0021-972X/07/$15.

00/0 The Journal of Clinical Endocrinology & Metabolism 92(8):3141–3147

Printed in U.S.A. Copyright © 2007 by The Endocrine Society
doi: 10.1210/jc.2007-0238

High Risk of Congenital Hypothyroidism in

Multiple Pregnancies
Antonella Olivieri, Emanuela Medda, Simona De Angelis, Herbert Valensise, Mario De Felice,
Cristina Fazzini, Isabella Cascino, Viviana Cordeddu, Mariella Sorcini, Maria Antonietta Stazi,
and The Study Group for Congenital Hypothyroidism

Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018

Dipartimento di Biologia Cellulare e Neuroscienze (A.O., S.D.A., C.F., V.C., M.S.), Centro Nazionale di Epidemiologia (E.M.,
M.A.S.), Sorveglianza e Promozione della Salute - Istituto Superiore di Sanità, 00161 Roma, Italy; Dipartimento di
Chirurgia (H.V.), Università Tor Vergata, 00133 Roma, Italy; Dipartimento di Biologia e Patologia Cellulare e Molecolare
(M.D.F.), Università Federico II Napoli, 80126 Napoli, Italy; and Istituto di Biologia Cellulare (I.C.), Consiglio Nazionale
delle Ricerche, Monterotondo scalo, 00015 Roma, Italy

Context: In Italy, the surveillance of congenital hypothyroidism (CH)
is performed by the Italian National Registry of Infants with CH
(INRICH). Up to now, about 3600 infants with CH are recorded in the
INRICH, and a high number of twins are included.
Objective: Our objective was to estimate the risk of CH in multiple
and single deliveries and to compare neonatal features of CH twins
with twins from the general population.
Design: The Italian population of CH infants recorded in the INRICH
from 1989 –2000 was investigated.
Results: A more than 3-fold higher frequency of twins was found in
the CH population than in the general population, and for the first
time, it was possible to estimate the CH incidence in multiple (10.1
in 10,000) and single deliveries (3.2 in 10,000 live births). Signifi-
cantly higher frequencies of in situ gland as well as lower TSH mean
level at screening were found in twin than in singleton CH babies. The
concordance rate for permanent CH was very low (4.3%) and due to
only three concordant couples. However, a high recurrence risk for CH
was estimated in siblings of affected babies recorded in the INRICH,
including twins considered as siblings.
Conclusions: The high CH incidence observed in twins is worthy of
interest for the high number of induced pregnancies in Italy as well
as in other Western countries. Moreover, the low concordance rate for
CH among twins together with a high recurrence risk for the disease
among siblings indicates that environmental risk factors may act as
a trigger on a susceptible genetic background in the etiology of the
disease. (J Clin Endocrinol Metab 92: 3141–3147, 2007)

C ONGENITAL HYPOTHYROIDISM (CH) is the most

frequent endocrine disease that occurs in the early
phases of life (1). The early deficit of thyroid function causes, if
not promptly treated, serious and irreversible damage to the
nervous system with consequent mental retardation. The early
diagnosis and the prompt institution of the replacement ther-
apy has been achieved by the implementation of newborn
screening programs by using blood spotted on filter paper. As
a result of these screening programs, virtually all cases of CH
are now identified. In Italy, the incidence of CH in the last 10
yr has been about one in 2700 live births (www.iss.it/rnic/). In
our country, CH babies diagnosed by neonatal screening are
recorded in the Italian National Registry of Infants with CH
(INRICH). This was established in Italy in 1987 as a program of
the Health Ministry and is coordinated by the National Insti-
tutes of Health (2). Up to now, 3600 infants with confirmed
diagnosis of CH are recorded in the INRICH, and a high num-
ber of twins are included.
Twin birth is a relatively common event. In Europe and in
First Published Online May 8, 2007
Abbreviations: CH, Congenital hypothyroidism; CI, confidence in-
terval; CM, congenital malformations; DZ, dizygous; FT4, free T4;
INRICH, Italian National Registry of Infants with CH; MZ, monozygous;
SRR, sibling recurrence risk; TH, transient primary hypothyroidism.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the en-
docrine community.
North America, the twinning rate has averaged between 12–14
per 1000 pregnancies (3). However, the occurrence of twinning
is increasing in the general population and has been shown to
be due to the increasing use of techniques of assisted repro-
duction and drugs inducing ovulation (4). Twins are widely
reported to have more morbidity than singletons, mainly be-
cause of a higher preterm birth rate. In fact, preterm delivery is
the major cause of adverse outcomes (both short-term and long-
term) and is directly related to fetal number (5). It is well known
that prematurity can affect thyroid function. Fetal thyroid sys-
tem metabolism is characterized by relatively high circulating
TRH and TSH concentrations, immature T4 feedback control of
TSH secretion, immature thyroid gland T4 biosynthesis and
secretory processes, and limited effects of T4 on responsive gene
transcription events (6 – 8). All these processes mature progres-
sively with advancing gestational age.
Up to now, no extensive studies have been performed to eval-
uate the occurrence of CH in twin pregnancies. In this study, the
Italian population of CH infants recorded in the INRICH was
investigated with the aim of 1) estimating the risk of CH in mul-
tiple and single deliveries and 2) comparing neonatal features of
CH twins with twins from the general population.
Patients and Methods
Diagnosis ascertainment
In Italy, the neonatal screening for CH is performed by a network of
26 regional or interregional centers. A biochemical assessment of TSH,

