Maternal hypothyroidism early in pregnancy was associated with fetal distress later in labor requiring emergency c-section delivery. Women who were severely hypothyroid (FT1 <0.6) at initial presentation were more likely to experience fetal distress compared to those who were mildly hypothyroid or euthyroid (3% vs 56%). Fetal distress correlated with low FT1 and high TSH at initial presentation, as well as low hematocrit at delivery, but was independent of these factors later in pregnancy. Early adequate treatment of severe hypothyroidism in pregnancy is important to prevent fetal complications.

1. Clinical Endocrinology (1995) 42, 353-358
Perinatal consequences of maternal hypothyroidism in
early pregnancy and inadequate replacement
Nathan Wasserstrum and Carol A. Anania low maternal haematocrit on admission to labour and
Departments of Obstetrics and Gynecology and delivery (P < 0 0 5 ) , but was independent of haematocrit
Medicine, Baylor College of Medicine, Houston, Texas, and gestational age at initial presentation, neonatal
USA weight percentile, and gestational age at birth.
CONCLUSIONS Severe maternal hypothyroidism early in
(Received 25 May 1994;returned for revision 1 August lQ94;
finally revised 17 November 1994;accepted 3 January 1995) gestation is strongly assoclated with fetal distress in
labour. This suggests that (1) Inadequate maternal
replacement leads to fetal distress and (2) maternal
Summary thyroid status In early gestation may exert Irreversible
effects on the fetus, the placenta, andlor on subsequent
Maternal hypothyroidism may be associated with a maternal adaptations to pregnancy. Early adequate
variety of pregnancy compiicatlons. replacement therapy is especially prudent in pregnant
OBJECTIVE W have evaluated the perinatai con-
e women presenting with severe hypothyroidism.
sequences of maternal hypothyroidism in early and late
gestation.
DESIGN Retrospective study of pregnant women, wlth Since hypothyroidism is often complicated by anovulation,
quasl-experlmentaidesign comparlng dlfferent subJects. hypothyroidism in pregnancy has been considered an
SUBJECTS Forty-three pregnancies in 42 women with infrequent phenomenon (Hall et al., 1993). Thyroid
hypothyroidism - either biochemically hypothyroid or deficiency in pregnant women in the United States is
with a history of hypothyroidism on adequate replace- estimated at 2.5% (Klein et al., 1991). In addition, recent
-
ment and no other preexisting medical conditions. data indicate that a large proportion of women with type I
MEASUREMENT Free thyroxine index (FTI), TSH, and diabetes and normal thyroid function before conception
haematocrit at initial antepartum presentation and near develop subclinical hypothyroidism during pregnancy
term gestation. Pregnancy outcome variables including: (Jovanovic-Peterson & Peterson, 1988).
rate of Caesarean section performed for fetal distress in Maternal hypothyroidism has reportedly been associated
labour, neonatal welght percentlie, and gestational age at with higher incidence of abortions, still births, congenital
blrth. malformations, maternal anaemia, pregnancy induced
RESULTS Of 42 hypothyroid pregnancies, six were com- hypertension, and post-partum haemorrhage (Jones &
plicated by fetal distress in labour leading to Caesarean Man, 1969; Davis et al., 1988). Perinatal outcomes
section. Five of these six were severely hypothyroid reported in recent investigations (Montoro et al., 1981;
(defined as FTi < 0.6 (normal range 1.1-4.4)) on initial Davis et al., 1988; Balen & Kurtz, 1990; Leung et al., 1993)
antepartum presentation. in contrast, of the 36 preg- have been more favourable than in earlier studies (Jones &
nancies without fetal distress, only four initially pre- Man, 1969).
