The Treatment of Subclinical Hypothyroidism Is Seldom Necessary

James W. Chu and Lawrence M. Crapo

Division of Endocrinology, Stanford University School of Medicine and Santa Clara Valley Medical Center, San Jose, California 95128

Subclinical hypothyroidism (SH) is a common disorder with a prevalence ranging from 1–10% of the adult population in most community studies (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14). The risk of developing SH increases with female gender, advanced age, and greater dietary iodine intake. The view that most subjects with SH should be treated with L-thyroxine has gained increasing credence (15, 16, 17). A limited number of placebo-controlled randomized trials involving a small number of patients with SH have been performed (18, 19, 20, 21, 22), and several of these studies show that L-thyroxine therapy may reduce symptoms of hypothyroidism (18, 22). Other studies in subjects with SH have shown that L-thyroxine lowers low-density lipoprotein (LDL) cholesterol, improves cardiac function, and diminishes neuropsychiatric symptoms (15, 16, 17). A recent cross-sectional observational study noted an association between SH and atherosclerotic disease (13). On initial assessment then, it appears that the position advocating L-thyroxine therapy for most subjects with SH enjoys strong experimental support.

However, careful scrutiny of the entire spectrum of primary data bearing on the question of whether or not SH should be treated with L-thyroxine yields a legitimate contrary view. There are as many placebo-controlled randomized trials that observed no reduction in symptoms of SH (20, 21) as there are that note benefits of treatment. Other reports have noted the doubtful clinical significance and frequent statistical nonsignificance of L-thyroxine therapy on changes in LDL cholesterol, myocardial performance, and neuropsychiatric parameters. No association between SH and ischemic heart disease was shown in the Whickham survey (23), the most extensive longitudinal study of thyroid disease ever conducted. Such conflicting findings stem from inconsistencies of the reports in the variable definition of SH, the wide degree of thyroid failure examined, as well as the heterogeneous age, gender, and ethnicity of the subjects tested.

There is considerable evidence suggesting that the subjects with SH who would benefit most from L-thyroxine therapy are those with TSH levels exceeding 10 mU/liter. Such individuals constitute the minority of those with SH in all large-scale epidemiological studies that have stratified TSH levels (1, 2, 4, 5, 6, 7, 8, 9, 12, 14). The majority of subjects with SH, however, have slight elevations of TSH ranging between 5 and 10 mU/liter, and they have minimal, often nonsignificant, metabolic abnormalities. They are either affected by mild incipient thyroid failure for which L-thyroxine therapy has not been shown to convey recognizable benefits, or they may simply represent "euthyroid outliers" in the 2.5% tail above the upper limit of the normal TSH reference range, in which case L-thyroxine treatment would be inappropriate. The available evidence also calls into question the need to treat men with SH, who are much less prevalent than women with SH (1, 4, 6, 7, 10, 11, 14), and who manifest almost no metabolic differences compared with men with normal TSH levels.

Unfortunately no large-scale multicenter randomized trial has been performed to address and settle these issues. As a result, we are left to struggle with and sort through the conflicting results from a variety of small studies. Given this lack of definitive data, clinicians must weigh the benefit to risk ratio of initiating L-thyroxine therapy in subjects with SH, realizing that lifelong treatment is not without inconvenience or potential morbidity. In the absence of definitive outcome studies, we will conclude that the decision not to treat the majority of SH subjects is as strongly or even more strongly supported by the existent data as is the decision to treat. Thus, the burden of proof rests squarely on the shoulders of clinicians who initiate lifelong L-thyroxine treatment for SH to demonstrate that the goals of such therapy are clear and compelling.