3141
3142 J Clin Endocrinol Metab, August 2007, 92(8):3141–3147
and in 11 of the 26 centers also of T4, is performed within a few days from
birth. In all the centers, positive results of screening tests are confirmed
by definitive tests of thyroid function on serum, and thyroid ultrasound
and/or scintigraphy are generally performed to complete the CH di-
agnosis. According to international guidelines (9, 10), when the defin-
itive diagnosis is not established in the neonatal period and a suspicion
of transient primary hypothyroidism (TH) is present, a reevaluation of
diagnosis is performed at the age of 2–3 yr after a withdrawal of the
replacement therapy to ascertain the persistence of CH. At that time, T4,
free T4 (FT4), and TSH levels are measured, and ultrasound imaging,
scintigraphy, and clinical evaluation are performed to establish the
definitive diagnosis. Babies with transient hyperthyrotropinemia on the
basis of spontaneous normalization of TSH between screening and di-
agnosis are not recorded in the Register.
A registration form is filled in at diagnosis. It includes anonymous
data concerning CH infants such as screening and confirmatory labo-
ratory tests, information on demographic data and details on clinical
state in neonatal period, diagnostic investigations (biochemical deter-
minations, radiography of the knee, thyroid scan, and ultrasound), in-
formation regarding pregnancy, birth, and family background, and
starting and dose of the replacement therapy (2). Since 1991, the Register
started collecting specific data on the occurrence of congenital anomalies
(detected during neonatal period) other than those of the thyroid gland
by using a specific reporting form. In this study, the classification of
additional congenital malformations (CM) as minor or major malfor-
mations was performed as previously described (11).
Zygosity assessment
Because the INRICH is a pathology register performing the surveil-
lance of CH in our country, no information is regularly collected about
the co-twins found negative at screening. This did not allow us to know
how many twin pairs were the same or opposite sex. However, it was
possible to know zygosity in 24 of the 80 CH twins recorded in the
Register between 1989 and 2000. In 15 of these 24 twins, zygosity in-
formation was obtained by follow-up center, which they referred on the
basis of a strikingly similar physical appearance beyond the neonatal
period and placenta information. For the remaining nine cases, parents
gave consent to analyze DNA. Zygosity was assigned using fingerprint-
ing only in same-sex twin pairs on the basis of nine short tandem repeats
(microsatellites) all localized on different chromosomes. For typing of
microsatellites, PCR were performed using forward primers labeled at
the 5⬘ end with fluorescent dye. The diluted (1:200) amplified products
were run by capillary electrophoresis on an ABI prism 310 genetic
analyzer (Applied Biosystems, Foster City, CA). Genotyping was per-
formed by GENESCAN and GENOTYPER softwares (Applied Biosys-
tems). Probability of identity by chance is approximately 10⫺4.
Statistical analysis
Pairwise concordance rate of CH, calculated as the proportion of
concordant pairs over the sum of concordant and discordant pairs, and
95% confidence intervals (CI) were estimated in twins. The recurrence
risk for CH was also calculated among siblings of affected babies with
CH (SRR). For this estimate, twins were assumed as siblings. The SRR
represents the ratio between the probandwise concordance rate among
siblings of CH babies and the Italian CH prevalence.
Student’s t test was used to evaluate differences between mean values
of continuous variables, whereas differences between proportions were
examined statistically with ␹2 test. The log-normal transformation was
used for not normally distributed data. The percentage mean increase
in TSH levels between screening and diagnosis was estimated as
(TSHdiagnosis⫺ TSHscreening/TSHscreening) ⫻ 100. A multiple regression
analysis was also performed.
Results
Between 1989 and 2000, 2159 CH babies were diagnosed
in Italy by neonatal screening and recorded in the INRICH.
All of the babies received the replacement therapy with l-T4.
Among them, 80 were twins, mostly discordant for CH. Only
three couples (all boys with in situ gland and unknown
Olivieri et al. • Congenital Hypothyroidism in Twins
zygosity) and one triplet (two girls and one boy with in situ
gland) from a medically induced pregnancy were concordant
for CH at birth. The proportion of multiple deliveries ob-
served in the CH population between 1989 and 2000 (75 of
2154, 3.5%) was 3-fold higher than that (1.1%) estimated in
the Italian general population in the same period (12).
Between 1991 and 2003, a reevaluation of the diagnosis, at
the age of 2–3 yr after a withdrawal of the replacement
therapy, was performed in 267 CH infants who presented a
suspicion of TH during the first years of life. The prevalence
of twins was 1.9% (three females and one male) in the 214
Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018
reevaluated infants who were affected by permanent CH and
13.2% (five females and two males) in the 53 babies who were
affected by TH (P ⬍ 0.01). Among these 53 babies, there was
the triplet concordant at birth. Seventy-three permanent CH
twins were then considered for the statistical analysis pre-
sented in this study (Fig. 1). It is important to note that,
because statistical analysis was performed after 2003, virtu-
ally all babies with TH born between 1989 and 2000 were
diagnosed and excluded from the analysis.
After excluding the 53 TH babies, the frequency of mul-
tiple deliveries in the CH population was reestimated. It was
still higher (3.3%; 95% CI, 2.6 – 4.2%) than that found in the
general population. Moreover, it has been possible to esti-
mate the CH incidence both in singleton and multiple de-
liveries by comparing the INRICH data with data concerning
deliveries in the Italian general population, given by the
Italian Central Institute of Statistics. The estimated CH in-
cidence was 3.2 per 10,000 singleton and 10.1 per 10,000 twin
live births. A relative risk of CH occurrence in twin deliveries
of 3.1 (95% CI, 2.5–3.9) was also calculated. As already dem-
onstrated in other patient series, also in the Italian population
of CH twins, the pairwise concordance rate for permanent
CH was very low (4.3%; 95% CI, 0.93%–7.6%) and due to only
three concordant couples (all boys with in situ gland). Besides
these three couples of twins, in the INRICH were also present
eight siblings concordant for CH. Specifically, four were
concordant for thyroid dysgenesis (two couples with ectopic
thyroid in one brother and agenesis in the other) and four
concordant for in situ gland (two couples). Given the fact that
zygosity was unknown in most of the 73 twins with CH,
twins were considered as siblings to calculate the genetic risk
for CH. The estimated SRR was 35.4 (95% CI, 4.7–269.3),
indicating a higher risk for CH in siblings of affected babies
than in a baby of the general population. It is important to
note that assuming twins as siblings may represent a possible
underestimation of the real recurrence risk in twins.
Among the 24 twins with known zygosity (all discordant
for CH), 19 had permanent CH (four monozygous, MZ; 15
dizygous, DZ) and five were affected by TH. Among these,
one was MZ (female) and four DZ (3 females, 1 male). A
description of the 19 twins with permanent CH is shown in
Table 1.
Comparison between the Italian births and CH population
Comparison of neonatal features between CH population
(n ⫽ 2106) and the Italian births (available data for the period
1991–1996) is shown in Table 2. As expected, the sex ratio
(female/male) was significantly higher in singleton CH ba-
Olivieri et al. • Congenital Hypothyroidism in Twins J Clin Endocrinol Metab, August 2007, 92(8):3141–3147 3143

