- Email: [email protected]
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants Adi Auerbach, MD1,*, Smadar Eventov-Friedman, MD, PhD2,*, Ilan Arad, MD2, Ofra Peleg, MD2, Tali Bdolah-Abram, MSc3, Benjamin Bar-Oz, MD2, and David Haim Zangen, MD1 Objective To assess the association between severe intraventricular hemorrhage (IVH) and blood glucose variables during the first 96 hours of life in preterm infants.
Study design Preterm infants with IVH grade 3-4 (n = 70) were compared with matched infants of similar gestational age and birth weight, but with no IVH (n = 108). Studied variables included the frequency and duration of hyper/hypoglycemic (>6.9/<3.3 mmol/L, respectively) events, the extreme slope of an event evolution, the maximal glucose value observed, and the “hyper/hypoglycemic index” representing a weighted average of the hyper/hypoglycemic amplitude. Results The IVH group had significantly more hyperglycemic events (2.9 1.7 vs 2.4 1.8 events, P < .05) with longer duration (22.2 14.2 vs 14.1 12.5 hours, P < .001) and a higher hyperglycemic index (1.00.9 vs 1.41.0, P = .003) compared with the non-IVH controls. Respiratory distress syndrome, hypotension, and thrombocytopenia increased the adjusted OR for IVH. Hypoglycemia was not independently associated with IVH. Conversely, the increase in hyperglycemic duration was most prominently increasing the aOR for severe IVH (OR = 10.33, 95% CI = 10.0-10.6, P = .033). Conclusion Longer duration of hyperglycemia in the first 96 hours of life was most strongly associated with severe IVH in preterm infants. Consequently, interventional studies to determine the selective effect of continuous control of long-lasting hyperglycemia by appropriate and timed insulin treatment on the incidence of severe IVH are warranted. (J Pediatr 2013;163:388-93).
I
ntraventricular hemorrhage (IVH), occurring usually within the first 4 postnatal days, is a major cause of morbidity and mortality in preterm infants.1-3 Its pathophysiology involves bleeding from fragile blood vessels in the germinal matrix, which respond poorly to frequent changes in blood flow.1,4,5 Several risk factors, such as low gestational age and birth weight, perinatal stress, low Apgar score, and postnatal complications (eg, respiratory distress syndrome (RDS) or blood pressure instability),6 have been associated with IVH. Less studied is the association between the frequent alterations in blood glucose levels in preterm infants7-12 and the resultant changes in cerebral blood flow and osmolarity,7 which may challenge the fragility of blood vessels and contribute to the development of IVH. Hyperglycemia has lately been associated with increased mortality and morbidity rates, and insulin therapy and well-controlled blood glucose levels were found to be beneficial in adults and children in intensive care units.13-16 In extremely low birth weight infants (birth weight <1000 grams), high blood glucose concentration was associated with increased risk for IVH grade 3-4 and mortality.17 However, no causal relationship has been proven so far between the hyperglycemic amplitude and a detrimental outcome in interventional studies.18 Furthermore, recent attempts to treat all extremely low birth weight infants with continuous insulin infusion early in life (Neonatal Insulin Therapy in Europe Trial) failed to have a clinical benefit but, rather, increased the incidence of hypoglycemic episodes and mortality rate at 28 days.19 In this study, we aimed to identify which of the specific glucose homeostatic alterations (not necessarily the amplitude) may predict the occurrence of severe IVH (grade 3 or 4) and consequently be used to monitor an interventional protocol aiming to decrease the incidence of IVH.
Methods The data for this observational study were extracted from the Hadassah University Hospital’s neonatal intensive care unit (NICU) admission records of all neonates whose birth weights were below 2000 g during 5 consecutive years IVH NICU PDA RDS US
Intraventricular hemorrhage Neonatal intensive care unit Patent ductus arteriosus Respiratory distress syndrome Ultrasound
From the 1Division of Pediatric Endocrinology, Department of Pediatrics, 2Department of Neonatology, and 3School of Public Health, Hadassah Hebrew University Medical Center, Jerusalem, Israel *Contributed equally. This study was the MD thesis of A.A. at the Hadassah Hebrew University Medical School. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2013 Mosby Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.01.051
388
Vol. 163, No. 2 August 2013 (n = 907). The IVH study group included all neonates (n = 70) diagnosed with severe (grade 3 or 4) IVH by at least 2 cranial ultrasound (US) studies performed routinely within the first 3 days of life and subsequently between 5 and 10 days of life (subsequent US studies were performed as clinically indicated). IVH grading was determined independently by a radiologist and at least 2 neonatologists according to the Papile criteria.3 Grade 3 IVH was characterized by germinal matrix hemorrhage with ventricular dilatation (clot fills more than 50% of the ventricles), and in grade 4 IVH the hemorrhage was also intraparenchymal. The control non-IVH group (n = 108) consisted of neonates (1 or 2 according to availability) delivered immediately before or after a study group case with similar gestational age (10 days) and birth weight (150 g) but with no IVH (of any grade) by the routine US studies. One neonate from the IVH group and 5 from the control group were excluded from the study because of missing data. Based on previous studies examining risk factors for IVH in preterm infants,1,6 we studied the following background characteristics: gestational age, birth weight, multiple gestation, primiparity, etiology of early delivery, type of delivery, intrauterine growth retardation, pregnancy complications (preeclamsia, chrioamnionitis, bleeding) Apgar scores, first hematocrit, antenatal corticosteroids exposure, the presence of thrombocytopenia (platelets count <100 000), RDS, hypotension requiring therapy (dopamine or dobutamine), early onset sepsis (positive blood culture during the first 3 days of life), clinically significant patent ductus arteriosus (PDA), necrotizing enterocolitis, skin bruising, hyperbilirubinemia, and neonatal indomethacin treatment administered for closure of PDA. Glucose Homeostasis Calculations and Variables Soon after birth, continuous glucose infusion was administered to all preterm infants at a rate of 4-7 mg/kg/min to match basal requirements and to prevent hypoglycemia. Routine (NICU protocol) bedside glucometer monitoring was performed 8 times daily in all infants during the initial days of life. Higher blood glucose monitoring frequency (12-24 times daily) was performed when clinical stability or blood glucose levels were deranged. Hospital laboratory confirmation of glucometer readings was done at least once daily. Glucose values for the first 4 days of life were collected from the NICU standard records. Three infants in the control group missed 1-4 hours and 11 infants in the study group lacked 6-58 hours because of early death. Hyperglycemia and hypoglycemia were defined as glucose levels >6.9 mmol/L (>125 mg/dL) and <3.3 mmol/L (<60 mg/dL), respectively. Insulin was used whenever blood glucose levels were higher than 11.1 mmol/L (200 mg/dL) in 2-3 consecutive measurements. The glucose homeostasis variables studied for each neonate included: (1) the number of hyperglycemic and hypoglycemic events; (2) the duration of these events; (3) the percentage of time the neonates were hypoglycemic or hyperglycemic during the 4-day period or relative to the time of data collection for this infant (the “relative duration”); (4)
the maximal slope of an event evolution being the largest quotient calculated from the change in the glucose levels divided by the time of its evolution; (5) the maximal glucose value; and (6) the “hyper/hypoglycemic index,” which was calculated by multiplying the amplitude size of each abnormal glucose reading by its time duration. The cumulative sum of those products divided by total hyperglycemic or hypoglycemic duration per neonate gives a general index of glucose homeostasis representing glucose stability or instability for each infant. In order to correlate between the occurrence of hyperglycemic-associated IVH and the general illness severity of preterm infants, we created a semi-quantitative severity index score (0-5) based on death (5), days of ventilation (none = 0, up to 7 days = 1, more than 7 days = 2), bronchopulmonary dysplasia (O2 requirement at 36 weeks corrected gestational age, no = 0, yes = 1), need for dopamine treatment (no = 0, yes = 1), and day of enteral feeds commencement (up to 4 days = 0, more than 4 days = 1). Then we correlated the final severity score with the different hypo/hyperglycemic variables. Statistical Analyses Independent sample t test was used for comparison of quantitative variables between the IVH and control groups. The Pearson c2 test was applied for testing the association between the study groups and qualitative variables. The variables found to be significantly (P < .05) associated with the dependent variable (IVH or control) in the univariate analysis were entered in a stepwise manner into a multivariate logistic regression model. The simultaneous effect of different variables and their calculated aOR and 95% CI were then assessed. All tests applied were 2-tailed, and a P value of 5% was considered statistically significant. Kruskal–Wallis test (nonparametric ANOVA for small subgroups with abnormal distribution) was used for correlating the severity score to glycemic variables.
