Autoimmune thyroid disease is a relatively common disorder in women of childbearing age. The reported prevalence of thyroid autoantibodies in pregnant women ranges from 5.2% in Belgium to 12.5% in North America, and the reported prevalence of elevated maternal serum TSH concentrations approximates 2.2% in Belgium and 2.5% in North America.1–5 In 1980 we reported that there is no correlation between maternal or newborn serum thyroid autoantibodies and sporadic congenital hypothyroidism6; however, there are many reports relating maternal autoimmune thyroid disease to transient congenital hypothyroidism in newborn thyroid screening programs.7,8 The prevalence of transient congenital hypothyroidism in North America was reported to be low (1:180,000, or 2% of congenitally hypothyroid infants) relative to the prevalence of autoimmune thyroid disease, occurring only in association with high titers of maternal TSH receptor-blocking antibody, which is present in a fraction of women with autoimmune thyroid disease.7,8 In an ongoing study, we have collected serum from 259 mothers of infants with congenital hypothyroidism from the Quebec neonatal screening program, between 1984 and 1996, to assess the thyroid status of these women and the correlation of maternal autoimmune thyroid disease with transient congenital hypothyroidism.
Materials and Methods
The study population included 259 women who delivered an infant with hypothyroidism detected by the Quebec newborn screening program between 1984 and 1996. Blood samples were taken routinely from these women at follow-up serum sampling of the infant (1–3 weeks postpartum) for measurements of free thyroxine (T4), TSH, and antimicrosomal antibodies. Most measurements were conducted concurrent with the sampling. Repeat third-generation TSH measurements were conducted in spring 1997 on samples from mothers with abnormal TSH values or positive microsomal antibody results.
To develop a control population, pregnant women were recruited from the Obstetrics Department at The St. Francis d'Assise Hospital in Quebec City between 1991 and 1992. A total of 1773 women participated as part of a routine second-trimester prenatal screening program measuring serum alpha fetoprotein (AFP) and hCG levels. These serum samples also were used to measure free T4, TSH, and antimicrosomal antibodies. Women with abnormal AFP or hCG values were excluded from the study, leaving 1773 eligible women. Seven of them were treated for Graves disease with propylthiouracil, the rest received no medication other than multivitamin supplements. Samples from 28 of 1773 women with positive antimicrosomal antibody titers were obtained 3 months postpartum to compare second-trimester and postpartum results.
Newborn filter paper blood samples that had an elevated TSH level on screening (over 15 mIU/L) were repeated in duplicate (same sample) with a T4 measurement. Congenital hypothyroidism was suspected if the TSH remained elevated with a normal or low T4 value. Serum samples were taken between 1 and 2 weeks, and if the follow-up serum TSH level remained high or increased, a thyroid scan was done (unless parental permission was denied or logistic considerations interdicted scanning) and thyroxine treatment was started. If between 1 and 2 weeks the serum TSH decreased and the T4 level was normal, treatment was withheld, but a scan was done, if possible. Serum TSH and T4 concentrations were measured again at 2 months of age. Infants whose elevated TSH values were transient (under 2 months, usually in association with a normal thyroid scan) and infants who between 2 and 3 years of age had normal thyroid function without treatment for 1 month, with a normal thyroid scan, were considered to have transient congenital hypothyroidism.
Maternal thyroid function was assessed using commercial kits for immunoassay measurements of free T4 (Intermedico, Montreal, Canada, normal values 9–27 pmol/L), total T4 (Delphia-Wallac, Montreal, Canada, normal values 60–170 nmol/L), and TSH (BioRad, Mississauga, Ontario, Canada, normal values 0.25–4.5 mU/L, sensitivity 0.01 mIU/mL). Eighty percent of the TSH measurements in the mothers of congenitally hypothyroid infants were conducted using the second-generation assay. Samples with undetectable TSH values were recently retested using the third-generation assay. Thyroid antimicrosomal antibody was measured using an agglutination method and reported as titer. (Miles, Bayer, NY, normal result negative). Statistical analyses were conducted using χ2 tests for homogeneity.