CH infants recorded in the INRICH 1989 - 2000

2106 53
Permanent CH Transient CH

Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018

2033 73 46 7
Singletons Twins Singletons Twins

3 3
67 4
Concordant Concordant
Discordant Discordant
for CH for CH
for CH for CH
(3 couples) (one triplet)

FIG. 1. CH infants recorded in the INRICH from 1989 –2000.

bies than in singletons from the general population, whereas
no significant differences were observed between CH twins
and the Italian twin newborns. Mean values of birth weight
and gestational age were found significantly lower in sin-
gletons and twins with CH than in the Italian singletons and
twins, respectively. As concerns maternal age, it was signif-
icantly higher in CH singletons than in the Italian single
babies, whereas it was similar in both groups of twins.
Clinical features in twin and singleton CH babies
Thyroid scintigraphy and/or ultrasound evaluation were
performed before starting therapy in 51% of twins and in 61%
of singletons with CH. In Fig. 2, the distribution of CH
diagnoses are shown. Although a similar frequency of ec-
topic gland was found in both singletons and twins, a sig-
nificantly higher prevalence of in situ thyroid as well as a
lower frequency of agenesis were observed (␹2 ⫽ 8.02; P ⬍
0.02) in twin than in singleton CH babies. Among babies with
in situ gland, the frequency of hypoplasia, representing a
further manifestation of thyroid dysgenesis, was similar in
twins and singletons (16 and 19%, respectively). Among the
16 twins with known zygosity and CH diagnosis (Table 1),
eight had thyroid dysgenesis, and among these, four were
MZ (two agenesis, two ectopy). Concerning the mean age at

