Werner & Ingbar's The Thyroid: A Fundamental & Clinical Text, 9th Edition

67.Treatment of Hypothyroidism

Kenneth A. Woeber

Hypothyroidism was the first endocrine disorder to be treated with replacement of the deficient hormone. More than 100 years have elapsed since extracts of animal thyroid glandular tissue were found to be effective in treating hypothyroidism. Subsequently, chemically synthesized thyroid hormone preparations were developed, and these have largely replaced the use of thyroid extracts. In addition, the elucidation of thyroid hormone economy in humans and the ability to accurately measure serum thyrotropin (TSH) have made it possible to virtually mimic normal thyroid physiology (1,2). Accordingly, the treatment of hypothyroidism has become straightforward. Nevertheless, various clinical circumstances and drugs can affect the absorption, metabolism, and action of administered thyroid hormone and may necessitate adjustment of dosage. This chapter will review the treatment of hypothyroidism in the adult. The treatment of hypothyroidism in the infant and child is presented in Chapter 75, and the treatment of myxedema coma in Chapter 65.

PHARMACOLOGIC CONSIDERATIONS

We now know that in the normal adult about 100 µg of L-thyroxine (T4) is secreted by the thyroid daily and that about 30 µg of L-triiodothyronine (T3) is produced daily, 80% arising from the 5′-deiodination of T3 in the peripheral tissues and 20%, or about 6 µg daily, from thyroid secretion (1). Thus, T4 is the principal product of thyroid glandular secretion and serves as the prohormone for T3, which is the principal arbiter of thyroid hormone action (3).

Various thyroid hormone preparations are available for the treatment of hypothyroidism. They include levothyroxine sodium (L-thyroxine, T4), liothyronine sodium (L-triiodothyronine, T3), liotrix (a 1:4 combination of T3 and T4), and thyroid USP (a porcine thyroid glandular extract containing T3 and T4 in a ratio of ~1:4).

T4 is the preferred drug because its administration mostly closely mimics glandular secretion and because its conversion to T3 will be appropriately regulated in the tissues. Approximately 70% of an administered dose is absorbed, principally in the jejunum (4). The advantages of T4 include its long half-life in blood (~7 days), small (~13%) fluctuations in serum concentration between single daily doses, and ease of titration of dosage because of the availability of multiple tablet strengths. An additional important advantage of T4 is the stable serum T3 concentration between T4 doses (Fig. 67.1).

FIGURE 67.1. Schema depicting the effects of daily dosing with various thyroid hormone preparations on serum triiodothyronine concentration.

T3 and the T3-containing preparations have several drawbacks to their use. T3 is rapidly and virtually completely absorbed and has a half-life in blood of approximately 1 day. Accordingly, serum levels attain a peak in 2 to 4 hours, and a dose of 25 µg will produce supranormal serum concentrations for 6 to 8 hours, often accompanied by precordial palpitations. Moreover, currently available T3-containing preparations (liotrix and thyroid USP) contain T3 and T4 in a ratio of approximately 1:4, whereas the ratio of T3 to T4 in thyroid glandular secretion is much less, approximately 1:15.

The T4 content of tablets is standardized by high-performance liquid chromatography and, as stipulated by the USP, must conform to between 90% and 110% of the stated amount. Because reformulation of T4 tablets by various manufacturers had previously resulted in changes in stability and potency, the U.S. Food and Drug Administration recently enacted legislation to ensure that the amount of available drug is consistent with a given tablet strength and that the tablets from various manufacturers be bioequivalent (5). Numerous brand-name and some generic preparations are available in 12 strengths ranging from 25 to 300 µg, which allows precise replacement in most patients. The 12 strengths of the more common brand-name preparations in the United States are similarly color coded. T4 tablets are stable with a finite shelf-life at 15° to 30°C, but may lose potency if exposed to light or moisture. T4 is also available as a lyophilized preparation for injection.

T3 tablets are available in three strengths—5, 25, and 50 µg—and are not color coded. T3 is also available in an injectable preparation. Liotrix tablets are available in five color-coded strengths, ranging from tablets containing 3.1 µg of T3 and 12.5 µg of T4 to tablets containing 50 µg of T3 and 200 µg of T4. Thyroid USP is available in eight strengths, ranging from 0.25 gr to 5 gr and containing 9 µg of T3 and 38 µg of T4 per grain. The approximate bioequivalence of the various thyroid hormone preparations is 100 µg T4, approximately 37.5 µg T3, liotrix-1, and 1.5 gr (90 mg) thyroid USP.

