ARTICLE IN PRESS

Clinical Nutrition (2005) 24, 502–509

http://intl.elsevierhealth.com/journals/clnu

ORIGINAL ARTICLE

Negative impact of hypocaloric feeding and energy

balance on clinical outcome in ICU patients
Stéphane Villeta, René L. Chiolerob, Marc D. Bollmannb,
Jean-Pierre Revellyb, Marie-Christine Cayeux RNb,
Jacques Delaruec, Mette M. Bergerb,

a
Anesthésiologie, Centre Hospitalier Universitaire Vaudois (CHUV), 1011 Lausanne, Switzerland
b
Soins Intensifs Chirurgicaux et Centre des Brûlés CHUV-BH 08.660, CH 1011 Lausanne, Switzerland
c
EA-948 Oxylipides, Laboratoire Régional de Nutrition Humaine, CHU de Brest, 29200 Brest, France

Received 29 January 2005; accepted 30 March 2005

KEYWORDS Summary
Nutritional support; Background and aims: Critically ill patients with complicated evolution are
Enteral nutrition; frequently hypermetabolic, catabolic, and at risk of underfeeding. The study aimed
Critically ill; at assessing the relationship between energy balance and outcome in critically ill
Malnutrition; patients.
Outcome; Methods: Prospective observational study conducted in consecutive patients
Infection stayingX5 days in the surgical ICU of a University hospital. Demographic data, time
to feeding, route, energy delivery, and outcome were recorded. Energy balance was
calculated as energy delivery minus target. Data in means7SD, linear regressions
between energy balance and outcome variables.
Results: Forty eight patients aged 57716 years were investigated; complete data
are available in 669 days. Mechanical ventilation lasted 1178 days, ICU stay 1579
was days, and 30-days mortality was 38%. Time to feeding was 3.172.2 days. Enteral
nutrition was the most frequent route with 433 days. Mean daily energy delivery was
10907930 kcal. Combining enteral and parenteral nutrition achieved highest energy
delivery. Cumulated energy balance was between
12,600710,520 kcal, and
correlated with complications ðPo0:001Þ, already after 1 week.
Conclusion: Negative energy balances were correlated with increasing number
of complications, particularly infections. Energy debt appears as a promising tool
for nutritional follow-up, which should be further tested. Delaying initiation
of nutritional support exposes the patients to energy deficits that cannot be
compensated later on.
& 2005 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +41 21 31 42 095; fax: +41 21 31 43 045.

E-mail address: mette.berger@chuv.hospvd.ch (M.M. Berger).