Definition of SH

The biochemical state characterized by an elevated serum TSH level with a concomitant normal free T4 (FT4) level has received a variety of designations, including mild thyroid failure, as well as compensated, early, latent, mild, minimally symptomatic, and preclinical hypothyroidism (15, 17). The most widely applied designation for this biochemical state is subclinical hypothyroidism, and this is the one that we will use, despite the fact that its meaning is somewhat ambiguous. Not all subjects with a normal FT4 and elevated TSH are "subclinical"; some have symptoms or signs of hypothyroidism. Moreover, not all of such subjects have early thyroid failure destined for eventual overt hypothyroidism. Adding to the ambiguity, some authors have extended the definition of SH to include subjects with normal FT4 and TSH, but exaggerated TSH responses to TRH.

Despite the confusing nomenclature, we will define SH as a condition with a normal FT4 and elevated TSH. The state of SH may be found in at least five distinct situations (Table 1Go): mild unrecognized thyroid failure, undertreated overt hypothyroidism, overtreated overt hyperthyroidism, transient disturbances of the thyroid axis, and euthyroid outliers (this term will be applied to the 2.5% of individuals possessing TSH values above the 97.5 percentile of the euthyroid distribution). A cogent treatment plan for SH should rely on awareness of the underlying pathogenesis, because the five different clinical states clearly require separate therapeutic strategies. There is no dispute about the need for adjustment of medication or L-thyroxine doses as well as closer monitoring of thyroid function tests in patients with over- or undertreated overt thyroid dysfunction. Few would advocate L-thyroxine treatment of subjects with transient dysfunction of the thyroid axis. Euthyroid outliers, if distinguishable from other subjects, clearly do not merit treatment. The controversy surrounding treatment of SH then involves only those with incipient thyroid failure.


View this table:
[in this window]
[in a new window]
 
Table 1. Differential diagnosis of SH

 
We recognize the graded phenomenon of thyroid gland failure (24) and will consequently define four stages of hypothyroidism. The earliest form is identifiable only by a positive TRH test and is represented by the state of a normal FT4 accompanied by a TSH level slightly above a given individual’s set point, but within the population reference range (stage a). Further along the continuum of thyroid failure are the two stages of SH that involve mild elevations of TSH from 5–10 mU/liter (stage b) and prominent elevations above 10 mU/liter (stage c); subjects in these two groups exhibit dissimilar characteristics, as will be discussed. The final stage of this continuum is overt thyroid gland failure or overt hypothyroidism, which is defined by a low FT4 and an elevated TSH, usually much higher than 10 mU/liter (stage d).

Epidemiology of SH

In cross-sectional community studies involving at least 1000 subjects, the overall prevalence of an elevated TSH alone, or with a concomitant normal FT4, ranges from 0.2–5.7% in men and 1.2–13.6% in women (Table 2Go). Surveys that stratified TSH levels indicate a predominance of stage b subjects (TSH <10 mU/liter), who account for 55–85% of all cases of SH. Studies that have reported thyroid autoantibody tests on subjects with elevated TSH demonstrate seropositivity rates from 20–78%. The widely variable scope of the surveys in Table 2Go is evidenced by some determining prevalence of elevated TSH without testing FT4, and others lumping newly diagnosed SH together with patients possessing iatrogenic SH. This heterogeneous composition of subjects, along with variations in age groups, ethnic extraction, methods of subjects selection, and degrees of iodine repletion, all contribute to the broad spread of prevalence values.


View this table:
[in this window]
[in a new window]
 
