Mark P.J. Vanderpump
The most common cause of thyroid disorders worldwide is iodine deficiency, leading to goiter formation and hypothyroidism. In iodine-replete areas, most persons with thyroid disorders have autoimmune disease, ranging from thyrotoxicosis to hypothyroidism. The problems encountered in epidemiologic studies of thyroid disorders are those of definition (e.g., overt hypothyroidism and subclinical hypothyroidism), selection criteria, influence of age and sex, environmental factors, the use of different techniques for assessment of thyroid function, and the relative paucity of incidence data. Recent data from screening large population samples in the United States have revealed differences in the frequency of thyroid dysfunction and high serum antithyroid antibody concentrations in different ethnic groups, whereas studies from Europe have revealed the influence of dietary iodine intake on the epidemiology of thyroid dysfunction.
A decrease in serum thyrotropin (TSH) concentration is the earliest measure of thyroid overactivity (subclinical thyrotoxicosis), followed by an increase in serum thyroxine (T4) and triiodothyronine (T3) concentrations (overt thyrotoxicosis). The most common causes of thyrotoxicosis are Graves' disease, followed by toxic multinodular goiter; rarer causes include an autonomously functioning thyroid adenoma, thyroiditis, and excessive T4 therapy. In epidemiologic studies, however, the etiology of thyrotoxicosis is rarely ascertained.
Prevalence of Thyrotoxicosis
The prevalence of thyrotoxicosis in women is between 0.5% and 2%, and it is 10 times more common in women than in men in iodine-replete communities (1). A cross-sectional study of 2779 subjects in the 1970s, in the community of Whickham, a mixed urban and rural area in northeast England, first documented the prevalence of thyroid disorders (2). The prevalence of undiagnosed thyrotoxicosis, based on clinical features and high serum T4 concentrations and high serum-free T4 index values was 4.7 per 1000 women. Thyrotoxicosis had been previously diagnosed and treated in 20 per 1000 women, increasing to 27 per 1000 women when possible, but unproven cases were included, as compared with 1.6 to 2.3 per 1000 men, in whom no new cases were found during the survey. Subsequent studies from Europe, Japan, and the United States confirmed these results (1). A cross-sectional survey of 25,682 subjects aged over 18 years attending a health fair in Colorado found that overt thyrotoxicosis, defined as serum TSH concentration ≤0.01mU/L, was present in only 1 per 1000 of those not taking thyroid medication (3). In the Third National Health and Nutrition Examination Survey (NHANES III) in the United States, serum TSH and total T4 were measured in a representative sample of 16,533 subjects aged over 12 years (4). In those subjects who were neither taking thyroid medication nor reporting histories of thyroid disease, 2 per 1000 had “clinically significant” thyrotoxicosis, defined as a serum TSH concentration < 0.1 mU/L and a serum total T4 concentration ≥13.0 µg/dL (170 nmol/L). In a survey of 1210 elderly persons over the age of 60 years in a single general practice in Birmingham, England, only 1 subject (sex not identified) was found to have thyrotoxicosis (5). A recent U.S. cross-sectional study of 2799 healthy community-dwelling adults aged 70 to 79 years found evidence of thyrotoxicosis (defined biochemically as a serum TSH concentration ≤0.1m U/L and a serum free T4 concentration >1.8 ng/dL (23 pmol/L) in only 5 subjects (1 man and 4 women) (6).
The prevalence of undiagnosed thyrotoxicosis in Pescopagano, Italy, an area of mild iodine deficiency (median urinary iodine excretion, 55 µg/L), was higher, at 2%, with a further 1% of adults there having a history of toxic nodular goiter (7). Approximately one third had a diffuse goiter; the frequency in men and women was similar. In a population sample of 2656 from Copenhagen, another area of mild iodine deficiency (median urinary iodine excretion, 70 µg/L), newly diagnosed thyrotoxicosis was found in 1.2% of women and no men, and the prevalence of known thyrotoxicosis was 1.4% (8).
The introduction of assays for serum TSH that were sensitive enough to distinguish between normal and low concentrations allowed subjects with subclinical thyrotoxicosis to be identified. Subclinical thyrotoxicosis is defined as a low serum TSH concentration and normal serum T4 and T3 concentrations, in the absence of hypothalamic or pituitary disease, nonthyroidal illness, or ingestion of drugs that inhibit TSH secretion (see Chapter 79). The available studies differ in the definition of a low serum TSH concentration and whether the subjects included were receiving T4 therapy.
The reported overall prevalence ranges from 0.5% to 6.3%, with men and women aged over 65 years having the highest prevalence, approximately half of whom were taking T4 (3,4,5,9,10). Among these studies, the serum TSH cut-off value ranged from < 0.1 to < 0.5 mU/L; it is not clear how this difference affected the reported prevalence rates. In the Colorado study of 25,862 subjects (of whom 88% were white), and in which the serum TSH cut-off value was 0.3 mU/L, the overall prevalence of subclinical thyrotoxicosis was 2.1%) (3). In contrast, the NHANES III study, defining subclinical thyrotoxicosis using a more stringent serum TSH cut-off value of 0.1 mU/L, reported an overall prevalence of 0.7% in the total population and 0.2% in the thyroid disease-free population (n = 13,344) (4). The rates were highest in those subjects aged 20 to 39 years and in those older than 79 years. In this study, the percentage of subjects with serum TSH concentrations < 0.4 mU/L was significantly higher in women than men, and black subjects had significantly lower mean serum TSH concentrations, and therefore a higher prevalence of subclinical thyrotoxicosis (0.4%) than whites (0.1%) or Mexican Americans (0.3%).
