Iodine,
Thyroid
Function,
and
Very
Low
Birth
Weight
Infants
Elvira Parravicini, MD*; Corinna Fontana, MD*; Giuseppe L. Paterlini, MD*; Paolo Tagliabue, MD*;
Franco Rovelli, MD*; Kenneth Leung, MS; and Raymond I. Stark, MD
ABSTRACT. Objective. Research was undertaken to
test two hypotheses. First, during the early neonatal
pe-nod, thyroid function of very low birth weight (VLBW)
infants is suppressed by exposure to iodine-containing
antiseptic solutions and/or iodized contrast media.
Sec-ond, this suppression is more pronounced in small for
gestational age (SGA) infants.
Methods. Urinary iodine concentration and thyroid
function measurements were obtained prospectively
from 44 VLBW infants with gestational ages at birth of
30 ± 2.3 weeks and weights of 1223 ± 231 g. Eleven of
these infants were SGA. The infants were grouped
ac-cording to iodine exposure: 18 infants had no increased exposure and served as control infants; 9 infants were exposed to an iodine-containing antiseptic (povidone
io-dine); 12 infants were exposed to an iodized contrast
medium (iopamidol); and 5 infants were exposed to both
agents. Urinary iodine and serum free triiodothyronine,
free thyroxine, and thyrotropin were measured on days 1,
7, 14, 21, and 28 of life
Results. During the period of maximum exposure
(days 1 to 7), the concentration of iodine in the urine of
study infants was 2 to 4 orders of magnitude greater than
that in the urine of control infants (123 ± 141 .tg/L). During the subsequent 3 weeks, levels of urinary iodine
in study infants returned to levels that were not
signifi-cantly different from controls. On day 7 of life,
iodine-exposed infants had a significantly higher mean
thyro-tropin level than control infants, whereas on day 28, free
triiodothyronine and thyroxine levels were lower. Of the
26 iodine-exposed infants, 6 had transient
hyperthyro-tropinemia and 2 had transient hypothyroidism. When
exposed to iodine, SGA infants had more labile thyroid
function than normally grown iodine-exposed or control
infants. These SGA infants had significantly lower levels
of thyroid hormones in umbilical cord blood, increased
production of thyroid hormones on day 14 of life, and
lower levels again at 1 month.
Conclusion. In VLBW infants, the use of
iodine-con-taming antiseptic solutions and iodized contrast media
results in massive uptake of iodine that is associated
with alterations in thyroid function. It is reasonable to
suggest that, whenever possible, iodized products should
be avoided in VLBW infants, because their routine use
results in exposure to excessive loads of iodine, which
can be associated with hyperthyrotropinemia and
hypo-thyroidism. Pediatrics 1996;98:730-734; iodine,
thyro-From the *j5yfl of Neonatal Pathology and Laboratory of
Radioimmu-nology, Ospedale S. Gerardo dei Tintori, Monza, Milan, Italy; and
Depart-ment of Pediatrics, Division of Perinatology, College of Physicians and
Surgeons, Columbia University, New York, New York.
Received for publication Sep 15, 1995; accepted Dec 12, 1995.
Reprint requests to (E.P.) do Raymond I. Stark, MD, College of Physicians
and Surgeons, Columbia University, 630 W 168th St. New York, NY 10032.
PEDIATRICS (ISSN 0031 4005). Copyright © 1996 by the American
Acad-emy of Pediatrics.
tropin, thyroxine, triiodothyronine, povidone-iodine, iop-amidol, prematurity, very low birth weight, small for gestational age, average for gestational age.
ABBREVIATIONS. TSH, thyrotropin; SGA, small for gestational
age; AGA, average for gestational age; VLBW, very low birth
weight; Fr3, free triiodothyronine; FT, free thyroxine.