Table 2. Prevalence and distribution of elevated TSH and SH

 
Euthyroid outliers and SH

The manner in which a "normal" TSH reference range is constructed relates directly to the definition and epidemiology of SH. Numerous studies note that the distribution of serum TSH values across a healthy, ambulatory adult population without history of thyroid disease or antibodies is continuous yet non-Gaussian on a linear scale, with a marked skew toward the left (Fig. 1Go). Logarithmic transformation of the TSH values results in a normal or near-normal Gaussian distribution (25, 26). Consequently, the generally accepted method for constructing a reference range by either nonparametrically bracketing the central 95% of the linear TSH distribution, or by determining the mean ± 2 SD of the logarithmic TSH distribution will result in similar values for the upper and lower limits of normal. Such a methodology will classify (whether justifiably or not) 2.5% of euthyroid subjects as fulfilling biochemical criteria for SH due to TSH values occurring in the upper tail of the distribution (27). However, in several general population surveys, the percentage of subjects with an elevated TSH is well below the expected 2.5% for men (7, 10, 11, 14) and women (3, 11, 14) (Table 2Go). This suggests that the normal range for some TSH assays (established in small reference samples) may not be applicable to large general populations, especially those of men. In other words, the methodology used to establish the upper limit of normal for a given TSH assay will profoundly influence the prevalence of SH as well as the percentage of euthyroid outliers.



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Theoretical distributions of TSH values in populations of individuals with euthyroidism (solid line) and thyroid failure (dashed line), whose quantitative features are estimated from epidemiological surveys. The parameters of the TSH distribution in patients with thyroid failure are supported by data from the follow-up Whickham survey (9 ), which noted an increased risk for progression to overt hypothyroidism for TSH values exceeding 2 mU/liter. The peak of the euthyroid distribution lies between 1 and 2 mU/liter and is censored in this diagram due to its large magnitude. The shaded area encompassing TSH values of 5–10 mU/liter indicates an important overlap zone between the two distribution curves.

 
Consider the theoretical construct of overlapping TSH distributions representing the two populations of individuals with euthyroidism and mild thyroid failure (Fig. 1Go). Based on static T4, T3, and TSH measurements, it would be impossible to correctly distinguish any subjects falling within the shaded overlapping region (TSH 5–10 mU/liter) of Fig. 1Go as being derived from either the right-sided tail of the euthyroid population (euthyroid outliers) or from the left-sided tail of the population with thyroid failure (early thyroid dysfunction). In addressing such diagnostic dilemmas, one study suggests using subject-based rather than group-based reference values (28); another study proposes a bivariate representation of thyroid function as more accurate than the currently accepted univariate method for characterization of pituitary-thyroid axis integrity (29).

Some investigators have used dynamic TRH testing in attempts to distinguish early thyroid failure from normal thyroid function. Although variable results have been reported, several studies note that a substantial 26–66% of subjects with TSH values from 5–10 mU/liter manifest a normal, nonexaggerated TSH response to TRH (30, 31, 32, 33). In a comprehensive systematic study of TRH stimulation tests involving 1061 individuals with a broad range of baseline TSH values, Spencer et al. (31) showed that ~43% of subjects with TSH values of 4.5–10 mU/liter manifested normal responses to TRH. They also noted that ~26% of the subjects with TSH levels in the upper normal range (2.0–4.5 mU/liter) exhibited exaggerated responses to TRH (31). Thus, it is not at all clear how accurately the TRH stimulation test may differentiate euthyroid subjects from those with mild thyroid failure when the TSH value lies in the range of 2–10 mU/liter. Another study showed that 46% of 30 subjects with mild SH and an exaggerated TSH response to TRH did not progress to overt hypothyroidism after follow-up to 15 yr, suggesting that abnormal TRH-stimulation does not always predict eventual thyroid gland failure (34). Repeat testing of TSH levels in SH subjects demonstrated remission to normal in ~15% of individuals entering two different intervention trials (18, 19) and showed occurrences of intermittent normal TSH values in 42% of placebo-treated subjects in another study (20). Reports from iodine-rich regions indicate that some subjects with mild TSH elevations achieve remission after institution of a low-iodine diet (7, 8). Thus, subjects designated as having SH with mildly elevated TSH levels may show a negative response to TRH, a positive response to TRH with no progression to overt hypothyroidism, or complete remission of TSH to normal. These and other lines of evidence suggest that a substantial proportion of subjects with mild SH may consist of euthyroid outliers and that TRH stimulation may not definitively distinguish such subjects from those with mild thyroid failure.