TABLE 1. Subgroup of permanent CH twins with known zygosity

Twins Gender Zygosity Diagnosis TSH screening TSH diagnosis

a
1 F MZ Agenesis 61 199
2 F MZ Agenesisa 91 395
3 F MZ Ectopy 57 53
4 F MZ Ectopy 49 200
5 M DZ Ectopy 140 114
6 F DZ Ectopy 169 147
7 F DZ Ectopy 50 200
8 F DZ Ectopy 260 310
9 F DZ In situ 6 6b
10 F DZ In situ 350 100
11 M DZ In situ 30 20
12 F DZ In situ 11 30
13 F DZ In situ 14 525
14 F DZ In situ 47 210
15 F DZ In situ 220 380
16 F DZ In situ 408 414
17 M DZ Unknown 10 200
18 M DZ Unknown 177 100
19 M DZ Unknown 51 100
F, Female; M, male.
a
Thyroid scintigraphy confirmed by ultrasound evaluation.
b
With serum FT4 and FT3 below the normal range.
3144 J Clin Endocrinol Metab, August 2007, 92(8):3141–3147
TABLE 2. Neonatal features in permanent CH babies and the Italian births
Singletons
(n ⫽ 2033)
Sex ratio (F/M) 1.73
Birth weight, mean ⫾ SD (g) 3,233 ⫾ 658c
Male 3,205 ⫾ 638c
Female 3,286 ⫾ 681c
Gestational age, mean ⫾ SD (wk) 39.2 ⫾ 2.7c
Maternal age, mean ⫾ SD (yr) 29.3 ⫾ 5.3c
a
CH infants recorded in the INRICH between 1989 and 2000.
b
Available data concerning neonatal features in the Italian general population,1991–1996.
c
CH singletons vs. Italian singletons, P ⬍ 0.01.
d
CH twins vs. Italian twins, P ⬍ 0.01.
starting therapy, it was similar in both singleton and twin CH
babies (26.4 ⫾ 25.7 and 27.8 ⫾ 13.6 d, respectively) as well as
the median values (22 and 25.5 d, respectively).
Table 3 shows T4 and TSH mean levels at screening and
diagnosis. No significant difference between groups was
observed in T4 values either at screening or diagnosis.
However, a significant lower TSH mean concentration at
screening was found in the twin group than in singletons,
whereas no difference was present at diagnosis. The mean
percentage increase in TSH levels between screening and
diagnosis was then calculated. It was significantly higher
in the twin CH babies compared with singletons (308 vs.
129%, P ⬍ 0.01). To verify whether the low mean TSH
value found at screening in the twin group was a conse-
quence of the high frequency of in situ gland or the effect
of some factors related to the twin status, a multivariate
analysis was performed. The following variables were
considered: twin status, gender, gestational age, scintig-
raphy, and/or ultrasound diagnosis and age at screening.
Results of the multivariate analysis (Table 4) showed that
the variability of TSH levels found at screening was neg-
atively associated with the twin status independently from
the other considered variables (␤ coefficient ⫽ ⫺0.40; P ⫽
0.02).
Regarding additional CM, the frequency of CM was
estimated in the population of twin and singleton CH
babies diagnosed between 1991 and 2000 with the exclu-
sion of 15 babies with Down’s syndrome, given the known
association of this syndrome with transient thyroid dys-
function (13). Our data showed that the frequency of major
CH TWINS CH SINGLETONS
Eutopic
gland
Ectopy
(49%)
Ectopy (42%)
(38%)
Agenesis Agenesis
(13%) (29%)
FIG. 2. Thyroid scintigraphy and/or ultrasound diagnosis in twins
and singletons with permanent CH.
Olivieri et al. • Congenital Hypothyroidism in Twins
CH babiesa Italian birthsb
Twins Singletons Twins
(n ⫽ 73) (n ⫽ 3,194,887) (n ⫽ 71,891)
1.09 0.94 0.97
2,046 ⫾ 737d 3,290 ⫾ 496 2,371 ⫾ 567
2,011 ⫾ 744d 3,353 ⫾ 503 2,416 ⫾ 587
2,084 ⫾ 739d 3,221 ⫾ 479 2,323 ⫾ 559
33.6 ⫾ 4.4d 39.3 ⫾ 1.7 36.4 ⫾ 3.0
30.2 ⫾ 6.3 28.8 ⫾ 5.0 29.5 ⫾ 4.8
Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018
CM found in singletons and twins with CH was signifi-
cantly higher than that found in the Italian general pop-
ulation (1.0 –2.0%, including Down’s syndrome). How-
ever, differently from that found in the general population
in which twins are at greater risk of CM than singletons
(14), in the population of CH babies, the frequency of
major CM was similar in the two groups. It was 8.2% in
singletons and 9.2% in twins with CH. The CH diagnosis
in the six twins with additional CM were two in situ gland,
one ectopy, and three unknown. All the CH twins with
additional CM had unknown zygosity and were discor-
dant for CH. Nevertheless, a same sex pair (two girls) was
concordant for CM occurrence (index case with unknown
CH diagnosis and bilateral microphthalmia, euthyroid co-
twin with congenital cerebral cysts).
Finally, an analysis of maternal anamnesis during preg-
