- Email: [email protected]
Serum glucose levels in term neonates during the first 48 hours of life
Volume 110 Number 1 pulmonary function in infants surviving with chronic lung disease. Pediatr Res 1984;392A:1976. 10. Watts J, Ariagno R, Brady J. Chronic pulmonary function in disease in neonates after artificial ventilation: distribution of ventilation and pulmonary interstitial emphysema. Pediatrics 1977;60:273-81. I 1. Rosen RC. Hyaline membrane disease and a related spectrum of neonatal pneumopathies: the relationship of normal to diseased neonatal lung, In: Rosenberg HS, Bolande RP, eds. Perspectives in pediatric pathology, vol 2. Chicago: Year Book, 1975:35. 12. Lindroth M, Johnson B, Svenningsen NW, Mortensson W.
Metaproterenol in chronic lung disease
119
Pulmonary mechanics, chest x-ray and lung disease after mechanical ventilation in low birth weight infants. Acta Paediatr Scand 1980;69:761-70. 13. Moyland FMB, Shannon DC. Preferential distribution of lobar emphysema and atelectasis in bronchopulmonary dysplasia. Pediatrics 1979;63:130-4. 14. Engel S. Lung structure. Springfield, I11.:Charles C Thomas, 1962:14-36. 15. Tal A, Bar-Yishay E, Eyal F, et al. Lack of response of airway obstruction after mechanical ventilation in the newborn. Crit Care Med 1982;10:361-2.
Clinical and laboratory observations Serum glucose levels in term neonates during the first 48 hours of life Louis J. Heck, M.D., a n d A l l e n E r e n b e r g , M.D. From the Department of Obstetrics and Gynecology, University of South Dakota School of Medicine, Yankton, and the Department of Pediatrics, The University of Iowa Hospitals and Clinics, Iowa City
The most common definition of hypoglycemia used today was derived from whole blood glucose levels of infants grouped by birth weight, and later adjusted upward mathematically to approximate plasma or serum levels. I Since these early studies, the importance of gestational age has become apparent, newborn infants are more frequently breast-fed and fed earlier, and intravenously administered dextrose solution is often given to women during labor and delivery., With these changes in perinatal care, clinical experience has led some authors to recommended that newborn serum glucose levels should be maintained at > 4 0 mg/dL. 2 However, Sexson 3 recently showed that using this value results in a 20% incidence of diagnosis of hypoglycemia in full-term infants. Srinivasan et al. 4 recently described plasma glucose values in healthy neonates, indicating that after the first 3 hours of life plasma glucose concentrations are normally >40 mg/dL.
Received for publication March 28, 1986; accepted Aug. 19, 1986. Reprint requests: Louis J. Heck, M.D., Department of Obstetrics and Gynecology, University of South Dakota School of Medicine, 1000 W. Fourth St., Yankton, SD 57078.
Our study was designed to define normal values of serum glucose levels during the first 48 hours of life in well term neonates cared for according to present standards, and to compare serum glucose levels in breast-fed and bottle-fed infants during this period. METHODS The study sample consisted of 113 mothers and their 114 term infants (37 to 42 weeks gestation) delivered at I
AGA LGA SGA
Appropriate for gestational age Large for gestational age Small for gestational age
The University of Iowa Hospital,s and Clinics between May 1982 and May 1983. Gestational age was estimated from available obstetric information and the neonatal Ballard examination? The prenatal course of each woman enrolled in the study was free of diabetes mellitus, hypertension, prenatal infection, fetal distress, or use of any drug during the last 4 weeks of pregnancy that might affect carbohydrate metabolism (corticosteroids,/3-mimetic tocolytic agents, theophylline, more than 1 oz alcohol per day,
120
Clinical and laboratory observations
The Journal of Pediatrics January 1987
Table I. Characteristics of patient samples Breast-fed (n = 6 4 ) n
Maternal IV dextrose Small for gestational age Large for gestational age Cesarean section
Gestational age 1-Minute Apgar score 5-Minute Apgar score Birth weight* Weight loss1Microhematocrit
27 4 6 4
Bottle-fed (n = 50) %
42.0 6.2 9.4 6.2
n
27 1 4 1
Combined groups (n = 114) %
54.0 2.0 8.0 2.0
n
52 5 10 5
%
45.6 4.4 8.8 4.4
Mean
SD
Mean
$D
Mean
SD
39.7 7.5 8.7 3545 212 55
1.2 1.0 0.5 432 67 6.3
39.9 7.3 8.7 3423 96 56
1.5 1.6 0.6 438 63 6.3
39.8 7.4 8.7 3491 160 56
1.3 1.3 0.6 437 87 6.3
*P <0.05, breast-fedvs bottle-fed,Student'sunpairedt test. I"P <0.001, breast-fedvs bottle-fed,Student'sunpairedt test.