Results The birth weights of the IVH and control groups were similar by multivariate analysis (Table I). Lower gestational age, lower Apgar score at 5 minutes, and the occurrences of RDS, mechanical ventilation, pneumothorax, thrombocytopenia, hypotension, early sepsis, and bruising were found to be significantly associated with the occurrence of severe IVH. Antenatal maternal corticosteroid therapy and neonatal indomethacin treatment were more common in the control group and appeared to be protective against IVH. No statistically significant association with IVH was found for multiple gestation, primiparity, type of delivery, first hematocrit level or the presence of chorioamnionitits, preeclampsia, PDA, necrotizing enterocolitis, or hyperbilirubinemia. When the statistically significant (IVH associated) background variables (P < .05) were introduced into a multivariate analysis, only RDS, thrombocytopenia, and 389
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 163, No. 2
Table I. Background characteristics of the study population and the incidence of IVH (univariate and multivariate analysis) Univariate analysis
Multivariate analysis*
Variable
IVH group (n = 70)
Control group (n = 108)
P value
OR (95% CI)
P value
Birth weight (mean SD), g Gestational age (mean SD), wk Etiology of early delivery Labor, % (n) Premature rupture of membranes, % (n) Pregnancy complications†, % (n) Fetal distress, % (n) Type of delivery Vaginal delivery, % (n) Cesarean delivery, % (n) Antenatal steroid therapy, % (n) Small for gestational age, % (n) Apgar -5 min (mean SD), n RDS, % (n) Mechanical ventilation, % (n) Pneumothorax, % (n) Thrombocytopenia, % (n) PDA, % (n) Hypotension, % (n) Early sepsisz, % (n) Indomethacin treatment, % (n) Bruising, % (n)
1026 385 27.6 2.4
1126 339 28.4 2.2
.07 .01 .31
1.00 (1.0-1.0) 0.89 (0.6-1.3)
.38 .58
45.7 (32) 5.7 (4) 31.4 (22) 17.1 (12)
41.7 (45) 13.9 (15) 31.5 (34) 13 (14)
31.4 (22) 68.6 (48) 57.1 (40) 11.4 (8) 8.1 1.6 81.4 (57) 95.7 (67) 15.7 (11) 22.9 (16) 45.7 (32) 55.7 (39) 15.7 (11) 11.4 (8) 24.3 (17)
39.8 (43) 60.2 (65) 74.1 (80) 11.1 (12) 9.1 0.8 46.3 (50) 68.5 (74) 5.6 (6) 2.8 (3) 35.2 (38) 13.9 (15) 0 (0) 24.1 (26) 12 (13)
0.99 (0.4-2.3)
.99
.41 .02 .95 <.001 <.001 <.001 .02 <.001 .2 <.001 <.001 .03 .033
x
0.3 (0.7-15.6) 3.3 (1.5-7.3)
.12 .003
2.6 (0.7-8.7) 7.9 (1.9-33.3)
.12 .005
5.5 (2.5-12.2) NA{ 0.36 (0.1-0.9) 1.45 (0.5-4.1)
<.001 NA{ .047 .48
NA, not applicable. Statistically significant P values in the univariate analysis are in italics and statistically significant P values in the multivariate analysis are in bold. *The variables for which there was a significant difference in the univariate analysis between the IVH and the control groups were entered into a multivariate logistic regression analysis. †Pregnancy complications-preeclampsia, placental abruption, chorioamnionitis, bleeding. zEarly sepsis-sepsis in the first 3 days after delivery. xOR for Apgar-5 min >7. {The aOR received was infinite and not significant because there was no early sepsis in the control group.
hypotension requiring pressor supportive therapy remained significantly associated with severe IVH (Table I). Indomethacin treatment (administered for closure of PDA) maintained significantly protective association against IVH. Glucose Variables Analysis Derangements in blood glucose levels were common in both the IVH and the control group. During their first 4 days of life, 86% of the premature neonates had 1 and 29% had more than 4 hyperglycemic events. Hypoglycemia was less common as the comparable values for hypoglycemia were
75% and 8%, respectively. Moreover, the total duration of hypoglycemia (defined conservatively as <3.3 mmol/L-60 mg/dL) was much shorter compared with hyperglycemia (5.1 6.7 vs 17.3 13.7 hours, respectively). Univariate analysis revealed statistically significant differences in glucose homeostatic variables between the IVH and the control groups (Table II). The IVH group had more frequent and longer hyperglycemic events. Furthermore, 15.7% of the infants in the IVH group had abnormal glucose values for more than 50% of the study duration versus only 4.6% of the infants in the control group (P = .008).
Table II. Glucose homeostasis parameters and incidence of IVH (univariate and multivariate analysis) Univariate analysis
Multivariate analysis*
Variable (mean ± SD)
IVH group (n = 70)
Control group (n = 108)
P value
OR (95% CI)
No. of hyperglycemic events No. of hypoglycemic events Duration of hyperglycemia, h Duration of hypoglycemia, h Relative duration of hyperglycemic readings, % Relative duration of hypoglycemic readings, % Maximal glucose value, mmol/L Extreme slope, mmol/L/h Hyperglycemic index, mmol/L Hypoglycemic index, mmol/L
2.9 1.7 1.6 1.2 22.1 14.2 6.1 8.1 25.3 16.0 7.6 11.9 10.6 2.9 1.7 1.8 1.4 1.0 0.3 0.3
2.4 1.9 1.5 1.4 14.1 12.5 4.5 5.5 14.7 13.0 4.7 5.7 9.1 2.3 1.3 1.6 1.0 0.9 0.3 0.2
.048 .633 .0001 .134 <.0001 .06 .0002 .124 .003 .413
1.13 (0.9-1.4)
P value .21
†
10.3 (10.0-10.6)
.033
10.4z (10.1-10.7)
.005
1.01 (0.99-1.01)
.74
1.01 (0.99-1.03)
.19
*The variables for which there was a significant difference in the univariate analysis between the IVH and the control groups were entered into a multivariate logistic regression analysis. †Increase in the adjusted OR when the duration increases by 10 hours. zIncrease in the adjusted OR when the relative duration increases by 10%.
