| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Pediatrics, University of Oxford (A.L.O.-S., S.J.H., C.J.A., A.R.W., D.B.D.), OX3 9DU Oxford; the Department of Surgery, University of Bristol (J.M.P.H.), BSZ 8HW Bristol; the Department of Diabetes and Endocrinology, Radcliffe Infirmary (D.R.M.), OX2 6HE Oxford; and the Department of Medicine, University College London Medical School, Whittington Hospital (V.M.-A., J.S.Y.), N19 3UA London, United Kingdom
Address all correspondence and requests for reprints to: Dr. A. L. Ogilvy-Stuart, Neonatal Unit, The Rosie Hospital, Addenbrooke’s NHS Trust, Robinson Way, Cambridge, United Kingdom CB2 2SW. E-mail: amanda.ogilvy-stuart{at}msexc.addenbrookes.anglox.nhs.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
IGF-I levels correlated with birth weight and length (r2 = 0.47; P = 0.004, and r2 = 0.5; P = 0.0005, respectively, after allowing for gestation), whereas mean GH levels were negatively related to birth size (r2 = -0.18; P = 0.04 and r2 = -0.2; P = 0.03 for weight and length, respectively). No direct relationship between mean GH levels and IGF-I was identified. IGF-I levels were higher in appropriate for gestational age (AGA; mean ± SD, 82 ± 61 ng/mL) than in SGA (34 ± 22 ng/mL; P = 0.03) babies. Baseline (mean ± SD, 25.9 ± 11.9), mean (33.9 ± 14.0), and peak (45.0 ± 18.1 µg/L) GH levels were higher in SGA than in AGA babies [17.1 ± 8.2 (P = 0.04), 22.5 ± 10.4 (P = 0.03), and 30.7 ± 15.4 µg/L (P = 0.04), respectively]. Mean IGFBP-1 levels were also higher in SGA than AGA babies (157.4 ± 90.7 vs. 62.7 ± 43.8 ng/mL; P = 0.01). A positive correlation was identified between changes in insulin and coincident pulses of GH (r = 0.147; P < 0.01), whereas there was an inverse relationship between insulin and IGFBP-1, with a lag time 120 min (r = -0.33; P < 0.0001).
In conclusion, these studies indicate that the GH-IGF-I axis is closely related to feeding in the newborn. In SGA babies, low IGF-I and elevated IGFBP-1 reflect the slow growth, but elevated GH and rapid GH pulsatility may be a signal for lipolysis.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Disruption of the insulin-like growth factor (IGF) genes has demonstrated the importance of IGF-I and IGF-II and their receptors in fetal growth in animals (10, 11) and humans (12). In the fetus, insulin also has a central role in regulating growth (13, 14). In postnatal life, nutrition, insulin, and IGF-I still largely regulate growth. Although the effects of GH on growth do not become prominent until about 18 months of age, GH levels are high in the fetus and newborn compared with those in later childhood and in adults (15, 16, 17, 18, 19, 20, 21, 22). An alternative metabolic role for GH during this period has been suggested (19, 23).
In older children and adults, the regulation of IGF-I is complex. Although the principal regulators of IGF-I are GH and nutrition, its bioavailability and bioactivity are determined by binding proteins. In older subjects, the majority of IGF-I is bound to IGF-binding protein-3 (IGFBP-3) in a ternary complex that retains IGF-I in the circulation. IGFBP-3 levels are low in the newborn, particularly in SGA babies. The ratio of IGF-I to IGFBP-3 and the increase in IGFBP-2 suggest that the bioavailability of IGF-I is increased in the fetus (24) and in infants (25). Levels of IGFBP-1 may also be important, as it is a potent inhibitor of IGF bioactivity. In older subjects, it is regulated by insulin (26, 27), but there are few data for the newborn.