Development of the central nervous system
de-pends on an adequate supply of thyroid hormones,
which in turn require iodine for biosynthesis.1 The
mechanisms regulating production and release of
thyroid hormones in premature infants are relatively immature at birth.2 During the first week of life, all
preterm infants, both healthy and sick, have low
serum concentrations of thyroxine, triiodothyronine,
and thyrotropin (TSH) in comparison with full-term
infants. There is also a relationship between growth
and thyroid function. Lower levels of thyroxine and
higher levels of TSH are found in small for
gesta-tional age (SGA) infants compared with normally
grown (average for gestational age [AGA]) infants.3
Although iodine deficiency can cause goiter and
cre-tiism,4 hypothyroidism and related disorders may
also occur in infants exposed to excess iodine,
espe-cially in areas of low environmental iodine, as is
found in certain regions of Europe.5
Earlier studies have reported that the use of
io-dine-containing solutions for asepsis during delivery
can cause increased TSH levels in umbilical cord
blood.6 In infants undergoing cardiac surgery,
signif-icant transcutaneous absorption of iodine results
from topical use of iodine-containing antiseptics.7 It
has been proposed that preterm infants, exposed in
utero or during the neonatal period, may be
par-ticularly susceptible to iodine-induced
hypothy-roidism.2 Cases of increased iodine exposure after
intravenous administration of iodinated contrast
material have been reported to cause transient
hy-pothyroidism.8 For these reasons, we conducted a
prospective study to investigate the effect of iodine
exposure on the thyroid function of very low birth
weight (VLBW) infants. We measured the
concen-trations of iodine in the urine and the
concentra-tions of thyroid hormones in the plasma of
prema-turely born SGA and AGA infants exposed to an
iodine-containing antiseptic solution and/or an
io-dized contrast medium. We compared these
mea-surements with those from a control group that
METHODS Statistics Patient Population
The study population was born at S. Gerardo Dei Tintori
Hos-pital (Monza, Milan, Italy) and included 44 infants with weights at
birth of less than 1500 g. Infants were divided into four groups
defined by postnatal exposure to iodine. Eighteen infants were
never exposed to an iodine load, and all aseptic procedures were
performed with 0.5% chlorhexidine in alcohol. In 9 infants the
exposure to iodine was as an antiseptic solution of 10%
povidone-iodine (Betadine solution; Chinoin Spa., Milan, Italy) used for all
aseptic procedures. Examples of these procedures include
veni-puncture, lumbar puncture, and arterial and venous line
place-ment. The skin was treated only at the procedure site. The
solu-tions were left on the skin and were removed during routine daily
baths. Infants were never bathed in antiseptic solutions. An
addi-tional 12 infants were exposed to iodine by injection of 0.3 mL of
30% iopamidol, a nonionic iodinated contrast medium (Iopamiro;
Bracco, Milan, Italy). This agent is used for radiologic
confirma-lion of the placement of percutaneous central venous catheters.
This procedure was done between days 3 and 7 after birth. The
fourth group of 5 infants was exposed to both povidone-iodine
and iopamidol. Moreover, to study the interaction between
growth status and iodine exposure, infants were divided into
three additional categories. Fifteen AGA infants, never exposed to
an iodine load, served as control infants, and 8 SGA and 18 AGA
infants were exposed to iodine. Three SGA infants had no
expo-sure to increased iodine, but these infants were excluded from the
analyses, because the sample size was too small for statistical
comparison. The demographic characteristics of the infants are
described in Table 1. Small amounts of enteral nutrition with
breast milk or formula were introduced after the first week of life.
There was no supplemental iodine in the intravenous solutions or
the formula given to these infants. Consent was obtained for this
study from the parents of all infants.
Sampling and Assay Methods
To estimate iodine uptake from the antiseptic solution and/or
contrast material, 1- to 5-mL samples of urine were collected on
days 1, 3, 7, 14, 21 and 28 after birth, and the iodine content was
detected by means of a colonmetric reaction using a Technicon
AutoAnalyzer (Bayern Group, Tarrytown, NY). The amount of
iodine was also measured in the breast milk of eight mothers of
infants in this study and in Prenidina (Nestl#{233}), the formula used in
the nursery at the time of the study for nourishment of premature
infants. For evaluation of thyroid function, TSH, free
triiodothy-ronine (Fr3), and free thyroxine (FF4) were measured.
Two-milli-liter blood samples were collected by venipuncture on days I
(umbilical cord blood), 7, 14, 21, and 28 of life in glass tubes. After
clotting, the samples were centrifuged at 2500 rpm for 5 minutes,
and serum was separated. Samples were either immediately
as-sayed or stored at -20#{176}C.TSH was measured in 200-s.d samples by
an immunoassay that has a high affinity and specificity for the
hormone (TSH Radioisotopic Assay; Nichols Diagnostic Institute,
San Juan Capistrano, CA). This assay has a sensitivity of 0.04
U/mL. VF3 and VF4 were measured in 400-s.d aliquots of serum
by competitive radioimmunoassay by a two-step method in a
solid phase (FT and FT Myria; Techno Genetics, Cassina de
Pecchi, Milan, Italy). The sensitivity of these assays is 0.03 ng/dL.
