Journal of Perinatology (2006) 26, 730736

r 2006 Nature Publishing Group All rights reserved. 0743-8346/06 $30

www.nature.com/jp

ORIGINAL ARTICLE
Hyperglycemia and morbidity and mortality in extremely
low birth weight infants
LS Kao1, BH Morris2, KP Lally3, CD Stewart4, V Huseby5 and KA Kennedy2
1
Department of Surgery, Lyndon Baines Johnson General Hospital, University of Texas, Houston Medical School, Houston, TX, USA;
2
Department of Pediatrics, Division of Neonatal-Perinatal Medicine, University of Texas, Houston Medical School, Houston, TX, USA;
3
Department of Surgery, Division of Pediatric Surgery, University of Texas, Houston Medical School, Houston, TX, USA; 4Dauterive
Hospital and Iberia Medical Center, New Iberia, LA, USA and 5Department of Pediatrics, University of Texas, Houston Medical School,
Houston, TX, USA

Keywords: hyperglycemia; infant, newborn; infant, premature; infant,

Objective: The purpose of this study was to determine the association very low birth weight; sepsis, mortality
between hyperglycemia and mortality and late-onset infections (>72 h) in
extremely low birth weight (ELBW) infants.

Study design: Retrospective analysis of a prospective cohort study of 201

ELBW infants who survived greater than 3 days after birth. Mean morning
glucose levels were categorized as normoglycemia (<120 mg/dl), mild-
moderate hyperglycemia (120 to 179 mg/dl) and severe hyperglycemia
(X180 mg/dl). Hyperglycemia was further divided into early (first 3 days
of age) and persistent (first week of age). Logistic regression was
performed to assess whether hyperglycemia was associated with either
mortality or late-onset culture-proven infection, measured after 3 and 7
days of age.
Results: Adjusting for age, the odds ratio (OR) for either dying or
developing a late infection was 5.07 (95% confidence interval (CI): 1.06 to
24.3) for infants with early severe hyperglycemia and 6.26 (95% CI: 0.73
to 54.0) for infants with persistent severe hyperglycemia. Adjusting for
age, both severe early and persistent hyperglycemia were associated with
increased mortality. Among survivors, there was no significant association
between hyperglycemia and length of mechanical ventilation or length of
hospital stay. Persistent severe hyperglycemia was associated with the
development of Stage II/III necrotizing enterocolitis, after adjusting for
age and male gender (OR: 9.49, 95% CI: 1.52 to 59.3).
Conclusion: Severe hyperglycemia in the first few days after birth is
associated with increased odds of death and sepsis in ELBW infants.
Journal of Perinatology (2006) 26, 730736. doi:10.1038/sj.jp.7211593;
published online 24 August 2006
Correspondence: Dr LS Kao, Department of Surgery, Lyndon Baines Johnson General
Hospital, University of Texas, Houston Medical School, 5656 Kelley Street, Suite 30S 62008,
Houston, TX 77026, USA.
E-mail: Lillian.S.Kao@uth.tmc.edu
Received 31 March 2006; revised 17 July 2006; accepted 2 August 2006; published online 24
August 2006
Introduction
Hyperglycemia in premature infants has been associated with
increased morbidity including retinopathy of prematurity1,2 and
intra-ventricular hemorrhage.3,4 In infants with necrotizing
enterocolitis (NEC), hyperglycemia has also been associated with
increased late mortality.5 However, studies examining the impact of
hyperglycemia on infectious morbidity and mortality in low birth
weight infants are sparse.
In adults, there is increasing evidence that even mild or
moderate hyperglycemia is associated with increased morbidity
and mortality. Van den Berghe et al.6 demonstrated that a strategy
of strict glycemic control with continuous insulin infusion to
maintain glucose levels less than or equal to 110 mg/dl decreased
mortality in patients requiring critical care for more than 5 days,
particularly in patients with multiple organ failure owing to a
septic focus. Other large observational cohort studies have
suggested that mild or moderate hyperglycemia in traumatically
injured (>135 mg/dl)7 and cardiac surgery (>150 mg/dl)8
patients also results in worse outcome.
In very low birth weight (VLBW) and extremely low birth weight
(ELBW) infants, there is a lack of consensus regarding the
threshold level defining hyperglycemia and the need for treatment
with insulin. Definitions have included but are not limited to blood
glucose associated with X0.5% glycosuria,3 >9.9 mmol/l
(178 mg/dl) with glycosuria,9 >8 mmol/l (144 mg/dl),5 a single
blood sugar >13.3 mmol/l (239 mg/dl) or repeated blood sugars
>8.8 mmol/l (158 mg/dl) for at least 4 h,10 and two consecutive
blood sugars >12 mmol/l (>216 mg/dl) for at least 4 h.11 The
incidence of hyperglycemia in VLBW and ELBW infants also varies
based on the definition used and ranges from 40 to 80%.12,13