At the 20-year follow-up of the Whickham survey cohort, 4% had serum TSH values < 0.5 mU/L (normal, 0.5 to 5.2 mU/L), decreasing to 3% if those subjects were taking T4; those with newly diagnosed overt thyrotoxicosis were excluded (9). When serum TSH was measured in the same subjects using a more sensitive TSH assay (detection limit of 0.01 mU/L, coefficient of variation of 10% at 0.08 mU/L, and a normal range of 0.17 to 2.89 mU/L), approximately 2% had subnormal serum TSH concentrations (≥0.01 but < 0.17 mU/L), and 1% had undetectable serum TSH concentrations (< 0.01 mU/L). In subjects over 60 years of age in the Framingham Heart Study, 4% had a low serum TSH concentration (< 0.1 mU/L), of whom half were taking T4 (10).
Among subjects with subclinical thyrotoxicosis, those with low but detectable serum TSH values may recover spontaneously when retested. In the community survey in Birmingham, UK, 6% had low serum TSH concentrations, and 2% of women and 1% of men had undetectable values (< 0.05 mU/L) (5). One year later, 88% of those with undetectable serum TSH values continued to have a subnormal value, and 76% of those with a value of 0.05 to 0.5 mU/L had normal values.
The prevalence of subnormal serum TSH concentrations (detection limit 0.01 mU/L and excluding those subjects taking T4) was higher in the iodine-deficient population of Pescopagano (6%), due to functional autonomy from a nodular goiter (7). In Jutland, an area of mild iodine deficiency in Denmark, 10% of a random sample of 423 subjects had low serum TSH concentrations, as compared with 1% of 100 subjects of similar age in iodine-rich Iceland (11). Subclinical thyrotoxicosis was not detected in a group of elderly nursing home residents in an iodine-rich region of Hungary (12).
Incidence of Overt Thyrotoxicosis
The annual incidence data available for overt thyrotoxicosis from large population studies are comparable, at 0.4 per 1000 women and 0.1 per 1000 men, but the age-specific incidence varies considerably (1). The peak age-specific incidence of Graves' disease was between 20 and 49 years in two studies (13,14), but increased with age in Iceland (15) and peaked at 60 to 69 years in Malmö, Sweden (16). The peak age-specific incidence of thyrotoxicosis caused by toxic nodular goiter and autonomously functioning thyroid adenomas in the Malmö study was over 80 years. The only available data in a black population, from Johannesburg, South Africa, also suggest a tenfold-lower annual incidence of thyrotoxicosis (0.09 per 1000 women and 0.007 per 1000 men) than in whites (17). In a prospective study of 12 towns in England and Wales, the annual incidence of thyrotoxicosis strongly correlated with the prevalence of endemic goiter among schoolchildren 60 years earlier (18). Subsequent to this survey, serum samples from 216 of the 290 cases identified were assayed for TSH receptor antibodies. The frequency of antibody-positive thyrotoxicosis, an indicator of Graves' disease, did not correlate with goiter in the past (19).
In a 20-year follow-up of the Whickham cohort, the mean annual incidence of thyrotoxicosis in women was 0.8 per 1000 survivors (95% confidence interval, 0.5 to 1.4) (9). The incidence rate was similar in the women who had died during the 20-year interval. No new cases were detected in men. An estimate of the probability of the development of thyrotoxicosis in women at a particular time averaged 1.4 per 1000 between the ages of 35 and 60 years (Fig. 19.1). Neither serum thyroid antibody status nor the presence of a goiter during the first survey was associated with the development of thyrotoxicosis at follow-up. Other cohort studies provide comparable incidence data, which suggests that many cases of thyrotoxicosis remain undiagnosed in the community unless routine testing is undertaken (1).
FIGURE 19.1. The age-specific hazard rates for the development of overt thyrotoxicosis (hyperthyroidism) and hypothyroidism in women followed for 20 years in the community survey in Whickham, England. (From Vanderpump MPJ, Tunbridge WMG, French JM, et al. The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol (Oxf) 1995;43:55–69, with permission.)
Data on the risk of progression of subclinical to overt thyrotoxicosis are limited. At the 1-year follow-up of the subjects over 60 years of age in Birmingham, of 50 who initially had serum TSH values below normal, only 1 subject developed overt thyrotoxicosis (5). In short follow-up studies, the incidence has been calculated at 5% per year (20).
CHRONIC AUTOIMMUNE THYROIDITIS
The presence of high serum concentrations of thyroid antibodies [antithyroid peroxidase (microsomal) and antithyroglobulin] correlates with the presence of focal thyroiditis in thyroid tissue obtained by biopsy and at autopsy from patients with no evidence of hypothyroidism during life (21). Patients with hypothyroidism caused by either atrophic or goitrous autoimmune thyroiditis usually have high serum concentrations of these same antibodies. High serum concentrations of these antibodies, especially thyroid peroxidase antibodies, also are often found in patients with Graves' disease and other thyroid diseases, but not as often, and the serum concentrations are usually lower (see Chapter 15).