All statistical analyses were performed using SYSTAT (SYSTAT
Inc. Evanston, IL). Urinary concentrations of iodine were not
normally distributed; thus, the log transform of values was used
for statistical analysis. For comparing groups of patients with the
same gestational ages and birth weights, one-way analyses of
variance were used. For all biochemical and hormonal
measure-ments, independent t tests were used to compare iodine-exposed
infants with nonexposed (control) infants at each of the time
points from birth to 28 days of age. Least squares linear regression
was used to evaluate the trend in urinary iodine during the first
month of life in the control infants. Unless specified, results were
considered significant at P < .05.
RESULTS Urinary Iodine
The results of urinary iodine level measurements
are summarized in Fig I. The mean ± SD value of
urinary iodine in the group of control infants was
123 ± 141 g/L at birth and increased to 381 ± 326
p.g/L at day 28 (r = .44; P < .004). In the group
exposed to povidone-iodine, the mean concentration
in urinary iodine was significantly higher than that
in control infants on days 1, 3, and 7 of life. On the first day of life, the group of infants, subsequently
exposed to iopamidol, had a mean concentration of
urinary iodine similar to that of the control group. On the third to fifth days of life, when infants were
exposed to iopamidol, the urinary level increased to
more than 1000 times that of control infants (range,
3060 to 298 000 g/L). Although the mean value
remained elevated through the seventh day of life,
the variability was large (133 701 ± 346 215 g/L)
because of differences in the timing of exposure to
the contrast solution. The group of infants exposed to
both iopamidol and povidone-iodine had mean
val-ues of urinary iodine that were significantly greater than those of the control group on days I and 3 after birth. On the 14th day of life and subsequently, there were no significant differences in urinary iodine con-centrations among these four groups of infants.
Iodine in Breast Milk and Premature Infant Formula
In eight samples of breast milk fed to these infants,
the iodine content was 123.1 ± 93.6 tg/L; in the
premature infant formula it was 84.8 g/L.
Thyroid Hormones
The results of thyroid function studies of these
infants are summarized in Fig 2. In Fig 2, A, it can be
seen that the mean TSH values were significantly
TABLE 1. Characteristics of Infants
Iopamidol, or Both*
Not Exposed to a n Increased lodin e Load (Control) and Groups Exposed to Povidone-Iodine,
Characteristic Control Povidone lopamidol Povidone and lopamidol P
Infants 18 9 12 5
Gestational age, wk (mean ± SD) 30.1 ± 2.7 30.5 ± 2.1 29.4 ± 1.9 30.6 ± 1.5 NSt
Birth weight, g (mean ± SD) 1219 ± 251 1215 ± 212 1252 ± 229 1183 ± 259 NS
SCA 3/18 3/9 2/12 3/5 ...
RDS 5/18 3/9 3/12 1/5 ...
Sepsis 0/18 1/9 0/12 1/5 ...
IVH 0/18 0/9 1/12 0/5 ...
*Included are the number of children small for gestational age (SGA) and those with respiratory distress syndrome (RDS), sepsis, or
intraventricular hemorrhage (IVH).
-J N C) :t C -C 0 > Co C 1000000
r
0c000 10000 1000 100 10 * --#{149}1 /_ \ -- _ ,/// Time (day\ Ccrtrc
\ - oarnidoIo
* \\ - -
_
-. \ Ioparnidoio &
Povidone-odne
0 10 20
Fig 1. Mean values of urinary iodine in three groups of infants
exposed to iodine as povidone-iodine, iopamidol, or both agents
and a group of infants not exposed to iodine (control). Levels of
significance refer to comparison between iodine groups and the
control group (* P < .05).
higher on day 7 in the two groups exposed to
povi-done-iodine. Taken together, at the end of I week of
life, infants exposed to an increased iodine load had
a significantly greater mean TSH value than control
infants (16.53 ± 24.7 vs 3.11 ± 2.3 U/mL). The
mean values of plasma VF3 and FF4 in the four
groups were similar throughout the study (Fig 2, B
and C), with one exception. On day 28, infants
ex-posed to both iopamidol and povidone-iodine had
mean FF3 and VF4 values that were significantly less
than the values for the control infants.