sented severely hypothyroid (P<0001). Conversely, The gestational age at which maternal illness occurs can
56% (519) of .pregnancies which initially presented determine subsequent outcome. In maternal diabetes
severely hypothyroid were subsequently complicated by mellitus, perinatal consequences depend not only on the
Caesarean section for fetal distress in labour. This degree of hyperglycaemia,but also on the gestational age at
compared to 3% (1133) among those who presented which it occurs (O’Sullivan et al., 1992). In hypothyroidism,
either mildly hypothyroid or euthyroid on replacement pregnancy induced hypertension appears to be associated
(P < 00001). Fetal distress correlated with low n l with inadequate replacement in late, rather than in early,
(P < 0001) and high TSH at initial presentation. How- gestation (Leung et al., 1993). In contrast, the importance of
ever, It was independent of FTI near term. A relation maternal hypothyroidism in early gestation is supported by
between fetal distress and TSH near term did not reach observations that the progeny of hypothyroid women on
statistical significance. Fetal distress also correlated wlth adequate replacement only later in gestation, have ocular
Correspondence: Dr Nathan Wasserstrum, 6437 Mercer Drive, and other neurological deficits (Man et al., 1991).
Houston, Texas 77005, USA. Fax: 713-432-7758. We compared the effect of maternal thyroid status at
0 1995 Blackwell Science Ltd 353

2. 354 N. Wasserstrum and C. A. Anania
initial antepartum presentation with that near term on included women who were biochemically euthyroid on
perinatal outcome in hypothyroid pregnant women. replacement throughout pregnancy, we refer to the full set of
subjects simply as 'hypothyroid'.
Haematocrit was measured on initial antepartum
Materlals and methods
presentation to the clinic and on admission to labour and
We reviewed the records of 43 pregnancies in 42 delivery. Routine antepartum management included
predominantly indigent women with hypothyroidism, initiation of iron and prenatal vitamin supplements at the
referred to the Baylor-Hams County Hospital District first visit. Labour and delivery were managed by the same
obstetric medical high risk clinic. Pregnant women were resident physicians for all patients and this management was
studied who on initial antepartum presentation had a unaffected by the diagnosis of maternal hypothyroidism.
history of hypothyroidism and were biochemically Fetal distress was defined as a clinical diagnosis made by the
hypothyroid and/or on thyroid hormone supplements at attending physician that was based on an abnormal fetal
presentation. Patients with a history of other medical heart pattern in labour and led to emergency Caesarean
problems, including diabetes and chronic hypertension, section. Neonatal weight at the time of delivery was
were excluded. The study was approved by the institutional recorded and the neonatal weight percentiles were calcu-
review board. lated (NCHS, 1976). Gestational ages (GA) were deter-
Free thyroxine index (FTI) and TSH were measured at mined by last menstrual period and confirmed by
presentation and several times throughout pregnancy. FTI, ultrasound. Results were expressed as mean f standard
rather than FT4 by direct measurement, was used because of deviation, unless otherwise noted. Statistical analysis was by
institutional cost considerations at the time patients were unpaired t-test and chi-squared as indicated (P< 0.05 taken
seen. Serum thyroxine was measured by radioimmunoassay as significant).
(RIA) kits (Diagnostic Products, Los Angeles; and Baxter Given the limitations inherent in retrospective studies, we
Travenol Diagnostics, Cambridge, Mass: normal non- carefully chose for analysis only those variables that were
pregnant values 4.4- 125 pg/dl (60- 160nmo1/1)) and T3 reliable and unambiguous, and did not analyse data that did
resin uptake (T3RU) with a solid-phase quot;'I test (Coat-A- not meet these criteria (e.g. estimates of the occurrence of
Count kit, Diagnostic products; normal non-pregnant mild preeclampsia were not made). Since the mean age at
values 25535%). FTI (normal non-pregnant values 1.1- referral was 19 weeks, data regarding spontaneous
4.4) was calculated by the laboratory as the product of the abortions would be biased by the absence of patients who,
serum thyroxine concentration (in pg/dl) and the T3RU, having already suffered early losses, were not referred to this
and the values of both components and the product were clinic. Hence, such data were not collected.