390
Auerbach et al
August 2013 The mean maximal glucose value measured in each infant during the study period was also higher in the IVH group compared with the control group. The hyperglycemic index indicating a time-weighted average of the hyperglycemic amplitude was also higher in the IVH group. Nevertheless, no difference was found between the groups in the rates of hypoglycemic or hyperglycemic event evolution and, intriguingly, rapid osmolar changes due to alterations in blood glucose levels were not associated with IVH occurrence. Hypoglycemic variables were not significantly associated with severe IVH in the univariate analysis. There was no statistically significant difference between hypoglycemic index and the minimal glucose value between the 2 groups. Multivariate analysis of glucose homeostasis variables including the significant background neonatal risk factors (eg, RDS) showed that only longer duration of hyperglycemia and longer relative duration of hyperglycemic readings remained strongly associated with IVH occurrence. A 10-hour increase in the total duration of even mild hyperglycemic events (close to the actual difference between IVH and control groups) resulted in a very prominent OR of 10.33 for the occurrence of severe IVH. When correlating the occurrence of hyperglycemia associated IVH and the general illness severity of the preterm infants, using the clinical severity score we found that in general the infants with severe IVH had higher severity scores and more variables of hyperglycemia. All hyperglycemic variables where significantly correlated with a higher clinical severity score (P < .05). Out of the different variables the duration of hyperglycemia (P < .0001) was most strongly correlated to the clinical severity score.
Figure. Box-plot representation of the distribution of the relative durations of hyperglycemic readings between the study groups. The line in the middle of the box is the median. The lower and the upper limit of the box are the lower and upper quartile, respectively. The whiskers extend to the lowest/ highest measurement, which is not an outlier. Outliers are individually presented.
ORIGINAL ARTICLES The Figure illustrates the distribution of the relative durations of hyperglycemic readings in the two groups. This box plot representation shows that in the control group most of the infants were hyperglycemic for a relative shorter duration than in the IVH group, and that the relative duration distribution was more homogeneous than in the IVH group. An infant with hyperglycemia exceeding 20 hours during the first 4 days of life was more prone to have severe IVH than an infant with shorter duration of hyperglycemia.
Discussion We examined the association between hyperglycemic alterations and severe IVH occurrence in preterm infants. Eighty-six percent of the infants experienced hyperglycemia in the first 96 hours of life, a higher prevalence compared with most previous studies (5%-60%).11,17 This finding may reflect our lower threshold for definition of hyperglycemia (above 6.9 mmol/L-125 mg/dL), our relative longer duration of glucose follow-up, and the relative low gestational age of the infants. In the multivariate analysis, out of many other risk factors studied here and in previous studies,6,20 only RDS, hypotension, and thrombocytopenia were independently associated with IVH. Interestingly, of the various glycemic variables, only the hyperglycemic duration was independently and strongly associated with severe IVH occurrence. A 10-hour increase in the duration of even mild hyperglycemia, (>6.9 mmol/L = 125 mg/dL) during the first 4 days resulted in a high aOR of 10.33 for the occurrence of severe IVH. The definitions of hypoglycemia or hyperglycemia in newborn infants are controversial21 because no correlations between plasma glucose values and clinical symptoms or long-term neurodevelopmental sequelae have been established.18,22-24 Furthermore, hypoglycemia and hyperglycemia reflect a continuum of biological abnormalities ranging from mild to severe.22 Thus, any definition should also consider the effect of duration and frequency of the abnormal glucose values. In this study, hypoglycemia was defined as blood glucose levels below 3.3 mmol/L (60 mg/dL). This relatively high threshold, (higher than the threshold for intervention, 45-50 mg/dL-2.5-2.7 mmol/L), was chosen in order to estimate whether we should treat hypoglycemia also in values above the accepted treatment threshold.22 Despite this high threshold, we found a lower incidence and shorter duration of hypoglycemia compared with hyperglycemia. Furthermore, the association between hypoglycemic variables and severe IVH was not statistically significant. It seems that the routine intravenous glucose infusion in premature infants and the higher awareness of the medical team to prevent hypoglycemia, compared with hyperglycemia21 have contributed to the low hypoglycemic duration in general and to the lower incidence of IVH in this subgroup. In contrast to hypoglycemia, hyperglycemia in preterm infants has been infrequently studied as a factor influencing their illness severity, IVH occurrence, and mortality.12,17,24
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants
391
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Furthermore, most of the studies related to hyperglycemic levels, which were significantly higher than the currently accepted threshold (above 180 mt/dL) for insulin therapy. The recent NIRTURE multicenter interventional study19,25 evaluated the response of all preterm infants to insulin treatment and concluded that so far no evidence supports the routine insulin administration in very low birth weight infants (birth weight <1500 g) to decrease mortality and morbidity rates. Nevertheless, studies in adults and children in intensive care units indicate an association between mild hyperglycemia and increased morbidity and mortality.16,17,26 We asked if there is a specific hyperglycemic variable that is strongly associated with IVH occurrence in order to define which subgroup of preterm infants may benefit from interventional glucose monitoring and treatment. As no clear hyperglycemic threshold definition is known for preterm infants, and as mild hyperglycemia for term infants is significantly higher than the 50th percentile of glucose values of preterm infants born at <32 weeks,18 we chose, similarly to previous studies,7,12,27 to use a relatively conservative glucose threshold, being above 6.9 mmol/L (125 mg/dL)—the statistical definition for term infants.18 Interestingly, the hyperglycemic amplitude or the rate of an event evolution that could lead to sharp changes in blood osmolality and, thus, to changes in cerebral blood flow and consequent cerebral bleeding,28 were not independently associated with IVH; supporting the observations that changes in blood glucose concentration from 90 mg/dL to 180 mg/dL result in a physiologically insignificant change in blood osmolality, 5 mOsm/L.29 Conversely, the finding that the duration of hyperglycemia is so strongly associated with IVH is intriguing and suggests that alterations in glucose homeostasis are important as reflecting a continuum of biological abnormalities. The need for interventional studies to evaluate the possible advantage of monitoring for hyperglycemia in preterm infants is further supported by the recent retrospective studies demonstrating that early hyperglycemia in preterm infants is independently associated with increased mortality and reduction of cerebral white matter at term,30 and with adverse neurodevelopmental and behavioral outcome at two years of age.31 A limitation of our study is its retrospective nature, allowing us to examine only associations but not cause and effect relationships. The prolonged hyperglycemic episodes observed in the IVH group could have been a result of clinical morbidity rather than a cause of severe clinical complications,11,18 as exemplified by the correlation between the clinical severity score and hyperglycemic variables. Likewise, we could not determine the time relation between hyperglycemia and IVH development, as cranial US was periodically performed. Nevertheless, the glucose metabolism variables assembled from the original NICU charts consisted of unbiased data that were independently and strongly associated with cranial US documentation of IVH. In conclusion, fluctuating glucose values and IVH are both common in stressed, small, and sick preterm 392
Vol. 163, No. 2 neonates. A strong independent association was found between hyperglycemia and IVH. This association was related mainly to the duration rather than to the amplitude of the hyperglycemic events. Based on a favorable outcome of a rigorous glucose monitoring in intensive care units, interventional studies are indicated also in preterm infants. Such studies should aim to selectively decrease the duration of hyperglycemia in infants at lower thresholds than the ones accepted today, under a strict and tuned protocol of insulin infusion. n Submitted for publication Aug 31, 2012; last revision received Dec 20, 2012; accepted Jan 23, 2013. Reprint requests: David Haim Zangen, MD, Division of Pediatric Endocrinology, Department of Pediatrics, Hadassah Hebrew University Medical Centre, P.O. Box 24035, Jerusalem 91240, Israel. E-mail: zangend@ hadassah.org.il