Although continuous insulin infusion has been utilized safely
for managing hyperglycemia in ELBW infants,4,10,11,14 the optimal
glucose target has not yet been determined. Further information
from observational studies regarding the association between
hyperglycemia and outcome in ELBW infants is necessary before
proceeding to a potential trial of conventional versus more
aggressive glucose control in these infants. We therefore conducted
an observational study to evaluate the association between
hyperglycemia and mortality and late-onset infections in ELBW
infants using data prospectively collected for a randomized trial
(Morris, unpublished data).
Materials and methods
Data were prospectively collected on 213 ELBW infants enrolled at
two institutions as part of a randomized trial of sterile water gavage
drip versus routine fluid administration in ELBW infants. This
analysis was performed as a retrospective observational study,
undertaken as a subset of the larger completed prospective clinical
trial (Morris et al., personal communication). The original
prospective study was conducted at two centers, Memorial Hermann
Childrens Hospital and Lyndon Baines Johnson (LBJ) General
Hospital in Houston, TX, USA; randomization was stratified by
center. Infants were eligible for the trial if they met the following
criteria: (1) <31 weeks gestation by Ballard examination, (2) birth
weights X400 and p1000 grams and (3) enrollment before 36 h
of age. Infants were excluded if they had (1) severe asphyxia
(defined as an initial infant pH<7.1 and evidence of end-organ
damage), (2) major congenital anomalies, (3) hypotension treated
with dobutamine or dopamine at greater than 15 mg/kg/minute or
(4) anticipated death (pH<6.80 or hypoxia with bradycardia for
>2 h). These infants were excluded given their poor prognosis.
Infants were enrolled between March 2000 and November 2003.
There were 180 infants who were eligible but not enrolled owing to
the following reasons: parental refusal (54%), consent not
requested (33%) and other (12%). A total of 12 infants were
excluded from the analysis 11 infants were enrolled in the trial
but did not survive more than 3 days of age and one infant who
transferred at 3 days of age did not have available outcome
data. Data collection was approved by the Institutional Review
Board at University of Texas Health Science Center, and informed
consent was obtained from the parents before participation
in the study.
The main predictor variables of interest were mean morning
glucose levels over the first 3 and 7 days after birth. In order to
ensure consistent sampling among subjects, only morning glucose
levels were analyzed. Inclusion of all serum laboratory values could
increase the potential for confounding by severity of illness as
sicker infants have their laboratory tests drawn more frequently,
which increases the likelihood of identifying extreme values.
Therefore, serum electrolytes and serum glucose levels were
Hyperglycemia in ELBW infants
LS Kao et al
731
measured daily after enrollment at approximately the same time
each day (around 0400 hours). The mean glucose levels were
divided into the following categories: normoglycemia (<120 mg/
dl), mild-moderate hyperglycemia (120 to 179 mg/dl) and severe
hyperglycemia (X180 mg/dl).
Confounding variables analyzed included: gestational age,
gender, race, birth weight, use of antenatal steroids and Apgar
scores at 1 and 5 min. Gestational age was estimated using the
modified Ballard test because of the unreliability of using dates
given that many of the women at LBJ General Hospital had no
prenatal care. Outcome measures included: development of
systemic infection (positive blood culture at >3 and 7 days of life
treated with >3 days antibiotics), NEC (XBells Stage II), length
of stay and death before discharge. Infections that occurred before
72 h of age were excluded so as to exclude early infection,
which may be more reflective of a congenital infection15 and
would be unlikely to be prevented by improved glycemic control
after birth.
The primary outcome measure was a composite outcome of
either mortality or late-onset culture-proven infection after 3 days
and after 1 week of age. As there is no consensus on the cutoff for
defining late-onset infection and given that there is no prior data
regarding duration of hyperglycemia and outcome in ELBW
infants, outcome at both time points was analyzed in this study.
Secondary outcome measures were incidence of hypoglycemia
(<40 mg/dl), development of NEC, total number of days of
mechanical ventilation among survivors and length of stay for
survivors not transferred to another facility.
Descriptive statistics were utilized to describe the demographic
data. Univariate analyses were performed to assess the association
between risk factors, including hyperglycemia, and mortality, late
culture-proven infection and length of stay. For the primary
outcome measure, w2 or Fishers exact test was utilized for
categorical independent variables, and the MannWhitney U test
or univariate logistic regression was used for continuous
independent variables. For length of mechanical ventilation and
length of stay, linear regression was used. Multivariate logistic and
linear regression analyses were performed for categorical and
continuous outcome measures, respectively. Factors that had a
P-value of less than or equal to 0.25 and known risk factors based
on the literature were included in the initial multivariate analyses.
Because birth weight and age by Ballard exam were correlated, the
variable with the strongest relationship with the outcome was
included where applicable. All of the final reduced models included
mean glucose categories as the independent variable of interest.
Hierarchical forward selection with switching, with a maximum of
four variables, was used to select the other independent variables
included in the reduced models. Significance in the multivariate
model was based on the Wald and deviance tests, Pp0.05. All
statistical analyses were performed using Number Cruncher
Statistical Systems (NCSS, 2004, Kaysville, UT, USA).
Journal of Perinatology
 Hyperglycemia in ELBW infants
LS Kao et al
732

