Hyperglycemia in extremely low birth weight infants in a predominantly Hispanic population and related morbidities

Transcription

1 ORIGINAL ARTICLE Hyperglycemia in extremely low birth weight infants in a predominantly Hispanic population and related morbidities CL Blanco 1, JG Baillargeon 2, RL Morrison 1 and AK Gong 1 (2006) 26, r 2006 Nature Publishing Group All rights reserved /06 $ Department of Pediatrics, Division of Neonatology, University of Texas Health Science Center, San Antonio, TX, USA and 2 University of Texas Medical Branch Galveston, Galveston, TX, USA Objective: This study describes the incidence, correlates and subsequent morbidities of hyperglycemia, a highly prevalent condition in extremely low birth weight (ELBW) infants. Study design: A retrospective chart review of 169 infants with birth weight (BW)<1000 g was conducted. Hyperglycemia was defined as plasma glucose level X150 mg/dl during the first 2 weeks of life. Data were analyzed by logistic regression, multivariate analysis and Fisher exact test. Results: Overall, 88% of the study sample developed hyperglycemia in the first 2 weeks of life. Both gestational age (GA) (odds ratio (OR) 0.11, 95% confidence interval (CI) ¼ ) and chorioamnionitis (OR 0.10, 95% CI ¼ ) were inversely associated with hyperglycemia, whereas BW, sepsis and postnatal steroid exposure were not. After adjusting for GA, BW and postnatal steroids, hyperglycemia was associated with a statistically significant increase in retinopathy of prematurity (ROP) (OR 4.6, 95% CI ). No association was found with bronchopulmonary dysplasia, intraventricular hemorrhage, death or prolonged hospital stay. Conclusion: Lower GA was identified as the main factor associated with hyperglycemia in ELBW infants during the first 2 weeks of life. Hyperglycemia was associated with an increased incidence of ROP; further studies need to determine if this association is causal. (2006) 26, doi: /sj.jp ; published online 24 August 2006 Keywords: hyperglycemia; Hispanic; extremely low birth weight; retinopathy of prematurity; bronchopulmonary dysplasia Correspondence: Dr CL Blanco, Department of Pediatrics, Division of Neonatology, University of Texas Health Science Center San Antonio, MSC 7812, 7703 Floyd Curl Drive, San Antonio, TX , USA. blanco@uthscsa.edu Received 21 April 2006; revised 11 July 2006; accepted 1 August 2006; published online 24 August 2006 Introduction Extremely low birth weight (ELBW) infants (birth weight (BW) <1000 g) have become the largest clinical population in the neonatal intensive care unit as a result of improved clinical care and enhanced survival. 1 3 Many of these infants have glucose intolerance resulting in hyperglycemia, a condition that has become increasingly important with reports of up to 68% of ELBW infants during the first few weeks of life. 4 In fact, the neonate with BW less than 1100 g is 18 times more likely to develop hyperglycemia than one who weighs more than 2000 g. 5 Several conditions have been implicated in the development of neonatal hyperglycemia, including intravenous fluids, parenteral nutrition, stress and drug treatment, in particular steroids Some investigators have evaluated the short-term adverse effects of hyperglycemia in the neonate; 1,11 however, few have focused on long-term outcomes such as intraventricular hemorrhage (IVH), retinopathy of prematurity (ROP) and death. 9,10,12 The majority of previous studies were performed when the survival of ELBW infants was significantly lower. Furthermore, no study has investigated the role of Hispanic ethnicity in the development of hyperglycemia in the premature infant. Thus, the aim of this study was to evaluate the incidence of hyperglycemia in ELBW infants in a predominantly Hispanic population, correlates and subsequent morbidities associated with this condition. Materials and methods Study population A retrospective chart review of live-born infants with BW less than 1000 g admitted to University Hospital, San Antonio, TX, from January 1998 to December 2001, was conducted. All data were obtained by reviewing existing patient records. A total of 230 charts were reviewed. Pre-viable infants were excluded (45 infants); they were defined as infants who received no resuscitative measures due to extreme prematurity and survival was limited to a few hours with no interventions provided. Of the remaining 185 eligible patients, 16 were excluded due to incomplete medical records;

