(
J
Tokyo Wom Med CoH) 55 (10・11)971-978 (1985))69
A Research Report Supported by
I
t
oe Okamoto AwardINSULIN-LIKE GROWTH FACTOR-I AND CPR LEVELS
IN THE UMBILICAL CORD BLOOD OF NEWBORNS
FROM DIABETIC MOTHERS
Yasue OMORI
，
Satomi MINEI，
Meimi SHIMIZU，
Keiko AZUMA，
Rima AKIHISA and Yukimasa HIRAT ADiabetes Center (Director: Prof.Yukimasa HIRAT A) Tokyo Women's Medical College
Kae W AKAI and Toshio TSUSHIMA Institute of C1inical Endocrinology Tokyo Women's Medical College (Received 14th August， 1985) Abstract Fetal hyperinsulinemia has been implicated in the macrosomia of infants born to diabetic mothers. We measu陀dInsulin-like Growth Factor (IGF)司I and C-peptide immunoreactivity (CPR) levels in umbilical
∞
rd blood obtained from 62 newborns born to diabetic mothers. Out of 62 babies， 45 showed normal birth weight (appropriate for date; AFD)， 12 were heavy for date (HFD) and the remaining 5 were small for date (SFD). Cord blood CPR levels were significantly elevated in infants of diabetic mothers compared to those in control babies of non-diabetic mothers and the CPR levels showed a positive correlation with birth weight. IGF-I levels also correlated with birth weight. Both CPR and IGF-I levels in umbilical cord blood were significantly higher in 12 HFD babies than in con -trol babies， suggesting that IGF-I as well as insulin may be involved in excessive growth of fetuses of diabetic mothers. However， some of the HFD babies showed IGF-I values lower than normal. Maternal HbA，
values showed a positive correla -tion with CPR levels in umbilical∞
rd blood， but not with IGF・1levels. Most of HFD babies with Key Words: Pregnant diabetic， Insulin-like growth factor-I in umbilical cord blood， Macrosomi泊fromdiabetic mothers， CPR in umbi1ical cord blood. high CPR and IGF-I levels were born to diabetic mothers with high HbA，
values， but such babies were also born to diabetic mothers with normal Hb A，
value. We suggest that not only fetal hyperinsu・ linemia， but also fetal IGF-I， tissue sensitivity to the growth factors or possibly other factors may be responsible for macrosomia of neonates of diabetic mothers. Introduction There are a number of observations suggesting that insulin may function asa stimulator of fetal growth. The fetal macrosomia (heavy-for-date baby) is frequently associated with fetal hyperinsu -linemia stimulated by maternal hyperglycemia1). However， such babies are occasionaly born to non -diabetic， healthy mothers or to diabetic mothers with good control. In the present study， we measured insulin-like growth factor(IGF)ー1andCPR (C-peptide immunoreactivity) in umbilical cord blood of infants born to diabetic mothers. The aim of the study is to clarify the role of insulin and IGF-I in the fetal macrosomia of infants born to diabetic mothers. IGF-I is one of the peptides of somatomedin family， and structural analysis has demonstrated that somatomedin C and IGF-I are the same protein2).Recent studies have shown that this
-971-;
て
¥
〈
同
:
:
¥ ・ ¥ ¥
¥ ¥
dependent growth factor is a prerequisite to normal postnatal growth. Also it has been suggested that IGF-I may play a role in the regulation of fetal growth in utero3). Very recent1y， Susa et a14) reported that IGF-I and IGF-II levels in umbilical plasma in infants of 7 diabetic mothers did not di任erfrom those in control infants， but umbilical plasma levels of CPR were significant1y elevated. They suggested that insulin plays the predominant role in stimulating fetal gain of the infant of the diabetic mother during the latter part of gestation. We examined a larger number of infants (62 in -fants) born to the diabetic mothers. Here we show that both umbilical CPR and IGF-I levels correlate well with birth weight of infants born to diabetic mothers.Materials and Methods Subjects