The aim of this study was to determine the dynamic relationships among insulin, GH, IGFBP-1, and feeding and thus characterize the normal physiology in the newborn. These studies were made possible by the development of a unique microsampling system that has overcome the practical and ethical difficulties of serial blood sampling in the newborn (28). Comparison of hormone levels between SGA and appropriate for gestational age (AGA) babies was also undertaken to explore the hypothesis that variations in substrate availability or insulin secretion in utero determine changes in the GH-IGF-IGFBP axis, and thus growth, in newborn babies.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Twenty-six babies, patients in the Neonatal Unit at the John
Radcliffe Hospital in Oxford, UK [15 boys and 11 girls; 13 SGA,
i.e. weight and length <10th percentile, by
Gairdner-Pearson charts (29, 30)], underwent a 12-h hormone profile at
1–13 days of postnatal age (median, 4.5 days). The median gestational
age of the SGA group was 33 (range, 28–36) weeks, and that of the AGA
group was 32 (range, 27–40) weeks. Their birth weights were between
0.822–3.27 kg. Two SGA babies and 2 AGA baby were receiving total
parenteral nutrition at the time of the study. Four SGA and 2 AGA
babies were receiving a combination of enteral feeds and total
parenteral nutrition; all other babies were receiving enteral milk
feeds (90–180 mL/kg·day) at 1- to 3-h intervals. All babies were
clinically well at the time of the study. Six infants at risk of
infection were receiving parenteral antibiotics, but the subsequent
results of the infection screen were negative. All babies were
normoglycemic and euthyroid. Ethical approval from the central Oxford
research ethics committee and informed parental consent were obtained.
A summary of the patient characteristics in the 2 groups of babies is
given in Table 1
.
|
Using a microsampling technique in the 26 study babies, samples of venous blood (40 µL) were collected through a 25-gauge cannula into 1:4 heparinized saline and aliquoted at 10-min intervals over a 12-h period (28). Validation of the system has shown that the error attributable to dilution is less than 4%. The cannula was sited in a peripheral vein, either for the administration of antibiotics or when a blood sample was required as part of routine care, and was left in situ for the duration of the study period. The total blood volume taken over the 12 h was less than 2.9 mL. GH levels were measured at 10-min intervals, whereas insulin and IGFBP-1 were measured at 20-min intervals. A single undiluted sample (0.5 mL) was taken for IGF-I levels at the time of the profile.
Assays
GH. Plasma GH concentrations in the diluted plasma samples were determined by an immunoradiometric assay (NETRIA, St. Bartholomew’s Hospital, London, UK). Fifty microliters of the diluted plasma samples were added to 350 µL working assay buffer [10 mL 0.54 mol/L phosphate buffer (pH 7.4), 5 mL 10% Tween-20, and 1 g BSA up to 100 mL distilled water]. Iodinated (125I) anti-GH was reconstituted in 1 mL distilled water and diluted in 5 mL working assay buffer to give 100,000 cpm. Fifty microliters of the GH tracer and 50 µL solid phase were added to the sample and incubated overnight on a rotator at room temperature. One milliliter of 0.1% Triton X-100 was added to each tube, spun at 3,000 rpm for 15 min, and decanted, and the latter stage was repeated. Standards ranging from 0.125–100 µg/L were reconstituted in 2 mL distilled water and similarly used in a volume of 50 µL in the immunoradiometric assay. The sensitivity of the assay was 0.05 µg/L, and intraassay coefficients of variation were 14.5%, 8.9%, 4.2%, and 5.7% at concentrations of 0.25, 1.25, 2.5, and 125 µg/L, respectively. The working range of the assay reproducible on diluted samples was between 0.25–6.25 µg/L. Plasma GH levels between 2.0–100 µg/L can be detected in the samples obtained using the microsampling technique.
IGF-I. Plasma IGF-I levels were determined in undiluted plasma samples by RIA after acid-ethanol extraction as previously described (27). The maximum sensitivity of the assay was 2 ng/mL. Inter- and intraassay coefficients of variation were less than 10 and 8%, respectively, at analyte levels of 600, 240, and 40 ng/mL.