nancy has also been performed. No significant differences
between CH singletons and twins were found in the fre-
quency of maternal diabetes (3.6 and 3.5%, respectively),
hypothyroidism (included autoimmune etiology), and/or
goiter (6.1 and 4.5%, respectively), hyperthyroidism (2.1 and
1.5%, respectively), iodine exposure during pregnancy (3.3
and 5%, respectively) and smoking (10.3% and 13.0%,
respectively).
Discussion
In this study, the analysis of the INRICH data showed
a more than 3-fold higher frequency of twins in the CH
population than in the general population. According to
previously reported highly discordant twin series, also in
the Italian CH twin population, the concordance rate for
the disease was very low (4.3%). Specifically, among the
73 permanent CH twins diagnosed during the study pe-
Eutopic riod, only three couples (all boys with in situ gland and
gland
unknown zygosity) were concordant for CH. These find-
(29%)
ings strongly suggest that the sporadic occurrence of the
disease is likely due to noninheritable postzygotic events
that may include epigenetic modifications and early so-
matic mutations. However, a genetic risk for the disease
cannot be excluded. In fact, the high recurrence risk for CH
estimated in siblings of affected babies recorded in the
INRICH indicates that environmental risk factors may act
as a trigger on a susceptible genetic background. More-
over, the fact that one (with bilateral microphthalmia) of
the six CH twins with additional malformations had the
Olivieri et al. • Congenital Hypothyroidism in Twins
TABLE 3. T4 and TSH levels at screening and diagnosis in twins and singletons with permanent CH
Screening
T4, mean ⫾ SD (␮g/dl)
TSH, mean ⫾ SD (␮IU/ml)
Diagnosis
T4, mean ⫾ SD (␮g/dl)
TSH, mean ⫾ SD (␮IU/ml)
% increase in TSH levels between screening and diagnosis
NS, Not significant.
a
Log normal TSH.
same-sex co-twin (F-F) concordant for CM occurrence
(congenital cerebral cysts) supports the hypothesis that
genes involved in the development of thyroid and other
organs may be affected during the early stages of embry-
ogenesis. The recently reported significant association be-
tween CH and congenital anomalies of nervous system,
eyes, and heart, which represent the most precocious
structures in the developing embryo (11), further supports
this hypothesis. Again, the possible involvement of genes
controlling development of thyroid and other organs in
the etiology of CH is further supported by the fact that no
difference was found in the frequency of CM between
twins and singletons in the CH population. This finding is
different from that found in the general population in
which twins are more likely than singletons to be affected
by congenital malformations, events that represent one
major cause of mortality and serious morbidity in twin
early life (14, 15). Finally, the fact that eight of the 16
permanent CH twins with known zygosity and CH diag-
nosis (Table 1) had thyroid dysgenesis, and among these,
four were MZ (two agenesis, two ectopy), seems to suggest
a higher risk of thyroid dysgenesis in MZ than in DZ twins.
However, additional studies are needed to confirm this
hypothesis and to verify whether the frequency of MZ
twins in the CH population is the same as that observed
in the general population.
The analysis of the INRICH data showed that the oc-
currence of in situ thyroid was significantly more frequent
in twin than in singleton CH babies. This finding can be
explained by the fact that iodine deficiency, which is still
present in our country as mild to moderate deficiency (16),
can have a role in the increasing risk for CH with in situ
gland in twins. It is well known that the physiological
increase in maternal thyroid activity during the first tri-
TABLE 4. Multiple linear regression of the effect of clinical and
neonatal features on log TSH levels at screening
Coefficient ␤ 95% CI
Twins ⫺0.40 ⫺0.75
Gender (male) ⫺0.10 ⫺0.22
Gestational age (wk) 0.02 ⫺0.004
Age at screening (d) 0.01 ⫺0.001
Diagnosis
Ectopy 0.95 0.80
Agenesisa 0.76 0.62
The analysis was performed on infants with thyroid scintigraphy
and/or ultrasound evaluation (number of observations ⫽ 970).
a
Thyroid scintigraphy confirmed by ultrasound evaluation.
J Clin Endocrinol Metab, August 2007, 92(8):3141–3147 3145
CH singletons CH twins P