Combination T4 and T3 treatment in the form of thyroid USP was used before chemically synthesized T4 became available. Although T4 has become the recommended form of replacement therapy, interest in the use of a combination of T4 and T3 has resurfaced. This interest has its basis in the observation that in thyroidectomized rats normal serum and tissue T3 concentrations can only be achieved with either supraphysiologic doses of T4 or a combination of T4 and T3(6). One short-term study in hypothyroid patients has suggested that partial substitution of T3 for T4 resulting in a T3 to T4 ratio of approximately 1:10 in the total replacement dosage improves mood and cognitive performance (7). This observation, however, was not corroborated in more recent studies (8,9,9a).

APPROACH TO TREATMENT

The goal of treatment is to restore a normal (euthyroid) state of the tissues with resolution of the symptoms and signs of hypothyroidism. This goal is accomplished by administering sufficient T4 so that the serum TSH in patients with primary hypothyroidism is reduced to within the normal range. The dosage of T4 that results in a normal serum TSH will increase the serum free T4 concentration to well within the normal range and the serum free T3 concentration into the low normal range, this discordance reflecting the lack of the thyroidal contribution to serum T3 concentration (10). Serum TSH bears a negative logarithmic-linear relationship to serum free T4 and is therefore a more sensitive index of the thyroid state than serum free T4 (2). At present, no sufficiently sensitive measurements exist for assessing the metabolic effects of T4 on the tissues. Thus, measurement of serum TSH is the best way to identify patients who are taking too much T4 (low serum TSH) or too little T4 (high TSH).

Serum TSH concentration does not reflect the thyroid state in patients with central hypothyroidism, and measurement of serum free T4 must be used to monitor treatment in this circumstance. Similarly, in patients recently treated for hyperthyroidism, delayed recovery of pituitary TSH secretion may result in a low serum TSH concentration for several months, necessitating measurement of serum free T4 to assess the thyroid state.

Patients who have taken thyroid USP or combined T4 and T3 preparations for many years may be reluctant to switch to T4 despite its advantages. However, such preparations can be continued provided that measurement of serum TSH, but not serum free T4 or free T3, is used to monitor therapy

INITIATION OF TREATMENT

The initial dosage of T4 should be based on the age of the patient, the severity and duration of hypothyroidism, and the presence of concurrent disorders. In healthy young patients with mild disease, treatment can be initiated with a dosage that approaches a full replacement dosage of approximately 1.6 µg/kg/day (11). In patients over 60 years of age without clinically overt cardiac disease and in patients with disease of long duration, treatment should be initiated with 50 µg of T4 per day. Treatment of patients with cardiac disease is discussed in a subsequent section.

T4 is taken in a single daily dose. Ideally, it should be taken alone on arising and at least 30 minutes before breakfast because some fiber and bran products as well as certain medications and dietary supplements can impede its absorption.

Because it takes about 2 months for serum TSH to attain a new steady state, the initial dosage of T4 should not be increased earlier. Thereafter, the dosage of T4 is gradually increased at about 2-month intervals until the patient is clinically euthyroid with serum TSH in the normal range.

MAINTENANCE AND MONITORING OF TREATMENT

With the exception of certain disorders that lead to self-limited hypothyroidism, such as recovery from thyroiditis or resolution of iodine-induced hypothyroidism, treatment will be lifelong. Because of the negative logarithmic-linear relationship between serum TSH and serum free T4, measurement of serum TSH is used to monitor replacement. Because serum TSH and free T4 levels are influenced by T4 administration, blood should be collected before the T4is taken (12). Dosage adjustments should be limited to 25 µg/day or less and should not occur more frequently than at 2-month intervals to permit serum TSH to stabilize. Similarly, if the brand of T4 is changed, measurement of serum TSH is advisable about 2 months later. Because normal values for serum TSH display a logarithmic normal distribution between approximately 0.5 and 5.0 mU/L with a tail at the upper end, a target range of 0.5 to 2.0 mU/L is the desired goal of treatment in otherwise healthy patients and is associated with optimal general well-being (13). Thereafter, the patient should be followed up at 6- to 12-month intervals with measurement of serum TSH and free T4 levels.