Table 1 Baseline characteristics of 201 ELBW infants who survived Normoglycemic infants (mean glucose
greater than 3 days of life < 120 mg/dL)

Mean Glucose (mg/dL)

500
Characteristic 400

Postnatal age at enrollment, hourss.d. 19.39.5 300

Birth weight, gs.d. 729127 200
Gestational age, weeks.d. (Ballard exam) 26.21.9 100
0
Race or ethnic group, n (%) 1 2 3 4 5 6
African-American 95 (47%) Day
Caucasian 33 (16%)
Mildly hyperglycemic infants (mean
Hispanic 64 (32%)
glucose 120-179 mg/dL)

Mean Glucose (mg/dL)

Other 9 (4%) 500
400
Male, n (%) 88 (44%)
300
Antenatal steroids, n (%) 115 (57%)
200
Apgar score p3, n (%) 100
At 1 min 85 (42%) 0
At 5 min 21 (10%) 1 2 3 4 5 6
Day
Abbreviations: ELBW, extremely low birth weight; s.d., standard deviation.

Severely hyperglycemic infants (mean

glucose
180 mg/dL)
Results
Mean Glucose (mg/dL)

500
Demographic data
400
The baseline characteristics of the 201 ELBW infants included in
300
the study are listed in Table 1. The majority of infants had at least
two morning glucose measurements for the first 3 days of age 200