0261-5614/$ - see front matter & 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.clnu.2005.03.006
 ARTICLE IN PRESS
Impact of hypocaloric feeding on ICU outcome
Introduction
Critically ill patients with prolonged and compli-
cated evolution have intense metabolic responses,
generally characterized by hypermetabolism and
protein catabolism. During the acute phase of their
illness, these patients are also exposed to under-
feeding, and to negative energy balances resulting
from insufficient feed delivery intake, which favor
the development of progressive malnutrition. Ent-
eral nutrition (EN), which is the recommended
method of artificial feeding in intensive care units
(ICU)1 is frequently associated with insufficient
energy delivery.2 Moreover, compared with other
ICU treatments, feeding continues to be considered
lower priority, due to lack of specific education.3,4
The deleterious effects of underfeeding on out-
come have long been recognized,5 and the role
of hospital in its development and worsening
has newly been re-emphasized by the European
Council.6
The optimal way to quantify malnutrition in the
ICU is still a matter of debate, as the criteria are
not unanimously accepted. Surgical scores, such as
the prognostic nutritional index (PNI), have not
been validated in ICU settings.7 The actual weight
is a source of confusion as is not reliable for
nutritional assessment due to the enormous fluid
shifts observed in critically ill patients. Indeed non-
nutritional factors such as fluid balance, and
inflammatory status are more important determi-
nants of body weight early on the clinical course:
the actual body mass index (BMI) becomes unreli-
able. Protein indicators of malnutrition (plasma
albumino35 g/l, plasma transferrine o2 g/l), are
altered by critical illness. Finally, anthropometric
determinations such as the cutaneous skin-fold are
rendered useless by edema. The patient’s physical
aspect, his nutritional history and the presence of
acute disease on admission to the ICU appear as the
best tool to assess nutritional status.8 Such a
clinical assessment should be part of routines,
particularly in the sickest ICU patients.
’’
The difficulty of assessing beginning’’ malnutri-
tion prompts for search of other variables to detect
malnutrition. Using indirect calorimetry, a few
trials have shown that underfeeding is indeed
frequent in the ICU,9–11 such patients may be
hypo-, iso- or hypermetabolic, which is difficult to
predict on a clinical basis.10,12,13 While energy
intake can easily be recorded in the ICU setting, the
measurement of total energy expenditure is pro-
blematic: 24-h indirect calorimetry, direct calori-
metry or doubly labeled water techniques are not
available in the ICU14; determining fat stores is only
possible under strict investigation conditions. Even
503
indirect calorimetry is not available in every ICU: it
enables a fairly accurate estimation of the 24-h
resting energy expenditure (REE)15 and its extra-
polation to total energy expenditure. In addition,
indirect calorimetry is not always possible to carry
out, e.g. in a patient with an O2 inspiratory fraction
460%, or extubated and non-cooperative, or in
those requiring fluidized beds for nursing (air
leaks).
This present study aimed at testing the relation-
ship between energy balance and clinical outcome
in severely ill surgical patients with prolonged ICU
stay, and to confront the results with the usual
biological markers of malnutrition.
Material and methods
The study was designed as a prospective observa-
tional study in consecutive patients staying for
more than 5 days in the surgical ICU of the tertiary
University Hospital of Lausanne. Exclusion criteria
were major burns, or short stay. The study was
conducted with approval of our institutional Ethics
Committee which delivered a 5waiver of con-
sentb provided for this observational study. Data
were made anonymous for analysis.
Patient data: Age, sex, pre-event weight, height,
and BMI were recorded; nutritional status on
admission was assessed as: malnourished if 2 of
the 3 below criteria were present: weight loss
410% within the last 6 month, BMI o19, wasting
’’
disease; they were considered normal’’ with
BMI ¼ 20229, and obese if BMI 430. Patients were
weighted weekly during the stay when possible.
Severity of disease was assessed using the simpli-
fied acute physiologic score (SAPS II) score during
the first 24 h.16 Organ failure was quantified
by the SOFA score (sepsis-related organ failure
assessment)17 on admission, and on weekly basis:
this score defines 6 organs/systems and attributes
0 (no failure) to 4 points (maximal failure) to
each according to defined criteria of severity:
stratification was 1–2 points for one organ ¼ mode-
rate failure, 3–4 points ¼ severe failure. The
cardiac SOFA score escalates from 1 ¼ mean
arterial pressure o70 mmHg, to 4 ¼ continuous
dopamine/dobutamine 415 mg/kg or, epinephr-
ine/norepinephrine 40.1 mg/kg/min.
Feeding protocol: Artificial nutritional support
was considered in patients unable to resume oral
feeding within 4–5 days.18 Indications for early
enteral nutritional support, i.e. within 48 h, were:
multiple injury, major gastro-intestinal surgery for
solid tumor, and pre-existing malnutrition. In other
 ARTICLE IN PRESS
504
ICU patients, nutritional support was considered if
they were predicted to be unable to feed them-
selves for 5 days or more; EN was then initiated
between days 2 and 4. EN is recommended as the
technique of choice: the ICU protocol defines the
stepwise increase of feed delivery over the first 4
days (objective: reaching target at 4th day of
initiation of enteral feeding). Both gastric and post-
pyloric feeding may be used. Parenteral nutrition
(PN) is used when EN is contraindicated, or when
difficult EN causes obvious energy deficit (com-
bined nutrition). By internal restrictive policy, no
albumin infusions was given in patients with plasma
albumin levels 418 g/l.
Nutritional data and calculations
 Determination of energy requirements: REE was
measured whenever possible by indirect calori-
metry (Deltatracs Metabolic Monitor, Datex,
Helsinki, Finland) connected to the ventilator in
mechanically ventilated patients, or using a
system with head canopy in spontaneously
breathing patients: measurements were made
over 40–50 min periods in resting conditions.
Energy target was then set at 1.3 times REE. In
the absence of indirect calorimetry, target was
set at 30 kcal/kg/day.19 Energy targets were set
at 100% from the first day of admission.
 Energy delivery: total delivery includes energy
from enteral and parenteral feeds, from non-
nutritional sources (glucose and gluco-saline
infusions used for drug dilution and fluid support)
and lipids delivered with sedatives (propofol).
 Energy balance was calculated as energy deli-
very—energy target, on daily basis: the energy
stored was ignored. In patients without calori-
metry, this calculation was an estimation.
Energy deliveries and balances of 7 days starting
on admission were pooled as weekly data:
weeks 1, 2, 3 and 4 after admission were
analyzed. Cumulated energy balance was calcu-
lated on discharge. Four levels of energy balance
were considered for analysis9: (1) no deficit:
+1000 to
5000 kcal, (2) moderate deficit:
5001
to
10000 kcal, (3) severe deficit:
10001 to