IGFBP-1. IGFBP-1 was measured by RIA (31). The minimum detection limit of the assay was 5 ng/mL. The intraassay coefficients of variation were 6.1% at 22 ng/mL, 4.5% at 50 ng/mL, and 3.7% at 105 ng/mL. The interassay coefficients of variation were 10.6% at 22 ng/mL, 8.8% at 50 ng/mL, and 7.4% at 105 ng/mL.
Insulin. Insulin was measured using an amplified, end-point enzymoimmunometric assay for specific measurement of human insulin. Recovery of insulin from spiked plasma was 96% (range, 88–105%) (32). The assay detected down to 3 pmol/L, with intra- and interassay coefficients of variation of 7.9% and 14.3%, respectively.
Analysis
To determine the levels of undiluted hormone in plasma, the concentration in the diluted plasma was multiplied by the dilution factor, calculated according to the venous hematocrit expressed as a fraction of 1: plasma dilution factor = [(4 - hematocrit)/(1 - hematocrit)], when whole blood is diluted 1:4.
The GH data were analyzed using the Pulsar peak detection program that identifies the baseline and then uses defined criteria to recognize peaks in the data array (33). Variations in hormone levels are associated with variations in the assay, fluctuations in the baseline, and variations in the pulse profiles themselves. The method screens out long term changes and then uses a moving average to calculate the baseline. Peaks are identified as individual subseries elevated above the baseline, of duration n. All the points have a magnitude at least G(n), where the values of G are cut-off criteria based on the width of the peak. The G(n) are calculated using statistical methods. The nonpeak values, or noise, are taken as being random. A probability distribution is taken that fits this random series, and Bonferroni confidence intervals are used to identify the upper bounds of the noise.
Fourier transformation, serial array averaging, and cross-correlation were performed on the first 17 babies studied (10 AGA and 7 SGA). Pulse frequency characteristics were determined using the Fourier transform, which defines the dominant and subdominant harmonics together with an estimate of power and amplitude (34, 35). The power of each oscillation is displayed as a histogram of power vs. frequency (power spectrum). These data can be pooled using parametric statistics to yield spectra that are composite for the group data.
Postprandial changes in GH levels were analyzed using serial array averaging that determines the change in GH from a fixed point within a profile (36). Here the fixed point is the time of a feed. By combining data from a number of feeds, other factors affecting GH levels randomly are averaged out. Postprandial changes in GH were averaged at 10-min intervals over 90 min after 58 feeds given to 13 infants. Four infants were excluded from analysis because they were fed less frequently than every 2 h or were receiving parenteral nutrition.
Cross-correlation was used to determine the interrelationships between different hormones (37). This is an iterative technique for establishing whether there are statistically coincident recurring waveforms (of any shape) within a data array. One data array is serially correlated against another with progressive step changes in the time relationship between the data. The result is dependent both on the relative amplitude of such waves or pulses (i.e. whether large pulses of one array are associated with large pulses of the other) and on the phase relationships between the arrays (i.e. when one data array is time-lagged with respect to the other). It is independent of the absolute concentration of hormone and is unaffected by the regularity or irregularity of waveform (e.g. a large prolonged secretory episode in one array mirrored by a similar episode in the other will correlate well). The correlation coefficient is at its zenith at a time lag equal to the mean phase difference between the hormone profiles. Cross-correlation yields an unbiased estimate of this phase difference and an assessment of its significance. These correlograms can be pooled (using Fisher’s Z transformation) to obtain an estimate of the overall likelihood of the significance and phase relationships of all available data.
The Fourier transform of GH and postprandial GH changes in a subset of these babies has previously been reported (28).