The replacement dosage of T4 may be affected by several factors. Patients with hyperthyroidism rendered hypothyroid with iodine 131 or subtotal thyroidectomy may initially require a lower dosage of T4 than patients with hypothyroidism due to chronic thyroiditis or following total thyroidectomy because of persistence of an autonomously functioning thyroid remnant. Progressive failure of a functioning thyroid remnant will necessitate an increase in dosage. Patients over 60 years of age may require a lower dosage of T4 because of its reduced metabolic disposition (14). Substantial changes in body weight may require an adjustment in T4 dosage.

Patients who forget or are unable to take T4 for several days should not be concerned because of the approximately 7-day half-life of T4. However, in patients who are unable to take T4 orally for prolonged periods, T4 can be given intravenously in a dosage of approximately 70% of the oral dose, reflecting the fractional absorption of an oral dose.

The response to treatment should be assessed both clinically and biochemically. Resolution of the symptoms and signs of hypothyroidism lags behind the return to normal of serum TSH concentration, and complete resolution requires at least 3 months. Many patients with hypothyroidism complain of weight gain and expect to lose substantial weight after treatment is initiated. Because most of the weight gain is due to fluid retention, the diuresis following treatment results in weight loss that seldom exceeds 10% of body weight.

FACTORS AFFECTING BIOAVAILABILITY OF THYROXINE

Various clinical circumstances and drugs can affect the bioavailability of T4 and may necessitate an adjustment of dosage (Table 67.1). Pregnancy results in an increased dosage requirement for T4 in women being treated for hypothyroidism. The increased requirement is due to an increase in serum thyroxine-binding globulin (TBG) concentration and to placental transfer of T4 to the fetus as well as placental 5′-deiodination of T4 to inactive reverse T3 (see Chapter 80). Oral, but not transdermal, estrogen, tamoxifen, and raloxifene also induce an increase in TBG and may result in an increased dosage requirement for T4 (15, 16, 17). Conversely, oral androgens, which induce a decrease in TBG, may result in a decreased requirement for T4.

TABLE 67.1. FACTORS THAT INCREASE THYROXINE REQUIREMENT


Pregnancy, estrogen, tamoxifen, raloxifene

Small bowel disease

Drugs or dietary supplements that reduce absorption

   Fiber, bran, soy protein

   Calcium carbonate

   Ferrous sulfate

   Aluminum hydroxide

   Cholestyramine

   Sucralfate

   Sodium polystyrene sulfonate

Drugs that increase metabolic disposition

   Rifampin

   Carbamazepine

   Phenytoin

   Phenobarbital

Mechanism not known

   Sertraline

   Lovastatin


Patients who have extensive small bowel disease or who consume large quantities of fiber, bran, or soy protein may absorb T4 less effectively and require more T4 (18, 19, 20). Many drugs and dietary supplements can reduce the bioavailability of T4 through various mechanisms. Ferrous sulfate and calcium carbonate reduce T4 absorption and are especially noteworthy because many older women consume these supplements and are the group in which hypothyroidism is most prevalent (21,22). Moreover, these compounds are present in prenatal vitamins and may contribute to the increased requirement for T4in pregnancy if concurrently taken with T4 (23). Accordingly, T4 should be taken at least 4 hours apart from the ingestion of drugs or dietary supplements that impede its absorption.

Drugs that induce hepatic microsomal enzymes, such as rifampin and anticonvulsants, increase the metabolic disposition of T4 and may increase its dosage requirement (24, 25, 26). Other drugs have been reported to increase T4 requirements, but these reports have been limited to a single patient or very few patients.

TREATMENT IN SPECIAL CIRCUMSTANCES

Central Hypothyroidism

In patients with central hypothyroidism, measurement of serum TSH is not useful for monitoring treatment with T4. Instead, measurement of both serum free T4 and free T3 before the daily dose of T4 is taken should be used to assess the adequacy of replacement, the target being normal serum values for both hormones (27). Because patients with central hypothyroidism have limited or no adrenal reserve, treatment with T4 alone may precipitate adrenal insufficiency by accelerating the metabolic disposition of cortisol as a euthyroid state is approached. Accordingly, hydrocortisone should be given until pituitary–adrenal function can be properly evaluated.

Pregnancy

As noted earlier, pregnancy results in an increased dosage requirement for T4, which becomes evident early in the first trimester. During the first trimester, serum TSH normally decreases because of the presence of high concentrations of chorionic gonadotropin, which has thyroid-stimulating activity. Accordingly, measurement of serum free T4 before the daily dose of T4 is taken should be used in conjunction with serum TSH to monitor treatment. Patients should be monitored at 6-week intervals until they are well into the second trimester, the target being the maintenance of a high normal serum free T4 and a low normal serum TSH. After parturition, the dosage of T4 can be reduced to the dosage before pregnancy.