20000 kcal, (4) extreme deficit, larger then

20000 kcal.
Clinical follow-up: Length of mechanical ventila-
tion, infections according to pre-defined criteria,20
use of antibiotics, duration of ICU stay, complica-
tions and mortality were recorded. Complications
were: neurological (coma, seizure, metabolic en-
cephalopathy), respiratory (length of mechanical
S. Villet et al.
ventilation, acute failure assessed by the PaO2/FiO2
ratio, pneumothorax), gastro-intestinal (GI bleed-
ing, ileus), cardiovascular (myocardial infarct,
heart failure, tamponade, vasoplegia, atrial fibril-
lation), hepatic failure (by SOFA), renal failure (by
SOFA), and coagulation disorder (hemorrhagic
shock, surgical site bleeding, thrombosis, embolic
event, thrombopenia). Outcome was recorded at
30 days.
Analysis: Blood samples were collected on
weekly basis to determine plasma albumin, pre-
albumin and C-reactive protein.
Statistics: Data are presented as means, med-
ians, and ranges. Analysis of daily and weekly
values of relevant variables was done using one-
way ANOVA. Non-parametric test was applied.
Multiple linear regressions were performed to
identify the predictors of outcome, and to analyze
the impact of energy balance using standard least
squares with statistical package JMP 4.0 (SAS
Institute, USA). Significance was considered at the
level of Po0:05.
Results
Fifty-five surgical ICU patients were enrolled out
of 962 consecutive admissions (5.7%): 7 patients
were excluded due to incomplete data, leaving
48 patients and 669 ICU days for analysis. The
patients’ characteristics are reported in Table 1.
The length of stay was variable with a mean length
of stay of 15 days: 24 patients stayed for more than
2 weeks, 5 for more than 3 weeks, and 4 for 4 weeks
or longer. Organ failure (single or multiple) was
present in all the patients. The ICU mortality was
35%, in agreement with the mortality predicted by
the high SAPS II scores (median SAPS II ¼ 45, with
predicted mortality 0.38).
Energy requirement determination: REE was
measured in 33/48 patients (69%), repeated mea-
surements being available in 7 patients. Mean REE
was 17007325 kcal/day, resulting in a mean energy
target of 22107500 kcal/day, which corresponds to
2977 kcal/kg/day.
Nutritional support: Feeding was started within
3.172.2 days after admission (Table 2A), resulting
in 101 days without any feeding (15.1%). Patients
with gastro-intestinal surgery and surgical jeju-
nostomy were started earliest. Of the 18 patients
identified as requiring early enteral feeding (10
with multiple injury, 3 with gastrointestinal surgery
and 5 with malnutrition), 11 (61%) were indeed fed
early. The majority of non-early fed were multiple
injury patients (feeding was started only between
 ARTICLE IN PRESS
Impact of hypocaloric feeding on ICU outcome 505

Table 1 Patient characteristics.