reported. TSH was measured by immunoradiometric assay
(IRMA; normal non-pregnant levels at our institution: 0.4-
Results
4.8 mU/l). The highest TSH reported by the laboratory was
> 60, and higher titres were not quantified. For purposes of The nine pregnancies that initially presented to the High
statistical analysis, > 60 was treated as equal to 60. Risk Clinic with severe hypothyroidism (FlT 0.4f0.2
Since the appearance of sensitive assays, TSH has been (mean fSD)) were compared (Table 1) to the 34 (33
considered superior for defining the severity of hypothyr- patients) with milder hypothyroidism (FTI 2.3 f 0.9). Fifty-
oidism after the first trimester (Carr et al., 1988; Surks et al., six per cent (5/9) of women who presented with severe
1990; Mandel et al., 1993). However, we used FTI to initially hypothyroidism subsequently underwent Caesarean section
categorize patients according to severity of hypothyroidism for fetal distress in labour, compared to only 3% (1/33) of
because (1) although the mean gestational age at presenta- the mildly hypothyroid and euthyroid group (P< 0.0001,
tion was 19 weeks, several patients presented in the first chi-squared, Table 1). The 3% rate of Caesarean section for
trimester, when high hCG may cause suppression of fetal distress equalled that of our general population. The
pituitary TSH; (2) the laboratory reported high TSH severity of hypothyroidismat presentation, herein defined in
values simply as > 60, without further quantification and terms of FTI, was also reflected in the concomitant TSH
(3) preliminary retrospective analysis had suggested marked (P< 0.0001, Table 1). Severity of hypothyroidism at initial
differences in outcome in the subset of patients defined by a presentation did not correlate with gestational age at either
FTI ( . . Hence, a FTI G0.6(normal range 1-1-4.4) was
06 presentation or birth, with neonatal weight percentile, or
chosen to define the set of severely hypothyroid subjects, with FTI in late gestation. In contrast, more severe
and a FTI of 0-7-4.4 to define the remaining set of mildly hypothyroidism at initial presentation was associated with
hypothyroid and euthyroid subjects. Although subjects a higher abnormal TSH in late gestation.
0 1995 Blackwell Science Ltd, Clinical Endocrinology,42,4

3. Perinatal consequences of maternal hypothyroidism 355
Table 1 Comparison of patients who were severely hypothyroid at antepartum presentation with those who were mildly hypothyroid or
euthyroid
Severely hypothyroid at Mildly hypothyroid or euthyroid
presentation at presentation P
Caesarean section for fetal distress 56% (5/9) 3% (1/33)
Fetal distress or still birth 56% (5/9) 6% (2/34)
Gestational age at presentation (weeks) 19.8 f 6.3 19.0 f 8.3
TSH at presentation (mu/]) 53.0 f 11.0 11.3f 13.0
Haematocrit at presentation (%) 31.6f5.5 33.9 f 4.0
FTI near term 1.7 f 0.04 1.8 f 0.8
TSH near term (mU/1) 17.8 f 8.8 7.3 f 6.1
Haematocrit on admission to labour and delivery (%) 31.3 f 4 . 9 34.9 f 4.9
Gestational age at birth (weeks) 39.2 f 2.2 39.1 f 2.5
Neonatal wt percentile at delivery 58 f 20 51 f 2 6
Data are expressed as mean f S D .
* Chi-squared.
NS, Non-significant.
t c-test.
FTI, Free thyroxine index; TSH, thyroid stimulating hormone (highest TSH reported by lab was >60; this occurred in 5 patients); fetal
distress, Caesarean section performed for fetal distress in labour.
Severe hypothyroidism defined as FTI G0.6.