2 738 the majority was of patients transferred into our facility after the study time period (first 2 weeks of life). w 2 analyses indicated that ineligible subjects exhibited comparable distribution of demographic and clinical characteristics relative to those included in the study sample. All infants were followed until death (if occurring in hospital), or until hospital discharge. Data collection Clinical and demographic data extracted from medical chart included gestational age (GA), BW, anthropometric data, gender, ethnicity, length of stay and death. GA was determined by the attending neonatologist based on clinical data along with maternal dates and/or prenatal ultrasound when available. GA data were dichotomized with an assigned value of 0 if GA was <26 weeks gestation, and a value of 1 if X26 weeks; BW was dichotomized with the following values: 0 if <750 g, and 1 if X750 g. Ethnicity was defined by mothers self-designation at the time of admission; if that data were not available, then designation was assigned by maternal Spanish surname. Prenatal data collected included mode of delivery, antenatal corticosteroid use and chorioamnionitis as defined by the clinical assessment of the obstetrician and/or placental pathology report. The following data were recorded if present within the first 15 days of life: sepsis as defined by at least one positive blood, urine or cerebrospinal fluid culture; any major surgical intervention; and the use of corticosteroids, insulin and catecholamines with regard to the day of initiation, the number of days used and the dose. Hyperglycemia was defined as a plasma glucose concentration of X150 mg/dl (8.3 mmol/l) on at least two different occasions during the first 15 days of life; this level has been previously identified as being clinically relevant. 4,9,13 Measurements of glucose levels were performed using the glucose oxidation technique (AU 640 Immuno Analyzer, Olympus Inc., Center Valley, PA, USA) at a frequency determined by the clinical status of each infant. The highest glucose value per day was recorded in all infants (G peak ). If the glucose value was X150 mg/dl, the number of days with hyperglycemia was recorded. Glucose, protein and lipid infusion rates were generated from parenteral nutrition information sheets; the number of days to reach maximum lipid and protein infusion concentrations were recorded when information was available (95 charts). For each patient, the following nutritional variables were obtained: discharge anthropometrics, weight gain by 28 days of age (g/day), the day enteral feeds were initiated, the number of days required to reach full volume feeds and volume of enteral feeds on the day of G peak. Outcomes reviewed included: bronchopulmonary dysplasia (BPD), defined as the need for oxygen at 36 weeks corrected GA; IVH noting the highest degree during hospitalization; and ROP, recording the most severe ROP examination staged by an ophthalmologist according to the International Classification of ROP. 14 The University of Texas Health Science Center San Antonio Institutional Review Board approved this study. Statistical methods All statistical analyses were performed with the Statistical Analysis System version 9.1. (SAS Institute, Cary, NC). Distributions and means of demographic and clinical variables were compared across study subgroups using w 2 tests and t-tests, respectively. Standard deviations are shown for continuous variables. Logistic regression was employed to assess correlates of hyperglycemia, and to assess the association of hyperglycemia with several clinical outcomes, after adjusting for GA, BW and postnatal steroid use. Owing to limited statistical power, only covariates of primary conceptual interest were included in the final multivariate analysis. Fisher s exact test was utilized when indicated. Results One hundred and fifty (88%) of the 169 patients included in this study developed hyperglycemia. Hispanics accounted for 71% of the total population, and males for 55%. No differences in gender and ethnicity were found between hyperglycemic and euglycemic infants. Being small for GA (SGA) at birth was not associated with hyperglycemia, although there were no SGA infants in the euglycemic group. The demographic details are summarized in Table 1. Infants with GA X26 weeks and BW X750 g were less likely to have hyperglycemia according to bivariate analysis, but only GA continued to reach statistical significance (odds ratio 0.11 ; 95% confidence interval ) after adjusting for other study covariates (Table 2). Infants born to a mother with Table 1 Demographic characteristics Study characteristics Euglycemic infants (n ¼ 19) Hyperglycemic P-value infants (n ¼ 150) GA (weeks) 26.8± ±2.2 <0.04 BW (g) 843± ±134 <0.002 Small for GA 0 18 (12%) 0.10 Female 10 (52%) 67 (45%) 0.64 Hispanic 16 (85%) 106 (71%) 0.63 C-section 12 (63%) 93 (61%) 0.92 Chorioamnionitis 6 (31%) 32 (26%) 0.27 Prenatal steroids 7 (36%) 47 (31%) 0.62 Postnatal Steroids 2 (10%) 42 (28%) 0.10 Catecholamines 5 (27%) 44 (29%) 0.58 Sepsis 5 (27%) 46 (31%) 0.77 Length of stay (days) 61±28 66± Death 3 (15%) 43 (28%) 0.39 GIR during G peak (mg/kg/min) 8.1± ± Abbreviations: GA, gestational age; GIR, glucose infusion rate; G peak, maximum glucose level.