In all the early studies, regardless of the methodology, a progressive increase in prevalence of positive serum tests for thyroid antibodies was found with age in women, as compared with a uniformly low prevalence and no age trend in men. In the Whickham survey, mean serum TSH concentrations were significantly higher in both men and women with positive serum antithyroid antibody tests, and 3% of the subjects (5% of women, 1% of men) had both positive antibody tests and a serum TSH value >6 mU/L (2). In the NHANES III survey, the percentage of subjects with high serum antithyroid peroxidase and antithyroglobulin antibody concentrations increased with age in both men and women, and high concentrations were more prevalent in women than in men and less prevalent in blacks than in other ethnic groups (4). High serum antithyroglobulin antibody concentrations alone were not significantly associated with thyroid disease in this study.
At the 20-year follow-up of the Whickham survey, 19% of the survivors had high serum antithyroid peroxidase antibody concentrations and 5% had high serum antithyroglobulin antibody concentrations (9). Seventeen percent of women and 7% of men who initially had normal values now had high values, 9% of women and 2% of men had high values on both occasions, and 2% of women and 0.5% of men had high values initially but not at follow-up. Over 50% of the women in whom the serum antithyroid antibody concentrations changed from high to normal were receiving T4 treatment for hypothyroidism. There was no evidence that a high serum antithyroid antibody concentration at the original survey was a risk factor for premature death in this cohort (22).
The earliest biochemical abnormality in hypothyroidism is an increase in serum TSH concentration associated with normal serum T4 and T3 concentrations (subclinical hypothyroidism). This is followed by a decrease in serum T4 concentration, at which stage most patients have symptoms and benefit from treatment (overt hypothyroidism). In persons living in iodine-replete areas, the cause is usually either chronic autoimmune thyroiditis (atrophic autoimmune thyroiditis or goitrous autoimmune thyroiditis [Hashimoto's thyroiditis]) (see Chapter 47) or destructive treatment for thyrotoxicosis, but the cause was not often determined in the large cohort studies.
Prevalence of Hypothyroidism
In iodine-replete communities, the prevalence of spontaneous hypothyroidism is between 1% and 2%, and it is more common in older women and about ten times more common in women than in men (23). In the Whickham survey, the prevalence of newly diagnosed overt hypothyroidism was 3 per 1000 women (2). The prevalence of previously diagnosed and treated hypothyroidism was 14 per 1000 women, increasing to 19 per 1000 women when possible, but unproven, cases were included. The overall prevalence in men was < 1 case per 1000. One third had been previously treated by surgery or radioiodine for thyrotoxicosis. Excluding iatrogenic causes, the prevalence of hypothyroidism was 10 per 1000 women, increasing to 15 per 1000 when possible, but unproven, cases were included (Fig. 19.1). This is comparable with other studies, including the Colorado and NHANES III studies, in which the prevalence of newly diagnosed hypothyroidism was 4 per 1000 and 3 per 1000, respectively (3,4). In Pescopagano, the prevalence of newly diagnosed overt hypothyroidism was 0.3% of 573 women (chronic autoimmune thyroiditis confirmed as etiology) (there were no cases among 419 men), and no subject had been diagnosed and treated for hypothyroidism (7). In Copenhagen, 6 per 1000 of the women and 2 per 1000 men had overt but undiagnosed hypothyroidism, and 1% were taking T4 (8).
In the original Whickham survey, 8% of women (10% of women over 55 years of age) and 3% of men had subclinical hypothyroidism (2). Serum TSH concentrations did not change as a function of age among adult men, but in women over 40 years of age the concentrations increased. If, however, women with high serum antithyroid antibody concentrations were excluded, there was no age-related increase (2). In the Colorado study, 9.4% of the subjects had a high serum TSH concentration, of whom 9.0% had subclinical hypothyroidism (3). Among those with a high serum TSH concentration, 74% had a value between 5.1 and 10 mU/L, and 26% had a value >10 mU/L. The percentage of subjects with a high serum TSH concentration was higher for women than men in each decade of age, and ranged from 4% to 21% in women and 3% to 16% in men (Fig. 19.2). In the NHANES III study, serum TSH concentrations increased with age in both men and women and were higher in whites than blacks, independent of serum antithyroid antibody concentrations (4).
FIGURE 19.2. The percentage of 25,682 subjects with a high serum thyrotropin concentration, by sex and decade of age, in the Colorado Thyroid Study. (From Canaris GJ, Manowitz NR, Mayor G, et al. The Colorado Thyroid Disease Prevalence Study. Arch Intern Med 2000;160:526–534, with permission.)
Community studies of elderly persons have confirmed the high prevalence of subclinical hypothyroidism in this age group, with approximately 10% of subjects over the age of 60 years having serum TSH values above the normal range (5,24). Recent data from a cohort of people aged 70 to 79 years in the United States found that black subjects had a significantly lower prevalence of subclinical hypothyroidism (2% in men, 3% in women), as compared with white subjects (4% in men, 6% in women) (6). In iodine-deficient Pescopagano, there was a slightly lower prevalence of subclinical hypothyroidism (4% of women and 3% of men), but high serum antithyroid antibody concentrations were as prevalent, although lower, as in iodine-replete communities (7). In Copenhagen (mild iodine deficiency), only 0.7% of subjects had subclinical hypothyroidism, and 83% of them had serum antithyroid antibody concentrations >200 kU/L (8). Among elderly persons living in iodine-deficient areas, approximately 10% of subjects over age 60 years had serum TSH values above the normal range (10,11). Subclinical hypothyroidism is found at higher frequency (18% in Iceland and 24% in Hungary) in areas where iodine intake is high, but most cases are not of autoimmune origin (11,12).