Cardiac Disease

Hypothyroidism is a risk factor for accelerated atherosclerosis and coronary artery disease. These adverse consequences result from the dyslipidemia, increased total homocysteine, increased C-reactive protein, and increased prevalence of diastolic hypertension (28, 29, 30). Accordingly, in patients over 60 years of age with long-standing hypothyroidism, treatment should be initiated with 50 µg or less of T4 daily. In patients with known coronary heart disease, the initial dosage of T4 should not exceed 25 µg daily. Treatment will aggravate angina in about one fifth of patients with preexisting angina and result in no change or improvement in the remainder. In some patients without preexisting angina, treatment will provoke angina. Angina should be managed with aspirin and lower than normal dosages of nitrate and a β-adrenergic blocker and the dosage of T4 gradually increased at about 2-month intervals. If angina cannot be controlled, percutaneous transluminal coronary angioplasty or coronary artery bypass grafting should be considered. A recent study has shown that the mortality rate in men on T4 replacement undergoing bypass grafting was not increased, but that the mortality rate in women was increased, and this increase appeared to be due to insufficient replacement (31).

The dosage requirement for warfarin is increased in hypothyroidism because of the reduced metabolic disposition of the vitamin K–dependent clotting factors (32). Accordingly, in hypothyroid patients who are taking warfarin, the international normalized ratio should be closely monitored because warfarin dosage will need to be lowered as the euthyroid state is approached.

ADVERSE EFFECTS OF THYROXINE

Although hypersensitivity reactions to the inactive ingredients of a specific brand of T4 tablet or to the dye have been reported, these are exceedingly rare. On the other hand, adverse effects resulting from an excessive dosage of T4 are very common. It has been estimated that more than one fifth of patients taking thyroid medication for hypothyroidism have a subnormal serum TSH and therefore are either clinically or subclinically thyrotoxic (33). The adverse consequences of a low serum TSH are principally manifested on the skeleton and heart. In women over 65 years of age, a low serum TSH (< 0.1 mU/L) has been shown to result in a threefold increased risk for hip fracture and a fourfold increased risk for vertebral fracture (34). In persons over 60 years of age, a low serum TSH (< 0.1 mU/L) has been shown to be associated with a threefold increased risk for atrial fibrillation during the ensuing 10 years of follow-up (35). Moreover, subclinical thyrotoxicosis due to an excessive dosage of T4 has been reported to lead to an increase in resting heart rate, to a decrease in exercise capacity, and to an increase in left ventricular mass index with the potential for reduced diastolic function (36). Whether lesser degrees of TSH suppression are associated with fewer adverse consequences is uncertain. Nevertheless, the foregoing observations indicate the importance of avoiding a suppressed TSH value, the only exception being the patient who has undergone ablative treatment for thyroid cancer.

APPARENT FAILURE OF THYROXINE TREATMENT

Persistent Elevation of Serum Thyroid-Stimulating Hormone

Some patients display a persistently elevated serum TSH despite being prescribed a dosage of T4 that should be adequate. The most common reason for this phenomenon is partial compliance, and is suggested by a discordance between the serum TSH and free T4 values. As the TSH response to treatment is delayed, serum TSH will remain elevated in the presence of a normal serum free T4 if the daily dosage of T4 is taken on an irregular basis.

Persistent elevation of serum TSH despite patient compliance may be due to one or more of the factors affecting bioavailability of the T4 preparation described earlier. In their absence, changing the brand of T4 should be considered as the dissolution time, which is a function of the inactive ingredients, varies among different preparations, and this phenomenon could be a factor in the patient with rapid intestinal transit.

Persistence of Symptoms Despite Normal Serum TSH

Some patients continue to display symptoms suggesting a persistent hypothyroid state after serum TSH has been restored to within the normal reference range with treatment. Several reasons may account for this phenomenon. First, it is important to recognize that each patient has a unique set point for free T4in relation to TSH within the reference ranges for the laboratory (37). Accordingly, adjustment of T4 dosage to maintain the serum TSH between 0.5 and 2.0 mU/L, which is the reference range for a rigorously screened normal population (38), may serve to attenuate the patient's symptoms. Second, there is evidence to suggest that some patients on an optimal replacement dosage of T4 fail to attain a sense of normal general well-being (39). This observation might be due to preexisting comorbidity, such as depression. Indeed, the presence of antithyroid peroxidase antibodies has been shown to be an independent risk factor for depression (40).