Comparisons of the thyroid function of AGA and
SGA infants are summarized in Fig 3. On the seventh
day of life, the mean values of TSH for both AGA
and SGA iodine-exposed infants were much greater
than those of AGA control infants (Fig 3, A);
none-theless, these differences did not reach statistical sig-nificance (P = .081 and .057, respectively). However,
from the second week of life until 1 month, the SGA
infants, who were exposed to iodine, had mean TSH
values that were significantly greater than those of
AGA infants who were also exposed to iodine. There
were no significant differences in mean TSH levels
between the group of AGA infants exposed to iodine
and the AGA control infants. The mean values of VF3
and VF4 in SGA and AGA infants are summarized in
Fig 3, B and C. In SGA infants, VF4 values in
umbil-ical cord blood were lower than those of AGA
in-fants. After iodine exposure, FT3 and FF4 values were
greater in SGA infants than in AGA infants on day 14
of life, whereas at 28 days, the mean values of VF3
and FF4 were less in SGA infants.
DISCUSSION
A major result of our study is the finding that
VLBW premature infants have massive uptake of
iodine. This was reflected by concentrations of iodine in the urine of infants exposed to topical antiseptic
and/or contrast media containing iodine that were 3
orders of magnitude greater than levels in the urine
of control infants. The increased iodine load was
associated with alterations of the thyroid function.
More than 20% of the infants exposed to iodine had
evidence of transient hyperthyrotropinemia (TSH
>20 p.U/mL), and one third of these infants had
transient hypothyroidism. None of the VLBW
con-trol infants had comparable alterations in thyroid
function. Finally, unlike AGA infants, SGA infants
responded to an iodine load with increases in plasma
levels of both Vf3 and FT, immediately after the
period of exposure.
At birth the mean urinary iodine concentration in
the group of control infants was similar to that
pre-viously found in other neonatal populations in
Eu-rope; however, this mean level is slightly lower when
J0 compared with that of North American infants.9 This
condition of mild iodine deficiency did not have a
demonstrable influence on thyroid function, in that
no cases of hypothyroidism were detected in the
control group. Urinary iodine concentrations of
in-fants not exposed to iodine-containing agents did
increase significantly by I month. This modest
in-crease was probably a consequence of iodine
pro-vided by enteral nutrition. In breast milk there was,
on average, 123 p.g/L, and in formula for premature
infants there was about 85 tg/L. Thus, it seems
reasonable that VLBW infants receiving total
paren-teral nutrition should be treated with an iodine
sup-plement.
The mean values of urinary iodine in
povidone-iodine-exposed infants were 10 times more than
those of control group in the first week of life. This
increase was prominent during this period because
procedures requiring skin asepsis were more
fre-quent in this early neonatal period. The finding of
massive iodine absorption through the skin is in
agreement with the results of other investigators.7’9’10
Infants exposed to the iodine-containing contrast
me-dium iopamidol between the third and seventh days
of life also excreted huge amounts of iodine in their
urine during the period of exposure (see Fig 1). It
seems that iodine is rapidly excreted by the kidney of
the premature infant, because by 2 weeks of life, the
urinary iodine concentrations in exposed infants
were comparable to those of the control infants who
had no increased exposure to iodine.
One week after birth, the mean concentration of
TSH in iodine-exposed infants was significantly
higher than that of control infants. This suggests
that the relative hyperthyrotropinemia of these
in-fants was a direct effect of iodine. No infant from
the control group had a TSH level of more than 7.6
tU/mL, whereas six iodine-exposed infants had
serum TSH values that were greater than 20 U/
mL. This level has been used as a cutoff value for
definition of hyperthyrotropinemia.5 Other
inves-tigators have reported cases of an association
be-tween increased TSH levels in premature infants
and exposure to exogenous iodine.6’1012 As shown
in Table 2, transient hypothyroidism developed in
two of the six infants with hyperthyrotropinemia.
