Recent studies in critically ill adults (JAMA 2003) and children (Pediatr Crit Care Med 2004) have demonstrated an association between hyperglycemia and increased mortality. Hyperglycemia is very common (prevalence 50-80%) in infants with birth weights below 1000g, especially those who are sick and who receive glucose infusion above their normal glucose turnover rate (~6mg/kg/min). It has been demonstrated that the risk of hyperglycemia increases when the infant receives glucose at rates exceeding the normal turnover rate of newborns. Thus, one approach to prevent or treat hyperglycemia would be to maintain or reduce the glucose infusion rate to ~ 6mg/kg min. In our previous studies in premature infants, we reduced the glucose infusion rate to 3mg/kg min (half normal glucose turnover rate) to test whether these infants were capable of activating their gluconeogenic pathway and, if so, which endogenous and exogenous substrates were more important in supporting gluconeogenesis. Hyperglycemia could be the result of insufficient glucose uptake to match the inflow of glucose, or of incomplete suppression of glucose production or a combination of both, and these factors could, in turn, be the result of insufficient insulin secretion with or without insulin resistance. Thus, an alternative approach to the treatment of hyperglycemia is infusion of exogenous insulin, that could potentially normalize glucose by suppressing glucose production from gluconeogenesis and/or glycogenolysis, or by increasing glucose utilization, or a combination of both. If insulin increases glucose utilization, that would allow us to maintain a glucose infusion rate exceeding the normal glucose turnover rate and, thus, a higher energy intake. However, in newborn infants, 90% of glucose utilization is by the brain. This uptake is independent of insulin. Therefore, only 10% of basal glucose utilization is by insulin dependent tissue i.e. fat and muscle. As a consequence, the capacity of newborn infants to increase glucose utilization in response to insulin infusion might be limited, particularly, in the ELBW infants, who have very small muscle and fat mass. Information about the mechanisms by which exogenous insulin affects glucose metabolism in sick newborn infants is crucial to choosing the most appropriate treatment of hyperglycemia in various clinical situations.

1) Rates of gluconeogenesis do not change in response to a decrease in blood levels of glucose and insulin, resulting from a decrease in the glucose infusion rate to 3mg/kg/min.

2) A fixed infusion of insulin sufficient to achieve a plasma insulin concentration in the high physiologic range (~80-100 uU/ml) will:

a) Not acutely suppress rates of gluconeogenesis
b) Increase glucose utilization regardless of body weight, gestational age and clinical status, but
increasing the glucose utilization will be proportional to body weight
c) Increase protein synthesis

The infants will be studied on one occasion during their first 7 days of life since this is the most critical period for the issues to be addressed. The patients will be assigned to one of two studies (Protocol 1 or Protocol 2) depending on their whole blood glucose (WBG) level. If they have a WBG lower than 150 mg/dL, they will be assigned to Protocol 1 in which the infants will be studied during infusion of routine TPN and TPN with reduced glucose infusion rate. If their WBG is higher than 150mg/dL, they will be randomized to either a study design identical to Protocol 1 or to Protocol 2, where they will receive a continuous low dose insulin infusion under conditions of routine TPN. In all infants assigned to Protocol 1, the pre-study infusion rates of lipids (Intralipid) and amino acids (Trophamine) will be continued (as ordered by the physician on service) throughout the entire 11 h study period. During the first 5 h of the 11 h study period, the glucose infusion will be given at the pre-study rate. In an identical fashion to that used in our previous studies, at study hour five, the glucose infusion rate will be reduced step wise (to minimize a counter regulatory response) first to 6 mg/kg min, and then at study hour 6, to 3 mg/kg min (i.e. half the infant normal glucose turnover rate). The glucose infusion will be maintained at this rate for the final 5 h of the study. Beginning at study hour zero and continued throughout the 11 h study period, 3-mg/kg min of the glucose infusion will be replaced by [U-13C] glucose, which is metabolically equivalent to natural glucose. Thus, all glucose given during the final five hours will be represented by [U-13C] glucose. This tracer will be used to measure rates of appearance of glucose, and glucose production from gluconeogenesis and glycogenolysis. Blood samples will be obtained at baseline and at study hours 4.5, 5, 10.5 and 11h. In addition, to prevent hypoglycemia (blood glucose < 40 mg/dL; 2.2mM), in all infants, during the final 6 h period, when the glucose infusion rate will be reduced first to 6 and then to 3 mg/kg min, blood glucose will be measured hourly. In the unlikely event that blood glucose falls bellow 40 mg/dL, the glucose infusion rate will be increased to maintain blood glucose above 40 mg/dl.

In protocol 2, all subjects will be studied in an identical fashion to that described above for protocol 1, except 1) the glucose infusion will not be reduced; 2) beginning at study hour 5, they will receive a fixed rate insulin infusion (see details below); and 3) in addition to [U-13C] glucose, the infants will receive i.v. infusions of [1-13C] leucine (to measure appearance rate of leucine, an indicator of protein turnover) and [15N2] urea (to measure appearance rate of urea an indicator of protein oxidation). At study hour 5, an I.V. bolus of Regular insulin (0.25 units) followed by a continuous infusion of Regular insulin (0.05 units/h) in 5% glucose (according to the clinical routines in the nursery) will be started and continued for 6 h. established and the insulin concentration will have To prevent hypoglycemia, if the glucose concentration falls below 60 mg/dL (3.3 mM), the glucose infusion rate will be increased to clamp blood glucose at this concentration. Since infusion of insulin might result in a more rapid fall in blood glucose as compared to decreased the glucose infusion rate to 3 mg/kg min (as will be done in protocol 1), we have set the threshold for increasing the glucose infusion rate at a blood glucose concentration of 60 mg/dL (3.3mM) rather than 40 mg/dL (2.2 mM) (described above for protocol 1). Blood samples will be drawn in an identical fashion to that described in protocol 1, except that in all infants small blood samples (5 uL each) will be obtained every 15 min during the first 2 h of insulin infusion and then every 30 min for the remaining 4 h of the study period for analysis of blood glucose. Urine will be collected during the entire study period using a diaper pad lined with a plastic layer to prevent leakage to the diaper.

32

Infants with birth weight of less than 1200 g appropriate for gestational age and with a postnatal age of 7 days or less. Infants eligible for Protocol 2 (fixed insulin infusion) must have blood glucose above 150 mg/dL at the time of the study.

Six babies have been studied. Preliminary data demonstrate that glucose production and gluconeogenesis are not acutely regulated by plasma insulin, plasma glucose or glucose infusion rates. Abstract submitted to SPR.