or more than 6 cups coffee per day). After selection of the mother, all neonates were retained in the study unless complications precluded further study, the infant was dismissed from the hospital, or the parent(s) chose to withdraw the infant. Informed consent was obtained from the parent(s). A minimum sample size of 35 for each group was calculated based on 95% confidence limits for the mean of 5 mg/dL and a standard deviation of 15 mg/dL. 6,7 A planned sample size of 50 allowed for a 30% dropout rate. Parental feeding preference determined the assignment of the subjects to bottle-fed or breast-fed groups. Feedings were scheduled at postnatal ages of 2 hours, 5 to 6 hours and subsequently according to the nursery routine (1:00, 5:00, and 9:00 AM and 1:00, 5:00, and 9:00 I'M). Daily weights of each subject were recorded for later analysis. Starting with the first feeding at 2 hours of age, bottle-fed infants received 20 kcal/oz infant formula, and breast-fed infants were nursed; Breast-fed infants who were offered feedings of dextrose water or formula after nursing were not given these extra feedings before specimen collection. Although the investigators made no decisions regarding patient care, the serum glucose levels were available to the responsible physician. A single maternal serum glucose level was measured in a specimen collected by venipuncture within 30 minutes of delivery or from the episiotomy incision immediately before delivery (using a sterile plastic syringe). The infant blood samples were cbtained from the cord at birth (mixed arterial and venous) and by heel stick at the following predetermined times: 1 hour _+ 10 minutes, 2 hours + 10
minutes, 3 to 4 hours after first feeding, and 3 to 4 hours after three subsequent feedings (i.e., at ages 10 to 18 hours, 20 to 28 hours, and 44 to 52 hours). A heel stick microhematocrit value was determined within the first 48 hours of life. Serum glucose levels were measured (Beckman Glucose Analyzer 2, Beckman Instruments Inc., Fullerton, Calif.) following the standard procedure. The characteristics and serum glucose levels of the two groups were compared statistically using the Student t test, chi-square test, Fisher exact probability test, and Statistical Analysis Systems general linear models procedure (SAS-GLM). Mean, standard deviation, and selected percentiles of serum glucose values were determined. RESULTS Two infants sustained complications precluding further study, two were dismissed before completion of the study, and 16 infants were withdrawn from the study by their parent(s) at various times in the study. Based on Iowa normal birth weight curves, the study sample included five (4.4%) small for gestational age infants and 10 (8.8%) large for gestational age infants? Significant differences were found between the breastfed and bottle-fed groups with respect to birth weight and weight loss. Breast-fed infants were larger and lost twice as much weight as bottle-fed infants during the study (Table I). Using SAS-GLM, statistically significant differences between the serum glucose levels of breast-fed and bottlefed groups were detected at 5 to 6 and 44 to 52 hours; however, the differences in means (4 and 3 mg/dL, respectively) are probably not clinically significant. With the two groups combined, the 5th percentile postnatal
Volume 110 Number 1
Clinical and laboratory observations
12 1
Table II. Maternal and neonatal serum glucose values (mg/dL): Combined groups Percentile
Maternal
Cord
1 Hr
2 Hr
5th 50th 95th
73 104 188
63 90 158
36 56 99
39 58 89
34 56 77
33 56 74
46 60 81
48 65 79
n Mean SD
97 112 37
110 97 29
113 60 18
107 61 15
105 56 11
102 56 12
101 61 10
92 64 10
