Clinical Endocrinology (1992) 37, 331-334

Thyroid function tests are rarely abnormal in patients with severe hyperemesis gravidarum R. Wilson, J. H. McKlllop, M. MacLean', J. J. Walker', W. D. Frasert, C. Grayt, F. Dryburght and J. A. Thomson University Departments of Medicine, 'Obstetrics and tChemical Pathology, Glasgow Royal Infirmary, Glasgow, UK (Received 17 March 1992; returned for revision 76 April 1992; finally received 23 April 1992; accepted 13 May 1992)

Summary OBJECTIVES There is considerable controversy In the literature as to the cause of hyperemesisgravldarum. The aim of this project was to measure a range of thyrold hormone levels In a group of hyperemetic pregnant women. PATIENTS The study was carried out In 10 flrst trimester pregnant women wlth hyperemesls gravidarum. All had been admltted to hospltal due to the severlty of their symptoms. Fifty age matched, healthy flrst trlmester pregnant women were used as controls. MEASUREMENTS Blood samples from the women were analysed for total T3 (Tr3), total T4 (lT4), free 1 4 (Fr4), TSH, thyrotrophln receptor antibodies (TRAb), thyroid stlmulatlngantibodies(TSAb) and thyrold mlcrosomaland thyroglobullnantlbodles. Human chorlonlc gonadotrophin (hCG) levels were also measured. RESULTS Whlle Individual patients were found to have some abnormalthyroid function tests the group as a whole showed no conslstent pattern of abnormality and did not differ significantly from a group of healthy flrst trlmester pregnant women. hCG levels were also wlthln the normal range In the hyperemetic patients. DISCUSSION None of the women in thls study recelved any antlthyrold medication and thelr symptoms Improved as the pregnancy progressed. These results would suggest that there Is no underlylng thyrold abnormality In patients wlth hyperemesis gravldarum. It would appear that neither thyrold hormones, nor hCG contribute to the pathogenesls of the condltion.

nancy. In women suffering from hyperemesis gravidarum, which affects approximately 0.1 % of all pregnancies, there have been many reports of transient abnormal thyroid hormone levels (Thomson et al., 1989; Chin & Lao, 1988; Bruun et al., 1976; Fairweather et al., 1968). No consistent abnormality has emerged from these studies. In some instances the symptoms subsided without treatment, whilst in others the patients received a course of antithyroid drugs. The normal rise in serum total T4 which occurs in pregnancy is due to the effects of human chorionic gonadotrophin (hCG). As TSH and hCG share the same common a-subunit it has been suggested, but not proved, that hyperemesis may be due to elevated hCG levels. This is supported by the fact that hyperemesis occurs mainly in early pregnancy when hCG levels are at their highest. In spite of these many previous studies the mechanism responsible for hyperemesis gravidarum is not yet fully understood. This may be because, with few exceptions (Chin & Lao, 1988; Bober et al., 1986) the previous reports have been based on single cases suggesting selection bias. In addition, comparison of the reports has been made difficult by the wide variation in the severity of symptoms of the patients who were included. In an attempt to improve our understanding of the mechanism responsible for hyperemesis gravidarum we have studied hCG, thyroid hormone and thyroid antibody levels in 10 pregnant women with hyperemesis gravidarum. Patients and methods

Ten first trimester pregnant women, all of whom had been admitted to hospital due to the severity of the hyperemesis gravidarum, were enrolled into the study. Details of the patients' obstetric histories are given in Table 1. All patients remained in hospital for between 3 and 7 days. A normal range for each parameter was constructed from data obtained from 50 healthy, first trimester pregnant women. Blood (10 ml) was removed and serum stored at -20°C until analysis. Hormone and antibody measurement

In a normal pregnancy changes in thyroid hormone levels occur in the first trimester and continue throughout pregCorrespondence: Dr R. Wilson, Department of Medicine, Queen Elizabeth Building, Glasgow Royal Infirmary, 10 Alexandra Parade, Glasgow G31 2ER, UK.