Fifty seven diabetic women were examined at 62 occasions of delivery. Twenty three of the patients were c1assified as IDDM and the remaining 39 as NIDDM. To exc1ude differences in birth weight as a factor， only women delivering after 37 weeks of gestation were studied. Average age at the onset of D M， and that at delivery were 22.5::t:7.0 years old (the mean ::t:SD)，and 29.0::t:4.5 years old， respec -tively. The average gestational period of pregnant diabetics was 38.7 weeks. That of normal healthy pregnant women was 39.9 weeks. Five out of 57 1。目
¥
内 H U 内 MU ( E コ E一
M 司 F E 』 o ヌ 60 t ι l' '7 40 てコ E O 20 m patients were treated with diet alone(羽Thite'sc1ass A) throughout pregnancy. All other patients were treated with insulin. Our goal in the treatment of D M is to control fasting blood glucose to below 100 mg/dl， postprandial blood glucose to below 120 mg/dl， and to keep blood HbA1 below 9% through -out pregnancy. One third of our patients were within this criteria， and the remaining patients were controlledc10sely to this goa1. Cord blood was collected at the time of delivery， and the sera were stored at -20oC until analyzed fro CPR and IGF-I. Cord blood from 15 infants born to normal mothers after 37 weeks of gestation served as controls (20 infants as controls for CPR).Assays for CPR
，
IGF-I and HbA1CPR levels in cord blood were determined using commercial kits (Shionogi). IGF-I was purified from Cohn Fraction IV of human plasma according to the method of Svoboda et a15) and iodinated to a specific activity of 50μCi似gby the Chloramine T method. Antibody to somatomedin C/lGF-I was obtained from the NIH National Hormone and Pituitary Program (a gift from Drs. L. Underwood and ].]. Van Wyk， University of North Carolina). RIA for IGF-I was performed as described by Fur1anetto et a16) except that separation of bound
125I-IGF-I was separated from unbound by poly -ethylene glyco Serum s.1 amples were extracted prior to assay with acid-ethanol according to the procedure of Daughaday et a]7).A typical standard ¥ ¥
•
0.01 0.03 0.1 0.3 Hormones (ng/tube) Fig.1 Standard curve of RIA forIGF-I -972curve in IGF-I RIA is presented in Fig. 1. Soma-tomedin A and MSA (multiplication stimulating activity; rat IGF-II)crossreacted with the antibody with 10% and 1% of potency of IGF-I. Crossreaひ
tion of IGF-II was approximately 5%. Neither insu1in nor proinsulin crossreacted with the anti -body. The percentage of HbA
，
was determined by high performance liquid chromatography using Auto A，
T M (Kyoto Daiichi Kagaku， Kyoto)Results
Fig. 2 shows serum IGF-I (left panel)and CPR (right panel) levels in cord blood of newborns from normal and diabetic mothers. The mean :tSD of IGF-I in cord blood of infants from 15 normal mothers was 67.3士 21.6 ng/ml (range 36-93 ng/ml)， which was significant1y lower than the value in normal adult women (range 200-400 ng/ml).These values are slight1y lower than those reported by Bennet et a13).They reported 113:t32 ng/ml of IGF-I in cord blood of normal infants. The difference may be due to the di妊erenceof standard
of IGF-I. IGF-I level in 62 newborns from diabetic mothers ranged widely from 11 to 214 ng/ml， and the mean土SDwas 81.0 :t41.0 ng/ml.The value
was not statistically different from that in controls. Cord blood CPR level in 62 babies from diabetic mothers was 2.63:t1.84 ng/ml， and was signifi
-ng/ml IGF-I 160r t_t 140ト
-
.
120ト 勿量耳 100卜 80ト 60ト 40ト.
r
.
20ト，
.