serum glucose values did not exceed 40 mg/dL until 24 hours of age (Table II). From the combined data, 33 (29%) of 114 infants had at least one serum glucose concentration ~<40 mg/dL in the first 2 days of life; nine (7.9%) infants had two or more serum glucose levels --<40 mg/dL. When data obtained before the first feeding are excluded, 18 (16%) of 105 infants had at least one serum glucose value _<40 mg/dL; seven (6.7%) infants had two or more serum glucose values -<40 mg/dL. Nine (7.9%) of 114 infants had at least one serum glucose level -<30 mg/dL; one infant had two serum glucose values -<30 rag/alL. After the first feeding, five (4.8%) of 105 infants each had one serum glucose level -<30 mg/dL. The few infants who had symptoms had serum glucose values >40 mg/dL (i.e., their symptoms were caused by other problems). DISCUSSION The small but statistically significant differences in mean serum glucose levels between the breast-fed and bottle-fed infants at 5 to 6 and 44 to 52 hours may be related to differences in intake. The larger intake of the bottle-fed infants at the first feeding produces significantly higher postprandial insulin levels and insulin/glucagon ratios (Heck and Erenberg, unpublished data). The lower 3- to 4-hour postprandial serum glucose concentration in the bottle-fed infants may represent a slight "rebound" hypoglycemic effect. The mean intake of bottle-fed infants at 44 to 52 hours of age was 52 ml (SD 18 ml). We did not determine the intake of milk in the breast-fed infants; however, the amount of milk ingested by nursing infants has been shown to be relatively small during the first 2 days of life (mean 13.8 g, SD ~ 15 g in the second 24 hours of life). This low intake (25% to 30% that of bottle-fed infants) may account for the significantly lower 3- to 4-hour postprandial serum glucose levels in this group noted at 44 to 52 hours of age. 9 Serum glucose levels of the SGA and LGA infants were not statistically different from those of. t h e AGA infants. Because only five infants in this study were delivered by cesarean section, no conclusions can be drawn about data
5-6 Hr
10-14 Hr
20-28 Hr
44-52 Hr
from these subjects. However, one mother received a large amount of dextrose-containing solution intravenously prior to cesarean section delivery and had a serum glucose concentration of 276 mg/dL at the time of delivery. The infant's cord serum glucose level was 237 mg/dL, and by 2 hours of age decreased to 14 mg/dL. After a feeding, the infant's serum glucose level returned to normal, and remained normal through the rest of the study. Srinivasan et al. 4 recently published a similar study with somewhat different results. Their data from the serial group of neonates (blood samples at 0, 1, 2, and 3 hours) correspond closely to our initial data (0, 1, and 2 hours). However, data from their cross-sectional group (after 3 hours) demonstrated mean plasma glucose levels 5 to 10 mg/dL greater than our mean serum glucose values after the first feeding. Sixteen percent of our sample had at least one serum glucose level -<40 mg/dl between 3 and 52 hours of life, similar to the results of Sexson.3 Our study differed from Srinivasan's in study design, time of starting infant feedings, type and possibly amounts of feedings, method of calculating normal limits, incidence of mothers receiving intravenously administered dextrose, and sample size. After reviewing each of these variables, we concluded that most would not likely account for the different results. The variables that can affect the results, in order of our perception of their importance, are as follows: 1. The feeding method and quantity significantly affect serum/plasma glucose levels, as shown in our study. If infants in Srinivasan's study received larger or more frequent feedings of formula than our subjects, this may account for the difference in results. 2. Sampling time relative to feedings is very important, especially with bottle-feeding. For example, the postprandial serum glucose levels at 2 days of age peak on average 25 mg/dL above basal levels with bottle-feeding, but only 15 mg/dL above basal values with breast-feeding, and the elevated serum glucose levels also persist longer with bottle feeding (Heck and Erenberg, unpublished data). After the first feeding, Srinivasan collected samples "before (the) routine morning feeding." Depending on the feeding