(193/201, 96%), but eight infants whose initial serum chemistries
were not drawn until after 52 h of age only had one blood glucose
measurement. The median value for mean morning glucose level
for the first 3 days after birth was 105 mg/dl (interquartile range
(IQR): 81 to 133 mg/dl) and the average was 11453 mg/dl. Sixty-
five percent of the infants had an average glucose level less than
120 mg/dl, 28% in the mild-moderate range and 7% in the severe
range. The median value for mean morning glucose level for the
first week after birth was 113 mg/dl (IQR, 97 to 132 mg/dl).
Figure 1 illustrates the trend for the mean daily glucose level
over time for each category of infants (normoglycemic, mildly
hyperglycemic and severely hyperglycemic infants). Sixty-one
percent of infants had a mean glucose in the normal range for the
first week, whereas 35 and 4% had averages that were in the mild-
moderate and severe groups, respectively. Overall, 77 (38%) infants
had at least one glucose measurement of 180 mg/dl or higher and
10 infants (5%) had at least one episode of hypoglycemia.
The median value for mean morning glucose for the first 3 days
after birth was 104 mg/dl (93 to 110 mg/dl) for infants randomized
to the control or routine fluids group (n 96) and 106 mg/dl
(97 to 118 mg/dl) for infants in the sterile water gavage group
(n 109). This difference was not statistically significant using
the MannWhitney test (P 0.24). However, the median glucose
level for the first week after birth was significantly higher in the
100
0
1 2 3 4 5 6
Day
Figure 1 The daily mean morning glucose level for the infants in
each category stayed relatively stable within the normoglycemic and
mildly hyperglycemic groups. The daily mean morning glucose level
trended downwards in the severely hyperglycemic group, but remained
above high (>180 mg/dl) even after 1 week.
control (n 94) than in the sterile water gavage group (n 104)
118 mg/dl (113 to 127 mg/dl) versus 108 mg/dl (103 to 114 mg/
dl), respectively (P 0.03). There was no statistically significant
difference in distribution of mean glucose levels at either 3
or 7 days across the strata of mild, moderate and severe
hyperglycemia based on group randomization.
Primary outcome
Overall, 100/201 (50%) infants experienced the primary outcome
of either mortality or a late-onset infection after 3 days of age.
On univariate analysis, there was a greater likelihood of having
a worse outcome in infants with severe hyperglycemia when
compared to normoglycemic infants with a mean glucose level less
than 120 mg/dl (Table 2), but there was no statistically significant

Journal of Perinatology
 Hyperglycemia in ELBW infants
LS Kao et al
733

Table 2 Multivariate analysis of risk factors for composite outcome of mortality or late culture-proven infection
Factor Adjusted OR 95% CI P-value Unadjusted OR 95% CI P-value

Outcome assessed after 3 days of age

Gestational age, per week 0.77 0.650.91 0.002 0.74 0.630.87 <0.001
Mean glucose (1st 3 days after birth) 120179 mg/dl 1.12 0.592.13 0.73 1.25 0.672.33 0.49
Mean glucose (1st 3 days after birth) X180 mg/dl 5.07 1.0624.3 0.04 7.22 1.5533.6 0.01

Outcome assessed after 7 days of age

Gestational age, per week 0.79 0.670.94 0.008 0.76 0.650.89 <0.001
Mean glucose (1st 7 days after birth) 120179 mg/dl 1.34 0.722.51 0.35 1.66 0.923.01 0.09
Mean glucose (1st 7 days after birth) X180 mg/dl 6.26 0.7354.0 0.10 9.80 1.1782.2 0.034
Abbreviations: CI, confidence interval; OR, odds ratio.
Referent category for glucose categories is mean glucose level <120 mg/dl.

Table 3 Multivariate analysis of risk factors for mortality

Factor Adjusted OR 95% CI P-value Unadjusted OR 95% CI P-value

Outcome assessed after 3 days of age

Gestational age, per week 0.66 0.510.084 <0.001 0.61 0.480.87 <0.001
Mean glucose (1st 3 days after birth) 120179 mg/dl 1.26 0.532.98 0.60 1.49 0.653.39 0.35
Mean glucose (1st 3 days after birth) X180 mg/dl 15.7 3.7465.9 <0.001 22.8 5.8089.8 <0.001

Outcome assessed after 7 days of age

Gestational age, per week 0.69 0.540.89 0.005 0.63 0.490.80 <0.001
Mean glucose (1st 7 days after birth) 120179 mg/dl 1.67 0.733.80 0.22 2.24 1.024.94 0.04
Mean glucose (1st 7 days after birth) X180 mg/dl 30.4 3.37274 0.002 53.0 6.06463 <0.001
Abbreviations: CI, confidence interval; OR, odds ratio.
Referent category for glucose categories is mean glucose level <120 mg/dl.

