum
3.
Feliciano
Moore
(IL-6)
1996
P,
G,
FA,
DV,
RJ,
J,
Systemic
Andrassy
correlates with intestinal
with
reduces
47:7,
septic
postoperative
meta-analysis.
permeability. Surg
ing,
compared
parenteral,
Ann
216:172-
26. Li
Kudsk
Hamidian
et al: Bombesin affects mucosal
tissue in IV-fed mice.
lymphoid
a
The results of
KA,
and
M,
J,
Surg
plications :
183, 1992
Kudsk
gut-associated
1995
immunity
4.
5.
al: Enteral versus
after blunt and
Arch
Surg
Croce
Fabian
et
TC,
130:1164-1170,
Li et al: Effect of bombesin on
J,
KA,
MA,
paren-
pene-
teral
on
Kudsk
Effects
27. Janu
P,
KA,
morbidity
septic
impairment
feeding:
upper respiratory tract immunity
1997
by
abdominal trauma. Ann
1992
induced
TPN.
of
Arch
Surg
trating
Kudsk
Surg 215:165-173,
et
A
al: randomized trial of
Minard
Croce
G,
enteral diets
KA,
MA,
132:89-93,
King
severe trauma: An
complications. Ann
Li
28.
Kudsk
KA,
influence mucosal
229:272-278,
et al: Route and
to bacterial
of
nutrition
type
isonitrogenous
J,
following
BK,
immune-enhancingdiet
reduces
Ann
(IED)
septic
immunity
pneumonia.
Surg
1996
1999
Surg 224:531-543,
Downloaded from pen.sagepub.com at Bobst Library, New York University on May 27, 2015