Variable Mean7SD (median) Range

Age years 57716 (61) 18/82

Sex ratio M/F 30 M/18 F
SAPS II 45716 (46) 14/94
SOFA 1st week 7.574.5 (7.5) 1/19.0
Failing organs n 3.271.5 (3) 1/4
Mechanical ventilation days 1178 (10) 0/35
Complications n 675 (5) 0/23
Length of ICU stay days 1579 (13) 5/49
Mortality
ICU n (%) 17 (35%)
30 days 21 (38%)

Abbreviations: SAPS: simplified acute physiologic score; SOFA: sepsis-related organ failure assessment; BMI: body mass index;
ICU: intensive care unit.

Table 2 Time to feeding by diagnostic category used in 433 days (including 17 days of transition to
(A), routes of feeding and energy delivery during oral feeding), parenteral only in 36 days, combined
the 669 days of artificial feeding (B). EN and PN in 81 days. Oral feeding was used in the
majority of patients at the end of their stay, but
(A) only for a few days: energy delivery was generally
Patients N Days low during these days. Considering the low number
All 48 3.172.2 (3) of days on exclusive oral feeding (n ¼ 18; 2.7%),
Trauma 10 3.870.7 (3.5) the detailed analysis includes 669 days of artificial
Cardiac surgery 13 3.470.6 (3) nutrition with 433 days of enteral feeding, 36
Respiratory failure 7 2.770.9 (2.5) days of pure PN and 81 days of combined feeding
Gastro-intestinal 3 1.771.3 (2) (Table 2B).
Sepsis 3 2.571.6 (2.5) Energy delivery and balances (Fig. 1, Table 2B):
Transplantation 4 3.071.2 (3) Energy delivery was lowest during the first week.
Other 8 2.970.8 (1) The days without feeding were characterized by the
(B) unintentional delivery of 150–600 kcal from glucose
Routes Days Energy delivery and from sedative lipids. With prolongation of
the ICU stay, reducing the energy deficit, delivery
No feeding 101 3457410 (225) increased to target for each subsequent week. The
(4/3/1.75) difference between energy target and delivery
Oral Feeding 18 8057490 (880)
decreased from
1270 kcal/day as a mean during
Enteraly 433 13657770 (1320)z
Combinedy 81 21607650 (2175)z the first week, to a mean of
625 kcal/day during
Parenteral 36 19157625 (1710)z the fourth week. Energy delivery was significantly
larger during combined enteral and parenteral
Results as mean7SD (median). feeding, reaching energy target. As expected, the
 As defined: days without oral or artificial feeding.
y
largest negative balances were observed during the
Enteral feeding includes 416 with pure enteral and 17
days with transition to oral feeding.
first week: delay between admission and initiation
z
Po0:0001 between enteral and either parenteral or of feeding resulted in a negative balance gap, which
combined nutrition. was not filled thereafter.
y
Combined ¼ combination of EN and PN. Plasma proteins (Table 3): albumin and pre-
albumin were low in all patients without any
significant change over time, and independent of
days 2 and 5), due to trauma to the small bowel, energy balance; CRP decreased.
attempted extubation, etc. Enteral nutrition was Evolution of patient weight and BMI: On admis-
the primary choice: 36 patients (75%) were fed sion 5 patients had a BMIo20, 23 were normal, and
exclusively by the enteral route, 11 (23%) received 20 had a BMI427. Repeated weekly weight deter-
combined EN and PN, while only 1 patient (2%) was minations were only available in 11 patients. During
fed exclusively intravenously. Enteral nutrition was the first week, 4 patients gained weight, 3 were
 ARTICLE IN PRESS
506 S. Villet et al.