(15) (62) neonates 01 neonates 01 normal molhers diabelic molhers cantly higher than that in 20 infants from non -diabetic mothers (1.34:t0.48ng/ml).There was a positive correlation between cord blood IGF-I and CPR levels in newborns from 62 diabetic mothers (r=0.270， p<0.05)， as shown in Fig. 3. In Fig. 4， birth weight of newborns from 23 IDDM (left panel) and 39 NIDDM mothers are plotted as a function of gestation periods. The mean :t3/2 SD birth weight of infants from non-diabetic mothers is indicated by soid 1lines. As can be seen， the majority of infants from either type of diabetics showed normal birth weight. On the whole， 12 babies (4 from IDDM and 8 from NIDDM) were judged heavy-for-date (HFD)， and only 5 babies were small-for-date (SFD). The mean birth weight of newborns was 3138 :t550g for IDDM and 3365 :t592g for NIDDM， respectively. There was no difference between birth weight of babies of both groups.
Correlation between birth weight and either cord blood CPR or IGF-I is shown in Fig. 5 and 6. There was a significant correlation (r=0.585， P<O.OOl) be -tween CPR levels and birth weight. Cord blood IGF-I levels also showed a positive correlation with birth weight (Fig. 6; r=0.473， p<0.05)， but the cor -relation was less than that between cord blood CPR levels and birth weight. The data presented in Fig. 5 and 6 suggested that ng/ml CPR 8.0 。 。 7.0 。 自 6.0 。 自 5.0 8 O 4.0 3。目 。 2.0 。
雪
1。目 。 (20) (62) neonates 01 neonates 01 normal molhers diabelic molhers Fig.2 IGF.I and CPR in umbilical cord blood of diabetic and normal pregnant womenng/ml 8.0 7.0 6.0 5.0 広 止 4.0
。
3.0 2.0 1じ
こ
日 50“
‘
n = 62 r = 0.27日 Y=0.012X+1.658••
•
•
•
-250 ng/ml 100 150 IGF-I 200 Fig_3 Correlation between cord blood IGF-I and CPR levels in pregnant diabetic women “ 4.日 ~ 閉 ω E Z HS3
日 kgl n= 39 n=23 5.0 。 4.0 cb +MSD/ニて
τ
ふ一一
~匂o
0， 00 問 ヘ Mean〉片二〕午一
一χsoュ
ク
ヤ
-
-
-
-
-
-
3.0/ イ プ ー
-二~一一一
O kg 5.0 2.0卜 2.0L 1
37 38 39 40 41 37 38 39 40 41 Gestation (Week) Gestation (Week) Neonates 01 Type 1 d旧beticmothers N回nates01 Typen diabetic mothers Fig.4 Birth weight of newborns born to diabetics after 37 weeks of gestation both fetal insulin and IGF-I are involved in theregulation of fetal growth. To further analyze the relationship among cord blood CPR， IGF-I and fetal growth， 62 babies were divided into three groups depending on their birth weight， and cord blood CPR and IGF-I concentrations were com-pared among the three goups， which consisted of SFD (small-for-date)， AFD (appropriate-for-date) and HFD (heavy-for-date) groups. As shown in Table 1， both cord blood CPR and IGF・1levels in
AFD group were not different from those in controls， but the vlaues in SFD group was signifi -cantly (p<O.Ol) lower than the control values. On
the other hand， there was a significant increment in either of CPR or IGF-I in the babies of HFD group. Thus， weight gain of neonate was c10sely asso -ciated with increased levels of both CPR and IFG-I in umbilical cord blood as a whole. However， when we looked at the individual result， CPR or IGF-I was not always related with birth weight. In the babiesc1assified as AFD， for example， some showed a higher than normal CPR or IGF-I or both. The majority of HFD babies showed higher than normal values both in CPR and IGF・1.Two out of 12 babies of this group， however， had higher than normal CPR， but lower than normal IGF-I， and one 一
ngl、町1 8.0 7.0 8.0 a:5.0 仏 4.0
。
3.0 2.0 ム一日2」0-
•
.
