Maternal Thyroid Deficiency during Pregnancy and Subsequent Neuropsychological Development of the Child

To the Editor

Haddow et al. (Aug. 19 issue)¹ suggest that screening pregnant women for hypothyroidism by measuring serum thyrotropin may be worthwhile and that treating women with serum thyrotropin concentrations at or above the 98th percentile could lead to “an increase of approximately 4 points in IQ scores” in their children. However, the results of their study should be interpreted with caution. The difference in full-scale IQ scores between children whose mothers had hypothyroidism during pregnancy and the children of matched control women is not statistically significant (103 vs. 107, P=0.06), and the results of only 2 of the 13 neuropsychological tests were significantly worse in the children of the women with hypothyroidism during pregnancy.

Many factors have a role in the neuropsychological development of children, apart from maternal thyroid function. We are not given details of the women's marital status after delivery, even though parental separation and divorce or serial monogamy on the part of one parent can impair a child's development.^2,3 The effect of siblings, though variable, should also be considered^2,4; 73 percent of the women who had hypothyroidism during pregnancy were multiparous, as compared with 65 percent of the control women and 51 percent of the remaining cohort of pregnant women. Maternal thyroid status may be important in the development of a healthy child, but on the basis of the data presented by Haddow et al., it is not clear that widespread screening is justified.

Christian Herzmann
Leipzig University, 04103 Leipzig, Germany

James K. Torrens, M.R.C.P.
Seacroft Hospital, Leeds LS14 6UH, United Kingdom

To the Editor

The paper by Haddow et al. has drawn attention to an often disregarded and possibly preventable cause of poor neuropsychological development in children — namely, hypothyroidism in pregnant women. Is the principal underlying factor in this poorer development maternal hypothyroidism, as defined by a high serum thyrotropin concentration, or maternal hypothyroxinemia itself, whether or not serum thyrotropin concentrations are high? In the study by Haddow et al., the children were selected according to the mothers' serum thyrotropin concentrations in the second trimester. The control children may therefore have included children whose mothers had hypothyroxinemia, which might have decreased the magnitude of the differences in neuropsychological development in the two groups, which was smaller than that reported in other studies.^1-3

In those studies, low maternal serum thyroxine or free thyroxine concentrations in the first or second trimester were correlated with poor postnatal neuropsychological development. During the first trimester, maternal thyroxine is the only source of thyroid hormone for the fetus. Normal maternal serum triiodothyronine concentrations, which mitigate hypothyroidism in pregnant women, do not prevent poorer cognitive development in their infants.² In preterm infants, who are prematurely deprived of maternal thyroxine and iodine, poor cognitive development and an increased risk of cerebral palsy are also correlated with the presence of perinatal hypothyroxinemia, not with high serum thyrotropin concentrations,⁴ indicating the continuing importance of thyroxine for later stages of brain development.

As Utiger noted in the accompanying editorial,⁵ an inadequate iodine intake may be a crucial factor contributing to the frequency of hypothyroidism during pregnancy, and it is also a crucial determinant of relative hypothyroxinemia. In Brussels, Belgium (with a median urinary iodine excretion of 56 μg per liter), 30 percent of pregnant women had low serum free thyroxine values in the first trimester and 2.3 percent had high serum thyrotropin values.⁶ Even in Madrid, an area where mild iodine deficiency is prevalent (median urinary iodine excretion, 90 μg per liter), 20 percent of pregnant women had first-trimester serum free thyroxine values below the 10th percentile of values for women with adequate iodine intake. These data support Utiger's recommendation that measures be taken to ensure that pregnant women have an adequate intake of iodine (>200 μg per day). Thyroxine therapy should be limited to pregnant women who have persistent hypothyroxinemia.

Gabriella Morreale de Escobar, Ph.D.
Francisco Escobar del Rey, M.D., Ph.D.
Instituto de Investigaciones Biomédicas Alberto Sols, 28039 Madrid, Spain

To the Editor

Haddow et al. suggest that screening for thyroid deficiency during pregnancy may be warranted. We would like to report the results of screening for thyroid disease among pregnant women in Sapporo, Japan.