stable, and 4 lost weight (BMI varied similarly). Discussion

Only 4 patients had lost weight on their transfer to
the ward.
Complications: There was no correlation be-
tween the total number of complications and the
nutritional status on admission. Fig. 2 shows the
relation between the progressive negative energy
balance and the number of infectious complica-
tions. Length of ICU stay was not correlated
with either SAPS II score, SOFA score or the number
of failing organs on admission, while cumulated
energy deficit strongly was ðPo0:0001Þ. Fig. 3
shows the multiple regression analysis between
length of ICU stay and patient variables and
cumulated energy balances of the ICU stay. Energy
deficit after 7 days and cumulated during the stay
were correlated with both total and infectious
complications (P ¼ 0:048 with F ¼ 4:14, and P ¼
0:0049 with F ¼ 8:81, respectively). Table 4 shows
the correlations between the total energy deficit
and the total number of complications during the
ICU stay. The correlations were also strong with the
length of mechanical ventilation, the total number
of complications, the infectious complications, the
antibiotic days, and the length of ICU stay. Energy
deficit was not correlated with mortality. Plasma
proteins were not correlated to either of the above
variables.
The hypothesis that persistent hypocaloric feeding
and negative energy balances are associated with
poor outcome in the sickest critically ill patients
was verified. This study investigated the nutritional
support over 669 ICU days in 48 very sick patients
with prolonged ICU stay. It confirms the difficulty to
keep a positive energy balance using the enteral
route. Only 11 patients (23%) were in positive
balance at the end of their stay, while 37 were in
negative balance throughout. These results were
observed in an ICU, which has been applying
feeding protocols for a prolonged period of time.2
Infections are a classical complication of mal-
nutrition and underfeeding: a recent study includ-
ing 138 medical ICU patients,21 showed that
hypocaloric feeding was associated with increasing
blood stream infections. Another study including
200 medical ICU patients, observed a reduction
in length of mechanical ventilation associated
with improved nutritional support.22 The negative
energy balance was strongly correlated with
complications, and particularly with the infectious
complications. The multiple regression analysis
showed that energy balance at the end of the first
week, and the cumulated energy balance of the ICU
stay were the strongest predictors of prolonged ICU

5000
3000 Target Delivery *p < 0.001 0
Energy – daily mean (kcal)

–5000
tot energy balance

2000
–10000
* –15000
1000 *
–20000
0 –25000
–30000
–1000 Balance
–35000
n = 48 16 11 7
–2000 –40000
1 2 3 4 –1 0 1 2 3 4 5 6 7 8 9 10
Weeks after admission infec tot

Figure 1 Progression of energy delivery compared to Figure 2 Relation between the progressive negative
energy target over 4 weeks: the figure shows that energy energy balance and the number of infectious complica-
delivery increases with time, reducing daily deficit. tions.

Table 3 Plasma proteins.

Protein Week 1 Week 2 Week 3 Week 4 Ref. value

P/n 33/48 16/16 11/11 7/7

Albumin 22.275.2 22.474.6 22.975.3 22.075.5 35–45 g/l
Prealbumin 0.1770.08 0.1570.08 0.1570.06 0.1770.06 0.2–0.4 mg/l
CRP 142798 96760 91782 80760 o10 mg/l

P/n: number of patients with visceral albumin/prealbumin determinations among patients present during the same period.
 ARTICLE IN PRESS
Impact of hypocaloric feeding on ICU outcome 507

50 50 50
Energy Balance:Cumulated Energy Balance: 1st week N organ failures
P<.0001 P=0.0134 P=0,6534
40 40 40
ICU stay (d)
30 30 30

20 20 20

10 10 10

0 0 0
-40 -30 -20 -10 0-5 -3 -2 -1 0 .5 1 2 3 4 5 6
Kcal/1000 Kcal/1000 N
50 50 50
Age (years) BMI SOFA score
40 P=0.2984 40 P=0.2622 40 P=0,8998
ICU stay (d)

30 30 30

20 20 20

10 10 10

0 0 0
10 20 30 40 50 60 70 80 90 15 20 25 30 35 40 0 5 10 15 20

Figure 3 Multiple regression analysis showing the influence of energy balance, number of organ failure, age, BMI and
SOFA score on length of ICU stay.