3.7 ⫾ 2.7 3.2 ⫾ 2.3 NS
229 ⫾ 185 178 ⫾ 210 ⬍0.01a
3.4 ⫾ 2.9 3.2 ⫾ 2.6 NS
299 ⫾ 264 276 ⫾ 274 NS
129% 308% ⬍0.01
Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018
mester of pregnancy, reflected by a decrease in the circu-
lating concentrations of TSH and an increase in FT4 and
FT3 levels (17), is more profound in twin than in singleton
pregnancy. This is probably because peak concentrations
of human chorionic gonadotropin are greater and remain
increased for longer in twin pregnancies (18). Moreover, it
has been demonstrated that in multiple pregnancy, ma-
ternal thyroid activity increases with fetal numbers (19). It
is known that an even mild iodine deficiency may result
in relative hypothyroxinemia during pregnancy (20) and
that during the first trimester, the fetus is exposed to
concentrations of FT4 ultimately depending on the circu-
lating maternal T4 and FT4 levels (21). Therefore, given the
high discordance rate for CH, we cannot exclude the pos-
sibility that competitive conditions regarding metabolic
factors in utero may occur and that differences in the sup-
ply of blood, oxygen, or other nutrients (including iodine)
may alter or disrupt the developmental processes of the
thyroid gland.
Our study demonstrated an increased risk for both per-
manent and transient CH in multiple than in single de-
liveries, and for the first time it was possible to estimate
the incidence of CH in multiple and single deliveries sep-
arately. It was found 3-fold higher in multiple (10.1 in
10,000 live births) than in single deliveries (3.2 in 10,000
live births). Moreover, the analysis of the reevaluated in-
fants with high suspicion of TH recorded in the INRICH
showed a high prevalence of twins among infants who
were affected by TH. This finding confirms an additional
risk of TH in multiple compared with single deliveries. It
is well known that newborn twins are at greater risk of
morbidity than singletons because of their increased
chance of prematurity. Also, thyroid function can be af-
fected by prematurity. In our study, the comparison of the
INRICH data set with available data of the Italian births
showed a lower birth weight and gestational age in twins
with permanent CH than in twins from the general pop-
ulation. This finding is consistent with the fact that the risk
⫺0.05 of not only transient but also permanent CH is higher in
⫺0.01 babies with low birth weight (22–24). Therefore, as im-
0.53 provements in perinatal and neonatal care have increased
0.02 the survival rate of an increasing number of preterm ba-
1.11 bies, to improve the long-term developmental outcome of
0.89 these babies, either permanent or transient forms of hy-
pothyroidism need to be treated as soon as they are iden-
tified to avoid consequences of the early thyroid deficit.
This implies rapid and effective screening procedures.
3146 J Clin Endocrinol Metab, August 2007, 92(8):3141–3147
At present, the mass screening procedures allow us to
virtually diagnose all infants with transient or permanent
forms of CH. However, a second sample for CH screening
at 14 d of age in at least same-sex twins has been suggested
(25). This is because fetal blood mixing between twins may
occur. Specifically, the vascular connections present in at
least 70% of monozygotic placentas may result in delayed
or missed CH diagnoses during the first days of life (25).
In fact, vascular connections between a hypothyroid and
a euthyroid twin could allow the transfer of T4 between
fetuses and, hence, maintain a normal TSH level in the
hypothyroid one until a few days after delivery, causing
a lowering of TSH at screening in the twin with transient
or permanent form of CH. Our findings confirm a high
discordance rate in the CH population, and a significantly
lower TSH mean level at screening was found in the twin
group than in singletons. After adjustment by multivariate
analysis, a significant association of low TSH levels at
screening with the twin status was confirmed. Moreover,
the mean percentage increase in TSH between screening
and diagnosis was higher in twins as compared with sin-
gletons. Although we could not separate monochorionic
twins from the others, taken together, these results may
suggest the possible occurrence of fetal blood mixing be-
tween a hypothyroid and a euthyroid fetus and underline