The prevalence of abnormal maternal serum TSH and antimicrosomal antibody measurements in normal pregnant women and in the selected group of mothers delivering infants with congenital hypothyroidism are summarized in Table 1. The prevalences of positive antimicrosomal antibodies were similar (12.1% and 11.6%) in the two groups. Suppressed and elevated TSH concentrations were more frequent (6.1% versus 2.9% and 7.0% versus 0.9%; P = .01 and P < .001, respectively) in the mothers of hypothyroid newborns. Of the mothers of hypothyroid newborns, 19.7% had abnormal TSH levels or positive antimicrosomal antibodies compared with 13% of the normal pregnant women (P = .001).
Twenty-eight of the normal pregnant women with positive antimicrosomal antibodies titers were available for blood sampling 3 months postpartum. All 28 had positive postpartum antimicrosomal antibody values, and the titers were similar or higher than the second-trimester values in 15 of 28 samples (second-trimester range 1:40 to 1:1.3 million; postpartum range 1:40 to 1:327,680; the median titer was 1:2,560 at both times). Of these 28 women, three (10.7%) had elevated TSH and low free T4 values postpartum, whereas second-trimester results were normal (except for one TSH value of 8 mU/L). These three women were asymptomatic and postpartum hypothyroidism was diagnosed. These results extrapolate to an incidence of 1.3% of postpartum hypothyroidism among the 1773 control women, a value similar to that (1.2%) recently reported by Walfish et al from Toronto.9
Table 2 shows the types of thyroid abnormalities in infants of mothers with autoimmune thyroid disease compared with the overall group of mothers whose infants had congenital hypothyroidism. The prevalence of newborn transient hypothyroidism was significantly higher (27% versus 15%, P = .04) in the mothers with autoimmune thyroid disease. The transient congenital hypothyroidism of three (23%) of 13 infants whose mothers had autoimmune thyroid disease was secondary to maternal propylthiouracil treatment of maternal Graves disease. In the remainder, the hypothyroid state resolved within 2 months.
Table 3 summarizes the abnormalities of thyroid function detected in the 259 women who delivered infants with congenital hypothyroidism and the types of congenital hypothyroidism in the affected infants. Of the 30 mothers whose antimicrosomal antibodies titers were positive, nine had increased serum TSH and four had suppressed serum TSH concentrations. Of those without antimicrosomal antibodies, nine had increased serum TSH, and 12 had suppressed serum TSH levels. The mothers of ten (77%) of 13 infants with transient hypothyroidism had antimicrosomal antibodies, and all 13 mothers of the transiently hypothyroid infants had abnormal TSH levels or antimicrosomal antibodies. The mothers of congenitally hypothyroid infants with thyroid agenesis or ectopic glands also demonstrated high prevalence of abnormal TSH levels or antimicrosomal antibodies (10 of 12 and 7 of 15, respectively).
Tables 4 and 5 summarize the thyroid function test results in the mothers with elevated and suppressed TSH concentrations, respectively (from Table 3). As noted in Table 4, the two mothers with elevated TSH and low free T4 concentrations did not have antimicrosomal antibodies. In Table 5, three of the four mothers with elevated free T4 and suppressed TSH values did not have antimicrosomal antibodies.
Table 6 summarizes thyroid function test results in the 30 women with positive antimicrosomal antibody test results; 13 of 30 had abnormal serum TSH concentrations, and of these three had abnormal free or total T4 concentrations. There was no correlation between antimicrosomal antibody titers and serum TSH concentrations. Antibody titers varied from 1/20 to 1/81,920.
At the beginning of our screening program in 1980, we investigated the association of maternal thyroid auto-antibodies with congenital hypothyroidism, by measuring antibody titers in maternal samples and in eluates from the filter paper blood spots used for newborn screening. There was a significant positive correlation between the maternal titer and the newborn titer of antimicrosomal antibodies, but no correlation between the presence of maternal antimicrosomal antibodies and congenital hypothyroidism.6 The present results concur. With a prevalence of congenital hypothyroidism approximating one in 4000 newborns, we expect no more than one congenitally hypothyroid infant from the population of normal pregnant women compared with 259 from the specific cohort of 259 mothers of congenitally hypothyroid infants. The prevalence of antimicrosomal antibodies was similar (12.1% and 11.6%) in the two groups (Table 1).