association between mild-moderate hyperglycemia and outcome.
Male gender, antenatal steroids, African-American race and sterile
water gavage were not significant on univariate analysis. Birth
weight was associated with worse outcome but because of the
correlation with gestational age, the better predictor of the two was
used in the final model. After adjustment for age, severe
hyperglycemia was still associated with worsened outcome
(Table 2). For mild-moderate hyperglycemia, there was not a
significant increase in odds of either having an infection or dying.
The r2 for the model including hyperglycemia as a categorical
value and age was 0.59.
One hundred and ninety-eight infants survived greater than
1 week, and of those infants, 95 (48%) either died or developed
a subsequent infection. Using univariate logistic regression, the
odds of developing the composite outcome if the mean glucose
level was greater than or equal to 180 mg/dl as compared to less
than 120 mg/dl was increased significantly (Table 2). On
univariate analysis, only age and birth weight were associated with
the primary outcome. Adjusting for age, neither mild-moderate
hyperglycemia nor severe hyperglycemia was positively associated
with the outcome. The model r2 was 0.51.
We used the composite outcome as our primary outcome
because death is a competing outcome for infection, but we also
analyzed each of the components of the primary outcome
separately. The overall mortality of the series was 20% (40/201). On
univariate analysis, severe hyperglycemia in the first 3 days of age
was significantly associated with death, but not mild-moderate
hyperglycemia (Table 3). On univariate analysis, age and birth
weight were the only other factors associated with increased
mortality. Adjusting for age (model r2 0.71), severe hyperglycemia
was still significantly associated with mortality (Table 3). Of the
198 infants surviving more than 7 days, 37 (19%) subsequently
died. Using univariate logistic regression analysis, the odds of dying
if the infant had mild-moderate hyperglycemia during the first
week after birth as compared to normoglycemia was increased,
as was the odds ratio (OR) for dying if the infant had severe
hyperglycemia. After adjusting for age (model r2 0.69), the odds
of dying were still significantly increased for infants with severe
hyperglycemia over the first week after birth, but not for infants
with mild-moderate hyperglycemia.
Seventy-six infants developed a culture-proven infection after
72 h of life, and 74 infants developed an infection greater than

Journal of Perinatology
 Hyperglycemia in ELBW infants
LS Kao et al
734

Table 4 Multivariate analysis of risk factors for sepsis

Factor Adjusted OR 95% CI P-value Unadjusted OR 95% CI P-value

Outcome assessed after 3 days of age

Gestational age, per week 0.84 0.711.00 0.04 0.85 0.731.00 0.05
African-American race 0.52 0.290.93 0.03 0.53 0.300.95 0.03
Mean glucose (1st 3 days after birth) 120179 mg/dl 0.96 0.491.85 0.90 1.04 0.551.97 0.91
Mean glucose (1st 3 days after birth) X180 mg/dl 0.92 0.282.97 0.89 1.24 0.413.79 0.71

Outcome assessed after 7 days of age

Gestational age, per week 0.85 0.721.01 0.06 0.85 0.731.00 0.05
African-American race 0.52 0.290.95 0.03 0.53 0.300.97 0.05
Mean glucose (1st 7 days after birth) 120179 mg/dl 0.94 0.491.78 0.84 1.19 0.652.19 0.57
Mean glucose (1st 7 days after birth) X180 mg/dl 0.56 0.122.53 0.45 1.07 0.244.72 0.92
Abbreviations: CI, confidence interval; OR, odds ratio.
Referent category for glucose categories is mean glucose level <120 mg/dl.