After destructive treatment of thyrotoxicosis with either radioiodine or surgery, the incidence of overt hypothyroidism is greatest in the first year (25,26). In an audit of 813 consecutive patients treated for thyrotoxicosis in Birmingham, United Kingdom, there was an increase in the incidence of hypothyroidism at 1 year in those given higher doses of radioiodine (61% among those given 10 mCi (370 MBq) versus 41% among those given 5 mCi (185 MBq) (27). The incidence of hypothyroidism in patients with Graves' thyrotoxicosis was higher than that in patients with nodular goiter (55% vs. 32%). If the patient had subclinical hypothyroidism 1 year or more after radioiodine or surgical treatment, then the rate of progression to overt hypothyroidism was 2% to 6% per year (25,26).
The incidence of hypothyroidism after surgery, external radiation therapy of the neck, or both, in patients with head and neck cancer (including lymphoma) is as high as 50% within the first year after treatment, particularly in patients who underwent surgery and received high doses of radiation (28,29). The effect is dose dependent, the onset is gradual, and subclinical hypothyroidism can be present for many years before the development of overt hypothyroidism.
The 20-year follow-up of the Whickham cohort provided incidence data and allowed the determination of risk factors for spontaneous hypothyroidism in this period (9). The mean annual incidence of spontaneous hypothy roidism in the surviving women during the 20-year follow-up period was 3.5 per 1000 (95% confidence interval, 2.8 to 4.5), increasing to 4.1 per 1000 (95% confidence interval, 3.3 to 5.0) if all cases, including those who had received destructive treatment for thyrotoxicosis, were included. The hazard rate (i.e., the estimate of the probability of a woman developing hypothyroidism at a particular time) increased with age to 13.7 per 1000 in women 75 to 80 years of age (Fig. 19.1). The mean annual incidence during the 20-year follow-up period in men (all spontaneous except for 1 case of lithium-induced hypothyroidism) was 0.6 per 1000 (95% confidence interval, 0.3 to 1.2). The risk of having developed hypothyroidism was examined with respect to risk factors identified in the first survey. In the surviving women, the annual risk of spontaneous overt hypothyroidism was 4% in those who had both high serum TSH and antithyroid antibody concentrations, 3% if only their serum TSH concentration was high, and 2% if only their serum thyroid antibody concentration was high; at the time of follow-up, the respective rates overall of hypothyroidism were 55%, 33%, and 27%. The probability of developing hypothyroidism was higher in those women who had serum TSH concentrations above 2.0 mU/L and high serum titers of antithyroid microsomal antibodies during the first survey (Fig. 19.3). Neither a positive family history of any thyroid disease, the presence of a goiter at either the first or the follow-up survey, nor parity at first survey was associated with an increased risk of hypothyroidism.
FIGURE 19.3. The probability of development of overt hypothyroidism in 20 years among 912 women of the original Whickham cohort, as a function of their serum thyrotropin values at baseline. (From Vanderpump MPJ, Tunbridge WMG, French JM, et al. The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol (Oxf) 1995;43:55–69, with permission.)
The other incidence data for hypothyroidism are from short (and often small) follow-up studies. In a follow-up study of 437 healthy women 40 to 60 years of age in the Netherlands, 24% of those who initially had a positive serum test for antithyroid microsomal antibodies and normal serum TSH concentrations had a high serum TSH concentration (>4.2 mU/L) 10 years later, as compared with 3% in the women who had a negative test for the antibodies (30). As in the 20-year follow-up of the Whickham cohort, serum TSH concentrations in the upper part of the normal range in this study also appeared to have a predictive value. In a 9-year follow-up of a cohort of 82 women with subclinical hypothyroidism, the cumulative incidence of overt hypothyroidism was 0% in those with serum TSH concentrations of 4 to 6 mU/L, 43% in those with serum TSH concentrations >6 to 12 mU/L, and 77% in those with serum TSH concentrations >12 mU/L (31).
In this study, the incidence of overt hypothyroidism also was higher in those women with high serum antithyroid microsomal antibody concentrations at baseline (59% vs. 23%; p = 0.03), but a high serum antithyroid antibody concentration contributed much less to the risk of overt hypothyroidism than did a high baseline serum TSH concentration, in contrast to the Whickham data. All studies indicate that the higher the serum TSH value, the greater the likelihood of development of overt hypothyroidism in subjects with chronic autoimmune thyroiditis. High serum antithyroid antibody and TSH concentrations have prognostic importance in elderly subjects as well (5,32,33).
The most common thyroid disease in the community is simple (diffuse) goiter. The clinical grading of thyroid size is subjective and imprecise (34,35). Ultrasonography has been used in epidemiologic studies to assess thyroid size (36,37,38), leading to much higher estimates of goiter prevalence than in studies in which goiter size was assessed by physical examination.
