The New England Journal of Medicine, Vol. 344, No. 23 

In this issue of the Journal, Arafah presents data indicating that some women with hypothyroidism receiving thyroxine therapy needed more thyroxine when they received estrogen therapy - a change caused by an estrogen-induced increase in the serum concentration of thyroxine-binding globulin (1).  The increased need for thyroxine occurred in women with hypothyroidism who were receiving moderate doses of thyroxine, intended to replace normal thyroxine secretion and restore thyrotropin secretion to normal, and in women with thyroid carcinoma who were receiving higher doses of thyroxine, intended to suppress thyrotropin secretion to below normal.  

    Thyroxine-binding globulin is a glycoprotein produced in the liver that binds thyroxine with high afinity.  Approximately 75 percent of the thyroxine in serum is bound to it; nearly all the rest is bound to transthyretin or albumin, so that less than 01 percent remains free or unbound.  The physiological function of thyroxine-binding globulin is not known, but it may serve to distribute thyroxine evenly within tissues, particularly within large solid organs like the liver (2).  This function is not critical, however, because patients with a complete absence of thyroxine-binding globulin, as a result of mutations in the gene for this glycoprotein, are clinically euthyroid and have normal serum concentrations of free thyroxine and thyrotropin concentrations, although their serum concentrations of thyroxine are low (3). 

    Serum concentrations of thyroxine-binding globulin are similar in men and women, including post-menopausal women, indicating that its production is not affected by the different rates of estrogen production in normal men and younger and older women.  The concentrations are high in pregnant women, patients with estrogn-secreting tumors, and women who are being treated with estrogen; this is because of an increase in the glycosylation of the protein, which slows its clearance, and to a lesser extent, to an increase in its production (4).  In the study by Arafah, the women were treated with 0.625 mg of conjugated estrogens daily, which raised serum thyroxine-binding globulin concentrations by approximately 50 percent in both the women with normal thyroid function and the women with hypothyroidism.  The concentrations reached a plateau between 6 and 12 weeks after treatment was initiated.  The dose of estrogen given orally is not sufficient to replace estrogen systemically in a pre-menopausal women, but the liver is exposed to a high concentration of estrogen and undoubtedly clears some of it. 

    The estrogen-induced increase in serum thyroxine-binding globulin concentrations is dose-dependent, and it occurs with any orally administered estrogen (5,6) whether given alone, in combination with a progestin in the form of an oral contraceptive, or in combination with medroxyprogesterone in post-menopausal women.  Serum thyroxine-binding globulin concentrations also increase in women treated with tamoxifen and raloxifene, although these compounds may be less potent than standard estrogens (7).  Transdermal estradiol therapy does not raise serum thyroxine-binding globulin concentrations (8), providing further evidence of a first-pass hepatic effect of a high serum concentration of estrogen in the portal vein.  in normal women, the estrogen-induced increased in the serum thyroxine-binding globulin concentration has no physiological effect other than raising the serum thyroxine concentration. 

    It might have been suspected that women treated with thyroxine in whom estrogen therapy is begun might need more thyroxine.  In pregnant women, serum thyroxine-binding globulin concentrations increase to the same extent as they do in women treated with 0.625 mg of conjugated estrogens per day, and they also lose some thyroxine to their fetuses and as a result of the placental deiodination of thyroxine to reverse triiodothyronine, which is biologically inactive.  As a result of these changes, women with hypothyroidism who are being treated with thyroxine need approximately 50 percent more thyroxine when they are pregnant (9).  Conversely, women with hypothyroidism treated with thyroxine need lower doses of thyroxine when they are treated with androgens, which lower the serum thyroxine binding globulin concentration (10).  

    How does an increase in the serum thyroxine-binding concentration lead to an increase in the need for thyroxine in women who are dependent on exogenous thyroxine?  In normal women, the estrogen-induced increase in the serum thyroxine-binding globulin concentration leads to a parallel increase in the serum thyroxine concentration, but the serum free thyroxine concentration remains normal.  The immediate effect of an increase in the serum thyroxine-binding globulin concentration must be to enhance the binding of thyroxine and therefore to decrease the amount of free thyroxine.  This decrease should result in a small increase in thyrotropin secretion, and therefore in thyroxine secretion, until the serum free thyroxine concentration is back to normal.  Neither the postulated transient fall in the serum free thyroxine concentration nor the postulated transient rise in the serum thyrotropin concentration has been documented well - probably because the changes are gradual and their magnitude is small - but given the physiochemical and physiologic considerations, both must occur.  In women with hypothyroidism, thyroxine secretion cannot increase, and therefore, as documented in the study by Arafah (1), their serum free thyroxine concentrations do fall and their serum thyrotropin concentrations increase.  