Table 4 Relationship between complications and
cumulated energy deficit by regression analysis.
Variables
Length of stay
Complications
Infections
Days on antibiotics
Start of nutrition
Days of mechanical ventilation
stay. Correlation does still not mean causality. It is
obvious for any ICU specialist, that the longer the
stay, the likelier a patient is to have experienced
complications and to have received antibiotics at
some stage: the most severe patients are also the
most difficult to feed by the enteral route. The fact
that this negative correlation was already present
at the end of the first week is another argument
supporting the role of the negative balances in
favoring complications; finally the strong relation-
ship between the complications and the initiation
time of feeding also supports the hypothesis.
The conventional nutritional variables were of
little help, confirming other studies.23 Plasma
albumin and pre-albumin were uniformly de-
pressed, and inversely related to the inflammatory
status: no relation was observed between these
proteins and either energy balance or outcome.
Repeated weight determinations were only avail-
F P able in 11 patients: this reflects the difficulty in
getting weight determinations in ICU patients. In
25.18 0.0001
15.15 0.0003
these patients, weight changes reflected mostly
9.14 0.0042 fluid balance changes associated with the hemody-
17.48 0.0003 namic and inflammatory problems, and not the
17.17 0.0002 energy balances. The body weight changes are
17.12 0.0002 important information for the clinician, but this
information is indeed not only nutritional, since it
seldom reflects the change of lean body mass
during the early phase. The BMI calculated on the
actual body weight was also unreliable: its value is
limited to pre-admission assessment based on the
’’
dry’’ pre-illness weight.
It has recently been shown that evidence-based
nutritional support promoting the use of early EN is
able to improve clinical outcome and even to
reduce death risk in critically ill patients.22 But ’’
there is yet no answer to the question how long
can an ICU patient be starved without deleterious
consequences’’. Considering the results of the
present study, which show that the initial energy
deficits cannot be compensated, this time appears
to be limited.
How the negative energy balance built up is
important to consider. The energy balances were
calculated considering the full 100% target from the
 ARTICLE IN PRESS
508
first day of admission. Although the energy bal-
ances became more negative with prolonged
ICU stay, the most important part of the energy
deficit was built up during the first week in the
ICU, creating an energy debt. This is the logical
consequence of depressed gastro-intestinal activity
during the first 2–3 days after injury or surgery and
the delay of restoration of organ function; it is also
the consequence of respecting the guidelines on
artificial feeding, i.e. initiating feeding within 48 h
in case of indication to early feeding, but between
days 5 and 7 in the other critically ill patients. This
resulted in an energy gap of roughly 5000–9000 kcal
at the end of the first week, which was never to be
filled thereafter. Pushing early EN seems to be the
most appropriate tool to limit such deficit in
patients at risk; alternatively, early administration
of concentrated intravenous glucose or full PN
under tight glycemic control may be considered in
the sickest patients.24
Combined nutritional support, which is consid-
ered an optional tool to avoid energy deficiency
during prolonged hypocaloric EN, was indeed
efficient in overcoming the energy deficits. Par-
enteral supplementation was only initiated when
the energy deficit with EN was persistent (45
days), and obvious (430% of target, or cumulated
deficit 48000 kcal). Altogether, there were only 81
days of combined nutrition, which reflects the
strong preference for enteral support in our ICU,