1 week after birth. Neither mean glucose level in the first 3 days
of life nor in the first week was predictive of infection, both on
univariate analysis and adjusting for African-American race and
age (Table 4).
Secondary outcomes
Eighteen infants who survived greater than 1 week developed Stage
II or III NEC. There was a significant association between a mean
7-day glucose level greater than or equal to 180 mg/dl as compared
to the referent category of a mean less than 120 mg/dl and
development of Stage II/ III NEC (OR: 7.40, 95% confidence
interval (CI): 1.52 to 36.1). After adjusting for age and male
gender, the association was still significant (OR: 9.49, 95% CI: 1.52
to 59.3). The r2 for the model including age, gender and
hyperglycemia was 0.31. For infants who subsequently developed
NEC, severe persistent hyperglycemia in the first week after birth
was associated with increased mortality (OR: 13.1, 95% CI: 1.20 to
143) but the sample size was very small.
Among survivors, there was no correlation between mean
glucose category and number of days of mechanical ventilation
or length of stay in the hospital on univariate or multivariate
analyses. In infants surviving longer than 1 week, the main
determinants of length of stay (accounting for 45% of the
variability) were a diagnosis of chronic lung disease, birth weight
and male gender.
Discussion
This study demonstrates a relationship between increased mean
glucose levels greater than or equal to 180 mg/dl over the first 3
days of life and worsened outcome in ELBW infants. Furthermore,
both early and persistent hyperglycemia over the course of the first
week of life was associated with increased odds of dying, although
the CIs were wide. These results are consistent with studies in other
pediatric16,17 and adult6,1820 critically ill populations.
In hospitalized adults, the definition of stress hyperglycemia has
been challenged over the past several years based on at least one
large trial demonstrating a survival benefit to maintaining a
glucose target of less than 110 mg/dl in the intensive care unit
(ICU).6 However, there is a paucity of literature regarding the
impact of mild or moderate hyperglycemia on outcome in
pediatric populations. In a retrospective review of 932 patients,
Faustino and Apkon16 reported a significantly increased risk of
death in non-diabetic children with elevated glucose levels above
150 mg/dl, in the first 24 h of admission (relative risk (RR): 2.50,
95% CI: 1.26 to 4.93). For children with maximum glucose levels
of 120 mg/l, there may also have been an increased risk of
mortality (RR: 2.10, 95% CI: 0.89 to 5.00). Srinivasan et al.17
reported an association between peak blood glucose level (OR: 1.01,
95% CI: 1.00 to 1.02 for each mg/dl increase) and duration of
hyperglycemia defined as percentage of days with blood glucose
greater than 126 mg/dl (OR: 1.03, 95% CI: 1.01 to 1.05) with
mortality in a cohort of 152 pediatric ICU patients. The results
from this study in ELBW infants did not demonstrate a statistically
significant increase in the risk of dying with mild-moderate
hyperglycemia after birth. However, the small number of infants in
this category who had the outcomes of interest may have reduced
the power of the study to detect a difference. Likewise, further
stratification of mean glucose levels, particularly between 120 and
180 mg/dl, was also limited by sample size considerations.
Although the results of this study did not demonstrate an
association between early or persistent hyperglycemia and
culture-proven infection, one mechanism by which strict glycemic
control has been postulated to impact outcome is by decreasing
the risk of septic complications. Hyperglycemia has been previously
linked in animal and human studies to alterations in the immune
response owing to modulation of both pro- and anti-inflammatory