3 739 Table 2 Multivariate analysis with factors associated with hyperglycemia Factor OR 95% CI GA BW Chorioamnionitis Postnatal steroids Sepsis Abbreviations: BW, birth weight; GA, gestational age chorioamnionitis were less likely to develop hyperglycemia, whereas sepsis and exposure to postnatal steroids were not associated with hyperglycemia (Table 2). Forty-four (26%) of the patients were exposed to postnatal steroids, dexamethasone accounting for 77% of the cases and hydrocortisone for the remainder. Glucose levels The mean plasma glucose level for the hyperglycemic group was 248±83 and 128±16 mg/dl for the euglycemic group. On average, hyperglycemic infants reached a G peak on day of life 6 (range 1 15 days), and remained hyperglycemic for 3.7 days (1 15 days). Iatrogenic hyperglycemia was found in 36 (21%) hyperglycemic infants who had a mean G peak of 194 mg/dl. This type of hyperglycemia resolved within 1 day; the glucose infusion rate (GIR) was significantly higher than the other hyperglycemic infants (8.0 vs 6.1 mg/kg/min, respectively, P ¼ 0.01). Insulin was utilized in 14% of the infants with hyperglycemia. The criterion for insulin therapy was determined by the attending neonatologist. In general, this group of infants was younger by GA (insulin group 24.4±1.1 weeks, non-insulin group 26.0±2.2 weeks, P ¼ 0.002), remained hyperglycemic for a longer period of time (insulin group 5.2 days, non-insulin group 3.5 days, P ¼ 0.01) and had high mortality rate (90%). Catecholamine use (dopamine, dobutamine and epinephrine) was not associated with hyperglycemia (Table 1). Nutritional data No significant difference was found in the volume of enteral feeds during G peak, initiation of enteral feeds, number of days to reach full feeds, GIR during G peak (Table 1) or days to reach maximum parenteral protein and lipid infusion rates between groups. The median number of days to reach maximum parenteral protein and lipid infusion in our cohort was 5 and 6.5 days, respectively. During the study period, the usual practice for advancing parenteral protein supplementation was to initiate 0.5 g/kg/day on the first day of life, and then advance by 0.5 g/kg every day to a maximum of 3 g/kg/day. Discharge anthropometric parameters and weight gain by 28 days of life were similar in both groups (data not shown). Table 3 Outcome variables adjusted for GA, BW and postnatal steroid exposure Outcome OR 95% CI BPD ROP IVH Death before discharge Duration of hospital stay >90 days Abbreviations: BPD, bronchopulmonary dysplasia; BW, birth weight; CI, confidence interval; GA, gestational age; IVH, intraventricular hemorrhage; OR, odds ratio; ROP, retinopathy of prematurity. Clinical outcomes BPD was diagnosed in 29%, ROP (any stage) in 59% and IVH was present in 24% of the population studied. Hyperglycemia was associated with a 4.5-fold increase in ROP even after correction for GA, BW and postnatal steroid use (Table 3). Threshold ROP did not reach statistical significance, although laser treatment was required in 15 of 66 patients with ROP in the hyperglycemic group and in none of the three patients with ROP in the euglycemic group (P ¼ 0.21). Hyperglycemia was not found to be associated with death, prolonged length of hospital stay (>than 90 days) or incidence of BPD or IVH (Table 3). Discussion Despite the high prevalence of hyperglycemia in ELBW infants, little is known about factors associated with this condition and the impact on subsequent morbidities. 12,15,16 Although there are known complications of high glucose levels in conditions such as diabetes mellitus, the significance of hyperglycemia in ELBW infants is questionable. Most studies to date have linked hyperglycemia with short-term outcomes such as sepsis, glucosuria, osmotic diuresis and dehydration. 1,3,4,9 Therefore, the aim of this study was to evaluate the incidence of hyperglycemia in ELBW infants, correlates and associated morbidities. In our predominantly Hispanic population, the overall incidence of hyperglycemia during the first 2 weeks of life was 88%. Although higher than previously reported, other factors besides Hispanic ethnicity may lead to the development of this condition. Extreme prematurity as evidenced by lower GA, not BW, was the most important factor for the development of hyperglycemia. This supports the existing data that suggest hepatic glucose production, pancreatic beta cell response and insulin sensitivity are physiologically related and developmentally regulated. 3,17 19 Contrary to previous reports, sepsis and postnatal steroid exposure during the first 2 weeks of life were not associated with hyperglycemia, although a trend existed with postnatal steroid exposure. The lack of association with steroids may be due to the