Sapporo has a population of 1.8 million, and 16,000 infants are born every year. Our screening program for thyroid function in early pregnancy was initiated in June 1986 and consists of measurements of serum thyrotropin, free thyroxine, and antithyroid peroxidase and antithyroglobulin antibodies in blood specimens collected on filter paper. By the end of March 1997, 70,632 pregnant women had been screened. At first, only 11 percent of pregnant women agreed to testing, but the percentage has been in the range of 45 to 55 percent since 1991. The mean (±SD) age of the women was 28±5 years, and the specimens were collected at 12±5 weeks of gestation.

The overall rate of reexamination because of abnormal results was 2.0 percent (1409 of 70,632 women), and 671 women (1.0 percent) were referred for further medical evaluation. Among these 671 women, 171 (25.5 percent) had hyperthyroidism (162 had Graves' disease, and 9 had other causes of hyperthyroidism) and 102 (15.2 percent) had hypothyroidism (67 had chronic autoimmune thyroiditis, and 35 had other causes of hypothyroidism). The overall incidence of the disorders was 1 in 413 and 1 in 692, respectively. In addition, 220 women (32.8 percent) had transient hyperthyroxinemia, and 121 women (18.0 percent) had low serum thyrotropin concentrations. None of the pregnant women with thyroid disease had symptoms or signs of thyroid dysfunction.

We subsequently obtained the results of neonatal screening for congenital hypothyroidism in the infants whose mothers had been referred for medical evaluation during pregnancy. We found six cases of transient hypothyroidism (incidence, 1 in 11,772), eight cases of transient hypothyroxinemia (incidence, 1 in 8829), three cases of neonatal Graves' disease, and one case of transient subclinical hypothyroidism. Thyroid function was not controlled during pregnancy in the mothers of these infants, and all the infants whose mothers' thyroid disorder was well controlled during pregnancy were normal. (The overall incidence of congenital hypothyroidism as detected by neonatal screening is about 1 in 3000 in Sapporo.)

The frequency of thyroid disease among pregnant women may be lower in Japan than elsewhere.¹ Nonetheless, we believe that our voluntary program of screening for thyroid disease in early pregnancy is useful because women who have thyroid dysfunction are so often asymptomatic.

Masaru Fukushi, Ph.D.
Kaori Honma, B.S.
Kozo Fujita, M.D., Ph.D.
Sapporo City Institute of Public Health, Sapporo 003-8505, Japan

To the Editor

Haddow et al. provide evidence of an important link between hypothyroidism in early pregnancy and a delay in neuropsychological development among children at the ages of seven to nine years. In the accompanying editorial, Utiger discussed the importance of thyroid hormones for fetal and infant development and pointed out that thyroid deficiency in most pregnant women is presumed to be due to autoimmune thyroid disease. He noted that hypothyroidism can also be caused by iodine deficiency, a preventable condition, and recommended that the dietary intake of iodine be increased in this country through vitamin supplements, by increasing the amount of iodine in salt, and by adding iodine to other foods. We are concerned that our study of iodine nutrition in the United States,¹ which revealed a significant decrease in the urinary concentration of iodine between the period from 1971 to 1974 and the period from 1988 to 1994, was the basis for that recommendation. The median urinary iodine concentration in the period from 1988 to 1994 was 145 μg per liter, which reflects an intake that exceeds the recommended daily allowance.² We could not, therefore, conclude from our study that iodine intake in the United States was inadequate. Rather, we concluded that iodine intake had fallen over a 20-year period and should be monitored carefully. As important as it is to ensure that iodine intake is adequate, it is equally important to avoid excessive intake of iodine. Excessive iodine intake can increase the risk of hypothyroidism through several mechanisms, including the development of autoimmune thyroid disease.³ Given this risk, we do not believe that a general recommendation to increase iodine intake in this country in order to ensure that pregnant women have an adequate intake is prudent.

Joseph G. Hollowell, Jr., M.D., M.P.H.
Paul L. Garbe, D.V.M., M.P.H.
Dayton T. Miller, Ph.D.
Centers for Disease Control and Prevention, Atlanta, GA 30341