Journal of Perinatology

cytokines,2125 impairment of the neutrophil response26 and
depression of cell-mediated immunity.27 Clinically, elevated glucose
levels have been associated with increased risk of infectious
complications, particularly after surgery,8,28 and improved glycemic
control has been associated with a reduction in surgical site8 and
bloodstream infections.6 In our study, there did not appear to be a
correlation between hyperglycemia and development of a culture-
proven infection after either 3 or 7 days among survivors. Owing to
the small and unbalanced sample size, the study may have been
underpowered to detect a small difference in the rates of infection.
For example, there was only a 5% difference in infection rates after
3 days between the infants with early severe hyperglycemia and
normoglycemic infants (43% or 49/130 versus 38% or 6/14).
This study demonstrated an association between severe persistent
hyperglycemia over the first week of life and Stage II/III NEC. In a
retrospective study of premature infants with NEC, 69% of infants
(n 95) were hyperglycemic above 144 mg/dl (8 mmol/dl) at
some stage during their admission. Their mean birth weight was
1310 g and mean gestational age was 29 weeks. In this population,
a maximum blood glucose level above 215 mg/dl (11.9 mmol/dl)
during an ICU admission was associated with an increased risk of
death (33 versus 18%, P 0.13), particularly in the subgroup of
neonates who died after 10 days in the ICU (29 versus 2%,
P 0.0009).5 In our study, 75% of ELBW infants who subsequently
developed NEC were hyperglycemic above 144 mg/dl within the first
week after birth. Their mean birth weight was 713 g and mean
gestational age was 26 weeks. Data were not collected regarding
their glucose levels for the entire length of ICU stay. Nonetheless,
severe persistent hyperglycemia over the first week after birth was
associated with increased mortality. One limitation however of
retrospective cohort studies is that causal relationships cannot be
defined, but one possibility is that the more severely ill infants had
a higher likelihood of both developing NEC and becoming
hyperglycemic.
This study is limited by all of the inherent biases of a
retrospective analysis. Although the data were originally collected
for a randomized trial of sterile water gavage versus routine fluid
administration in ELBW infants, the group assignment did not
have any effect on either univariate or multivariate analysis on
outcome. Furthermore, although the control group had a higher
mean morning glucose level than the gavage group, the
classification of infants by strata was not significantly different.
Lastly, one advantage of using these data is the rigor in planning
and data collection as a result of the randomized study design.
Another significant limitation of this study is the potential
identification of a statistically significant correlation between
hyperglycemia and outcome in the absence of a true causal
relationship. Furthermore, even if a true correlation exists, as
would be supported by prospective studies in other populations, this
study does not elucidate the nature of the relationship that is,
whether hyperglycemia contributes to increased mortality or
Hyperglycemia in ELBW infants
LS Kao et al
735
whether infants who are more severely ill are more likely to be
hyperglycemic. Additionally, because of the retrospective nature
of the study, we were unable to adjust for all of the potential
confounders such as administration of total parenteral nutrition,
insulin administration and glucose load, which were not
standardized among neonatologists. However, the results of this
study are in concordance with those demonstrating worse outcome
with hyperglycemia in other patient populations and at the very
least warrant further investigation.
In conclusion, severe hyperglycemia is associated with worsened
outcome in ELBW infants. Further studies to elucidate the specific
relationship between hyperglycemia, particularly in the mild-
moderate range and outcome and to evaluate the effects of
improved glycemic control in this patient population are necessary.
Acknowledgments
We acknowledge Dr Fernando Moya, Director of Neonatology at New Hanover
Regional Medical Center and Professor of Pediatrics at University of North
Carolina, for his valuable comments and suggestions on the study.
References
1 Ertl T, Gyarmati J, Gaal V, Szabo I. Relationship between hyperglycemia and
retinopathy of prematurity in very low birth weight infants. Biol Neonate
2005; 89(1): 5659.
2 Garg R, Agthe AG, Donohue PK, Lehmann CU. Hyperglycemia and
retinopathy of prematurity in very low birth weight infants. J Perinatol 2003;
23(3): 186194.
3 Binder ND, Raschko PK, Benda GI, Reynolds JW. Insulin infusion with
parenteral nutrition in extremely low birth weight infants with
hyperglycemia. J Pediatr 1989; 114(2): 273280.
4 Mena P, Llanos A, Uauy R. Insulin homeostasis in the extremely low birth
weight infant. Semin Perinatol 2001; 25(6): 436446.
5 Hall NJ, Peters M, Eaton S, Pierro A. Hyperglycemia is associated with
increased morbidity and mortality rates in neonates with necrotizing
enterocolitis. J Pediatr Surg 2004; 39(6): 898901; discussion 898901.
6 van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz
M et al. Intensive insulin therapy in the critically ill patients. N Engl J Med
2001; 345(19): 13591367.
7 Yendamuri S, Fulda GJ, Tinkoff GH. Admission hyperglycemia as a
prognostic indicator in trauma. J Trauma 2003; 55(1): 3338.
8 Furnary AP, Wu Y, Bookin SO. Effect of hyperglycemia and continuous
intravenous insulin infusions on outcomes of cardiac surgical procedures:
the Portland Diabetic Project. Endocr Pract 2004; 10(Suppl 2): 2133.
9 Collins Jr JW, Hoppe M, Brown K, Edidin DV, Padbury J, Ogata ES. A
controlled trial of insulin infusion and parenteral nutrition in extremely low
birth weight infants with glucose intolerance. J Pediatr 1991; 118(6):
921927.
10 Meetze W, Bowsher R, Compton J, Moorehead H. Hyperglycemia in
extremely-low-birth-weight infants. Biol Neonate 1998; 74(3): 214221.
Journal of Perinatology
 Hyperglycemia in ELBW infants
LS Kao et al
736