.
•
.
ーノ
・
6・
o -y，
....- ー・・ .
〆
・
・ .
.
.ー司 " 0・
.
・
.
・
，
.
3.0 Birth weight 73 nニ61 r=0.585 Y=1.85X-3.45 4.0 5.0kg Fig. 5 Correlation between cord blood CPR and birth weight of neonates from diabetic mothers ng/ml 200 150.
125一
100 k o-.
・
-a・•
.
・・.
75 0 0・
・
〈 ・
.
o _ 0・ .
.
・
・-・
50 25.
.
.・・
.
トτ
。
3.0 ot•
，
n = 62 r =日473 Y=33.34X-28.38 4.0 内 u e bR 角 川 u ' 4 ・ F h d v Birth weight Fig.6 Correlation between cord blood IGF-I and birth weight of neonates from diabetic mothersTable 1 Cord blood CPR and IGF-I in newborns from diabetic mothers. Number Birth CPR IGF.I W〔egつ1ht (ng/ml) (ng/ml) Small for date 5 2406土208 0.94:t0.34 32.2土13.1 Adaptpe ropriate for 45 3128:t277 2.35士1.55 79.0士35.8 Heavy for date 12 4220土392 4.37士2.05 108.7士46.6 Control 10 3145士362 1.34士0.48 67.3:t21.6 (n=20)
baby showed a higher than normal IGF・1but normalCPR.
Finally， we examined the correlation among maternal mean HbA1 at the third trimester， fetal CPR and IGF-I levels. HbA1 was measured once a month throughout pregnancy to monitor the diabetic state. Out of the 62 babies born to diabetic mothers， 20 newborns with normal CPR (within the range of 0.7-1.9 ng/ml) and normal IGF・1(within the range of 41-81 ng/ml) were selected on one hand， and 18neonates with higher than normal
-975-• Heavy for date o Adequate for date
•
12.0.
一隻ー。
O 11.0 10.日 誌 ぬ ・7
8
京
αフ0 α由 。 写b 一吉一 9.0 8.0 ︽ DZ 7.0 。 n =18 n = 15 6.0 t t IGF-I CPR I G F - I ¥ . Inor、町al CPR /Fig. 7 Mean HbA] at third trimester， cord blood CPR and IGF.I
observations suggestthat SMs may be involved in the regulation of fetal growth. Itis well recognized that oversized infants are frequently born to diabetic mothers， especially when diabetes is poorly controlled. Fetal hyper -glycemia induced by maternal hyperglycemia has been implicated in the macrosomia in these over -sized infants.)9). There are few data available on the role of SMs in the macrosomia of infants of diabetic mothers. Very recently Susa et a14)
measured IGF-I， IGF-II， and CPR levels in 7 in -sulin-dependent diabetic pregnant women and their infants at delivery. They showed that IGF-I and II levels in infants of diabetic mothers did not differ from those in control infants， but that CPR levels are significantly elevated. Their observations that infant birth weight ratio is logarithmically cor -related with umbilical cord CPR levels would sug -gest the predominant role of insulin in stimulating fetal weight gain of infants of diabetic mothers. We have conducted similar studies with a larger number of infants of diabetic mothers. Our results agree well with those of Susa et a14) and Sosenko et a19) in that CPR levels in umbilical cord blood are significantly elevated in the newborns of diabetic mothers. The CPR levels also showed a positive correlation with birth weight. On the other hand， IGF-I levels in newborns of diabetic mothers were not different from those in infants of normal 976 CPR (more than 2.0ng/ml)and IGF-I (more than 87 ng/ml) were chosen on the other hand. Maternal HbA1 values are shown in Fig. 7. In the majority(13 out of 15) of mothers of new-borns with normal CPR and IGF-I levels， HbA. was less than 9%， indicating good control of their diabetes throughout the third trimester of preg -nancy， but in two of the 15 mothers， the values ex -ceeded 10%. Out of 18 diabetic mothers of new-borns with high CPR and IGF-I， eight subjects showed high HbA. values of more than 9%. Six of th巴mothersproduced newborns with macrosomia.