These two infants had the highest levels of urinary
1 7 14 21 28
I,
Control
- - lopamidolo
A
B
---- Povidone-iodineC
. lopamidolo I.
.iO 20 0.5
Povidcne-iodine
0.4 15
20
N N N
C 0) 0)
3 .s i.o .5
I 0)
? : 0.2
10
05
0.1
0 1 1 I I I. ) 00
-1 7 14 21 28 1 7 14 21 25
Time (days) Time (days) Time (days)
Fig 2. Mean serum concentrations of thyrotropin (TSH; A), free thyroxine (FF4; B), and free triiodothyronine (FT3; C) in three groups of
infants exposed to iodine as povidone-iodine, iopamidol, or both agents and a group of infants not exposed to iodine (control). Levels of
significance in A refer to the comparison between the two povidone-iodine-exposed groups and the control group and in B and C to the
comparison between the iopamidol and povidone-iodine groups and the control group (* P < .05).
Control
A B --- l-SGA
C
30 2 0 1-AGA 0.5
15
20
I’ 03
N ‘ N N
C I C) C)
. / \ .510 .5
I I’ Co 0))
? .\\* 02
i: 17142128 : 14 21 1 7 14 21 25
Time (days) Time (days) Time (days)
Fig 3. Mean serum concentrations of thyrotropin (TSH; A), free thyroxine (Ff4; B) and free triiodothyronine (FF3; C) in iodine-exposed
average for gestational age (AGA) and small for gestational age (SGA) infants compared with AGA non-iodine-exposed infants (control).
Levels of significance refer to comparison between iodine-exposed AGA and SGA infants (* P < .05).
TABLE 2. Infants Exposed to a n Increased Iodine Load Ha ving Hyperthyrotropinemi a on the Seventh Day of Life
Case Thyrotropin, U/mL Free Thyroxine, ng/dL Urinary Iodine, g/L Group Growth Status*
I 76.8 0.45 48 740 Iopamidol AGA
2 53.5 0.57 76 900 Iopamidol AGA
3 82.6 1 .49 7270 Povidone-lopamidol SCA
4 20.5 1.43 I 910 Povidone SGA
5 57.6 0.85 4 140 Povidone AGA
6 28.6 1.10 3 783 Povidone AGA
* AGA indicates average for gestational age; and SGA, small for gestational age.
iodine-containing contrast medium. This result dine have an inhibitory effect on the synthesis of
contrasts with that of other researchers who re- thyroid hormones as long as plasma iodine levels
ported that the use of iodized contrast media was remained elevated.14 This process, known as the
not associated with alteration in thyroid function; Wolff-Chaikoff effect, may play a role in the
regula-however, only full-term infants were studied.’3 It tion of thyroid function of premature infants
ex-may be that premature infants are more suscepti- posed to marked increases in plasma iodine.9”0 In
ble to the effects of an iodine load.10 Table 2 it can be seen that the urinary iodine levels of
elevated during the period of time when levels of VF4 were depressed.
There are other factors that are causally related to suppression of thyroid function in infancy. Transient
neonatal hypothyroidism is frequently reported in
preterm infants and is associated with severe
ill-nesses, especially respiratory distress syndrome.3’15
As shown in Table 1, the incidence of respiratory
distress syndrome, sepsis, and intraventricular
hem-orrhage was similar among the study groups. Thus,
these pathologic conditions were probably not
re-sponsible for differences in thyroid hormone levels
among the groups of infants.
Intrauterine growth retardation is another factor
that has been shown to affect thyroid function, with
SGA infants having higher TSH and lower thyroid
hormone levels.3’16 The TSH data in Fig 3 are in
agreement with these findings. The TSH levels of
SGA infants exposed to iodine were significantly
higher than those of AGA infants also exposed to
iodine. As shown in Fig 2, the majority (three of five)
of infants who had significantly lower values of VF3
and FF4 on day 28 were SGA. In addition, before
exposure to exogenous iodine, the mean level of VF4
in the umbilical cord blood of SGA infants was
sig-nificantly lower than that of AGA infants (Fig 3, B).
Thus, the results of this study provide evidence that
growth status at birth, independent of iodine
expo-sure, has an effect on the thyroid function of prema-ture infant.