11 Ng SM, May JE, Emmerson AJ. Continuous insulin infusion in
hyperglycaemic extremely-low-birth-weight neonates. Biol Neonate 2005;
87(4): 269272.
12 Dweck HS, Cassady G. Glucose intolerance in infants of very low birth
weight. I. Incidence of hyperglycemia in infants of birth weights 1100 grams
or less. Pediatrics 1974; 53(2): 189195.
13 Louik C, Mitchell AA, Epstein MF, Shapiro S. Risk factors for neonatal
hyperglycemia associated with 10% dextrose infusion. Am J Dis Child 1985;
139(8): 783786.
14 Vaucher YE, Walson PD, Morrow III G. Continuous insulin infusion in
hyperglycemic, very low birth weight infants. J Pediatr Gastroenterol Nutr
1982; 1(2): 211217.
15 Hodgman JE, Barton L, Pavlova Z, Fassett MJ. Infection as a cause of death
in the extremely-low-birth-weight infant. J Matern Fetal Neonatal Med
2003; 14(5): 313317.
16 Faustino EV, Apkon M. Persistent hyperglycemia in critically ill children.
J Pediatr 2005; 146(1): 3034.
17 Srinivasan V, Spinella PC, Drott HR, Roth CL, Helfaer MA, Nadkarni V.
Association of timing, duration, and intensity of hyperglycemia with
intensive care unit mortality in critically ill children. Pediatr Crit Care Med
2004; 5(4): 329336.
18 Gore DC, Chinkes D, Heggers J, Herndon DN, Wolf SE, Desai M. Association
of hyperglycemia with increased mortality after severe burn injury.
J Trauma 2001; 51(3): 540544.
19 Krinsley JS. Association between hyperglycemia and increased hospital
mortality in a heterogeneous population of critically ill patients. Mayo Clin
Proc 2003; 78(12): 14711478.
20 Stranders I, Diamant M, van Gelder RE, Spruijt HJ, Twisk JW, Heine RJ et al.
Admission blood glucose level as risk indicator of death after myocardial
infarction in patients with and without diabetes mellitus. Arch Intern Med
2004; 164(9): 982988.
21 Weekers F, Giulietti AP, Michalaki M, Coopmans W, Van Herck E, Mathieu C
et al. Metabolic, endocrine, and immune effects of stress hyperglycemia in a
rabbit model of prolonged critical illness. Endocrinology 2003; 144(12):
53295338.
22 Esposito K, Nappo F, Marfella R, Giugliano G, Giugliano F, Ciotola M et al.
Inflammatory cytokine concentrations are acutely increased by hyperglyce-
mia in humans: role of oxidative stress. Circulation 2002; 106(16):
20672072.
23 Krogh-Madsen R, Moller K, Dela F, Kronborg G, Jauffred S, Pedersen BK.
Effect of hyperglycemia and hyperinsulinemia on the response of IL-6,
TNF-alpha, and FFAs to low-dose endotoxemia in humans. Am J Physiol
Endocrinol Metab 2004; 286(5): E76672.
24 Losser MR, Bernard C, Beaudeux JL, Pison C, Payen D. Glucose modulates
hemodynamic, metabolic, and inflammatory responses to lipopolysaccharide
in rabbits. J Appl Physiol 1997; 83(5): 15661574.
25 Yu WK, Li WQ, Li N, Li JS. Influence of acute hyperglycemia in human sepsis
on inflammatory cytokine and counterregulatory hormone concentrations.
World J Gastroenterol 2003; 9(8): 18241827.
26 Rassias AJ, Givan AL, Marrin CA, Whalen K, Pahl J, Yeager MP. Insulin
increases neutrophil count and phagocytic capacity after cardiac surgery.
Anesth Analg 2002; 94(5): 11131119 table of contents.
27 Speert DP, Silva Jr J. Abnormalities of in vitro lymphocyte response to
mitogens in diabetic children during acute ketoacidosis. Am J Dis Child
1978; 132(10): 10141017.
28 Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic
control and the risk of infectious complications in a cohort of adults with
diabetes. Diabetes Care 1999; 22(9): 14081414.

Journal of Perinatology