The converse analysis, with fetal distress as the indepen- Similarly, four of six pregnant women with fetal distress had
dent variable (Table 2), revealed a strong correlation initially presented with TSH > 60, while only 1/36 without
between fetal distress and severity of hypothyroidism at fetal distress had such a high value (P< 0.001;chi-squared).
presentation, assessed both by low FTI (P< 0.001; r-test) A relation between TSH near term and fetal distress is
and high TSH (P< 0.0001; t-test, Table 2). Indeed, 5/6 suggested but did not reach statistical significance. Fetal
pregnant women who underwent Caesarean sections for distress correlated with lower haematocrit on admission to
fetal distress had initially presented with FTI GO.6; only labour and delivery, but was independent of haematocrit on
4/36 of those without fetal distress presented with such initial presentation (Table 2).
severe hypothyroidism (P< 0.001, chi-squared). In con- Other complications included a single intrauterine fetal
trast, fetal distress was not related to FTI near term. death at 28 weeks gestation in a patient who had presented
Table 2 Comparison of hypothyroid women who underwent Caesarean section for fetal distress, with those without this complication
Underwent Caesarean Did not develop
section for fetal distress fetal distress P
FTI at presentation 0.8 f 0.8 2.1 f 0.8
Gestational age at presentation (weeks) 17.2 f 3.6 19.0 f 8.3
TSH at presentation (mU/I) 47 f 20 11 f 13
Haematocrit at presentation (%) 30 f 2.2 33.8 f 4.0
FTI near term 1.5 f 0.3 1.9 f 0.8
TSH near term (mu/]) 21.2 f 9.9 7.3 f 6.1
Haematocrit on admission to labour and delivery (%) 29.2 f 3.9 34.9 f 4.9
Gestational age at birth (weeks) 40 f 0.53 39.0 f 2.5
Neonatal wt percentile at delivery 53 f 22 51 f 2 6
* Data expressed as mean fSD.
NS, Non-significant.
t t-test.
FTI, Free thyroxine index. The patient with a still birth at 28 weeks is not included in this table.
0 1995 Blackwell Science Ltd, Clinical Endocrinology, 42,4

4. 356 N. Wasserstrum and C. A. Anania
euthyroid on replacement. Preterm rupture of membranes not sufficiently low to account by itself for fetal distress. The
occurred at 30 weeks gestation in an initially severely significant association between fetal distress and term
hypothyroid woman. However, she did not develop fetal haematocrit may simply illustrate magnification by the
distress in labour. None of the pregnancies were increased physiologic demands of pregnancy on the intrinsic
complicated by severe preeclampsia or eclampsia, placental delays in normalization of homeostatic systems after
abruption, or post-partum haemorrhage. initiation of thyroid replacement (Herbert, 1986). Alterna-
tively, this association may reflect decreased compliance
with routinely prescribed nutritive supplements, in the
Discussion
group that developed fetal distress.
In this study, very low FTI at initial antepartum pres- In rats, maternal hypothyroidism in early gestation
entation correlated strongly with Caesarean section for fetal compromises subsequent maternal catabolic adaptations
distress in labour. Fetal distress was not related to FTI near in late gestation (Bonet & Herrera, 1991) and thereby affects
term, to gestational age at presentation and delivery, or to fetal homeostasis (Bonet & Herrera, 1988; 1991). Maternal
neonatal weight percentile. If fetal distress can largely be hypothyroidism in the first but not the second half of
accounted for by the presence of severe hypothyroidism in gestation is associated with abnormal fetal pituitary growth
early pregnancy, maternal hypothyroidism at this stage may hormone secretion (Bonet & Herrera, 1988).
exert effects that are not reversed by subsequent replacement Before the onset of fetal thyroid function, embryonic
therapy. Alternatively, one may focus on the apparent tissues in both rats and humans contain thyroid hormone
persistently less adequate thyroid replacement even at term receptors (Bernal & Pekonen, 1984; Perez-Castillo et al.,
in those with early severe hypothyroidism, that is, women 1985) and contain thyroxine and triiodothyronine of
who presented with severe hypothyroidism continued to maternal origin (Obregon et al., 1984; Woods et al., 1984;
have markedly higher TSH even near term (P< 0.02; Table Bernal & Pekonen, 1984; Morreale de Escobar et al., 1985).