and the reluctance to administer PN. Indeed, the
place of PN remains controversial.25 If combined
enteral and parenteral feeding during the early
phase of the critical illness may improve energy
supply, there is yet no clear indication on its
nutritional value. Starting with combined EN and
PN support has not proven superior to EN alone in
patients without prior malnutrition.26 In a study
performed in a French medical ICU, 100 patients
were randomized to receive supplemental parent-
eral nutrition or not in addition to enteral
support27: there was no observable clinical benefit.
This later study deserves being repeated.
Limitations of the study: Our results show that
our feeding protocol was not respected particularly
in trauma patients: failure to early feed this
category of patients shows the difference between
protocols and real life. Although a median of 3 days
without nutritional support is acceptable both in
clinical terms, and considering actual guidelines,
it contributed to build up the early energy debt.
The study was designed as an observational study
aiming at detecting a relationship between energy
balance and occurrence of complications. Although
this hypothesis has been confirmed, the absence of
intervention does not definitively prove the causal
S. Villet et al.
relationship. The presence of a strong correlation
between the energy balance already at the end of
the first week and the number of complications
ðP ¼ 0:0049Þ, is an argument in favor of the
nutritional origin of complications. Another limita-
tion is that REE was measured in only 69% of the
patients: as a consequence, energy balance was
based on a 30 kcal/kg/day estimation of energy
expenditure in absence of indirect calorimetry,
which corresponds to a mean target value of
1.3  REE. Such approximation seems reasonable
in a surgical population. However, both the
between-patient and intra-patient daily variation
of REE may be substantial, as shown in trials
measuring 24-h REE by indirect calorimetry, or in
those using as doubly labeled water or other
isotopic methods to determine total energy ex-
penditure.14 This approximation thus corresponds
to clinical conditions.
Conclusion
The study confirms that negative energy balances
are very frequent during severe critical illness
despite nutrition protocols. It shows that under-
feeding is correlated with increasing number of
complications, and particularly with infections.
Analysis of timing shows that the energy debt is
initiated during the first week after admission, and
that delaying the initiation of nutritional support
exposes the patients to energy deficits that cannot
be compensated during the remaining ICU stay.
Combined nutritional support only moderately
reversed the building up deficiency: starting to
top up on EN as early as 4–5 days is probably
reasonable. No plasma protein marker was asso-
ciated with the growing deficit, and we therefore
’’
propose the energy debt’’ as the easiest marker of
increasing nutritional risk.
References
1. Jolliet P, Pichard C, Biolo G, Chiolero R, Grimble G, Leverve
X, Nitenberg G, Novak I, Planas M, Preiser JC, Roth E,
Schols AM, Wernermann J. Enteral nutrition in intensive
care patients: a practical approach. Clin Nutr 1999;
18:47–56.
2. Berger MM, Chioléro RL, Pannatier A, Cayeux MC, Tappy L.
A ten year survey of nutritional support in a surgical ICU:
1986-1995. Nutrition 1997;13:870–7.
3. Allison SP. The uses and limitations of nutritional support.
Clin Nutr 1992;11:319–30.
4. Beck AM, Balknäs UN, Fürst P, Hasunen K, Jones L, Keller U,
Melchior JC, Mikkelsen B, Schauder P, Sivonen L, Zinck O,
Oien H, Ovesen L. Food and nutritional care in hospitals:
 ARTICLE IN PRESS
Impact of hypocaloric feeding on ICU outcome
how to prevent undernutrition—report and guidelines from
the Council of Europe. Clin Nutr 2001;20:455–60.
5. Cahill GF, Owen OE. Starvation and survival. Trans Am Clin