the need for special guidelines for twin CH screening
procedures. However, if we assume that in the CH pop-
ulation there is a frequency of MZ twins similar to that
observed in the general population (about 30% of all
twins), the occurrence of fetal blood mixing may explain
the low TSH at screening only partially. Possible other
factors related to the twin status remain to be clarified.
In conclusion, this is the first report presenting such
population-based results on CH in multiple deliveries. The
high discordance rate found in the INRICH data set as well
as in other series of CH babies demonstrates the impor-
tance of environmental risk factors and suggests the oc-
currence of noninheritable postzygotic events in the eti-
ology of CH. Nevertheless, the results concerning the high
incidence of CH in multiple deliveries are worthy of in-
terest for the high number of induced pregnancies, in Italy
as well in other Western countries (26, 27), because of the
increasing use of techniques of assisted reproduction and
drugs inducing ovulation. In this view, great attention
should be paid to multiple pregnancies, which are at
greater risk for both permanent and transient CH than
single pregnancies.
Acknowledgments
The skillful technical assistance of Mrs. F. Latini and Mrs. D. Rotondi
is gratefully acknowledged.
Received January 31, 2007. Accepted May 2, 2007.
Address all correspondence and requests for reprints to: Antonella
Olivieri, Dipartimento di Biologia Cellulare e Neuroscienze, Istituto
Superiore di Sanità, Viale Regina Elena, 299, 00161 Roma, Italy. E-mail:
antonella.olivieri@iss.it.
The Study Group for Congenital Hypothyroidism includes R. Al-
tamura (Brindisi), U. Angeloni (Roma), I. Antonozzi (Roma), M.
Baserga (Catanzaro), R. Berardi (Siena), S. Bernasconi (Parma), G.
Bona (Novara), M. Burroni (Fano), E. Cacciari (Bologna), F. Calaciura
Olivieri et al. • Congenital Hypothyroidism in Twins
(Catania), R. Caldarera (Messina), M. Cappa (Roma), M. R. Casini
(Cagliari), A. Cassio (Bologna), L. Cavallo (Bari), V. Cherubini (An-
cona), G. Chiumello (Milano), L. Chiovato (Pavia), M. Cicchetti (Cam-
pobasso), M. P. Cicciò (Messina), A. Cicognani (Bologna), G. V. Coppa
(Ancona), A. Coppola (Napoli), C. Corbetta (Milano), R. Cordova
(Potenza), A. Correra (Napoli), P. Costa (Roma), F. Dammacco (Bari),
F. De Luca (Messina), C. De Santis (Torino), S. Di Maio (Napoli), G.
Gallicchio (Potenza), R. Gastaldi (Genova), G. Giovannelli (Parma), G.
Grasso (Caltanissetta), R. Gurrado (Taranto), L. Lasciarrea (Bari), A.
Lelli (Roma), D. Leonardi (Catania), A. Liotta (Palermo), S. Loche
(Cagliari), R. Lorini (Genova), G. Manente (Taranto), G. Minelli (Fog-
gia), F. Monaco (Chieti), L. Moschini (Roma), M. A. Musarò (Siena),
G. C. Mussa (Torino), T. Narducci (Foggia), S. Pagliardini (Torino),
L. Palillo (Palermo), G. Parlato (Catanzaro), E. Pasquini (Firenze), L.
Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018
Peruzzi (Siena), S. Piazzi (Bologna), A. Pinchera (Pisa), M. Pizzolante
(Lecce), R. Puggioni (Cagliari), A. Rizzo (Lecce), G. Saggese (Pisa), D.
Sala (Napoli), C. Salerno (Napoli), R. Salti (Firenze), L. Sava (Catania),
D. Scognamiglio (Napoli), V. Stoppioni (Fano), L. Tatò (Verona), M.
Tonacchera (Pisa), R. Vigneri (Catania), G. Vignola (Potenza), M. C.
Vigone (Milano), C. Volta (Parma), and G. Weber (Milano).
Disclosure Statement: The authors have nothing to disclose.
References
1. La Franchi S 1999 Congenital hypothyroidism: etiologies, diagnoses, and
management. Thyroid 7:735–740
2. Sorcini M, Balestrazzi P, Grandolfo ME, Carta S, Giovannelli G 1993 The
National Register of infants with congenital hypothyroidism detected by
neonatal screening in Italy. J Endocrinol Invest 16:573–577
3. Derom R, Orlebeke J, Eriksson A, Thiery M 1995 The epidemiology of
multiple births in Europe. In: Keith LG, Papiernik E, Keith DM, Luke B, eds.
Multiple pregnancy. Epidemiology, gestation and perinatal outcome. New
York: Parthenon Publishing Group; 145–162
4. Hall JG 1996 Twinning: mechanisms and genetic implications. Curr Opin
Genet Dev 6:343–347
5. Umstad MP, Gronow MJ 2003 Multiple pregnancy: a modern epidemic?
Med J Aust 178:613– 615
6. Fisher DA 1999 Hypothyroxinemia in premature infants: is thyroxine treat-
ment necessary? Thyroid 7:715–720
7. Fisher DA, Polk DH, Wu SY 1994 Fetal thyroid metabolism: a pluralistic