1). This apparent inadequate replacement may reflect poor Developmental abnormalities in fetuses and neonates of
compliance in this subset of patients. Indeed, since all hypothyroid pregnancies may reflect absent or low levels of
patients had a history of hypothyroidism and should, thyroid hormones in the fetus before activation of its own
therefore, have already been using exogenous thyroxine at thyroid (Bernal & Pekonen, 1984).
the time of presentation, it follows that those who were most Man et al. (1991) reported that hypothyroid women who
severely hypothyroid at the time of presentation could also received adequate replacement only later in gestation had
have been the least compliant. progeny with ocular and other neurological deficits. The
Several mechanisms exist whereby maternal hypothyroid- investigators reasoned that significant maternal-fetal
ism early in pregnancy might exert irreversible effects. First, transfer of thyroxine normally occurs in human gestation
even in non-pregnant hypothyroid patients, variable delays and that maternal hypothyroidism early in gestation, before
occur between the return to biochemically euthyroid status maturation of the fetal thyroid-pituitary axis, led to the
and the subsequent normalization of multiple thyroid neurological deficits (Man et al., 1991). In light of these
dependent physiological systems, including control of results, they encouraged prenatal and/or early gestational
haematopoiesis and of plasma volume. Anaemia in non- screening for maternal hypothyroidism.
pregnant hypothyroid patients is characterized by reduc- Thyroid hormones condition vascular responsivenessand
tions in both red blood cl mass and plasma volume
el profoundly affect both the developing and adult human
(Herbert, 1986), and thyroid replacement sometimes raises circulation (Gardner et al., 1987; Klein, 1990). Inadequate
plasma volume before red cell mass. Hence, a low thyroid hormone levels early in gestation may produce
haematocrit does not adequately reflect the reduction in irreversible changes in the fetoplacental and/or fetal
total blood volume and it may fall further in the early stages vascular beds that would impair subsequent circulatory
of thyroid replacement. The anaemia may disappear only responses to the stress of labour. If only the fetoplacental
gradually over months after chemically euthyroid status is vasculature - which plays an essential role in fetal
attained (Herbert, 1986). In a pregnancy complicated by cardiovascular homeostasis in utero (Le et al., 1993), but
decrease in either maternal red blood cell mass or plasma obviously has no post-natal role - is primarily involved,
volume, impaired uteroplacental oxygen delivery can circulatory impairment after birth should be absent. Such
contribute to fetal distress. Although in our study fetal abnormalities may decrease physiologic reserve without
distress was associated with lower haematocrit (29.2) on impairing fetal growth. They would therefore be manifest in
admission to labour and delivery at term (YS 34.9 without response to acute stress such as that imposed by labour or by
fetal distress, P < 0.05; Table 2), haematocrit was probably antepartum fetal assessment with the contraction stress test
0 1995 Blackwell Science Ltd, Clinical Endocrinology,42,4

5. Perinatal consequences of maternal hypothyroidism 357
(OCT). This differs from a more severe stage, and/or In conclusion, severe maternal hypothyroidism early in
different type, of chronic uteroplacental insufficiency pregnancy was followed by fetal distress in labour. Since this
characterized by fetal growth retardation and/or by adverse effect was not prevented by achievement of a normal
abnormalities revealed in antepartum fetal assessment FTI near term it may reflect the delay in the adjustments of
even under unstressed conditions (e.g. non-stress test, multiple thyroid-responsive maternal physiological systems.
biophysical profile). Notably, despite the extremely high Alternatively, the persistent elevations in TSH may reflect
rate of Caesarean section for fetal distress in labour in the undertreatment. In either case, early adequate replacement
severely hypothyroid group, we found no growth retarded therapy is especially prudent in pregnant women presenting
neonates. Furthermore, neonatal weight percentile was with severe hypothyroidism. In addition, our results suggest
unrelated to severity of hypothyroidism or to frequency of the possibility that maternal thyroid status in early gestation
fetal distress (Tables 1 and 2). - an interval when maternal thyroxine may reach the fetus -
Investigators at the University of Southern California may exert irreversible effects on the fetus, placenta or uterus
(USC) reported successful outcomes in 9 of 11 pregnancies that impair their subsequent ability to tolerate the stress of
that initially presented with elevated TSH (Montoro et al., labour.