Data are expressed as the mean ± SD unless otherwise stated. Comparison of mean hormone levels over the 12-h period between SGA and AGA babies were made using the Mann-Whitney U test. Differences between IGF-I values between SGA and AGA babies was determined using Student’s t test. A multiple regression model, including sex, gestation, and day of sampling, was used to determine relationships between size and IGF-I and GH levels.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
) and birth length
(r2 = 0.5; P = 0.0005). Mean GH levels over
the 12-h sampling period were high (median, 27.2; range, 11.3–63.8
µg/L) and, using a similar multivariate model, showed a weak negative
correlation with birth weight (r2 = -0.18;
P = 0.04; Fig. 2) and
birth length (r2 = -0.2; P = 0.03). No
direct correlation between GH and IGF-I was observed.
|
|
In all babies, including those receiving continuous iv nutrition,
independent pulsatile secretion of GH, insulin, and IGFBP-1 was
demonstrated. The representative profile of GH, insulin, and IGFBP-1
levels in an AGA, 34-week gestation baby studied at 1 day of age
demonstrates the relationships between hormones and feeds that was
identified (Fig. 3).
|
). The average postprandial change in GH
at 60 min was 9 µg/L.
|
),
suggesting that both GH and insulin increase in relation to feeds.
Coincident GH and insulin pulses were also observed when GH pulses were
lagged by 180 min with respect to insulin pulses (r = 0.232;
P < 0.0001), again perhaps reflecting feed-related
pulsatility, as the feed interval ranged from 60–180 min.
|
). These interrelationships are evident
in the representative profile (Fig. 3
).
|
IGF-I levels were significantly lower in SGA than in AGA babies
(SGA mean, 34 ± 22 ng/mL; AGA mean, 82 ± 61 ng/mL;
P = 0.03). Pulsar analysis of GH profiles identified
significantly higher baseline (25.9 ± 11.9 vs.
17.1 ± 8.2 µg/L; P = 0.04), mean (33.9 ±
14.0 vs. 22.5 ± 10.4 µg/L; P =
0.03), and peak (45.0 ± 18.1 vs. 30.7 ± 15.4
µg/L; P = 0.04) GH levels in SGA compared with AGA
babies (Table 2). There was no
significant difference in pulse amplitude or frequency attributes
between the two groups as detected by Pulsar, but Fourier
transformation indicated shorter pulse periodicities of 140 min in SGA
babies compared with a pulse frequency of 180 min in the AGA babies
(Fig. 7).
|
|
), yet there was no
significant difference in mean insulin levels during the 12-h sample
between the groups (mean insulin in AGA group, 40.5 ± 28.4
pmol/L; mean insulin in SGA group, 51.7 ± 48.7 pmol/L). |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The role of insulin in fetal growth has been clearly identified. Body weight at birth is related to the amount of functioning pancreatic tissue, with hyperinsulinemic babies being macrosomic and SGA babies having reduced amounts of pancreatic tissue (13). Insulin may augment fetal growth by stimulating the production of IGF-I (50, 51). The umbilical venous plasma insulin concentration and the fetal insulin/glucose ratio increase exponentially with gestation, reflecting maturation of the endocrine pancreas (14). Some SGA babies are relatively hypoglycemic and hypoinsulinemic; the degree of hypoinsulinemia correlates with the degree of hypoglycemia, presumably reflecting chronically diminished glycogen stores. This suggests a degree of ß-cell dysfunction in these fetuses (14, 52). There is a positive correlation between birth weight and umbilical vein insulin levels (45). Yet in our subjects, aged 1–13 days, we were unable to demonstrate any relationship between insulin levels and birth weight. In fact, insulin levels were higher in SGA babies, although differences did not reach statistical significance. This may be related to sample size or inability to detect postprandial rises with 20-min sampling despite the sensitive assay. Alternatively, this may reflect the very early development of insulin resistance, as recently suggested in neonatal rats (53), in older children born SGA (54), and in adults born SGA (55)
Insulin deficiency is associated with elevated IGFBP-1 levels in older children (27), as insulin regulates IGFBP-1 synthesis in the hepatocyte (56). Cord blood IGFBP-1 levels correlate inversely with birth size (57, 58). We were not able to detect any relationship between IGFBP-1 levels at 1–13 days and birth weight. However, we were able to confirm an inverse relationship between insulin and IGFBP-1 levels. Changes in IGFBP-1 occurred 120 min after changes in insulin as in older subjects (59). The relatively higher levels of IGFBP-1 that we noted in the SGA babies may result in an inhibitory effect on IGF-I bioactivity (27). The observations that IGFBP-1 levels were higher in SGA babies, but insulin levels were similar to those in AGA babies, may suggest dysregulation of IGFBP-1 or that IGF-I may be more important than insulin in the regulation of IGFBP-1 in the newborn. The association between IGFBP-1 and GH was also unexpected and requires further study.