4 740 small number of euglycemic patients or to the change in clinical practice of using lower dexamethasone doses and shorter regimens. Furthermore, the time of occurrence of both postnatal steroid exposure and diagnosis of sepsis did not occur prior to the development of hyperglycemia in many of the cases and were often coincidental, thus perpetuating glucose intolerance in infants already at high risk. Unfortunately, we were unable to sort out the time of occurrence due to overlapping of variables and the retrospective nature of the study. The protective nature of chorioamnionitis with hyperglycemia was a surprising finding that has never been reported; the nature of this association remains unclear and will need to be studied prospectively with pre-defined criteria for the diagnosis. The infants with hyperglycemia had a 4.5-fold increase in ROP; this association has been reported in a single study. 12 Whether the increase in ROP is secondary to hyperglycemia or is just an expression of severity of illness is unknown. This study was not designed to answer that question, although no differences in other variables that express severity of illness such as BPD, IVH, death and prolonged length of stay were found. No association was found between hyperglycemia and threshold ROP, although the sample size does not allow for any conclusions here. Many factors are involved in the development and progression of ROP. 20 Animal and human studies support the relationship between hyperglycemia and circulatory retinal changes, which, in conjunction with hypoxemia, can accelerate the onset and progression of diabetic retinopathy Furthermore, it is well established that the prevention of proliferative retinopathy in diabetic adults requires the tightest possible glucose control. 22,25 A number of study limitations need to be addressed. A retrospective design can demonstrate associations between variables but not causation. There was no control for the variation of parenteral glucose infusion, glucose monitoring and treatment among physicians. Our incidence of hyperglycemia might be higher due to intensive glucose monitoring and/or overly aggressive advancement of parenteral nutrition, as higher GIRs were observed on the infants classified as having iatrogenic hyperglycemia. Owing to the high incidence of hyperglycemia found in our population, the number of patients in the euglycemic group was lower than expected. Therefore, only covariates of primary conceptual interest were used in the multivariate analysis. Understanding the metabolic complications associated with ELBW infants is important due to the increased survival of these infants and their extensive morbidities. Of particular interest is the growing concern of cell programming during critical periods of development. In particular, we do not understand the programming of glucose regulation and how the pancreas functions during this critical stage of extra uterine development for extremely premature infants and the subsequent development of disease later in life. 16,26 28 Recent reports have shown that strict normalization of blood glucose levels among critically ill adults, most of them without a history of diabetes, achieved significant reductions in morbidity and mortality. Hence, it is important to elucidate if optimal glucose management may have an impact in the outcomes of these vulnerable infants. 29 In fact, current changes in clinical practice may have an impact on glucose control. An area of increasing interest is the early institution of parenteral amino-acid solutions in premature infants, which has shown to reduce hyperglycemia. 30 Therefore, future studies of glucose metabolism in ELBW infants are warranted. Conclusion The present study identifies lower GA as the main factor associated with hyperglycemia in ELBW infants during the first 2 weeks of life. Other factors such as chorioamnionitis need further investigation to elucidate their relationship to glucose metabolism. Hyperglycemia during the first weeks of life may be associated with an increased incidence of ROP. As new evidence relating to the adverse effects of hyperglycemia in ELBW infants arises, well-constructed, prospective studies must be performed to investigate the short- and long-term benefits of tighter glucose control. References 1 