1981). In 4 of 11 pregnancies, initial FTI was G0.6; of these
four, one suffered an intrauterine fetal death at 30 weeks
gestation, and one underwent Caesarean section for fetal
bradycardia. In a subsequent prospective study of a large References
cohort of hypothyroid pregnancies, the USC group detected Balen, A.H. & Kurtz, A.B. (1990) Successful outcome of pregnancy
no increase in prematurity, still birth, abruption, or post- with severe hypothyroidism. Case report and literature review.
partum haemorrhage (Leung er al., 1993). However, they British Journal of Obstetrics and Gynaecology, 97,536-539.
did find an increase in preeclampsia, of particular note since Bernal, J. & Pekonen, F. (1984) Ontogenesis of the nuclear
3,5,3’-triiodothyronine receptor in the human fetal brain.
none of the patients had chronic hypertension or other
Endocrinology, 114, 611-619.
confounding medical illness. When other preexisting Bonet, B. & Herrera, E. (1988) Different response to maternal
medical conditions were excluded, an increased prevalence hypothyroidism during the first and second half of gestation in
of intrauterine growth retardation was accounted for by the the rat. Endocrinology, 122,450-455.
pregnancies complicated by gestational hypertension. We Bonet, B. & Herrera, E. (1991) Maternal hypothyroidism during the
first half of gestation compromises normal catabolic adaptations
also found no associations between maternal hypothyroid-
of late gestation in the rat. Endocrinology, 129, 210-216.
ism and prematurity, still birth, post-partum haemorrhage, Carr, D., McLeod, D.T., Parry, G. & Thornes, H.M. (1988) Fine
or growth retardation. adjustment of thyroxine replacement dosage: comparison of the
Adverse effects reported in 16 pregnancies deemed either thyrotrophin releasing hormone test using a sensitive thyro-
clinically or chemically hypothyroid included a remarkable trophin assay with measurement of free thyroid hormones and
clinical assessment. Clinical Endocrinology, 28, 325-333.
44% (7/16) incidence of pregnancy induced hypertension,
Davis, L.E., Leveno, K.J. & Cunningham, F.G. (1988) Hypo-
and high incidence of anaemia, placental abruption, thyroidism complicating pregnancy. Obstetrics and Gynecology,
intrauterine fetal death and post-partum haemorrhage 72, 108-112.
(Davis et a[., 1988). Of note, four of the seven overtly Gardner, D.G., Gertz, B.J. & Hane, S. (1981) Thyroid hormone
hypothyroid patients who developed preeclampsia were increases rat atrial natriuretic peptide messenger ribonucleic acid
accumulation in vivo and in vitro. Molecular Endocrinology, 1,
chronically hypertensive (Davis et al., 1988).
260-265.
The incidence of hypertension is increased in both Hall, R., Richards, C.J. & Lazarus, J.H. (1993) The thyroid and
pregnant and non-pregnant hypothyroid patients. Up to pregnancy. British Journal of Obstetrics and Gynaecology, 100,
50% of older non-pregnant hypothyroid women suffer from 512-515.
hypertension that improves with thyroxine replacement Herbert, V. (1986) The blood. In The Thyroid (eds S.H. Ingbar &
(Saito et al., 1983). Among hypothyroid pregnant women, L.E. Braverman) pp. 1162-1 168. J.B. Lippincott, Philadelphia.
Jones, W.S. Man, E.B. (1969) Thyroid function in human
&
those who develop preeclampsia have significantly higher pregnancy. VI Premature deliveries and reproductive failures of
TSH levels both initially and shortly before delivery (Leung pregnant women with low serum butanol extractable iodines.
et al., 1993). Furthermore, pregnant women with pre- American Journal of Obstetrics and Gynecology, 104,909-914.
eclampsia show an increased incidence of hypothyroidism Jovanovic-Peterson, L. & Peterson, C.M. (1988) De novo clinical
hypothyroidism in pregnancies complicated by type I diabetes,
(Lao et al., 1988). None of our patients had chronic
subclinical hypothyroidism, and proteinuria: a new syndrome.