As previously reported, GH levels are high during the neonatal period, and this may result from a lack of negative feedback from relatively low levels of circulating IGF-I, particularly in SGA infants (60). We were able to detect a weak relationship between GH and birth weight, but there was no direct relationship between GH and IGF-I. Differences in free levels of IGF-I secondary to changes in IGFBP-1 may be a confounding variable in determining feedback drive for GH secretion. Overall baseline, mean, and peak GH levels were higher than those observed in older children. Furthermore, the baseline, mean, and peak values for GH were higher in the SGA than in the AGA infants, as reported by other investigators using single GH estimates (23, 61, 62). The role of elevated levels of GH in the fetus and in infancy is unclear. Although GH receptors are present in fetal life (63), studies in animals (64) and human fetuses (65) suggest that GH does not play a major role in fetal growth. An alternative metabolic role for the high GH levels has been proposed (19, 23). They may induce insulin resistance and thus protect the fetal brain from hypoglycemia, whereas the lipolytic effects could provide alternative fuels for metabolism (66).
Pulsatile GH secretion has previously been described in both term (67, 68, 69) and preterm (22, 67, 69) infants and was evident in all 26 babies who we studied. Previous studies indicate that preterm babies had higher pulse amplitudes but similar pulse frequency compared with term babies (67). Estimates of pulse frequency have ranged between 2.5–9.9/12 h. Our data indicate a pulse frequency per 12 h of 5.1 in the SGA babies and 4.0 in the AGA babies, thus falling within this range. A pulse periodicity of 180 min observed in AGA babies is similar to previous reports in older children and adults (70). The shorter pulse periodicity of 140 min in the SGA infants with elevated baseline GH levels is similar to that observed during fasting (71). Miller et al. (22) suggested that pulse frequency was greater in the first 48 h, but this is unlikely to account for the higher GH levels we noted in SGA infants, as the mean age at the time of sampling was greater in the SGA than in the AGA group (6.7 vs. 4.5 days), and mean gestational ages were similar in the SGA and AGA groups (32.3 vs. 33.6 weeks). The similarities to pulse characteristics during fasting (72) may be a signal for lipolysis in SGA infants. This needs to be confirmed with a larger sample size.
A novel finding of this study is the relationship among GH pulsatility, feeds, and insulin secretion. A clear postprandial rise in GH and the coincident pulses of GH and insulin suggest that the stimulus for both may be related to feeds. GH pulsatility did not appear to be entrained on postprandial sleep patterns, but further studies will be needed to examine this in more detail. Adrian et al. (19) also observed a relationship between feeds and GH secretion in term infants, but only noted a significant postprandial rise after 13–24 days in preterm infants. However, that study was based on cross-sectional data of GH measurements in single blood samples from 248 babies at variable time relationships to feeds. Our observations based on repeated sampling suggest that GH release is more closely linked to feeding in the newborn than in older children and adults. The mechanism is unknown. It may relate to a direct effect of insulin or to other secretagogues related to feeds (such as changes in amino acids or free fatty acids). A paradoxical rise in GH during hyperglycemia has been observed in the first 6 days of life, which is most pronounced in preterm infants (73). It was suggested that the increase in GH induced by a glucose load could be a signal for protein synthesis.
In conclusion, we provide data indicating that the secretion of GH, IGFBP-1, and insulin are closely related and may be entrained by feeding in the newborn. The higher GH levels and increased frequency of GH pulsatility may indicate a role for GH in the regulation of lipolysis. Persisting elevations in IGFBP-1 and reduced IGF-I levels in SGA infants may determine the rate of subsequent catch-up growth.
|
Acknowledgments |
|---|
Received February 23, 1998.