Among the mothers of hypothyroid newborns, 51 (20%) had antimicrosomal antibody or abnormal serum TSH concentrations (Tables 3–6). Of these, two of 51 were hypothyroid, (increased TSH and low free T4), four were hyperthyroid (suppressed TSH and increased free T4), 16 had increased serum TSH with normal free T4 concentrations, and 12 had suppressed serum TSH levels with normal free T4 values (Tables 4–6). Of the mothers with elevated serum TSH values, 11 of 18 had levels of 10 mIU/L or higher (range 6–67) (Table 4). In mothers with suppressed serum TSH, values ranged from under 0.01 to 0.22 mIU/L; seven of 16 were under 0.10 mIU/L (Table 5).
Of the 259 mothers who delivered infants with congenital hypothyroidism, 51 (20%) had elevated thyroid antimicrosomal antibody titers or thyroid dysfunction. This contrasts with a prevalence of positive thyroid autoantibody or thyroid dysfunction of 13% similarly assessed in a population of 1773 control pregnant women in Quebec. We did thyroid scans of 48 of the 51 infants, and with the clinical data and course of the infants, determined that 13 of 48 (27%) had transient hypothyroidism. Of these, antimicrosomal antibodies were present in 10 of 13 mothers and the other three had abnormal TSH concentrations (Table 3). These data indicate that there is a high prevalence of transient congenital hypothyroidism in Quebec (15%, Table 2) and that most cases are associated with maternal autoimmune thyroid disease. Antithyroid drugs accounted for the transient congenital hypothyroidism in three (23%) of the 13 infants of mothers with autoimmune thyroid disease.
The observations of Brown and coworkers8 suggest that the prevalence of congenital hypothyroidism resulting from TSH receptor blocking autoantibodies approximates one in 180,000 births, accounting for only about 2% of infants with congenital hypothyroidism. We reported previously that TSH receptor-binding inhibitory immunoglobulin was detectable in only 18% of mothers of infants with transient hypothyroidism, about 2% of pregnancies associated with congenital hypothyroidism.10 We have not measured TSH receptor autoantibodies in the mothers in the present study, but a 2% prevalence would account for only five of the ten infants with transient hypothyroidism not accounted for on the basis of antithyroid drugs. It is possible that low levels of TSH receptor-blocking antibody, undetectable by current assay methods, might account for some mild cases of transient congenital hypothyroidism. Additionally, antibodies thyrotoxic to thyroid cells have been found in infants with congenital hypothyroidism but have not been examined in a population of infants with transient congenital hypothyroidism.11 As-yet-undefined, intrathyroidal autoimmune thyroid disease-related events might also be involved. Glinoer and coworkers12 reported a decreased mean serum thyroglobulin concentration in the newborns of 45 women with significant thyroid peroxidase autoantibody titers. The mean serum T4 and TSH levels in these infants were similar to control values, indicating that TSH receptor-blocking autoantibodies were not involved. The high prevalence (ten of 12) of antimicrosomal antibodies in mothers of the infants with thyroid agenesis (Table 3) is also unexplained.
It is interesting that only one of six mothers with thyroid dysfunction (two with hypothyroidism [Table 4] plus four with mild hypothyroidism [Table 5]) had an increased antimicrosomal antibody titer. Of the remaining seven mothers with elevated TSH levels and normal free T4 values in the absence of antimicrosomal antibodies (Table 4), three had TSH levels of 10–17 mIU/L and four had TSH values of 6 to 7 mIU/L (normal range 0.25–4.5 mIU/L). Of the nine mothers with suppressed TSH concentrations and normal free T4 values (Table 5), two had TSH levels under 0.10 and seven had levels of 0.10 to 0.22 mIU/L. The five women with hypothyroidism or mild hypothyroidism in the absence of antimicrosomal antibodies, the three women with TSH levels of 10–17 mIU/L and normal free T4 in the absence of antimicrosomal antibodies, and the two women with TSH values under 0.10 mIU/L and normal free T4 in the absence of antimicrosomal antibodies presumably have autoimmune thyroid disease. The significance of the mildly elevated and suppressed TSH values in the mothers without antimicrosomal antibodies (n = 4, Table 4; n = 7, Table 5) remains unclear. We did not measure antithyroglobulin, anti-thyroid peroxidase, or anti-TSH receptor autoantibodies in these women.