Considerable regional variations in the incidence of goiter exist, even in nonendemic goiter areas. In cross-sectional surveys, the prevalence of diffuse goiter declines with age, the greatest prevalence is in premenopausal women, and the ratio of women to men is at least 4:1 (2). This decline in frequency of goiter with age is in contrast to the age-related increase in frequency of thyroid nodules and high serum antithyroid antibody concentrations. In the Framingham study of subjects over 60 years of age, the prevalence of single thyroid nodules was 3% and that of multiple nodules was 1%, as assessed by palpation (39). In a survey of 101 women 49 to 58 years of age, thyroid nodules were detected by ultrasonography in 36%, of which less than one third were detected by palpation (40). Ultrasonography as a screening test is too sensitive and will result in the unnecessary pursuit of findings that are so common that they rarely have any pathologic importance (see Chapter 16). A higher prevalence of multinodular goiter is found in areas of iodine deficiency (41).
Longitudinal studies confirm the decreasing frequency of goiter with age. In the 20-year follow-up of the Whickham cohort, 10% of women and 2% of men had a goiter, as compared with 23% and 5%, respectively, at the first survey (9). The presence of a diffuse goiter was not predictive of any clinical or biochemical evidence of thyroid dysfunction. In women, an association was found between the development of a goiter and thyroid antibody status at follow-up, but not initially. In a 20-year follow-up study of 11- to 18-year-old subjects in southwestern United States, 60% of the 92 subjects who had a diffuse goiter initially had spontaneous regression by the age of 30 years (42). Longitudinal data from Framingham suggests an annual incidence of thyroid nodules of 1 per 1000, and that, once formed, they tend to remain present for a long period of time; virtually all are benign (39).
SCREENING FOR THYROID DYSFUNCTION
Congenital hypothyroidism affects about one newborn in 3500 to 4500 births and is the most treatable cause of mental retardation. Iodine deficiency (intake below 25 µg/day), particularly in preterm infants, still accounts for many cases of congenital hypothyroidism in Europe, Asia, and Africa. Clinical diagnosis occurs in < 5% of newborns with hypothyroidism because symptoms and signs are often minimal (43). The value of screening for congenital hypothyroidism in heel-prick blood specimens is unquestioned, and it is now done routinely in many countries (see section on neonatal screening in Chapter 75).
Certain groups of adults who should have an assessment of thyroid function (e.g. measurement of serum TSH) at least once to detect thyroid dysfunction include those with atrial fibrillation (44) or hyperlipidemia (45). There is a high frequency of asymptomatic thyroid dysfunction in unselected patients with diabetes mellitus, and annual assessment of thyroid function in patients with diabetes is cost-effective (46). There is no consensus on whether healthy women should be screened for postpartum thyroiditis (see Chapters 27 and 80). However, women with type 1 diabetes are three times more likely to develop postpartum thyroid dysfunction than are normal women, and therefore all such diabetic women should be tested (47). Any woman with a past history of postpartum thyroiditis should be offered annual assessment of thyroid function, in view of their increased long-term risk of permanent hypothyroidism. Similarly, in view of the high prevalence of hypothyroidism in patients with Down's syndrome (48) and Turner's syndrome (49), these patients should also have an annual assessment of thyroid function. Assessment of thyroid function is indicated every 6 months in patients receiving amiodarone and lithium and every 12 months in those treated with head and neck radiation (see section of effects of drugs and other substances on thyroid hormone synthesis and the section on effect of excess iodide in Chapters 11 and 50) (23).
All patients with thyrotoxicosis who receive destructive treatment for thyrotoxicosis should be followed indefinitely for the development of hypothyroidism; this follow-up should begin 4 to 8 weeks after treatment, then at 3-month intervals for 1 year, and then annually thereafter (see Chapter 45). Among patients hospitalized for acute illness, the occurrence of thyroid disease is no more common than in the general population. Therefore, testing should be limited but with a high index of clinical suspicion, particularly in elderly women, and with an awareness of the difficulties in inter preting thyroid function tests in the presence of acute illness (see section on nonthyroidal illness in Chapter 11) (50).
Controversy exists as to whether healthy adults living in an area of iodine sufficiency benefit from screening for thyroid disease. The prevalence of unsuspected overt thyroid disease is low, but as described above a substantial proportion of subjects tested will have evidence of thyroid dysfunction, usually subclinical hypothyroidism. In the absence of the confounding effects of nonthyroidal illness or drugs, a normal serum TSH concentration has a high predictive value in ruling out thyroid disease in healthy subjects. In unselected populations, measurement of serum TSH has a sensitivity of 89% to 95% and specificity of 90% to 96% for overt thyroid dysfunction, as compared with cases confirmed by history, examination, and additional testing. Normal serum TSH concentrations are found in some patients with hypothyroidism caused by pituitary or hypothalamic disease, but both these situations are rare (see Chapter 13). In nearly all studies serum TSH value >5 or 6 mU/L is accepted as being raised (51).