122
Clinical and laboratory observations
schedule, this may have been less than 3 hours after feeding. 3. Larger sample size improves the reliability of normative data. Our sample sizes at corresponding neonatal ages were 1.5 to 2.5 times greater than Srinivasan's; however, we believe that this accounts for little of the difference between our results. 4. Inappropriate methods of calculating normal limits wil ! produce erroneous results. We used percentile calculations because of skewing of data at the earlier sampling times and adequate sample size for this method. Srinivasan used 95% confidence intervals of log-transformed data. The magnitude of the confidence interval is determined by sample standard deviation and sample size. This parameter represents the range within which the true population mean likely lies; it is not a valid estimate of population norms. We do not think that this difference contributes much to the differences in our data. On the basis of our findings in this study of a large sample of well term infants, we recommend that hypoglycemia in full-term infants be defined as a serum glucose concentration <30 m g / d L in the first day of life or <40 m g / d L in the second day of life. We thank Gary Leonardson, Ph.D., University of South Dakota School of Medicine, and John Lemke, Ph.D., University of Iowa Hospitals and Clinics, for assistance with statistical analysis; Horst Jordan for supplying the equipment and reagents; and all
The Journal of Pediatrics January 1987
the nurses in the Normal Newborn Nursery for their patience during this study. REFERENCES 1. Fanaroff AA, Martin RJ. Neonatal-perinatal medicine: diseases of the fetus and infant, 3rd ed. St. Louis: CV Mosby, 1983:849. 2. Wald M. Problems in metabolic adaptation: glucose, calcium, and magnesium. In: Klaus M, Fanaroff A, eds. Care of the high risk neonate. Philadelphia: WB Saunders, 1979:227. 3. Sexson WR. Incidence of neonatal hypoglycemia: a matter of definition. J PEDIATR 1984;105:149. 4. Srinivasan G, Pildes RS, Cattamanchi G, Voora S, Lillien LD. Plasma glucose values in normal neonates: a new look. J PEDIATR 1986;109:114. 5. Ballard J, Kazmier K, Driver M. A simplified assessment of gestational age. Pediatr Res 1977;11:374. 6. Brown BW, Hollander M. Statistics: a biomedical introduction. New York: John Wiley, 1977. 7. Baens GS, Lundeen E, Cornblath M. Studies of carbohydrate metabolism in the newborn infant. VI. Levels of glucose in blood in premature infants. Pediatrics 1963;31:580. 8. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. Am J Obstet Gynecol 1976;126:555. (Modification of curve from 38 to 42 weeks gestation by 749 infants born at the University of Iowa Hospitals and Clinics/unpublished.) 9. Houston M J, Howie PW, McNeilly AS. Factors affecting the duration of breast feeding. I. Measurement of breast milk intake in the first week of life. Early Human Dev 1983; 8:49.
Serum adrenal steroid levels in healthy full-term 3-day-old infants D o n n a Wiener, M.D., Jeffrey Smith, M.D., S t e p h e n D a h l e m , M.D., G r e g o r y Berg, B.S., a n d Thomas M o s h a n g , Jr., M.D. From the Divisions of Endocrinology/Diabetes and Neonatology, and Endocrine Science Laboratories, The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia
The diagnosis of congenital adrenal hyperplasia, a disorder involving a deficiency of one of several adrenal enzymes, is often based on an abnormally elevated serum concentraSupported in part by Grant 00240 from the National Institutes of Health. Submitted for publication June 26, 1986; accepted Sept. 22, 1986. Reprint requests: Thomas Moshang, Jr., M.D., Division of Endocrinology/Diabetes, The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA 19104.