Itshould be noted， however， that 3 HFD babies were born to the mothers whose HbA. levels were in the normal range.
Discussion
The Somatomedins (SMs) are a family of pep -tides which mediate the effects of growth hormone on postnatal growth. The human SMs inc1ude at least two types of insulinlike peptides， IGF-I and II.In contrast to postnatal growth， fetal growth seems independent of GH. However， a number of reports have shown that the concentrations of SMs in umbilical cord blood is related to birth size of infants3)8). Using a specific RIA for IGF-I， Bennet et a/3)showed that both IGF-I and II levels in cord blood positively correlated with birth weight in term infants， and also with gestational age. These
mothers when grouped data were compared. How-ever， regression analysis revealed a significant cor -relation between cord blood IGF-I levels and birth weight， which is similar to the results reported for infants of non-diabetic mothers3)8). Therefore， we divided 62 newborns into three groups (SFD， AFD and HFD) based on birth weight， and compared CPR and IGF-I levels among the groups.羽Te demonstrated that IGF-I as well as CPR levels are significantly elevated in the HFD group. Susa et a14) found no di妊erencein IGF-I levels between infants of diabetic mothers and those of normal mothers. The mean birth weight of infants of diabetic mothers in their study was 3278 :::!::218 g
and was not different from that of infants from normal mothers. This may account for the absence of difference in IGF-I levels between the two groups. These results would suggest that fetal IGF-I as well as insulin may be involved in the overgrowth of neonates of diabetic mothers. In some infants with overgrowth， however， IGF-I levels were lower than normal.Conversely， in spite of higher than normal CPR and IGF-I levels， there were infants of normal birth weight. The sensitivity of fetus to these growth factors may be important for over -growth. Alternatively， other factors may contri -bute to the regulation of fetal growth， possibly with insulin and/or IGF-I. It has been reported that IGF -11 levels in umbilical cord blood show a positive correlation with birth weight3)， but the physiolog -ical significance of IGF-I1 in fetal growth is still uncertain. The reason of the elevated cord blood IGF-I levels is also not clear.Insulin administration resulted in an increase of fetal SMSIO)lI)and SM production in vitro from rat liver cells was enhanced in the presence of insulinI2)13). We
showed a positive correlation between CPR and IGF-I levels in umbilical cord blood. These ob -servations are compatible with the hypothesis that fetal insulinemia is responsible for elevated levels ofIGF・1.However， Blethen et al14) reported that in children with normal GH secretion， hyperinsu -linemia does not result in an increase in either IGF -1 or IGF-II. 977 75 Finally we analyzed the corr巴lation between maternal diabetic state with fetal IGF-I， CPR且日d
birth weight. There was a positive correlation be -tween fetal CPR levels and maternal HbA1 values
(data not shown)， but fetal IGF-I levels did not cor -relate to maternal HbA1 indicating that maternal diabetic state during the third trimester of gesta -tion does not a丘ectfetal IGF-I levels. Out of 62 babies born to diabetic mothers， 12 were heavy for date， and 9 of the 12 infants showed higher than normal CPR and IGF-I levels (see Fig. 7). Among 9 diabetic mothers who delivered oversized babies， 6 subjects showed HbA1 values higher than 9%. But
the values were within normallimits in the other 3 mothers. Thus， heavy for date babies are born even to the diabetic mothers under good control.It is no doubt that a number of unidentified factors exist in th巴processof fetal overgrowth. Further studies
are required to clarify this point. Acknowledgement
The present work was supported by a grant from Okamoto Itoe Foundation and a grant from the Ministry of Health and Welfare.