Various physician organizations have made different recommendations as to whether subclinical thyroid disease is of sufficient clinical importance to warrant screening and therapy (1,52,53,54,55,56,57,58,59). A cost-utility analysis using a computer decision model suggested that the cost-effectiveness of screening for subclinical hypothyroidism compared favorably with other preventive medical practices, such as screening for hypertension or breast cancer in women in the same age group, while providing a similar increase in quality-adjusted life years (60,61). Over one half of the presumed benefit in the latter was accounted for by preventing progression to overt hypothyroidism, 30% by improving associated mild symptoms, and 2% by preventing cardiovascular disease. The cost of detecting subclinical hypothyroidism was $9,223 for women and $22,595 for men per quality-of-life year gained, but this cost was heavily dependent on the cost of the serum TSH assay. The cost-benefit analysis did not allow for the extra costs of detecting, investigating, and potentially treating subclinical thyrotoxicosis.
This analysis led the American Thyroid Association to recommend population-based screening for thyroid dysfunction by measurement of serum TSH, beginning at age 35 years and every 5 years thereafter (62). Those organizations that assess screening programs, such as the U.S. Preventive Services Task Force, have not recommended regular assessment of thyroid function in adults (55). The reason is that there have been no trials to determine whether identification and treatment of subjects with thyroid dysfunction, which would mostly be subclinical hypothyroidism, results in any long-term benefit. The potential benefits in terms of decreased symptoms or other systemic effects are generally small and may not enhance quality of life (see Chapters 78 and 79). In a recent observational 10-year study, a low serum TSH concentration (< 0.05 mU/L), but not a high serum TSH concentration, was associated with an increase in all-cause mortality and cardiovascular mortality (63).
If screening is done, and a high serum TSH concentration is found, the measurement should be repeated 1 or 2 months later, along with measurement of serum-free T4, after excluding nonthyroidal illness and drug interference. If the serum TSH concentration is high and the serum-free T4 concentration is low, then the subject has overt hypothyroidism, and, even if asymptomatic on further questioning, should be treated with T4. If the serum free T4 concentration is normal, but the serum TSH concentration is >10 or 15 mU/L, then treatment with T4 is reasonable, if only because of the likelihood of progression. If the serum TSH concentration is between 5 and 10 mU/L, then serum antithyroid peroxidase antibodies should be measured. If the serum antibody concentration is high, then serum TSH should be measured annually; T4 therapy should be started if the serum TSH concentration rises above 10 mU/L. If the serum antibody concentration is normal, then repeat measurement of serum TSH every 3 to 5 years may be all that is required (1).
Few subjects screened will have overt thyrotoxicosis, but the consequences of finding subclinical thyrotoxicosis have to be addressed. If a subject has a low screening serum TSH value, the first step is to repeat the measurement of serum TSH and measure serum free T4, to identify overt thyrotoxicosis (and also central hypothyroidism). Usually, subclinical thyrotoxicosis will be confirmed. In addition to the risk of overt thyrotoxicosis, the subject may be at risk for atrial fibrillation and osteoporosis (see Chapter 79) (64,65). No consensus exists regarding the treatment of subclinical thyrotoxicosis, although therapy with an antithyroid drug or radioiodine may be indicated (66). Any potential benefits of therapy in subclinical thyrotoxicosis must be weighed against the substantial morbidity associated with the treatment of thyrotoxicosis (see Chapter 45). If treatment is not undertaken, serum TSH should be measured at least annually, with follow-up measurements of serum-free T4 if the serum TSH value is low.
There is an urgent need for long-term studies of the effects of treatment of both subclinical hypothyroidism and subclinical thyrotoxicosis to determine if there is indeed benefit from screening for thyroid dysfunction in adults.
1. Tunbridge WMG, Vanderpump MPJ. Population screening for autoimmune thyroid disease. Endocrinol Metab Clin North Am 2000;29:239–253.
2. Tunbridge WMG, Evered DC, Hall R, et al. The spectrum of thyroid disease in the community: The Whickham Survey. Clin Endocrinol (Oxf) 1977;7:481–491.
3. Canaris GJ, Manowitz NR, Mayor G, et al. The Colorado Thyroid Disease Prevalence Study. Arch Intern Med 2000; 160:526–534.
4. Hollowell JG, Staehling NW, Flanders D, et al. Serum TSH, T4, and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Survey (NHANES III). J Clin Endocrinol Metab 2002;87:489–499.
5. Parle JV, Franklyn JA, Cross KW, et al. Prevalence and follow-up of abnormal thyrotrophin (TSH) concentrations in the elderly in the United Kingdom. Clin Endocrinol (Oxf) 1991;34: 77–83.
6. Kanaya AM, Harris F, Volpato S, et al. Association between thyroid dysfunction and total cholesterol level in an older biracial population: The Health, Aging and Body Composition Study. Arch Intern Med 2002;162:773–779.
7. Aghini-Lombardi F, Antonangeli L, Martino E, et al. The spectrum of thyroid disorders in an iodine-deficient community: The Pescopagano Survey. J Clin Endocrinol Metab 1999;84:561–566.
8. Knudsen N, Jørgensen T, Rasmussen S, et al. The prevalence of thyroid dysfunction in a population with borderline iodine deficiency. Clin Endocrinol (Oxf) 1999;51:361–367.
9. Vanderpump MPJ, Tunbridge WMG, French JM, et al. The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol (Oxf) 1995; 43:55–69.
10. Sawin CT, Geller A, Kaplan MM, et al. Low serum thyrotropin (thyroid-stimulating hormone) in older persons without hyperthyroidism. Arch Intern Med 1991;151:165–168.