References 1. Whitelaw A. Intraventricular haemorrhage and posthaemorrhagic hydrocephalus: pathogenesis, prevention and future interventions. Semin Neonatol 2001;6:135-46. 2. Sherlock RL, Anderson PJ, Doyle LW. Victorian Infant Collaborative Study Group. Neurodevelopmental sequelae of intraventricular haemorrhage at 8 years of age in a regional cohort of ELBW/very preterm infants. Early Hum Dev 2005;81:909-16. 3. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of the subependymal intraventricular hemorrhage: a study of infants with birth weights less than 1500 grams. J Pediatr 1978;92: 529-34. 4. Perlman JM, McMenamin JB, Volpe JJ. Fluctuating cerebral blood-flow velocity in respiratory-distress syndrome. Relation to the development of intraventricular hemorrhage. N Engl J Med 1983;309:204-9. 5. Tsuji M, Saul JP, du Plessis A, Eichenwald E, Sobh J, Crocker R, et al. Cerebral intravascular oxygenation correlates with mean arterial pressure in critically ill premature infants. Pediatrics 2000;106:625-32. 6. Shankaran S, Bauer CR, Bain R, Wright LL, Zachary J. Prenatal and perinatal risk and protective factors for neonatal intracranial hemorrhage. National Institute of Child Health and Human Development Neonatal Research Network. Arch Pediatr Adolesc Med 1996;150:491-7. 7. Sunehag AL, Haymond MW. Glucose extremes in newborn infants. Clin Perinatol 2002;29:245-60. 8. Ktorza A, Bihoreau MT, Nurjhan N, Picon L, Girard J. Insulin and glucagon during the perinatal period: secretion and metabolic effects on the liver. Biol Neonate 1985;48:204-20. 9. Kalhan SC, Oliven A, King KC, Lucero C. Role of glucose in the regulation of endogenous glucose production in the human newborn. Pediatr Res 1986;10:49-52. 10. Ziegler EE, O’Donnell AM, Nelson SE, Fomon SJ. Body composition of the reference fetus. Growth 1976;40:329-41. 11. Lilien LD, Rosenfield RL, Baccaro MM, Pildes RS. Hyperglycemia in stressed small premature neonates. J Pediatr 1979;94:454-9. 12. Dweck HS, Cassady G. Glucose intolerance in infants of very low birth weight, part I: incidence of hyperglycaemia in infants of birth weights 1100 grams and less. Pediatrics 1974;53:189-95. 13. 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;5:329-36. 14. Faustino EV, Apkon M. Persistent hyperglycemia in critically ill children. J Pediatr 2005;146:30-4. 15. Branco RG, Garcia PCR, Piva JP, Casartelli CH, Seibel V, Tasker RC. Glucose level and risk of mortality in pediatric septic shock. Pediatr Crit Care Med 2005;6:470-2.
Auerbach et al
ORIGINAL ARTICLES
August 2013 16. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-67. 17. Hays SP, O’Brian Smith E, Sunehag AL. Hyperglycemia is a risk factor for early death and morbidity in extremely low birth-weight infants. Pediatrics 2006;118:1811-8. 18. Hey E. Hyperglycaemia and the very preterm baby. Semin Fetal Neonatal Med 2005;10:377-87. 19. Beardsall K, Vanhaesebrouck S, Ogilvy-Stuart AL, Vanhole C, Palmer CR, van Weissenbruch M, et al. Early insulin therapy in very low birth weight infants. New Eng J Med 2008;359:1873-84. 20. Synnes AR, Macnab YC, Qiu Z, Ohlsson A, Gustafson P, Dean CB, et al, Canadian Neonatal Network. Neonatal intensive care unit characteristics affect the incidence of severe intraventricular hemorrhage. Med Care 2006;44:754-9. 21. Alsweiler JM, Kuschel CA, Bloomfield FH. Survey of the management of neonatal hyperglycemia in Australia. J Paediatr Child Health 2007;43: 632-5. 22. Cornblath M, Hawdon JM, Williams AF, Aynsley-Green A, WardPlatt MP, Schwartz R, et al. Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds. Pediatrics 2000;105: 1141-5. 23. Kalhan S, Peter-Wohl S. Hypoglycemia: what it is for the neonate? Am J Perinatol 2000;17:11-8.
24. Hawdon JM, Ward Platt MP, Aynsley Green A. Patterns of metabolic adaptation for preterm and term infants in the first neonatal week. Arch Dis Child 1992;67:357-65. 25. Sinclair JC, Bottino M, Cowett RM. Interventions for prevention of neonatal hyperglycemia in very low birth weight infants. Cochrane Database Syst Rev 2011;CD007615. 26. Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-61. 27. Kao LS, Morris BH, Lally KP, Stewart CD, Huseby CD, Kennedy KA. Hyperglycemia and morbidity and mortality in extremely low birth weight infants. J Perinatol 2006;26:730-6. 28. Yager JY. Hypoglycemia injury to immature brain. Clin Perinatol 2002; 29:651-74. 29. Kalham SC, Devaskar SU. Disorders of carbohydrate metabolism. In: Fanaroff and Martin’s Neonatal Perinatal Medicine Diseases of the Fetus and Infant, 9th ed. Mosby: Elsevier; 2010. p. 1518-9. 30. Alexandrou G, Skiold B, Karlen J, Tessma MK, Norman M, Aden U, et al. Early hyperglycemia is a risk factor for death and white matter reduction in preterm infants. Pediatrics 2010;125:e584-91. 31. van der Lugt NM, Smits-Wintjens VE, van Zwieten PH, Walther FJ. Short- and long-term outcome of neonatal hyperglycemia in very preterm infants: a retrospective follow-up study. BMC Pediatr 2010; 10:52.
50 Years Ago in THE JOURNAL OF PEDIATRICS Sudden and Unexpected Death in Infants Valdes-Dapena MA, Eichman MF, Ziskin L. J Pediatr 1963;63:290-4
A
t the time of this study, there was a growing recognition of, and an increasing incidence of, sudden unexpected deaths (SUD), also referred to as sudden infant death syndrome, in previously well infants. Hypothesized pathogenesis included hypogammaglobulinemia, viscerovisceral reflex, hypersensitivity to aspirated milk, and viral infection. In this study, serum gammaglobulin levels in heart blood obtained postmortem from 114 infants with SUD were compared with samples from 33 healthy living infants of approximately similar age. (Measurement of specific immunoglobulins had not yet been introduced clinically.) The mean gammaglobulin level was 665 mg/dL in SUD cases, compared with 595 mg/dL in controls, thus dispelling the notion that hypogammaglobulinemia induced vulnerability to SUD. Dr Dapena performed these autopsies at the Office of the Medical Examiner in Philadelphia. The Chair of Pediatrics at Temple Medical School and St Christopher’s Hospital for Children at the time, W.E. Nelson, hired Dr Dapena (with some hesitation and expressed concern that as a woman she might not “pull her weight”) to join James B. Arey on the faculty of the Department of Pathology at St Christopher’s. Dr Dapena went on to become a renowned international researcher in SUD, a lauded teacher, and a mother of 11 children. Dr Dapena died in 2012, at which time we remaining admirers remembered the heady times of our training when over the hospital loudspeaker came the announcement, “paging Dr Post.” All available residents and students scurried to the autopsy room to watch Drs Arey and Dapena—along with a posse of scary-smart people that included Angelo DiGeorge, John Kirkpatrick, David Smith, Harold Lischner, and Victor Auerbach—make amazing discoveries and connect many of our clinical dots into cherished strings of pearls. Sarah S. Long, MD Department of Pediatrics St Christopher’s Hospital for Children Philadelphia, Pennsylvania http://dx.doi.org/10.1016/j.jpeds.2013.02.025
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants
393
Study design Preterm infants with IVH grade 3-4 (n = 70) were compared with matched infants of similar gestational age and birth weight, but with no IVH (n = 108). Studied variables included the frequency and duration of hyper/hypoglycemic (>6.9/<3.3 mmol/L, respectively) events, the extreme slope of an event evolution, the maximal glucose value observed, and the “hyper/hypoglycemic index” representing a weighted average of the hyper/hypoglycemic amplitude. Results The IVH group had significantly more hyperglycemic events (2.9 1.7 vs 2.4 1.8 events, P < .05) with longer duration (22.2 14.2 vs 14.1 12.5 hours, P < .001) and a higher hyperglycemic index (1.00.9 vs 1.41.0, P = .003) compared with the non-IVH controls. Respiratory distress syndrome, hypotension, and thrombocytopenia increased the adjusted OR for IVH. Hypoglycemia was not independently associated with IVH. Conversely, the increase in hyperglycemic duration was most prominently increasing the aOR for severe IVH (OR = 10.33, 95% CI = 10.0-10.6, P = .033). Conclusion Longer duration of hyperglycemia in the first 96 hours of life was most strongly associated with severe IVH in preterm infants. Consequently, interventional studies to determine the selective effect of continuous control of long-lasting hyperglycemia by appropriate and timed insulin treatment on the incidence of severe IVH are warranted. (J Pediatr 2013;163:388-93).