References
1) Pedersen， J.: The pregnant diabetic and her newborn. Problems and management. p. 128， The Williams& Wilkins Company (1967)
2) Klapper， D.G.， Svoboda， M.E. and Van Wyk，
J.J.: Sequence analysis of somatomedin C: confirma -tion of identity with insulin-1ike growth factor-1. Endocrinology 112 2215-2217 (1983)
3) Bennett， A.， Wison， D.M.， Liu， F.， Naga-shima， R.， Rosenfeld， R.G. and Hintz， R.L.: Levels of insulin-like Growth Factors 1 and II in human cord blood.J Clin Endocrinol Metab 57 609-612 (1983)
4) Susa， J.B.， Widness， J.A.， Hintz， R.， Liu， F.，
Sehgal， P. and Schwartz， R.: Somatomedins and insulin in diabetic pregnancies: Effects on fetal macrosomia in the human and Rhesus monkey.J Clin Endocrinol Metab 581099-1109 (1984)
5) Svoboda， M.E.， Van Wyk， J.J.， Klapper，
D.G.， Fellows， R.E.， Grissom， F.E. and Schlueter， R.J.: Purification of somatomedin司C
from human plasma: Chemical and biological pro -perties， partial sequence analysis， and relationship to other somatomedins. Biochemistry 19 790-797
(1980)
6) Furlanetto， R.W.， Underwood， L.E.， Van Wyk， J.J. and D'Ercole， A.J.: Estimation of somatomedin-C levels in normals and patients with pituitary diseases by radioimmunoassay.J Clin Invest 60648-657 (1977)
7) Daughaday， W.H.， Mariz， I.K. and Blethen， S.L.:Inhibition of acsess of bound somatomedin to membrane receptor and immunobinding sites. A comaprison of radioreceptor and radioimmunoassay for somatomedin in native and acid-ethanol-extracted serum.J Clin Endocrinol Metab 51 781-788 (1980) 8) Gluckman， P.D.， Johnson-Barrett， J.J.，
But1er， J.H.， Edgar， B.W. and Gunn， T.R.: Studies of insulin-like growth factor-I and 11 by specific radioligand assays in umbilical cord blood. Clinical Endocrinology 19405-413 (1983)
9) Sosenko， I.R.， Kitzmi11er， J.L.， Loo， S.W.， Blix， P.， Rubenstein， A.H. and Gabbay， K.H.: The infant of the diabetic mother.Correlation of in -creased cord C-peptide levels with macrosomia and hypoglycemia. N EnglJ Med 301859-862 (1979) 10) Hi11， D.J. and Milner， R.D.G.: Increased soma -tomedin and cartilage metabolic activity in rabbit fetuses injected with insulin in utero. Diabetologia 19143-147 (1980) 11) Spencer， G.S.G.， Hill， D.J.， Garssen， G.J.， Macdonald， A.A. and Colenbrander， B.: Soma-tomadin activity and growth hormone levels in body fluids of the fetal pig: e妊ectof chronic hyperinsu -linemia.J Endocrinology 96107-114 (1983)
12) Daughaday， W.H.， Phi11ips， L.S. and Mueller， M.C.: The effects of insulin and growth hormone on the release of somatomedin by the isolated liver， Endocrinology 981214-1219 (1976)
13) Kogawa， M.， Takano， K.， Asakawa， K.， Hizuka， N.， Tsushima， T. and Shizume， K.: Insulin stimulation of somatomedin A production in monolayer cultures of rat hepatocytes. Acta En-docrinoll03 385-390 (1983)
14) Blethen， S.， White， N.H.， Santiago， J.V. and Daughaday， W.H.: Plasma somatomedins in chlidren with hyperinsulinism.J Clin Endocrinol Metab 52 748-750 (1981) 岡本糸枝賞研究報告 糖 尿 病 の 母 親 か ら 生 ま れ た 新 生 児 の 騎 帯 血