Hey E. Hyperglycaemia and the very preterm baby. Semin Fetal Neonat Med 2005; 10(4): Cowett RM, Farrag HM, Cowett RM, Farrag HM. Selected principles of perinatal neonatal glucose metabolism. Semin Neonatol 2004; 9(1): Kalhan SC, Cowett RM. Carbohydrate metabolism In: Polin RA, Fox WW, Abman SH, (eds). Fetal and Neonatal Physiology. Philadelphia: Saunders, 2004, pp Farrag HM, Cowett RM. Glucose homeostasis in the micropremie. Clin Perinatol 2000; 27(1): Louik C, Mitchell AA, Epstein MF, Shapiro S, Louik C, Mitchell AA et al. Risk factors for neonatal hyperglycemia associated with 10% dextrose infusion. Am J Dis Children 1985; 139(8): Fanaroff AA, Korones SB, Wright LL, Verter J, Poland RL, Bauer CR et al. Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J 1998; 17(7): Lilien LD, Rosenfield RL, Baccaro MM, Pildes RS, Lilien LD, Rosenfield RL et al. Hyperglycemia in stressed small premature neonates. J Pediatr 1979; 94(3): The Vermont Oxford Network Steroid Study Group. Early postnatal dexamethasone therapy for the prevention of chronic lung disease. Pediatrics 2001; 108(3): Pildes RS. Neonatal hyperglycemia. J Pediatr 1986; 109(5):

5 Zarif M, Pildes RS, Vidyasagar D, Zarif M, Pildes RS, Vidyasagar D. Insulin and growth-hormone responses in neonatal hyperglycemia. Diabetes 1976; 25(5): Finberg L, Finberg L. Dangers to infants caused by changes in osmolal concentration. Pediatrics 1967; 40(6): Garg R, Agthe AG, Donohue PK, Lehmann CU. Hyperglycemia and retinopathy of prematurity in very low birth weight infants. J Perinatol 2003; 23(3): Denne SC, Poindexter BB, Leitch CA, Ernst JA, Lemons PK, Lemons JA. Nutrition and metabolism in the high risk neonate In: Fanaroff AA, Martin RJ (eds). Neonatal Perinatal Medicine. Diseases of the Fetus and Infant. Philadephia: Mosby, 2006, pp Committee for the Classification of Retinopathy of Prematurity. An international classification of retinopathy of prematurity. Pediatrics 1984; 74: 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): Catalano PM, Kirwan JP, Haugel-de Mouzon S, King J. Gestational diabetes and insulin resistance: role in short- and long-term implications for mother and fetus. J Nutr 2003; 133(5, Suppl 2): 1674S 1683S. 17 Cowett RM, Rapoza RE, Gelardi NL. The contribution of glucose to neonatal glucose homeostasis in the lamb. Metab Clin Exp 1998; 47(10): Gelardi NL, Rapoza RE, Renzulli JF, Cowett RM. Insulin resistance and glucose transporter expression during the euglycemic hyperinsulinemic clamp in the lamb. Am J Physiol 1999; 277(6, Part 1): E1142 E Mitanchez-Mokhtari D, Lahlou N, Kieffer F, Magny JF, Roger M, Voyer M. Both relative insulin resistance and defective islet \{beta\}-cell processing of proinsulin are responsible for. Pediatrics 2004; 113(3): Early Treatment for Retinopathy of Prematurity Cooperative Group. The incidence and course of retinopathy of prematurity: findings from the early treatment for retinopathy of prematurity study. Pediatrics 2005; 116(1): Atherton A, Hill DW, Keen H, Young S, Edwards EJ. The effect of acute hyperglycaemia on the retinal circulation of the normal cat. Diabetologia 1980; 18(3): The Diabetes Control and Complications Trial Research Group. Lifetime benefits and costs of intensive therapy as practiced in the diabetes control and complications trial. JAMA 1996; 276(17): Padnick-Silver L, Linsenmeier RA. Effect of hypoxemia and hyperglycemia on ph in the intact cat retina. Arch Ophthalmol 2005; 123(12): Nyengaard JR, Ido Y, Kilo C, Williamson JR. Interactions between hyperglycemia and hypoxia: implications for diabetic retinopathy. Diabetes 2004; 53(11): Porta M, Allione A. Current approaches and perspectives in the medical treatment of diabetic retinopathy. Pharmacol Therap 2004; 103(2): Phillips DI, Barker DJ, Hales CN, Hirst S, Osmond C. Thinness at birth and insulin resistance in adult life. Diabetologia 1994; 37(2): Ong KK, Petry CJ, Emmett PM, Sandhu MS, Kiess W, Hales CN et al. Insulin sensitivity and secretion in normal children related to size at birth, postnatal growth, and plasma insulin-like growth factor-i levels. Diabetologia 2004; 47(6): Aynsley-Green A, Hawdon JM, Deshpande S, Platt MW, Lindley K, Lucas A. Neonatal insulin secretion: implications for the programming of metabolic homeostasis. Acta Paediatr Jpn 1997; 39(Suppl 1): S21 S 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(5): te Braake FWJ, van den Akker CHP, Wattimena DJL, Huijmans JGM, van Goudoever JB. Amino acid administration to premature infants directly after birth. J Pediatr 2005; 147(4):