tion of the steroid precursor. With each enzyme deficiency, the adrenal steroid just preceding the enzymatic block is most abnormal. 17-OH progesterone is most commonly elevated in 21 hydroxylase deficiency, but equally high
CAH ll-DC DHEA-S 17-OHP 17-OH preg 3-r
Congenital adrenal hyperplasia 11-Deoxycortisol Dehydroepiandrosterone sulfate 17-Hydroxyprogesterone 17-Hydroxypregnenolone 3-fl-Hydroxysteroid dehydrogenase
Metaproterenol in chronic lung disease
119
Pulmonary mechanics, chest x-ray and lung disease after mechanical ventilation in low birth weight infants. Acta Paediatr Scand 1980;69:761-70. 13. Moyland FMB, Shannon DC. Preferential distribution of lobar emphysema and atelectasis in bronchopulmonary dysplasia. Pediatrics 1979;63:130-4. 14. Engel S. Lung structure. Springfield, I11.:Charles C Thomas, 1962:14-36. 15. Tal A, Bar-Yishay E, Eyal F, et al. Lack of response of airway obstruction after mechanical ventilation in the newborn. Crit Care Med 1982;10:361-2.
Clinical and laboratory observations Serum glucose levels in term neonates during the first 48 hours of life Louis J. Heck, M.D., a n d A l l e n E r e n b e r g , M.D. From the Department of Obstetrics and Gynecology, University of South Dakota School of Medicine, Yankton, and the Department of Pediatrics, The University of Iowa Hospitals and Clinics, Iowa City
The most common definition of hypoglycemia used today was derived from whole blood glucose levels of infants grouped by birth weight, and later adjusted upward mathematically to approximate plasma or serum levels. I Since these early studies, the importance of gestational age has become apparent, newborn infants are more frequently breast-fed and fed earlier, and intravenously administered dextrose solution is often given to women during labor and delivery., With these changes in perinatal care, clinical experience has led some authors to recommended that newborn serum glucose levels should be maintained at > 4 0 mg/dL. 2 However, Sexson 3 recently showed that using this value results in a 20% incidence of diagnosis of hypoglycemia in full-term infants. Srinivasan et al. 4 recently described plasma glucose values in healthy neonates, indicating that after the first 3 hours of life plasma glucose concentrations are normally >40 mg/dL.
Received for publication March 28, 1986; accepted Aug. 19, 1986. Reprint requests: Louis J. Heck, M.D., Department of Obstetrics and Gynecology, University of South Dakota School of Medicine, 1000 W. Fourth St., Yankton, SD 57078.
Our study was designed to define normal values of serum glucose levels during the first 48 hours of life in well term neonates cared for according to present standards, and to compare serum glucose levels in breast-fed and bottle-fed infants during this period. METHODS The study sample consisted of 113 mothers and their 114 term infants (37 to 42 weeks gestation) delivered at I
AGA LGA SGA
Appropriate for gestational age Large for gestational age Small for gestational age
The University of Iowa Hospital,s and Clinics between May 1982 and May 1983. Gestational age was estimated from available obstetric information and the neonatal Ballard examination? The prenatal course of each woman enrolled in the study was free of diabetes mellitus, hypertension, prenatal infection, fetal distress, or use of any drug during the last 4 weeks of pregnancy that might affect carbohydrate metabolism (corticosteroids,/3-mimetic tocolytic agents, theophylline, more than 1 oz alcohol per day,
120
Clinical and laboratory observations
The Journal of Pediatrics January 1987
Table I. Characteristics of patient samples Breast-fed (n = 6 4 ) n
Maternal IV dextrose Small for gestational age Large for gestational age Cesarean section
Gestational age 1-Minute Apgar score 5-Minute Apgar score Birth weight* Weight loss1Microhematocrit
27 4 6 4
Bottle-fed (n = 50) %
42.0 6.2 9.4 6.2
n
27 1 4 1
Combined groups (n = 114) %
54.0 2.0 8.0 2.0
n
52 5 10 5
%
45.6 4.4 8.8 4.4
Mean
SD
Mean
$D
Mean
SD
39.7 7.5 8.7 3545 212 55
1.2 1.0 0.5 432 67 6.3
39.9 7.3 8.7 3423 96 56
1.5 1.6 0.6 438 63 6.3
39.8 7.4 8.7 3491 160 56
1.3 1.3 0.6 437 87 6.3
*P <0.05, breast-fedvs bottle-fed,Student'sunpairedt test. I"P <0.001, breast-fedvs bottle-fed,Student'sunpairedt test.