An unexpected finding was a significant increase
in mean levels of VF3 and VF4 at 2 weeks of age in
SGA infants who had been exposed to iodine (Fig 3,
B and C). This increased production of thyroid
hor-mones could be evidence of more mature
hypo-thalamopituitary regulation in these undergrown
in-fants. Thus, in the period immediately after exposure
to iodine, these SGA infants escape the
Wolff-Chaikoff effect, as do adults with mature thyroid
regulation.’0
In conclusion, exposure of VLBW premature
in-fants to exogenous iodine as either topical antiseptic
solutions or contrast media results in absorption of
iodine and markedly increased concentrations of
un-nary iodine. Altenations of thyroid function in these
infants can be found during the period of maximal
iodine load. These alterations are reflected by
tran-sient hyperthyrotropinemia and hypothyroidism in
some exposed infants. In VLBW infants who are SGA
as well, the values for plasma thyroid hormones are
more labile and may be increased as a result of
exposure to an iodine load. Thus, it is reasonable to
recommend that the routine use of iodized products
should be avoided in VLBW infants.
ACKNOWLEDGMENTS
This work was sponsored in part by grant 13063 from the
National Institute of Child Health and Human Development,
Na-tional Institutes of Health.
REFERENCES
1. Stanbury JB. Iodine and human development. Med Anthropol. 1992;13:
413-423
2. Fisher DA. Euthyroid low thyroxine (T4) and triiodothyronine (T3) states
in premature and sick neonates. Pediatr Clin North Am. 1990;37:
1297-1312
3. Uhrmann 5, Marks KH, Maisels MJ, et al. Thyroid function in the
preterm infant: a longitudinal assessment. I Pediatr. 1978;92:968-973
4. Xue-Yi C, Xin-Min 1’ Zhi-Hong D, et al. Timing of vulnerability of the brain to iodine deficiency in endemic cretinism. N Engl IMed. 1994;331:
1739-1744
5. American Academy of Pediatrics. Newborn screening for congenital
hypothyroidism: recommended guidelines. Pediatrics. I 993;91:
1203-1209
6. Novaes M Jr. Biancalana MM, Garcia SA, Rassi I, Romaldini JH.
Eleva-tion of cord blood TSH concentration in newborn infants of mothers
exposed to acute povidone iodine during delivery. JEndocrinol Invest.
1994;17:805-808
7. Mitchell IM, Pollok JCS, Jamieson MPG, Fitzpatrick KC, Logan RW.
Transcutaneous iodine absorption in infants undergoing cardiac
oper-ation. Ann Thorac Surg. 1991;52:1138-1140
8. Giroux JD, Sizun J, Rubio 5, et al. Hypothyroidie transitoire aprbs
opacification iodee des catheters epicutaneocaves en reanimation
neo-natale. Arch Fr Pediatr. 1993;50:273
9. Delange F, Dalhem A, Bourdoux F, et al. Increased risk of primary
hypothyroidism in preterm infants. IPediatr. 1984;105:462-469
10. Smerdely F, Boyages SC, Wu D, et al. Topical iodine-containing
anti-septics and neonatal hypothyroidism in very low birth weight infants.
faucet. 1989;16:661-664
I I. Vilain E, Bompard Y, Clement K, Laplanche 5, De Kermadec 5, Aufrant
C. Application breve d’antiseptique iode en soins intensifs neonatals:
consequences sur Ia fonction thyroidienne. Arch Pediatr. 1994;1:795-800
12. Davidovitch N, Linder N, Kuint J, et al. Topical iodine containing
antiseptics and subclinical hypothyroidism in preterm infants. Pediatr Res. 1995;37:75A. Abstract
13. Bona G, Zaffaroni M, Defilippi C, Gallina MR. Mostert M. Effects of
iopamidol on neonatal thyroid function. Eur IRadiol. 1992;14:22-25
14. Wolff J, Chaikoff IL. The inhibitory action of excessive iodide upon the
synthesis of diiodotyrosine and of thyroxine in the thyroid gland of the normal rat. Endocrinology. 1948;43:174-179
15. Wilson DM, Hopper AO, McDougall RI, et al. Serum free thyroxine values in term, premature and sick infants. / Pediatr. 1982;lOl:113-117 16. Uhrmann 5, Marks KH, Maisels MJ, Kuhn HE, Kaplan M, Utiger R.
Frequency of transient hypothyroxinaemia in low birth weight infants.