11. Laurberg P, Pedersen KM, Hreidarsson A, et al. Iodine intake and the pattern of thyroid disorders: a comparative epidemiological study of thyroid abnormalities in the elderly in Iceland and in Jutland, Denmark. J Clin Endocrinol Metab 1998;83:765–769.
12. Szabolcs I, Padoba J, Feldkamp J, et al. Comparative screening for thyroid disorders in old age in areas of iodine deficiency, long-term iodine prophylaxis and abundant iodine intake. Clin Endocrinol (Oxf) 1997;47:87–92.
13. Furszyfer J, Kurland LT, McConahey WM, et al. Graves' disease in Olmsted County, Minnesota, 1935 through 1967. Mayo Clin Proc 1970;45:636–644.
14. Mogensen EF, Green A. The epidemiology of thyrotoxicosis in Denmark. Incidence and geographical variation in the Funen region 1972–1974. Acta Med Scand 1980;208:183–186.
15. Haraldsson A, Gudmundsson ST, Larusson G, et al. Thyrotoxicosis in Iceland 1980–1982. An epidemiological survey. Acta Medica Scandinavica 1985;217:253–258.
16. Berglund J, Christensen SB, Hallengren B. Total and age-specific incidence of Graves' thyrotoxicosis, toxic nodular goitre and solitary toxic adenoma in Malmö 1970–1974. J Intern Med 1990;227:137–141.
17. Kalk WJ, Kalk J. Incidence and causes of hyperthyroidism in blacks. S Afr Med J 1989;75:114–116.
18. Barker DJB, Phillips DIW. Current incidence of thyrotoxicosis and past prevalence of goiter in 12 British towns. Lancet 1984; 2:567–570.
19. Phillips DIW, Barker DJ, Rees Smith B, et al. The geographical distribution of thyrotoxicosis in England according to the presence or absence of TSH-receptor antibodies. Clin Endocrinol (Oxf) 1985;23:283–287.
20. Wiersinga WM. Subclinical hypothyroidism and hyperthyroidism. I. Prevalence and clinical relevance. Neth J Med 1995; 46:197–204.
21. Dayan CM, Daniels GH. Chronic autoimmune thyroiditis. N Engl J Med 1996;335:99–107.
22. Vanderpump MPJ, Tunbridge WMG, French JM, et al. The development of ischemic heart disease in relation to autoimmune thyroid disease in a 20-year follow-up of an English community. Thyroid 1996;6:155–160.
23. Vanderpump MPJ, Tunbridge WMG. Epidemiology and prevention of clinical and subclinical hypothyroidism. Thyroid 2002;12:839–847.
24. Sawin CT, Bigos ST, Land S, et al. The aging thyroid: relationship between elevated serum thyrotropin level and thyroid antibodies in elderly patients. Am J Med 1985;79:591–595.
25. Franklyn JA, Daykin J, Droic Z, et al. Long term follow-up of treatment of thyrotoxicosis by three different methods. Clin Endocrinol (Oxf) 1991;34:71–76.
26. Berglund J, Christensen SB, Dymling JF, et al. The incidence of recurrence and hypothyroidism following treatment with antithyroid drugs, surgery or radioiodine in all patients with thyrotoxicosis in Malmö during the period 1970–1974. J Intern Med 1991;229:435–442.
27. Allahabadia A, Daykin J, Sheppard MC, et al. Radioiodine treatment of hyperthyroidism—prognostic factors for outcome. J Clin Endocrinol Metab 2001;86:3611–3617.
28. Grande C. Hypothyroidism following radiotherapy for head and neck cancer: multivariate analysis of risk factors. Radiother Oncol 1992;25:31–36.
29. Tell R, Sjodin H, Lundell G, Lewin F, et al. Hypothyroidism after external radiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 1997;39:303–308.
30. Geul KW, van Sluisveld ILL, Grobbee DE, et al. The importance of thyroid microsomal antibodies in the development of elevated serum TSH in middle-aged women: associations with serum lipids. Clin Endocrinol (Oxf) 1993;39:275–280.
31. Huber G, Staub J-J, Meier C, et al. Prospective study of the spontaneous course of subclinical hypothyroidism: Prognostic value of thytoropin, thyroid reserve, and thyroid antibodies. J Clin Endocrinol Metab 2002;87:3221–3226.
32. Sawin CT, Castelli WP, Hershman JM, et al. The aging thyroid: thyroid deficiency in the Framingham Study. Arch Intern Med 1985;145:1386–1388.
33. Rosenthal KJ, Hunt WC, Garry PJ, et al. Thyroid failure in the elderly: microsomal antibodies as discriminant for therapy. JAMA 1987;258:209–213.
34. Jarløv EA, Hegedüs L, Gjørup T, et al. Observer variation in the clinical assessment of the thyroid gland. J Intern Med 1991; 229:159–161.
35. Jarløv EA, Hegedüs L, Gjørup T, et al. Inadequacy of the WHO classification of the thyroid gland. Thyroidology 1992;4:107–110.
36. Brander A, Viikinkoski P, Nickels J, et al. Thyroid gland: US screening in random adult population. Radiology 1991;181: 683–687.
37. Hintze G, Windeler J, Baumert J, et al. Thyroid volume and goitre prevalence in the elderly as determined by ultrasound and their relationships to laboratory indices. Acta Endocrinologica 1991;124:12–18.