or more than 6 cups coffee per day). After selection of the mother, all neonates were retained in the study unless complications precluded further study, the infant was dismissed from the hospital, or the parent(s) chose to withdraw the infant. Informed consent was obtained from the parent(s). A minimum sample size of 35 for each group was calculated based on 95% confidence limits for the mean of 5 mg/dL and a standard deviation of 15 mg/dL. 6,7 A planned sample size of 50 allowed for a 30% dropout rate. Parental feeding preference determined the assignment of the subjects to bottle-fed or breast-fed groups. Feedings were scheduled at postnatal ages of 2 hours, 5 to 6 hours and subsequently according to the nursery routine (1:00, 5:00, and 9:00 AM and 1:00, 5:00, and 9:00 I'M). Daily weights of each subject were recorded for later analysis. Starting with the first feeding at 2 hours of age, bottle-fed infants received 20 kcal/oz infant formula, and breast-fed infants were nursed; Breast-fed infants who were offered feedings of dextrose water or formula after nursing were not given these extra feedings before specimen collection. Although the investigators made no decisions regarding patient care, the serum glucose levels were available to the responsible physician. A single maternal serum glucose level was measured in a specimen collected by venipuncture within 30 minutes of delivery or from the episiotomy incision immediately before delivery (using a sterile plastic syringe). The infant blood samples were cbtained from the cord at birth (mixed arterial and venous) and by heel stick at the following predetermined times: 1 hour _+ 10 minutes, 2 hours + 10
minutes, 3 to 4 hours after first feeding, and 3 to 4 hours after three subsequent feedings (i.e., at ages 10 to 18 hours, 20 to 28 hours, and 44 to 52 hours). A heel stick microhematocrit value was determined within the first 48 hours of life. Serum glucose levels were measured (Beckman Glucose Analyzer 2, Beckman Instruments Inc., Fullerton, Calif.) following the standard procedure. The characteristics and serum glucose levels of the two groups were compared statistically using the Student t test, chi-square test, Fisher exact probability test, and Statistical Analysis Systems general linear models procedure (SAS-GLM). Mean, standard deviation, and selected percentiles of serum glucose values were determined. RESULTS Two infants sustained complications precluding further study, two were dismissed before completion of the study, and 16 infants were withdrawn from the study by their parent(s) at various times in the study. Based on Iowa normal birth weight curves, the study sample included five (4.4%) small for gestational age infants and 10 (8.8%) large for gestational age infants? Significant differences were found between the breastfed and bottle-fed groups with respect to birth weight and weight loss. Breast-fed infants were larger and lost twice as much weight as bottle-fed infants during the study (Table I). Using SAS-GLM, statistically significant differences between the serum glucose levels of breast-fed and bottlefed groups were detected at 5 to 6 and 44 to 52 hours; however, the differences in means (4 and 3 mg/dL, respectively) are probably not clinically significant. With the two groups combined, the 5th percentile postnatal
Volume 110 Number 1
Clinical and laboratory observations
12 1
Table II. Maternal and neonatal serum glucose values (mg/dL): Combined groups Percentile
Maternal
Cord
1 Hr
2 Hr
5th 50th 95th
73 104 188
63 90 158
36 56 99
39 58 89
34 56 77
33 56 74
46 60 81
48 65 79
n Mean SD
97 112 37
110 97 29
113 60 18
107 61 15
105 56 11
102 56 12
101 61 10
92 64 10