Finally, the possibility exists that in some patients an optimal dosage of T4 does not provide through peripheral conversion a quantity of T3 that is sufficient to sustain a euthyroid state. In fact, in most patients on T4 replacement, serum free T3 is at the lower end of the normal reference range and in a small number of patients subnormal (10). One short-term study had suggested that partial substitution of T3 for T4 in the total replacement dosage of T4 improved mood and cognitive performance (7). This observation, however, has been refuted in 3 recent studies (8,9,9a). Thus, in most patients with hypothyroidism, combined treatment with T4 and T3 cannot be currently recommended over treatment with T4 alone.

REFERENCES

1. Pilo A, Iervasi G, Vitek F, et al. Thyroidal and peripheral production of 3,5,3′-triiodothyronine in humans by multicompartmental analysis. Am J Physiol 1990;258:E715–E726.

2. Spencer CA, Lopresti JS, Patel A, et al. Applications of a new chemiluminometric thyrotropin assay to subnormal measurement. J Clin Endocrinol Metab 1990;70:453–460.

3. Koenig RJ. Ubiquinated deiodinase: not dead yet. J Clin Invest 2003;112:145–147.

4. Hays MT, Nielsen KRK. Human thyroxine absorption: age effects and methodological analyses. Thyroid 1994;4:55–64.

5. Hennessey JV. In my view…Levothyroxine a new drug? Since when? How could that be? Thyroid 2000;13:279–282.

6. Escobar-Morreale HF, Escobar del Rey F, Obregon MJ, et al. Only the combined treatment with thyroxine and triiodothyronine ensures euthyroidism in all tissues of the thyroidectomized rat. Endocrinology 1996;137:2490–2502.

7. Bunevicius R, Kazanavicius G, Zalinkevicius R, et al. Effects of thyroxine as compared with thyroxine plus triiodothyronine in patients with hypothyroidism. N Engl J Med 1999;340:424–429.

8. Walsh JP, Shiels L, Mun Lim E, et al. Combined thyroxine/liothyronine treatment does not improve well-being, quality of life, or cognitive function compared to thyroxine alone: a randomized controlled trial in patients with primary hypothyroidism. J Clin Endocrinol Metab 2003;88:4543–4550.

9. Sawka AM, Gerstein HC, Marriott MJ, et al. Does a combination regimen of thyroxine (T4) and 3,5,3′-triiodothyronine improve depressive symptoms better than T4 alone in patients with hypothyroidism? Results of a double-blind, randomized, controlled trial. J Clin Endocrinol Metab 2003;88:4551–4555.

9a. Clyde PW, Harari AE, Getka EJ, et al. Combined levothyroxine plus liothyronine compared with levothyroxine alone in primary hypothyroidism. A randomized controlled trial. JAMA 2003; 290:2952–2958.

10. Woeber KA. Levothyroxine therapy and serum free thyroxine and free triiodothyronine concentrations. J Endocrinol Invest 2002;25:106–109.

11. Fish LH, Schwartz HL, Cavanaugh J, et al. Replacement dose, metabolism, and bioavailability of levothyroxine in the treatment of hypothyroidism. N Engl J Med 1987;316:764–770.

12. Ain KB, Pucino F, Shiver, T, et al. Thyroid hormone levels affected by time of blood sampling in thyroxine-treated patients. Thyroid 1993;3:81–85.

13. Hollowell JG, Staehling NW, Hannon WH, et al. Serum thyrotropin, thyroxine and thyroid antibodies in the United States population (1988–1994): NHANES III. J Clin Endocrin Metab 2002;87:489–499.

14. Sawin CT, Herman T, Molitch ME, et al. Aging and the thyroid. Decreased requirement for thyroid hormone in older hypothyroid patients. Am J Med 1983;75:206–209.

15. Arafah BM. Increased need for thyroxine in women with hypothyroidism during estrogen therapy. N Engl J Med 2001;344: 1743–1749.

16. Kostoglou-Athanassiou I, Ntalles K, Markopoulos C, et al. Thyroid function in postmenopausal women with breast cancer on tamoxifen. Eur J Gynaecol Oncol 1998;19:150–154.

17. Duntas LH, Mantzou E, Koutras DA. Lack of substantial effects of raloxifene on thyroxine-binding globulin in postmenopausal women: dependency on thyroid status. Thyroid 2001;11:779–782.