38. Nygaard B, Gideon P, Dige-Petersen H, et al. Thyroid volume and morphology and urinary iodine excretion in a Danish municipality. Acta Endocrinologica 1993;129:505–510.
39. Vander JB, Gaston EA, Dawber TR. The significance of nontoxic thyroid nodules: final report of a 15-year study of the incidence of malignancy. Ann Intern Med 1968;69:537.
40. Brander A, Viikinkoski P, Nickels J, et al. Thyroid gland: US screening in middle-aged women with no previous thyroid disease. Radiology 1989;173:507–510.
41. Smyth PPA, Darke C, Parkes AB, et al. Assessment of goitre in an area of endemic iodine deficiency. Thyroid 1999;9:895–901.
42. Rallison ML, Dobyns BM, Meikle AW, et al. Natural history of thyroid abnormalities: prevalence, incidence, and regression of thyroid diseases in adolescents and young adults. Am J Med 1991;91:363–370.
43. Anonymous. Newborn Screening for congenital hypothyroidism: recommended guidelines. Pediatrics 1993;91:1203–1209. American Academy of Pediatrics AAP Section on Endocrinology and Committee on Genetics and American Thyroid Association on Public Health.
44. Woeber KA. Thyrotoxicosis and the heart. N Eng J Med 1992;327:94–98.
45. Pirich C, Mullner M, Sinzinger H. Prevalence and relevance of thyroid dysfunction in 1922 cholesterol screening participants. J Clin Epidemiol 2000;53:623–629.
46. Perros P, McCrimmon RJ, Shaw G, et al. Frequency of thyroid dysfunction in diabetic patients: value of annual screening. Diabet Med 1996;12:622–627.
47. Alvarez-Marfany M, Roman S, Drexler A, et al. Long-term prospective study of postpartum thyroid dysfunction in women with insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1994;79:10–16.
48. Karlsson B, Gustafsson J, Hedov G, et al. Thyroid dysfunction in Down's syndrome: relation to age and thyroid autoimmunity. Arch Dis Child 1998;79:242–245.
49. Chiovato L, Larizza D, Bendinelli G, et al. Autoimmune hypothyroidism and hyperthyroidism in patients with Turner's syndrome. Eur J Endocrinol 1996;134:568–575.
50. Attia J, Margetts P, Guyatt G. Diagnosis of thyroid disease in hospitalized patients: a systematic review. Arch Intern Med 1999;159:658–665.
51. Baloch Z, Carayon P, Conte-Devolx B, et al. Laboratory medicine practice guidelines: laboratory support for the diagnosis and monitoring of thyroid disease. Thyroid 2003;13:3–126.
52. Surks MI, Chopra IJ, Mariash CN, et al. American Thyroid Association guidelines for use of laboratory tests in thyroid disorders. JAMA 1990;263:1529–1532.
53. Singer PA, Cooper DS, Levy EG, et al. Treatment guidelines for patients with hyperthyroidism and hypothyroidism. JAMA 1995;273:808–812.
54. Glenn GC. Practice parameter on laboratory panel testing for screening and case finding in asymptomatic adults: the Laboratory Testing Strategy Task Force of the College of American Pathologists. Arch Pathol Lab Med 1996;120:929–943.
55. US Preventive Services Task Force. Screening for thyroid disease. In: Guide to clinical preventive services. 2nd ed. Baltimore: Williams & Wilkins, 1996:209–218.
56. Vanderpump MPJ, Ahlquist JAO, Franklyn JA, et al. Consensus statement for good practice and audit measures in the management of hypothyroidism and hyperthyroidism. BMJ 1996;313: 539–544.
57. Weetman AP. Hypothyroidism: screening and subclinical disease. BMJ 1997;314:1175–1178.
58. Helfand M, Redfern CC. Clinical guideline, part 2: Screening for thyroid disease: an update. Ann Intern Med 1998;129:144–158.
59. Buskin HJ. American Association of Clinical Endocrinologists Medical Guidelines for clinical practice for the evaluation and treatment of hyperthyroidism and hypothyroidism. Endocr Pract 2002; 8:458–467.
60. Danese MD, Powe NR, Sawin CT, et al. Screening for mild thyroid failure at the periodic health examination: a decision and cost-effectiveness analysis. JAMA 1996;276:285–292.
61. Powe NR, Danese MD, Ladenson PW. Decision analysis in endocrinology and metabolism. Endocrinol Metab Clin North Am 1997;26:89–111.
62. Ladenson PW, Singer PA, Ain KB, et al. American Thyroid Association Guidelines for detection of thyroid dysfunction. Arch Intern Med 2000;160:1573–1575.
63. Parle JV, Maisonneuve P, Sheppard MC, et al. Prediction of all-cause and cardiovascular mortality in elderly people from one thyrotropin result: a 10-year cohort study. Lancet 2001;358: 861–865.
64. Sawin CT, Geller A, Wolf PA, et al. Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older persons. N Engl J Med 1994;331:1249–1252.
65. Uzzan B, Campos J, Cucherat M, et aDl. Effects of bone mass of long-term treatment with thyroid hormones: a meta-analysis. J Clin Endocrinol Metab 1996;81:4278–4289.
66. Toft AD. Subclinical hyperthyroidism. N Eng J Med 2001; 345:512–516.