The present results demonstrate increased prevalence of abnormal serum TSH concentrations, particularly increased levels, and of transient congenital hypothyroidism in the newborns of mothers with autoimmune thyroid disease. Antimicrosomal antibodies were found in ten (77%) of 13 mothers of infants with transient congenital hypothyroidism, and either antimicrosomal antibodies or abnormal TSH levels were found in 13 of 13 (100%), indicating that most, if not all, cases of transient congenital hypothyroidism were attributable to maternal autoimmune thyroid disease. Earlier studies of the prevalence of TSH receptor-blocking antibodies in infants with congenital hypothyroidism and their mothers suggest that about half of the transient congenital hypothyroidism in the present study was accounted for on this basis. The mechanism for the transient congenital hypothyroidism in the remaining infants remains unknown. The presence of thyroid auto-antibodies in pregnant women is associated with an increased incidence of miscarriage, thyroid dysfunction, and postpartum thyroid disease.2,12,13 The present study shows that a high prevalence of transient congenital hypothyroidism can be added to this list. On the basis of the 20% prevalence of thyroid dysfunction in mothers of congenitally hypothyroid infants, we recommend thyroid function assessment of such women, if this information is not already available. The measurement of TSH receptor-blocking autoantibodies in infants with transient congenital hypothyroidism provides helpful information regarding treatment of the infants and counseling for future pregnancies.
1. Klein RZ, Haddow JE, Faix JD, Brown RS, Hermos RJ, Pulkkinen A, et al. Prevalence of thyroid deficiency in pregnant women. Clin Endocrinol 1991;35:41–6.
2. Glinoer D. The regulation of thyroid function in pregnancy: Pathways of endocrine adaptation from physiology to pathology. Endocr Rev 1997;18:404–33.
3. Hidaka Y, Tamaki H, Iwatani Y, Tada H, Mitsuda N, Amino N. Prediction of post-partum Graves' thyrotoxicosis by measurement of thyroid stimulating antibody in early pregnancy. Clin Endocrinol 1994;41:15–20.
4. Bech K, Hertel J, Rasmussen NG, Hegedus L, Hornnes PJ, Feldt-Rasmussen U, et al. Effect of maternal thyroid autoantibodies and post-partum thyroiditis on the fetus and neonate. Acta Endocrinol (Copenh) 1991;125:146–9.
5. Pop VJM, deRooy HAM, Vader HL, Van der Heide D, Van Son MM, Komproe IH. Microsomal antibodies during gestation in relation to post-partum thyroid dysfunction and depression. Acta Endocrinol 1993;129:26–30.
6. Dussault JH, Letarte J, Guyda H, Laberge C. Lack of influence of thyroid antibodies on thyroid function in the newborn infant and on a mass screening program for congenital hypothyroidism. J Pediatr 1980;96:385–9.
7. Matsuura N, Yamada Y, Nohara Y, Konishi J, Kasagi K, Endo K, et al. Familial neonatal transient hypothyroidism due to maternal TSH-binding inhibitor immunoglobulins. N Engl J Med 1980;303:738–41.
8. Brown RS, Bellisario RL, Botero D, Fournier L, Abrams CAL, Couger ML, et al. Incidence of transient congenital hypothyroidism due to maternal thyrotropin receptor-blocking antibodies in over one million babies. J Clin Endocrinol Metab 1996;81:1147–51.
9. Walfish PG, Meyerson J, Provias JP, Vargas MT, Papsin FR. Prevalence and characteristics of post-partum thyroid dysfunction: Results of a survey from Toronto, Canada. J Endocrinol Invest 1992;15:265–72.
10. Dussault JH. Thyroid growth blocking antibodies and congenital hypothyroidism. In: Delange F, Fisher DA, Glinoer D, eds. Research in congenital hypothyroidism. New York: Plenum Press, 1989:135–40.
11. Bogner U, Gruters A. Sigle B, Helge H, Schleusnener H. Cytotoxic antibodies in congenital hypothyroidism. J Clin Endocrinol Metab 1989;68:671–5.
12. Glinoer D, Fernandez-Soto ML, Bourdoux P, Lejeune B, Delange F, Lemone M, et al. Pregnancy in patients with mild thyroid abnormalities: Maternal and neonatal repercussions. J Clin Endocrinol Metab 1994;73:421–7.
13. Daniels GH. Thyroid disease and pregnancy: A clinical overview. Endocr Pract 1995;1:287–301.