Smoking is associated with so many abnormalities of thyroid function that it is unlikely it has a single action on the thyroid gland. In this regard, it is like at least one substance that occurs naturally in the body -- iodine -- and several drugs.
The article by Muller et al. (1) in this issue of the Journal presents evidence that in women with hypothyroidism, smoking decreases both thyroid secretion and thyroid hormone action. Among women with subclinical hypothyroidism, smoking was associated with decreased thyroid secretion on the basis of findings of lower serum concentrations of free thyroxine and higher serum concentrations of thyrotropin in smokers than nonsmokers. In contrast, among both normal women and women with overt hypothyroidism, the serum concentrations of free thyroxine and thyrotropin were similar in those who smoked and those who did not, presumably because the antithyroid action of smoking is so weak that it is apparent only when thyroid secretion is marginal.
Smoking was associated with peripheral antithyroid actions in both groups of women with hypothyroidism, but these actions -- which resulted in more symptoms, signs, and biochemical and physiologic changes of hypothyroidism -- were greater in the women with overt hypothyroidism, who would be expected to be more sensitive to blockade of the peripheral actions of thyroid hormone. Although the tests of peripheral thyroid hormone action used in the study are not very specific or sensitive indicators of decreased thyroid hormone action, the fact that differences were detected and that there was a dose-response relation between smoking and serum cholesterol concentrations suggests that the effects are real.
The results do not indicate that smoking causes hypothyroidism, only that it increases the severity and peripheral effects of hypothyroidism. The small magnitude of both the thyroidal and peripheral antithyroid effects of smoking means that we do not need to change the way we think about hypothyroidism in smokers as compared with nonsmokers with respect to either diagnosis or treatment, except to be less reluctant to treat patients with subclinical hypothyroidism.
How might smoking affect thyroid secretion or action? One component of tobacco smoke is cyanide, which is converted to thiocyanate, which inhibits iodide uptake and hormone synthesis and increases iodide efflux from the thyroid gland. (2) There are many other components of smoke that might have antithyroid actions; decrease the binding of triiodothyronine to its receptors or its post-receptor actions in the liver, muscle, or other organs; or both.
The women with hypothyroidism studied by Muller et al. had identifiable thyroid disease. Among the normal women there were no differences in either the secretion or action of thyroid hormone between the smokers and nonsmokers. Previous studies of normal subjects revealed a higher frequency of thyroid enlargement among smokers than nonsmokers, (3) but the differences in frequency and degree of thyroid enlargement were small. The results of tests of thyroid function have varied, but most often serum triiodothyronine and thyroglobulin concentrations were slightly higher and serum thyrotropin concentrations slightly lower in the smokers. (3) These findings are most indicative of minimal stimulation of thyroid growth and a minor degree of thyroid autonomy, which could be caused by chronic sympathetic stimulation induced by nicotine, the effect of a component of smoke on thyroid cellular function, or the induction of subclinical Graves' disease (see below).
The most dramatic effect of smoking on the thyroid is its association with Graves' hyperthyroidism, and especially with Graves' ophthalmopathy. Whether smoking precedes Graves' hyperthyroidism (with or without ophthalmopathy) or not, there are more smokers than would be expected among those with these conditions. (4,5,6) In a case-control study, for example, the odds ratios for smoking among patients with Graves' hyperthyroidism with and without ophthalmopathy were 7.7 and 1.9, respectively, and the smokers had more severe eye disease. (5) Smoking may alter the structure of the thyrotropin receptor slightly, so that in a susceptible person it becomes more immunogenic and the resulting antireceptor antibodies are more reactive with retro-orbital tissue. Alternatively, smoking could augment immunologic responsiveness to whatever factor initiates Graves' hyperthyroidism, sensitize retro-orbital tissue to whatever factor causes ophthalmopathy, or both.
What might account for these multiple abnormalities of thyroid function and multiple thyroid diseases in smokers? Variations in iodine intake are one factor that might modulate the response to smoking, the predominant action of smoking being antithyroid when iodine intake is low and immunogenic when it is adequate. However, iodine intake is higher in Switzerland, where Muller et al. worked, than in Italy, the Netherlands, and Sweden, where the studies of smoking and Graves' disease were done. (4,5,6) Other explanations include quantitative and qualitative differences in the components of tobacco smoke, underlying differences in thyroid function or preexisting thyroid disease in the subjects, and differences in susceptibility to thyroid autoimmune disease.
Better known for their multiple effects on the thyroid gland are iodine and several drugs. Too little iodine -- less than 50 µg a day -- is of course the cause of endemic goiter and cretinism. Too much iodine -- it need be only a few milligrams per day -- can induce either hyperthyroidism or hypothyroidism, depending on the patient. (7) The former occurs in patients with thyroid nodular disease, and the latter in those whose thyroid is damaged, whether from chronic autoimmune thyroiditis, previous radioactive-iodine therapy, or something else. These effects occur in response not only to inorganic iodine but also to any iodine-containing compound, including expectorants, antiseptics, drugs, and radiographic contrast agents. All the latter are deiodinated in vivo, releasing inorganic iodine, and a few -- notably amiodarone -- also inhibit extrathyroidal conversion of thyroxine to triiodothyronine.
Lithium carbonate also has bidirectional effects on the thyroid system. It has antithyroid actions in normal subjects and patients with hyperthyroidism. (8) Among patients with psychiatric disorders receiving long-term lithium therapy, about 5 percent have overt hypothyroidism and 25 percent have subclinical hypothyroidism. (9) Many of the patients have antithyroid antibodies, suggesting that lithium induces not only hypothyroidism but also chronic autoimmune thyroiditis. Finally, lithium therapy may be associated with an increased risk of Graves' hyperthyroidism. (10)
Many other drugs and substances can affect some facet of thyroid hormone production, metabolism, or action, but none have the diverse effects on thyroid economy or thyroid autoimmunity or are as commonly used as cigarettes, iodine and iodine-containing compounds, and lithium. Whenever a patient with thyroid dysfunction is encountered, the possibility that one of these substances caused or at least contributed to it should be kept in mind.
Robert D. Utiger, M.D.