Revised June 10, 1998.
Accepted June 22, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Kelly, R. Tang, S. Becker, and C. A. Stanley Poor Specificity of Low Growth Hormone and Cortisol Levels During Fasting Hypoglycemia for the Diagnoses of Growth Hormone Deficiency and Adrenal Insufficiency Pediatrics, September 1, 2008; 122(3): e522 - e528. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Filippi, M. Pezzati, C. Poggi, S. Rossi, A. Cecchi, and C. Santoro Dopamine versus dobutamine in very low birthweight infants: endocrine effects Arch. Dis. Child. Fetal Neonatal Ed., September 1, 2007; 92(5): F367 - F371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Dunger, C. J. Petry, and K. K. Ong Genetics of Size at Birth Diabetes Care, July 1, 2007; 30(Supplement_2): S150 - S155. [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Hyatt, H. Budge, D. Walker, T. Stephenson, and M. E. Symonds Effects of maternal parity and late gestational nutrition on mRNA abundance for growth factors in the liver of postnatal sheep Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2007; 292(5): R1934 - R1942. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. De Blasio, K. L. Gatford, I. C. McMillen, J. S. Robinson, and J. A. Owens Placental Restriction of Fetal Growth Increases Insulin Action, Growth, and Adiposity in the Young Lamb Endocrinology, March 1, 2007; 148(3): 1350 - 1358. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. De Blasio, K. L. Gatford, J. S. Robinson, and J. A. Owens Placental restriction of fetal growth reduces size at birth and alters postnatal growth, feeding activity, and adiposity in the young lamb Am J Physiol Regulatory Integrative Comp Physiol, February 1, 2007; 292(2): R875 - R886. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Roelfsema, A. J. Gunn, B. H. Breier, J. S. Quaedackers, and L. Bennet The Effect of Mild Hypothermia on Insulin-like Growth Factors After Severe Asphyxia in the Preterm Fetal Sheep Reproductive Sciences, May 1, 2005; 12(4): 232 - 237. [Abstract] [PDF]
|
|
|
|
 |
|
 S. Tenhola, P. Halonen, J. Jaaskelainen, and R. Voutilainen Serum markers of GH and insulin action in 12-year-old children born small for gestational age Eur. J. Endocrinol., March 1, 2005; 152(3): 335 - 340. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. P. Geary, P. J. Pringle, C. H. Rodeck, J. C. P. Kingdom, and P. C. Hindmarsh Sexual Dimorphism in the Growth Hormone and Insulin-Like Growth Factor Axis at Birth J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3708 - 3714. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Bazaes, T. E. Salazar, E. Pittaluga, V. Pena, A. Alegria, G. Iniguez, K. K. Ong, D. B. Dunger, and M. V. Mericq Glucose and Lipid Metabolism in Small For Gestational Age Infants at 48 Hours of Age Pediatrics, April 1, 2003; 111(4): 804 - 809. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Kajantie, L. Dunkel, E.-M. Rutanen, M. Seppala, R. Koistinen, A. Sarnesto, and S. Andersson IGF-I, IGF Binding Protein (IGFBP)-3, Phosphoisoforms of IGFBP-1, and Postnatal Growth in Very Low Birth Weight Infants J. Clin. Endocrinol. Metab., May 1, 2002; 87(5): 2171 - 2179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Ogilvy-Stuart, M. A. Soos, S. J. Hands, M. Y. Anthony, D. B. Dunger, and S. O'Rahilly Hypoglycemia and Resistance to Ketoacidosis in a Subject without Functional Insulin Receptors J. Clin. Endocrinol. Metab., July 1, 2001; 86(7): 3319 - 3326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Westwood, J.M. Gibson, S. R. Sooranna, S. Ward, J. P. Neilson, and R. Bajoria Genes or placenta as modulator of fetal growth: evidence from the insulin-like growth factor axis in twins with discordant growth Mol. Hum. Reprod., April 1, 2001; 7(4): 387 - 395. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||
, SGA; 
, AGA.
, Feed.