I
ntraventricular hemorrhage (IVH), occurring usually within the first 4 postnatal days, is a major cause of morbidity and mortality in preterm infants.1-3 Its pathophysiology involves bleeding from fragile blood vessels in the germinal matrix, which respond poorly to frequent changes in blood flow.1,4,5 Several risk factors, such as low gestational age and birth weight, perinatal stress, low Apgar score, and postnatal complications (eg, respiratory distress syndrome (RDS) or blood pressure instability),6 have been associated with IVH. Less studied is the association between the frequent alterations in blood glucose levels in preterm infants7-12 and the resultant changes in cerebral blood flow and osmolarity,7 which may challenge the fragility of blood vessels and contribute to the development of IVH. Hyperglycemia has lately been associated with increased mortality and morbidity rates, and insulin therapy and well-controlled blood glucose levels were found to be beneficial in adults and children in intensive care units.13-16 In extremely low birth weight infants (birth weight <1000 grams), high blood glucose concentration was associated with increased risk for IVH grade 3-4 and mortality.17 However, no causal relationship has been proven so far between the hyperglycemic amplitude and a detrimental outcome in interventional studies.18 Furthermore, recent attempts to treat all extremely low birth weight infants with continuous insulin infusion early in life (Neonatal Insulin Therapy in Europe Trial) failed to have a clinical benefit but, rather, increased the incidence of hypoglycemic episodes and mortality rate at 28 days.19 In this study, we aimed to identify which of the specific glucose homeostatic alterations (not necessarily the amplitude) may predict the occurrence of severe IVH (grade 3 or 4) and consequently be used to monitor an interventional protocol aiming to decrease the incidence of IVH.
Methods The data for this observational study were extracted from the Hadassah University Hospital’s neonatal intensive care unit (NICU) admission records of all neonates whose birth weights were below 2000 g during 5 consecutive years IVH NICU PDA RDS US
Intraventricular hemorrhage Neonatal intensive care unit Patent ductus arteriosus Respiratory distress syndrome Ultrasound
From the 1Division of Pediatric Endocrinology, Department of Pediatrics, 2Department of Neonatology, and 3School of Public Health, Hadassah Hebrew University Medical Center, Jerusalem, Israel *Contributed equally. This study was the MD thesis of A.A. at the Hadassah Hebrew University Medical School. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2013 Mosby Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.01.051
388
Vol. 163, No. 2 August 2013 (n = 907). The IVH study group included all neonates (n = 70) diagnosed with severe (grade 3 or 4) IVH by at least 2 cranial ultrasound (US) studies performed routinely within the first 3 days of life and subsequently between 5 and 10 days of life (subsequent US studies were performed as clinically indicated). IVH grading was determined independently by a radiologist and at least 2 neonatologists according to the Papile criteria.3 Grade 3 IVH was characterized by germinal matrix hemorrhage with ventricular dilatation (clot fills more than 50% of the ventricles), and in grade 4 IVH the hemorrhage was also intraparenchymal. The control non-IVH group (n = 108) consisted of neonates (1 or 2 according to availability) delivered immediately before or after a study group case with similar gestational age (10 days) and birth weight (150 g) but with no IVH (of any grade) by the routine US studies. One neonate from the IVH group and 5 from the control group were excluded from the study because of missing data. Based on previous studies examining risk factors for IVH in preterm infants,1,6 we studied the following background characteristics: gestational age, birth weight, multiple gestation, primiparity, etiology of early delivery, type of delivery, intrauterine growth retardation, pregnancy complications (preeclamsia, chrioamnionitis, bleeding) Apgar scores, first hematocrit, antenatal corticosteroids exposure, the presence of thrombocytopenia (platelets count <100 000), RDS, hypotension requiring therapy (dopamine or dobutamine), early onset sepsis (positive blood culture during the first 3 days of life), clinically significant patent ductus arteriosus (PDA), necrotizing enterocolitis, skin bruising, hyperbilirubinemia, and neonatal indomethacin treatment administered for closure of PDA. Glucose Homeostasis Calculations and Variables Soon after birth, continuous glucose infusion was administered to all preterm infants at a rate of 4-7 mg/kg/min to match basal requirements and to prevent hypoglycemia. Routine (NICU protocol) bedside glucometer monitoring was performed 8 times daily in all infants during the initial days of life. Higher blood glucose monitoring frequency (12-24 times daily) was performed when clinical stability or blood glucose levels were deranged. Hospital laboratory confirmation of glucometer readings was done at least once daily. Glucose values for the first 4 days of life were collected from the NICU standard records. Three infants in the control group missed 1-4 hours and 11 infants in the study group lacked 6-58 hours because of early death. Hyperglycemia and hypoglycemia were defined as glucose levels >6.9 mmol/L (>125 mg/dL) and <3.3 mmol/L (<60 mg/dL), respectively. Insulin was used whenever blood glucose levels were higher than 11.1 mmol/L (200 mg/dL) in 2-3 consecutive measurements. The glucose homeostasis variables studied for each neonate included: (1) the number of hyperglycemic and hypoglycemic events; (2) the duration of these events; (3) the percentage of time the neonates were hypoglycemic or hyperglycemic during the 4-day period or relative to the time of data collection for this infant (the “relative duration”); (4)
the maximal slope of an event evolution being the largest quotient calculated from the change in the glucose levels divided by the time of its evolution; (5) the maximal glucose value; and (6) the “hyper/hypoglycemic index,” which was calculated by multiplying the amplitude size of each abnormal glucose reading by its time duration. The cumulative sum of those products divided by total hyperglycemic or hypoglycemic duration per neonate gives a general index of glucose homeostasis representing glucose stability or instability for each infant. In order to correlate between the occurrence of hyperglycemic-associated IVH and the general illness severity of preterm infants, we created a semi-quantitative severity index score (0-5) based on death (5), days of ventilation (none = 0, up to 7 days = 1, more than 7 days = 2), bronchopulmonary dysplasia (O2 requirement at 36 weeks corrected gestational age, no = 0, yes = 1), need for dopamine treatment (no = 0, yes = 1), and day of enteral feeds commencement (up to 4 days = 0, more than 4 days = 1). Then we correlated the final severity score with the different hypo/hyperglycemic variables. Statistical Analyses Independent sample t test was used for comparison of quantitative variables between the IVH and control groups. The Pearson c2 test was applied for testing the association between the study groups and qualitative variables. The variables found to be significantly (P < .05) associated with the dependent variable (IVH or control) in the univariate analysis were entered in a stepwise manner into a multivariate logistic regression model. The simultaneous effect of different variables and their calculated aOR and 95% CI were then assessed. All tests applied were 2-tailed, and a P value of 5% was considered statistically significant. Kruskal–Wallis test (nonparametric ANOVA for small subgroups with abnormal distribution) was used for correlating the severity score to glycemic variables.