18. d'Esteve-Bonetti L, Bennet AP, Malet D, et al. Gluten-induced enteropathy (coeliac disease) revealed by resistance to treatment with levothyroxine and alfacalcidol in a sixty-eight-year-old patient: a case report. Thyroid 2002;12:633–636.

19. Liel Y, Harman-Boehm I, Shany S. Evidence for a clinically important adverse effect of fiber-enriched diet on the bioavailability of levothyroxine in adult hypothroid patients. J Clin Endocrinol Metab 1996;80:857–859.

20. Bell DSH, Ovalle F. Use of soy protein supplement and resultant need for increased dose of levothyroxine. Endocr Pract 2001;7: 193–194.

21. Campbell NRC, Hasinoff BB, Stalts H, et al. Ferrous sulfate reduces thyroxine efficacy in patients with hypothyroidism. Ann Intern Med 1992;117:1010–1013.

22. Singh N, Singh PN, Hershman JM. Effect of calcium carbonate on the absorption of levothyroxine. JAMA 2000;283:2822–2825.

23. Chopra IJ, Baber K. Treatment of primary hypothyroidism during pregnancy: is there an increase in thyroxine dose requirement in pregnancy? Metabolism 2003;52:122–128.

24. Nolan SR, Self TH, Norwood JM. Interaction between rifampin and levothyroxine. South Med J 1999;92:529–531.

25. Curran PG, DeGroot LJ. The effect of hepatic enzyme-inducing drugs on thyroid hormones and the thyroid gland. Endocr Rev 1991;12:135–150.

26. Isojarvi JI, Turrka J, Pakarinen AJ, et al. Thyroid function in men taking carbamazepine, oxcarbazepine, or valproate for epilepsy. Epilepsia 2001;42:930–934.

27. Ferretti E, Persani L, Jaffrain-Rea M-L, et al. Evaluation of the adequacy of levothyroxine replacement therapy in patients with central hypothyroidism. J Clin Endocrinol Metab 1999;84:924–929.

28. Duntas LH. Thyroid disease and lipids. Thyroid 2002;12:287–293.

29. Christ-Crain M, Meier C, Guglielmetti M, et al. Elevated C-reactive protein and homocysteine values: cardiovascular risk factors in hypothyroidism? A cross-sectional and a double-blind, placebo-controlled trial. Atheroscloerosis 2003;166:379–386.

30. Dernellis J, Panaretou M. Effects of thyroid replacement therapy on arterial blood pressure in patients with hypertension and hypothyroidism. Am Heart J 2002;143:718–724.

31. Zindrou D, Taylor KM, Peder Bagger J. Excess coronary artery bypass graft mortality among women with hypothyroidism. Ann Thorac Surg 2002;74:2121–2125.

32. Stephens MA, Self TH, Lancaster D, et al. Hypothyroidism: effect on warfarin anticoagulation. South Med J 1989;82:1585–1586.

33. Canaris GJ, Manowitz NR, Mayor G, et al. The Colorado thyroid disease prevalence study. Arch Intern Med 2000;160:526–534.

34. Bauer DC, Ettinger B, Nevitt MC, et al. Risk for fracture in women with low serum levels of thyroid-stimulating hormone. Ann Intern Med 2001;134:561–568.

35. Sawin CT, Geller A, Wolf PA, et al. Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older patients. N Engl J Med 1994;331:1249–1252.

36. Biondi B, Palmieri EA, Lombardi G, et al. Effects of subclinical thyroid dysfunction on the heart. Ann Intern Med 2002;137: 903–914.

37. Andersen S, Pedersen KM, Bruun NH, et al. Narrow individual variations in serum T4 and T3 in normal subjects: a clue to the understanding of subclinical thyroid disease. J Clin Endocrinol Metab 2002;87:1068–1072,

38. Spencer C, Hollowell J, Nicoloff J, et al. NHANES III: Impact of TSH: TPOAb relationships on redefining the serum TSH normal reference range. Presented at the American Thyroid Association 74th Annual Meeting, 2002. Program No. 2, p. 111.

39. Saravanan P, Chau W-F, Roberts N, et al.Psychological well-being in patients on “adequate” doses of L-thyroxine: results of a large, controlled community-based questionnaire study. Clin Endocrinol 2002;57:577–585.

40. Pop VJ, Maartens LH, Leusink G, et al. Are autoimmune thyroid dysfunction and depression related. J Clin Endocrinol Metab 1998;83:3194–3197.