The following parameters were measured in all patients: (1) Total T3 and total T4 were measured by conventional radioimmunoassay. The normal reference for the total T3 assay was 0.8-2.8 nmol/l and 55-144 nmol/l for total T4. The intra and inter-assay coefficients of variation 331

332

Clinical Endocrinology (1992) 37

R. Wilson et al.

were 6 and 8% respectively for TT3 and 5 and 7% for TT4. (2) TSH was measured by our own two-site immunoradiometric assay. The normal reference range was < 5 mU/1. The intra and inter-assay coefficients of variation were 3 and 8% respectively. Free T4 was measured using our own assay (Wallace et al., 1990) The normal reference range was 8-18 pmol/l. The intra and inter-assay coefficients of variation were 6 and 10% respectively. Thyroid binding globulin (TBG) was measured by a Corning Immunophase radioimmunoassay. The normal reference range was from 10 to 30 mg/l. The intra and inter-assay coefficients of variation are 4 and 15% respectively. Thyrotrophin receptor antibodies (TRAb) were measured by radioreceptor assay (Shewring & Rees Smith, 1982).Values > 25 were taken to indicate the presence of TRAb (Wilson et al., 1989). The inter and intra-assay coefficients of variation'were 4 and 9% respectively. Thyroid stimulating antibody (TSAb) levels were measured by CAMPproduction from FRTL-5 cells (Rappaport, 1982). TSAb were considered to be present when the value was higher than mean+2SD of 60 normal subjects. This was found to be > 150% (Wilson et al., 1989). The inter and intra-assay coefficients of variation were 5 and 22% respectively. Thyroid microsomal and thyroglobulin antibodies were measured using a haemagglutination kit (Wellcome Diagnostics). hCG levels were measured using the Serono Serozyme method. The normal reference range was < 5 U/1 with intra and inter-assay coefficientsof variation of 4 and 6% respectively.

Table 1 Obstetrical data relating to hyperemetic patients (all had

i.v. therapy) Patient no.

Gestation

1 2 3 4 5 6 7 8 9 10

Parity

(weeks)

o+o

12 14 12 16 18 17 14 15 16 17

o+o 1+ O

1 +O

o+o 2+0 Of0

o+o 1 +o Of0

Table 2 Serum thyroid hormone and hCG levels in hyperemetic

patients

Total T3 (nmol/I) Total T4 (nmolil) TSH (mU/l) Free T4 @mol/l) TBG (mg/l) hCG (U/l)

Hyperemesis

Control

2.3 +0.7 I43 54 0.7 0.5 173+ 3.3 42.4 + 5.6 120333+84019

2.3+4.5 159+28 0.6 0.5 14.5+2.5 35.7+ 6.3 210000 + 53666

+ +

+

Results are expressed as mean+SD.

Statistical analysis

Results are given as meankone standard deviation. They were analysed for statistical significance using Student's ttest. Results

There were no significant differences between the hyperemetic patients and healthy pregnant controls for serum levels of total and free thyroid hormones, TSH, TBG and hCG (Table 1).

The results of the individual hyperemetic patients are shown in Table 2. Of the 10 patients studied, six showed one or more abnormality of thyroid function. TT4 levels were abnormal in four, TSH in one, FT4 in three and TBG in four. Although all patients were negative for thyroid microsomdl

Table 3 Individual results for hyperemetic patients

Patient

1 2 3 4 5 6 7 8 9 10

TT3

TT4

TSH

FT4

TBG

hCG

1.9 2.6 1.4

121 164 135 128 165 177 173 158 214' 160

0.8 0.5 1.1

14 15 20 16 23* 23* 15 15 17 16

34 46. 37

87000 75000 87000 56000 177000 191000 49000 49000 3 12000 56000

1.7

1.9 2.0 2.9 3.1 3.8 2.2

0.4 0.3 0.3 2,1* 0.8 0.3 0.5

40

38 40

52* 49+

46* 42

* Patients beyond the normal pregnant reference range

Thyroid tests in hyperemesis

Clinical Endocrinology (1992) 37

and thyroglobulin antibodies and TRAb, one (patient 10) was positive for TSAb. None of the hyperemetic patients had an abnormal hCG level. A weak but significant correlation (r=0.74, P