system. Thyroid 4:367–371
8. Thorpe Beeston JG, Nicolaides KH, McGregor AM 1992 Fetal thyroid
function. Thyroid 2:207–217
9. ESPE Working Group on Congenital Hypothyroidism 1993 Guidelines for
neonatal screening programs for congenital hypothyroidism. Eur J Pediatr
152:974 –975
10. American Academy of Pediatrics 1993 Newborn screening for congenital
hypothyroidism: recommended guidelines. Pediatrics 91:1203–1209
11. Olivieri A, Stazi MA, Mastroiacovo P, Fazzini C, Medda E, Spagnolo A,
De Angelis S, Grandolfo ME, Taruscio D, Cordeddu V, Sorcini M, The
Study Group for Congenital Hypothyroidism 2002 A population-based
study on the frequency of additional congenital malformations in infants
with congenital hypothyroidism: data from the Italian Registry for Con-
genital Hypothyroidism. J Clin Endocrinol Metab 87:557–562
12. Astolfi P, Ulizzi L, Zonta LA 2003 Changes in twinning rate: Italy 1950 –
1996. Hum Reprod 1:207–211
13. Fort P, Lifshitz F, Bellisario R, Davis J, Lanes R, Pugliese M, Richman R,
Post EM, David R 1984 Abnormalities of thyroid function in infants with
Down’s syndrome. J Pediatr 104:545–549
14. Mastroiacovo P, Castilla EE, Arpino C, Botting B, Cocchi G, Goujard G,
Marinacci C, Merlob P, Matneki J, Mutchinick O, Ritvanen A, Rosano A
1999 Congenital malformations in twins: an international study. Am J Med
Genet 83:117–124
15. Doyle PE, Beral V, Botting B, Wale CJ 1991 Congenital malformations in
twins in England and Wales. J Epidemiol Community Health 45:43– 48
16. Vitti P, Rago T, Aghini-Lombardi F, Pinchera A 2001 Iodine deficiency in
Europe. Pub Health Nutr 4:529 –535
17. Glinoer D, De Nayer P, Bordoux P, Lemone M, Robyn C, Van Steirteghem
A, Kinthaert J, Lejeune B 1990 Regulation of maternal thyroid during
pregnancy. J Clin Endocrinol Metab 71:276 –287
18. Grun JP, Meuris S, De Nayer P, Glinoer D 1997 The thyreotrophic role of
human chorionic gonadotropin (hCG) in the early stages of twin (versus
single) pregnancies. Clin Endocrinol 46:719 –725
19. Ogueh O, Hawkins AP, Abbas A, Carter GD, Nicolaides KH, Johnson MR
2000 Maternal thyroid function in multiple pregnancies before and after
fetal reduction. J Endocrinol 164:7–11
20. Glinoer D 2004 The regulation of thyroid function during normal preg-
nancy: importance of the iodine nutrition status. Best Pract Res Clin En-
docrinol Metab 18:133–152
Olivieri et al. • Congenital Hypothyroidism in Twins
21. Calvo RM, Janiaux E, Gulbis B, Asuncion M, Gervy C, Contempre B,
Morreale de Escobar G 2002 Fetal tissues are exposed to biologically
relevant free thyroxine concentration during early phases of development.
J Clin Endocrinol Metab 87:1768 –1777
22. Larson C, Hermos R, Delaney A, Daley D, Mitchell M 2003 Risk factors
associated with delayed thyrotropin elevations in congenital hypothyroid-
ism. J Pediatr 143:587–591
23. Medda E, Olivieri A, Stazi MA, Grandolfo ME, Fazzini C, Baserga M,
Burroni M, Cacciari E, Calaciura F, Cassio A, Chiovato L, Costa P, Leo-
nardi D, Martucci M, Moschini L, Pagliardini S, Parlato G, Pignero A,
Pinchera A, Sala D, Sava L, Stoppioni V, Tancredi F, Valentini F, Vigneri
R, Sorcini M 2005 Risk factors for congenital hypothyroidism: results of a
population case-control study (1997–2003). Eur J Endocrinol 153:765–773
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
J Clin Endocrinol Metab, August 2007, 92(8):3141–3147 3147
24. Waller DK, Anderson JL, Lorey F, Cunningham GC 2000 Risk factors for
congenital hypothyroidism: an investigation of infants birth weight, eth-
nicity and gender in California 1990 –1998. Teratology 62:36 – 41
25. Perry R, Heinrichs C, Bourdoux P, Khoury K, Szots F, Dussault JH,
Vassart G, Van Vliet G 2002 Discordance of monozygotic twins for thyroid
dysgenesis: implication for screening and for molecular pathophysiology.
J Clin Endocrinol Metab 87:4072– 4077
26. Corchia C, Mastroiacovo P, Lanni R, Mannazzu R, Curro V, Fabris C 1996
What proportion of multiple births are due to ovulation induction? A
register-based study in Italy. Am J Pub Health 86:851– 854
27. Reynolds MA, Schieve LA, Martin JA, Jeng G, Macaluso M 2003 Trends
in multiple births conceived using assisted reproductive technology,
United States, 1997–2000. Pediatrics 111:1159 –1162
Downloaded from https://academic.oup.com/jcem/article-abstract/92/8/3141/2598099 by guest on 28 November 2018