    Is the increase in the serum thyrotropin concentration that occurs when women taking thyroxine are given estrogen clinically important? Yes, because among the 18 women in the thyroxine-replacement group who were studied by Arafah (1), 7 (39 percent) had a decrease in their serum free thyroxine concentration that was sufficient to increase their serum thyrotropin concentration to more than 7 U per milliliter, although only 1 of the 7 had any clinical manifestations of hypothyroidism.  These seven women were then given more thyroxine, so we do not know whether their hypothyroidism would have become more severe, either clinically or biochemically, with continued estrogen therapy.  That it might not have done so is suggested by the fact that the serum thyrotropin concentrations in the other women in this group did not increase progressively with time. Among the seven women in the thyroxine-suppression group, three had increases in serum thyrotropin concentration to more than 1 U per milliliter, and two of the three had increases in their serum thyroglobulin concentration, possibly indicating the stimulation of thyroid-carcinoma tissue.  The need for more thyroxine does not seem to occur as often in women given estrogen as in those who become pregnant.  

    Consider the reverse sequence - women who are taking estrogen in whom hypothyroidism develops and who then need thyroxine therapy.  In such women, hypothyroidism could occur sooner or be more severe because of the extra demand on the failing thyroid gland.  Once they are being treated with thyroxine these women may need a higher dose than they would if they were not taking estrogen, and they may be at risk for hyperthyroidism after the estrogen therapy is discontinued.  

    Many women take estrogen, either for contraception or to ameliorate menopausal symptoms or other manifestations of estrogen deficiency.  Similarly, many women take thyroxine because they have hypothyroidism.  There should therefore be a substantial number of women taking both estrogen and thyroxine.  Because women with hypothyroidism who are taking thyroxine may need more thyroxine when they are treated with estrogen and may need less thyroxine after estrogen is discontinued, it is prudent to reassess their thyroid function several months after estrogen therapy is either initiated or discontinued.  


1. Arafah, BM. Increased need for thyroxine in women with hypothyroidism during estrogen therapy. N Engl J Med 2001;344:1743-9  

2. Mendel GM, Weisiger RA, Jones AL, Cavlieri RR. Thyroid hormone binding proteins in plasma facilitate uniform distribution of thyroxine within tissues: a perfused rat liver study.  Endocrinology 1987;120:1742-9. 

3. Stockigt JR. Transport protein variants. In: DeGroot LJ, Jameson JL, eds. Endocrinology. 4th ed. Vol. 2. Philadelphia: W.B. Saunders, 2001: 1603-8. 

4. Aln KB, Mori Y, Refetoff S. Reduced clearance rate of thyroxine-binding globulin (TBG) with increased sialylation: a mechanism for strogen induced elevation of serum TBG concentration. 

5. Geola FL, Frumar AM, Tataryn IV, et al. Biological effects of various doses of conjugated equine estrogens in postmenopausal women.  J Clin Endocrinol Metab 1989;51:620-5

6. Mandel FP, Geola FL, Lu JKH, et al. Biologic effects of various doses of ethinyl estradiol in postmenopausal women.  Obstet Gynecol 1982;59:673-9

7. Kostoglou-Athanassiou I, Ntalles K, Markopoulos C, Athanassiou P, Gogas J, Proukakis C. Thyroid function in postmenopausal women with breast cancer on tomixfen. Eur J Gynaecol Oncol 1998;19:150-4

8. Chetkowski RJ, Meldrum DR, Steingold KA, et al. Biologic effects of transdermal estradiol. N Engl J Med 1986;314:1615-50.

9. Kaplan MM, Management of thyroxine therapy during pregnancy. Endocr Pract 1996;2:281-6

10. Arafah BM. Decreased levothyroxine requirement in women with hypothyroidism during androgen therapy for breast cancer. Ann Intern Med 1994;121-247-51.