serum glucose values did not exceed 40 mg/dL until 24 hours of age (Table II). From the combined data, 33 (29%) of 114 infants had at least one serum glucose concentration ~<40 mg/dL in the first 2 days of life; nine (7.9%) infants had two or more serum glucose levels --<40 mg/dL. When data obtained before the first feeding are excluded, 18 (16%) of 105 infants had at least one serum glucose value _<40 mg/dL; seven (6.7%) infants had two or more serum glucose values -<40 mg/dL. Nine (7.9%) of 114 infants had at least one serum glucose level -<30 mg/dL; one infant had two serum glucose values -<30 rag/alL. After the first feeding, five (4.8%) of 105 infants each had one serum glucose level -<30 mg/dL. The few infants who had symptoms had serum glucose values >40 mg/dL (i.e., their symptoms were caused by other problems). DISCUSSION The small but statistically significant differences in mean serum glucose levels between the breast-fed and bottle-fed infants at 5 to 6 and 44 to 52 hours may be related to differences in intake. The larger intake of the bottle-fed infants at the first feeding produces significantly higher postprandial insulin levels and insulin/glucagon ratios (Heck and Erenberg, unpublished data). The lower 3- to 4-hour postprandial serum glucose concentration in the bottle-fed infants may represent a slight "rebound" hypoglycemic effect. The mean intake of bottle-fed infants at 44 to 52 hours of age was 52 ml (SD 18 ml). We did not determine the intake of milk in the breast-fed infants; however, the amount of milk ingested by nursing infants has been shown to be relatively small during the first 2 days of life (mean 13.8 g, SD ~ 15 g in the second 24 hours of life). This low intake (25% to 30% that of bottle-fed infants) may account for the significantly lower 3- to 4-hour postprandial serum glucose levels in this group noted at 44 to 52 hours of age. 9 Serum glucose levels of the SGA and LGA infants were not statistically different from those of. t h e AGA infants. Because only five infants in this study were delivered by cesarean section, no conclusions can be drawn about data
5-6 Hr
10-14 Hr
20-28 Hr
44-52 Hr
from these subjects. However, one mother received a large amount of dextrose-containing solution intravenously prior to cesarean section delivery and had a serum glucose concentration of 276 mg/dL at the time of delivery. The infant's cord serum glucose level was 237 mg/dL, and by 2 hours of age decreased to 14 mg/dL. After a feeding, the infant's serum glucose level returned to normal, and remained normal through the rest of the study. Srinivasan et al. 4 recently published a similar study with somewhat different results. Their data from the serial group of neonates (blood samples at 0, 1, 2, and 3 hours) correspond closely to our initial data (0, 1, and 2 hours). However, data from their cross-sectional group (after 3 hours) demonstrated mean plasma glucose levels 5 to 10 mg/dL greater than our mean serum glucose values after the first feeding. Sixteen percent of our sample had at least one serum glucose level -<40 mg/dl between 3 and 52 hours of life, similar to the results of Sexson.3 Our study differed from Srinivasan's in study design, time of starting infant feedings, type and possibly amounts of feedings, method of calculating normal limits, incidence of mothers receiving intravenously administered dextrose, and sample size. After reviewing each of these variables, we concluded that most would not likely account for the different results. The variables that can affect the results, in order of our perception of their importance, are as follows: 1. The feeding method and quantity significantly affect serum/plasma glucose levels, as shown in our study. If infants in Srinivasan's study received larger or more frequent feedings of formula than our subjects, this may account for the difference in results. 2. Sampling time relative to feedings is very important, especially with bottle-feeding. For example, the postprandial serum glucose levels at 2 days of age peak on average 25 mg/dL above basal levels with bottle-feeding, but only 15 mg/dL above basal values with breast-feeding, and the elevated serum glucose levels also persist longer with bottle feeding (Heck and Erenberg, unpublished data). After the first feeding, Srinivasan collected samples "before (the) routine morning feeding." Depending on the feeding
122
Clinical and laboratory observations