Results The birth weights of the IVH and control groups were similar by multivariate analysis (Table I). Lower gestational age, lower Apgar score at 5 minutes, and the occurrences of RDS, mechanical ventilation, pneumothorax, thrombocytopenia, hypotension, early sepsis, and bruising were found to be significantly associated with the occurrence of severe IVH. Antenatal maternal corticosteroid therapy and neonatal indomethacin treatment were more common in the control group and appeared to be protective against IVH. No statistically significant association with IVH was found for multiple gestation, primiparity, type of delivery, first hematocrit level or the presence of chorioamnionitits, preeclampsia, PDA, necrotizing enterocolitis, or hyperbilirubinemia. When the statistically significant (IVH associated) background variables (P < .05) were introduced into a multivariate analysis, only RDS, thrombocytopenia, and 389
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 163, No. 2
Table I. Background characteristics of the study population and the incidence of IVH (univariate and multivariate analysis) Univariate analysis
Multivariate analysis*
Variable
IVH group (n = 70)
Control group (n = 108)
P value
OR (95% CI)
P value
Birth weight (mean SD), g Gestational age (mean SD), wk Etiology of early delivery Labor, % (n) Premature rupture of membranes, % (n) Pregnancy complications†, % (n) Fetal distress, % (n) Type of delivery Vaginal delivery, % (n) Cesarean delivery, % (n) Antenatal steroid therapy, % (n) Small for gestational age, % (n) Apgar -5 min (mean SD), n RDS, % (n) Mechanical ventilation, % (n) Pneumothorax, % (n) Thrombocytopenia, % (n) PDA, % (n) Hypotension, % (n) Early sepsisz, % (n) Indomethacin treatment, % (n) Bruising, % (n)
1026 385 27.6 2.4
1126 339 28.4 2.2
.07 .01 .31
1.00 (1.0-1.0) 0.89 (0.6-1.3)
.38 .58
45.7 (32) 5.7 (4) 31.4 (22) 17.1 (12)
41.7 (45) 13.9 (15) 31.5 (34) 13 (14)
31.4 (22) 68.6 (48) 57.1 (40) 11.4 (8) 8.1 1.6 81.4 (57) 95.7 (67) 15.7 (11) 22.9 (16) 45.7 (32) 55.7 (39) 15.7 (11) 11.4 (8) 24.3 (17)
39.8 (43) 60.2 (65) 74.1 (80) 11.1 (12) 9.1 0.8 46.3 (50) 68.5 (74) 5.6 (6) 2.8 (3) 35.2 (38) 13.9 (15) 0 (0) 24.1 (26) 12 (13)
0.99 (0.4-2.3)
.99
.41 .02 .95 <.001 <.001 <.001 .02 <.001 .2 <.001 <.001 .03 .033
x
0.3 (0.7-15.6) 3.3 (1.5-7.3)
.12 .003
2.6 (0.7-8.7) 7.9 (1.9-33.3)
.12 .005
5.5 (2.5-12.2) NA{ 0.36 (0.1-0.9) 1.45 (0.5-4.1)
<.001 NA{ .047 .48
NA, not applicable. Statistically significant P values in the univariate analysis are in italics and statistically significant P values in the multivariate analysis are in bold. *The variables for which there was a significant difference in the univariate analysis between the IVH and the control groups were entered into a multivariate logistic regression analysis. †Pregnancy complications-preeclampsia, placental abruption, chorioamnionitis, bleeding. zEarly sepsis-sepsis in the first 3 days after delivery. xOR for Apgar-5 min >7. {The aOR received was infinite and not significant because there was no early sepsis in the control group.
hypotension requiring pressor supportive therapy remained significantly associated with severe IVH (Table I). Indomethacin treatment (administered for closure of PDA) maintained significantly protective association against IVH. Glucose Variables Analysis Derangements in blood glucose levels were common in both the IVH and the control group. During their first 4 days of life, 86% of the premature neonates had 1 and 29% had more than 4 hyperglycemic events. Hypoglycemia was less common as the comparable values for hypoglycemia were
75% and 8%, respectively. Moreover, the total duration of hypoglycemia (defined conservatively as <3.3 mmol/L-60 mg/dL) was much shorter compared with hyperglycemia (5.1 6.7 vs 17.3 13.7 hours, respectively). Univariate analysis revealed statistically significant differences in glucose homeostatic variables between the IVH and the control groups (Table II). The IVH group had more frequent and longer hyperglycemic events. Furthermore, 15.7% of the infants in the IVH group had abnormal glucose values for more than 50% of the study duration versus only 4.6% of the infants in the control group (P = .008).
Table II. Glucose homeostasis parameters and incidence of IVH (univariate and multivariate analysis) Univariate analysis
Multivariate analysis*
Variable (mean ± SD)
IVH group (n = 70)
Control group (n = 108)
P value
OR (95% CI)
No. of hyperglycemic events No. of hypoglycemic events Duration of hyperglycemia, h Duration of hypoglycemia, h Relative duration of hyperglycemic readings, % Relative duration of hypoglycemic readings, % Maximal glucose value, mmol/L Extreme slope, mmol/L/h Hyperglycemic index, mmol/L Hypoglycemic index, mmol/L
2.9 1.7 1.6 1.2 22.1 14.2 6.1 8.1 25.3 16.0 7.6 11.9 10.6 2.9 1.7 1.8 1.4 1.0 0.3 0.3
2.4 1.9 1.5 1.4 14.1 12.5 4.5 5.5 14.7 13.0 4.7 5.7 9.1 2.3 1.3 1.6 1.0 0.9 0.3 0.2
.048 .633 .0001 .134 <.0001 .06 .0002 .124 .003 .413
1.13 (0.9-1.4)
P value .21
†
10.3 (10.0-10.6)
.033
10.4z (10.1-10.7)
.005
1.01 (0.99-1.01)
.74
1.01 (0.99-1.03)
.19
*The variables for which there was a significant difference in the univariate analysis between the IVH and the control groups were entered into a multivariate logistic regression analysis. †Increase in the adjusted OR when the duration increases by 10 hours. zIncrease in the adjusted OR when the relative duration increases by 10%.