hypertension, and none developed severe preeclampsia. American Journal of Obstetrics and Gynecology, 159, 4 2 - 4 6 ,
Because of methodological issues alluded to earlier, we did Klein, I. (1990) Thyroid hormone and the cardiovascular system.
not estimate the occurrence of mild preeclampsia. American Journal o Medicine, 88,631-637.
f
0 1995 Blackwell Science Ltd, Clinical Endocrinology, 42,4

6. 358 N. Wasserstrum and C. A. Anania
KIein, R.Z., Haddow, J.E., Faix, J.D., Brown, R.S., Hermos, R.J., weight and thyroid hormone content of rat embryonic tissues,
Pulkkinen, A. & Mitchell, M.L. (1991) Prevalence of thyroid before and after onset of fetal thyroid function. Endocrinology,
deficiency in pregnant women. Clininical Endocrinology, 35,41- 117, 1890-1900.
46. NCHS (1976) National Center for Health Statistics (NCHS)
Lao, T.T., Chin, R.K. & Swaminathan, R. (1988) Thyroid function Growth Charts. Monthly Vital Statistics Report, 25 (HRA76-
in pre-eclampsia. British Journal of Obstetrics and Gynaecology, 1120).
95, 880-883. Osullivan, M.J., Skyler, J.S., Raimer, K.A. & Abu-Hamad, A.
Le, S.Q., Wasserstrum,N., Mombouli,J. & Vanhoutte, P.M. (1993) (1992) Diabetes and Pregnancy. In Principles and Practice of
Contractile effect of endothelin in human placental veins: role of Medical Therapy in Pregnancy, second edition (ed. N. Gleicher),
prostaglandins and thromboxanes. American Journal of pp. 357-378. Appleton & Lange, Norwalk, Connecticut.
Obstetrics and Gynecology, 169,919-924. Obregon, M.J., Mallol, J., Pastor, R., Morreale de Escobar, G. &
Leung, AS., Millar, L.K., Koonings, P.P., Montoro, M. & Escobar del Rey, F. (1984) L-thyroxine and 3,5,3’-triiodo-~-
Mestman, J. (1993) Perinatal outcome in hypothyroid thyronine in rat embryos before onset of fetal thyroid function.
pregnancies. Obstetrics and Gynecology, 81, 349-353. Endocrinology, 114, 305-307.
Man, E.B., Brown, J.F. & Senmian, S.A. (1991) Maternal Perez-CastiIlo,A., BernaI, J., Ferreiro, B. & Pans (1985) The early
hypothyroxinemia: psychoneurological deficits of progeny. ontogenesis of thyroid hormone receptor in the rat fetus.
Annals of Clinical and Laboratory Science, 21,227-239. Endocrinology, 117, 2457-2461.
Mandel, S.J.,Brent, G.A. & Larsen, P.R. (1993) Levothyroxine Saito, I., Ito, K. & Saruta, T. (1983) Hypothyroidism as a cause of
therapy in patients with thyroid disease. Annals of Internal hypertension. Hypertension, 5, 112-1 15.
Medicines, 119, 492-502. Surks, M.I., Chopra, I.J., Mariash, C.N., NicoIoff, J.T. & Solomon,
Montoro, M., Collea, J.V., Frasier, S.D. & Mestman, J.H. (1981) D.H. (1990) American Thyroid Association guidelinesfor use of
Successful outcome of pregnancy in women with hypothyroidism. laboratory tests in thyroid disorders. JAMA, 263, 1529-1532.
Annals of Internal Medicine, 94, 31-34. Woods, R.J., Sinha, A.K. & Ekins, R.P. (1984) Uptake and
Morreale de Escobar, G., Pastor, R., Obregon, M.J. & Escobar del metabolism of thyroid hormones by the rat foetus in early
Rey, F. (1985) Effects of maternal hypothyroidism on the pregnancy. Clinical Science, 67, 359-363.
0 1995 Blackwell Science Ltd, Clinical Endocrinology, 42,4