Climatol Assoc 1968;79:13–20.
6. Committee Public Health, Committee of Experts on
Nutrition Food Safety, and Consumer Health. In: Council E,
editor. Food and nutritional care in hospitals: how to
prevent undernutrition—report and recommendations.
Paris, 2002.
7. Buzby GP, Mullen JL, Matthews DC, Hobbs CL, Fischer JE.
Prognostic nutritional index in gastrointestinal surgery. Am J
Surg 1980;139:160–7.
8. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. Educational
and Clinical Practice Committee European Society of
Parenteral and Enteral Nutrition. ESPEN guidelines for
nutrition screening 2002. Clin Nutr 2003;22:415–21.
9. Bartlett RH, Dechert RE, Mault JR, Ferguson SK, Kaiser AM,
Erlandson EE. Measurement of metabolism in multiple organ
failure. Surgery 1982;92:771–9.
10. Cunningham JJ, Hegarty MT, Meara PA, Burke JF. Measured
and predicted calorie requirements of adults during recov-
ery from severe burn trauma. Am J Clin Nutr 1989;
49:404–8.
11. McClave SA, Lowen CC, Kleber MJ, Nicholson JF, Jimmerson
SC, McConnel JW, Jung LY. Are patients fed appropriately
according to their caloric requirements? JPEN 1998;
22:375–81.
12. Weissman C, Kemper M, Askanazi J, Hyman AI, Kinney JM.
Resting metabolic rate in the critically ill patient: measured
versus predicted. Anesthesiology 1986;64:673–9.
13. Weissman C. The metabolic response to stress: an overview
and update. Anesthesiology 1990;73:308–27.
14. Plank LD, Hill GL. Energy balance in critical illness. Proc
Nutr Soc 2003;62:545–52.
15. Astrup A, Thorbek G, Lind J, Isaksson B. Prediction of 24-h
energy expenditure and its components from physical
characteristics and body composition in normal-weight
humans. Am J Clin Nutr 1990;52:777–83.
16. LeGall JR, Lemeshow S, Saulnier F. A new simplified acute
physiology score (SAPS II) based of a European/North
American multicenter study. JAMA 1993;270:2957–63.
509
17. Vincent JL, Moreno R, Takala J, Willatts S, DeMedonc- a A,
Bruining H, Reinhart CK, Suter PM, Thijs LG. The SOFA
(Sepsis-related Organ Failure Assessment) score to describe
organ dysfunction/failure. Intens Care Med 1996;22:
707–10.
18. ASPEN Board of Directors and the Clinical Guidelines Task
Force. Guidelines for the use of parenteral and enteral
nutrition in adult and pediatric patients. JPEN 2002;
26(1 Suppl):1SA–138SA.
19. Chioléro R, Revelly JP, Tappy L. Energy metabolism in sepsis
and injury. Nutrition 1997;13(Suppl):45S–51S.
20. Bone RC, Balk RA, Cerra FB, Fein AM, Knaus WA, Schein RMH,
Sibbald WJ. Definition for sepsis and organ failure and
guidelines for the use of innovative therapies in sepsis.
Chest 1992;101:1644–55.
21. Rubinson L, Diette GB, Song X, Brower RG, Krishnan JA. Low
caloric intake is associated with nosocomial bloodstream
infections in patients in the medical intensive care unit. Crit
Care Med 2004;32:350–7.
22. Barr J, Hecht M, Flavin KE, Khorana A, Gould MK. Outcomes
in critically ill patients before and after the implementation
of an evidence-based nutritional management protocol.
Chest 2004;125:1446–57.
23. Lopez-Hellin J, Baena-Fustegueras JA, Schwartz-Riera S,
Garcia-Arumi E. Usefulness of short-lived proteins as
nutritional indicators in surgical patients. Clin Nutr
2002;21:119–25.
24. van den Berghe G, Wouters PJ, Bouillon R, Weekers F,
Verwaest C, Schetz M, Vlasselaers D, Ferdinande P, Lauwers
P. Outcome benefit of intensive insulin therapy in the
critically ill: Insulin dose versus glycemic control. Crit Care
Med 2003;31:359–66.
25. Jeejeebhoy KN. Total parenteral nutrition: potion or poison?
Am J Clin Nutr 2001;74:160–3.
26. Dhaliwal R, Jurewitsch B, Harrietha D, Heyland DK.
Combination enteral and parenteral nutrition in critically
ill patients: harmful or beneficial? A systematic review of
the evidence. Intens Care Med 2004;30:1666–71.
27. Bauer P, Charpentier C, Bouchet C, Nace L, Raffy F,
Gaconnet N. Parenteral with enteral nutrition in the
critically ill. Intens Care Med 2000;26:893–900.