schedule, this may have been less than 3 hours after feeding. 3. Larger sample size improves the reliability of normative data. Our sample sizes at corresponding neonatal ages were 1.5 to 2.5 times greater than Srinivasan's; however, we believe that this accounts for little of the difference between our results. 4. Inappropriate methods of calculating normal limits wil ! produce erroneous results. We used percentile calculations because of skewing of data at the earlier sampling times and adequate sample size for this method. Srinivasan used 95% confidence intervals of log-transformed data. The magnitude of the confidence interval is determined by sample standard deviation and sample size. This parameter represents the range within which the true population mean likely lies; it is not a valid estimate of population norms. We do not think that this difference contributes much to the differences in our data. On the basis of our findings in this study of a large sample of well term infants, we recommend that hypoglycemia in full-term infants be defined as a serum glucose concentration <30 m g / d L in the first day of life or <40 m g / d L in the second day of life. We thank Gary Leonardson, Ph.D., University of South Dakota School of Medicine, and John Lemke, Ph.D., University of Iowa Hospitals and Clinics, for assistance with statistical analysis; Horst Jordan for supplying the equipment and reagents; and all
The Journal of Pediatrics January 1987
the nurses in the Normal Newborn Nursery for their patience during this study. REFERENCES 1. Fanaroff AA, Martin RJ. Neonatal-perinatal medicine: diseases of the fetus and infant, 3rd ed. St. Louis: CV Mosby, 1983:849. 2. Wald M. Problems in metabolic adaptation: glucose, calcium, and magnesium. In: Klaus M, Fanaroff A, eds. Care of the high risk neonate. Philadelphia: WB Saunders, 1979:227. 3. Sexson WR. Incidence of neonatal hypoglycemia: a matter of definition. J PEDIATR 1984;105:149. 4. Srinivasan G, Pildes RS, Cattamanchi G, Voora S, Lillien LD. Plasma glucose values in normal neonates: a new look. J PEDIATR 1986;109:114. 5. Ballard J, Kazmier K, Driver M. A simplified assessment of gestational age. Pediatr Res 1977;11:374. 6. Brown BW, Hollander M. Statistics: a biomedical introduction. New York: John Wiley, 1977. 7. Baens GS, Lundeen E, Cornblath M. Studies of carbohydrate metabolism in the newborn infant. VI. Levels of glucose in blood in premature infants. Pediatrics 1963;31:580. 8. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. Am J Obstet Gynecol 1976;126:555. (Modification of curve from 38 to 42 weeks gestation by 749 infants born at the University of Iowa Hospitals and Clinics/unpublished.) 9. Houston M J, Howie PW, McNeilly AS. Factors affecting the duration of breast feeding. I. Measurement of breast milk intake in the first week of life. Early Human Dev 1983; 8:49.
Serum adrenal steroid levels in healthy full-term 3-day-old infants D o n n a Wiener, M.D., Jeffrey Smith, M.D., S t e p h e n D a h l e m , M.D., G r e g o r y Berg, B.S., a n d Thomas M o s h a n g , Jr., M.D. From the Divisions of Endocrinology/Diabetes and Neonatology, and Endocrine Science Laboratories, The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia
The diagnosis of congenital adrenal hyperplasia, a disorder involving a deficiency of one of several adrenal enzymes, is often based on an abnormally elevated serum concentraSupported in part by Grant 00240 from the National Institutes of Health. Submitted for publication June 26, 1986; accepted Sept. 22, 1986. Reprint requests: Thomas Moshang, Jr., M.D., Division of Endocrinology/Diabetes, The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA 19104.
tion of the steroid precursor. With each enzyme deficiency, the adrenal steroid just preceding the enzymatic block is most abnormal. 17-OH progesterone is most commonly elevated in 21 hydroxylase deficiency, but equally high
CAH ll-DC DHEA-S 17-OHP 17-OH preg 3-r
Congenital adrenal hyperplasia 11-Deoxycortisol Dehydroepiandrosterone sulfate 17-Hydroxyprogesterone 17-Hydroxypregnenolone 3-fl-Hydroxysteroid dehydrogenase