390
Auerbach et al
August 2013 The mean maximal glucose value measured in each infant during the study period was also higher in the IVH group compared with the control group. The hyperglycemic index indicating a time-weighted average of the hyperglycemic amplitude was also higher in the IVH group. Nevertheless, no difference was found between the groups in the rates of hypoglycemic or hyperglycemic event evolution and, intriguingly, rapid osmolar changes due to alterations in blood glucose levels were not associated with IVH occurrence. Hypoglycemic variables were not significantly associated with severe IVH in the univariate analysis. There was no statistically significant difference between hypoglycemic index and the minimal glucose value between the 2 groups. Multivariate analysis of glucose homeostasis variables including the significant background neonatal risk factors (eg, RDS) showed that only longer duration of hyperglycemia and longer relative duration of hyperglycemic readings remained strongly associated with IVH occurrence. A 10-hour increase in the total duration of even mild hyperglycemic events (close to the actual difference between IVH and control groups) resulted in a very prominent OR of 10.33 for the occurrence of severe IVH. When correlating the occurrence of hyperglycemia associated IVH and the general illness severity of the preterm infants, using the clinical severity score we found that in general the infants with severe IVH had higher severity scores and more variables of hyperglycemia. All hyperglycemic variables where significantly correlated with a higher clinical severity score (P < .05). Out of the different variables the duration of hyperglycemia (P < .0001) was most strongly correlated to the clinical severity score.
Figure. Box-plot representation of the distribution of the relative durations of hyperglycemic readings between the study groups. The line in the middle of the box is the median. The lower and the upper limit of the box are the lower and upper quartile, respectively. The whiskers extend to the lowest/ highest measurement, which is not an outlier. Outliers are individually presented.
ORIGINAL ARTICLES The Figure illustrates the distribution of the relative durations of hyperglycemic readings in the two groups. This box plot representation shows that in the control group most of the infants were hyperglycemic for a relative shorter duration than in the IVH group, and that the relative duration distribution was more homogeneous than in the IVH group. An infant with hyperglycemia exceeding 20 hours during the first 4 days of life was more prone to have severe IVH than an infant with shorter duration of hyperglycemia.
Discussion We examined the association between hyperglycemic alterations and severe IVH occurrence in preterm infants. Eighty-six percent of the infants experienced hyperglycemia in the first 96 hours of life, a higher prevalence compared with most previous studies (5%-60%).11,17 This finding may reflect our lower threshold for definition of hyperglycemia (above 6.9 mmol/L-125 mg/dL), our relative longer duration of glucose follow-up, and the relative low gestational age of the infants. In the multivariate analysis, out of many other risk factors studied here and in previous studies,6,20 only RDS, hypotension, and thrombocytopenia were independently associated with IVH. Interestingly, of the various glycemic variables, only the hyperglycemic duration was independently and strongly associated with severe IVH occurrence. A 10-hour increase in the duration of even mild hyperglycemia, (>6.9 mmol/L = 125 mg/dL) during the first 4 days resulted in a high aOR of 10.33 for the occurrence of severe IVH. The definitions of hypoglycemia or hyperglycemia in newborn infants are controversial21 because no correlations between plasma glucose values and clinical symptoms or long-term neurodevelopmental sequelae have been established.18,22-24 Furthermore, hypoglycemia and hyperglycemia reflect a continuum of biological abnormalities ranging from mild to severe.22 Thus, any definition should also consider the effect of duration and frequency of the abnormal glucose values. In this study, hypoglycemia was defined as blood glucose levels below 3.3 mmol/L (60 mg/dL). This relatively high threshold, (higher than the threshold for intervention, 45-50 mg/dL-2.5-2.7 mmol/L), was chosen in order to estimate whether we should treat hypoglycemia also in values above the accepted treatment threshold.22 Despite this high threshold, we found a lower incidence and shorter duration of hypoglycemia compared with hyperglycemia. Furthermore, the association between hypoglycemic variables and severe IVH was not statistically significant. It seems that the routine intravenous glucose infusion in premature infants and the higher awareness of the medical team to prevent hypoglycemia, compared with hyperglycemia21 have contributed to the low hypoglycemic duration in general and to the lower incidence of IVH in this subgroup. In contrast to hypoglycemia, hyperglycemia in preterm infants has been infrequently studied as a factor influencing their illness severity, IVH occurrence, and mortality.12,17,24
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants
391
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Furthermore, most of the studies related to hyperglycemic levels, which were significantly higher than the currently accepted threshold (above 180 mt/dL) for insulin therapy. The recent NIRTURE multicenter interventional study19,25 evaluated the response of all preterm infants to insulin treatment and concluded that so far no evidence supports the routine insulin administration in very low birth weight infants (birth weight <1500 g) to decrease mortality and morbidity rates. Nevertheless, studies in adults and children in intensive care units indicate an association between mild hyperglycemia and increased morbidity and mortality.16,17,26 We asked if there is a specific hyperglycemic variable that is strongly associated with IVH occurrence in order to define which subgroup of preterm infants may benefit from interventional glucose monitoring and treatment. As no clear hyperglycemic threshold definition is known for preterm infants, and as mild hyperglycemia for term infants is significantly higher than the 50th percentile of glucose values of preterm infants born at <32 weeks,18 we chose, similarly to previous studies,7,12,27 to use a relatively conservative glucose threshold, being above 6.9 mmol/L (125 mg/dL)—the statistical definition for term infants.18 Interestingly, the hyperglycemic amplitude or the rate of an event evolution that could lead to sharp changes in blood osmolality and, thus, to changes in cerebral blood flow and consequent cerebral bleeding,28 were not independently associated with IVH; supporting the observations that changes in blood glucose concentration from 90 mg/dL to 180 mg/dL result in a physiologically insignificant change in blood osmolality, 5 mOsm/L.29 Conversely, the finding that the duration of hyperglycemia is so strongly associated with IVH is intriguing and suggests that alterations in glucose homeostasis are important as reflecting a continuum of biological abnormalities. The need for interventional studies to evaluate the possible advantage of monitoring for hyperglycemia in preterm infants is further supported by the recent retrospective studies demonstrating that early hyperglycemia in preterm infants is independently associated with increased mortality and reduction of cerebral white matter at term,30 and with adverse neurodevelopmental and behavioral outcome at two years of age.31 A limitation of our study is its retrospective nature, allowing us to examine only associations but not cause and effect relationships. The prolonged hyperglycemic episodes observed in the IVH group could have been a result of clinical morbidity rather than a cause of severe clinical complications,11,18 as exemplified by the correlation between the clinical severity score and hyperglycemic variables. Likewise, we could not determine the time relation between hyperglycemia and IVH development, as cranial US was periodically performed. Nevertheless, the glucose metabolism variables assembled from the original NICU charts consisted of unbiased data that were independently and strongly associated with cranial US documentation of IVH. In conclusion, fluctuating glucose values and IVH are both common in stressed, small, and sick preterm 392
Vol. 163, No. 2 neonates. A strong independent association was found between hyperglycemia and IVH. This association was related mainly to the duration rather than to the amplitude of the hyperglycemic events. Based on a favorable outcome of a rigorous glucose monitoring in intensive care units, interventional studies are indicated also in preterm infants. Such studies should aim to selectively decrease the duration of hyperglycemia in infants at lower thresholds than the ones accepted today, under a strict and tuned protocol of insulin infusion. n Submitted for publication Aug 31, 2012; last revision received Dec 20, 2012; accepted Jan 23, 2013. Reprint requests: David Haim Zangen, MD, Division of Pediatric Endocrinology, Department of Pediatrics, Hadassah Hebrew University Medical Centre, P.O. Box 24035, Jerusalem 91240, Israel. E-mail: zangend@ hadassah.org.il
References 1. Whitelaw A. Intraventricular haemorrhage and posthaemorrhagic hydrocephalus: pathogenesis, prevention and future interventions. Semin Neonatol 2001;6:135-46. 2. Sherlock RL, Anderson PJ, Doyle LW. Victorian Infant Collaborative Study Group. Neurodevelopmental sequelae of intraventricular haemorrhage at 8 years of age in a regional cohort of ELBW/very preterm infants. Early Hum Dev 2005;81:909-16. 3. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of the subependymal intraventricular hemorrhage: a study of infants with birth weights less than 1500 grams. J Pediatr 1978;92: 529-34. 4. Perlman JM, McMenamin JB, Volpe JJ. Fluctuating cerebral blood-flow velocity in respiratory-distress syndrome. Relation to the development of intraventricular hemorrhage. N Engl J Med 1983;309:204-9. 5. Tsuji M, Saul JP, du Plessis A, Eichenwald E, Sobh J, Crocker R, et al. Cerebral intravascular oxygenation correlates with mean arterial pressure in critically ill premature infants. Pediatrics 2000;106:625-32. 6. Shankaran S, Bauer CR, Bain R, Wright LL, Zachary J. Prenatal and perinatal risk and protective factors for neonatal intracranial hemorrhage. National Institute of Child Health and Human Development Neonatal Research Network. Arch Pediatr Adolesc Med 1996;150:491-7. 7. Sunehag AL, Haymond MW. Glucose extremes in newborn infants. Clin Perinatol 2002;29:245-60. 8. Ktorza A, Bihoreau MT, Nurjhan N, Picon L, Girard J. Insulin and glucagon during the perinatal period: secretion and metabolic effects on the liver. Biol Neonate 1985;48:204-20. 9. Kalhan SC, Oliven A, King KC, Lucero C. Role of glucose in the regulation of endogenous glucose production in the human newborn. Pediatr Res 1986;10:49-52. 10. Ziegler EE, O’Donnell AM, Nelson SE, Fomon SJ. Body composition of the reference fetus. Growth 1976;40:329-41. 11. Lilien LD, Rosenfield RL, Baccaro MM, Pildes RS. Hyperglycemia in stressed small premature neonates. J Pediatr 1979;94:454-9. 12. Dweck HS, Cassady G. Glucose intolerance in infants of very low birth weight, part I: incidence of hyperglycaemia in infants of birth weights 1100 grams and less. Pediatrics 1974;53:189-95. 13. 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;5:329-36. 14. Faustino EV, Apkon M. Persistent hyperglycemia in critically ill children. J Pediatr 2005;146:30-4. 15. Branco RG, Garcia PCR, Piva JP, Casartelli CH, Seibel V, Tasker RC. Glucose level and risk of mortality in pediatric septic shock. Pediatr Crit Care Med 2005;6:470-2.
Auerbach et al
ORIGINAL ARTICLES
August 2013 16. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-67. 17. Hays SP, O’Brian Smith E, Sunehag AL. Hyperglycemia is a risk factor for early death and morbidity in extremely low birth-weight infants. Pediatrics 2006;118:1811-8. 18. Hey E. Hyperglycaemia and the very preterm baby. Semin Fetal Neonatal Med 2005;10:377-87. 19. Beardsall K, Vanhaesebrouck S, Ogilvy-Stuart AL, Vanhole C, Palmer CR, van Weissenbruch M, et al. Early insulin therapy in very low birth weight infants. New Eng J Med 2008;359:1873-84. 20. Synnes AR, Macnab YC, Qiu Z, Ohlsson A, Gustafson P, Dean CB, et al, Canadian Neonatal Network. Neonatal intensive care unit characteristics affect the incidence of severe intraventricular hemorrhage. Med Care 2006;44:754-9. 21. Alsweiler JM, Kuschel CA, Bloomfield FH. Survey of the management of neonatal hyperglycemia in Australia. J Paediatr Child Health 2007;43: 632-5. 22. Cornblath M, Hawdon JM, Williams AF, Aynsley-Green A, WardPlatt MP, Schwartz R, et al. Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds. Pediatrics 2000;105: 1141-5. 23. Kalhan S, Peter-Wohl S. Hypoglycemia: what it is for the neonate? Am J Perinatol 2000;17:11-8.
24. Hawdon JM, Ward Platt MP, Aynsley Green A. Patterns of metabolic adaptation for preterm and term infants in the first neonatal week. Arch Dis Child 1992;67:357-65. 25. Sinclair JC, Bottino M, Cowett RM. Interventions for prevention of neonatal hyperglycemia in very low birth weight infants. Cochrane Database Syst Rev 2011;CD007615. 26. Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-61. 27. Kao LS, Morris BH, Lally KP, Stewart CD, Huseby CD, Kennedy KA. Hyperglycemia and morbidity and mortality in extremely low birth weight infants. J Perinatol 2006;26:730-6. 28. Yager JY. Hypoglycemia injury to immature brain. Clin Perinatol 2002; 29:651-74. 29. Kalham SC, Devaskar SU. Disorders of carbohydrate metabolism. In: Fanaroff and Martin’s Neonatal Perinatal Medicine Diseases of the Fetus and Infant, 9th ed. Mosby: Elsevier; 2010. p. 1518-9. 30. Alexandrou G, Skiold B, Karlen J, Tessma MK, Norman M, Aden U, et al. Early hyperglycemia is a risk factor for death and white matter reduction in preterm infants. Pediatrics 2010;125:e584-91. 31. van der Lugt NM, Smits-Wintjens VE, van Zwieten PH, Walther FJ. Short- and long-term outcome of neonatal hyperglycemia in very preterm infants: a retrospective follow-up study. BMC Pediatr 2010; 10:52.
50 Years Ago in THE JOURNAL OF PEDIATRICS Sudden and Unexpected Death in Infants Valdes-Dapena MA, Eichman MF, Ziskin L. J Pediatr 1963;63:290-4
A
t the time of this study, there was a growing recognition of, and an increasing incidence of, sudden unexpected deaths (SUD), also referred to as sudden infant death syndrome, in previously well infants. Hypothesized pathogenesis included hypogammaglobulinemia, viscerovisceral reflex, hypersensitivity to aspirated milk, and viral infection. In this study, serum gammaglobulin levels in heart blood obtained postmortem from 114 infants with SUD were compared with samples from 33 healthy living infants of approximately similar age. (Measurement of specific immunoglobulins had not yet been introduced clinically.) The mean gammaglobulin level was 665 mg/dL in SUD cases, compared with 595 mg/dL in controls, thus dispelling the notion that hypogammaglobulinemia induced vulnerability to SUD. Dr Dapena performed these autopsies at the Office of the Medical Examiner in Philadelphia. The Chair of Pediatrics at Temple Medical School and St Christopher’s Hospital for Children at the time, W.E. Nelson, hired Dr Dapena (with some hesitation and expressed concern that as a woman she might not “pull her weight”) to join James B. Arey on the faculty of the Department of Pathology at St Christopher’s. Dr Dapena went on to become a renowned international researcher in SUD, a lauded teacher, and a mother of 11 children. Dr Dapena died in 2012, at which time we remaining admirers remembered the heady times of our training when over the hospital loudspeaker came the announcement, “paging Dr Post.” All available residents and students scurried to the autopsy room to watch Drs Arey and Dapena—along with a posse of scary-smart people that included Angelo DiGeorge, John Kirkpatrick, David Smith, Harold Lischner, and Victor Auerbach—make amazing discoveries and connect many of our clinical dots into cherished strings of pearls. Sarah S. Long, MD Department of Pediatrics St Christopher’s Hospital for Children Philadelphia, Pennsylvania http://dx.doi.org/10.1016/j.jpeds.2013.02.025
Long Duration of Hyperglycemia in the First 96 Hours of Life Is Associated with Severe Intraventricular Hemorrhage in Preterm Infants
393