The epidemiological data in Table 2Go indicates that a preponderance of SH subjects (72% by meta-analysis) in community studies are affected by mild TSH elevations (<10 mU/liter), or stage b SH. Modestly elevated TSH levels may represent the long right tail of the euthyroid distribution, whereas dramatically elevated TSH values would more likely portend thyroid gland failure. Do thyroid autoantibodies serve as useful discriminants in this setting? The meta-analysis of the studies in Table 2Go determines an overall community prevalence of 6.3% for elevated TSH or SH and that 49% of such individuals are negative for thyroid antibodies. Large-scale community studies support the use of high thyroid autoantibody titers as predictors for thyroid failure, and the view that subjects with seronegative SH are at lower risk for eventual overt hypothyroidism (6, 9). Thus, if 2.5% of the 6.3% prevalence figure (or 40% of SH cases) were due to euthyroid outliers, this proportion resembles the 49% seronegativity rate in SH subjects!

Of note, epidemiological studies report that approximately 17–50% of stage b SH subjects are thyroid antibody seropositive (1, 6, 7, 14). This is intermediate to reported rates of 10% in euthyroid subjects and to rates of 80% in stage c patients (1, 6). Expressed alternatively, 80–95% of all seronegative individuals with elevated TSH are stage b (1, 6, 7, 14). These calculations further support the notion that a number of subjects with SH, especially those seronegative in stage b, may represent euthyroid outliers who would not progress to thyroid failure and would not need L-thyroxine.

Do symptoms improve after treatment of SH?

Five placebo-controlled blinded studies have assessed the ability of L-thyroxine therapy to improve psychometric, quality-of-life, or hypothyroid symptoms associated with SH (Table 3Go) (18, 19, 20, 21, 22). Of these five trials, two showed statistically significant improvement (18, 22), but the average TSH level exceeded 11 mU/liter in both, indicating large numbers of stage c patients. One study indicated marginal, yet convincing, benefit, with a response rate difference of 24% between placebo and T4-treated subjects; however, no baseline characteristics predicted who would benefit from L-thyroxine (19). The remaining two studies showed no clear benefit of L-thyroxine (20, 21); one noted some improvement of psychometric memory scores with L-thyroxine compared with placebo (20), but the other, which was the only study to enroll solely stage b SH subjects noted surprising results. Significantly fewer hypothyroid symptoms were found in both placebo and T4-treated groups, but significantly improved quality-of-life measures were observed only in the placebo-treated group (21). Furthermore, stage b SH subjects were unable to predict whether they were taking placebo or L-thyroxine (21). Taken in concert, the conflicting results from these studies suggest that 1) stage b SH subjects do not benefit more from L-thyroxine than placebo, 2) only one fourth of patients with more severe SH may benefit from L-thyroxine, and 3) indiscriminate L-thyroxine treatment of SH would only infrequently reverse symptoms associated with potential thyroid deficiency, given the large proportion of community SH subjects who are classified as stage b. Moreover, the possibility exists that L-thyroxine may exert effects beyond that of placebo to improve the well-being of euthyroid outliers in these trials; consider the study that demonstrated overtly hypothyroid patients subjectively favoring the use of L-thyroxine doses that caused suppression of, rather than normalization of, their TSH levels (35).


View this table:
[in this window]
[in a new window]
 
Table 3. Treatment of symptoms in SH

 
The relationship between SH, hyperlipidemia, and atherosclerosis

Although the view that overt hypothyroidism causes secondary hyperlipidemia and promotes atherosclerosis has been generally accepted (36), studies examining the relationships between hyperlipidemia, atherosclerosis, and SH have yielded less convincing results. Some cross-sectional observational surveys note a higher prevalence of SH in hyperlipidemic patients, whereas others show that SH subjects manifest moderately (mostly up to 10%) higher average total cholesterol (TC) than controls (12, 14, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48). Such evidence is far from decisive, because more than half of these noninterventional cross-sectional studies report either nonstatistically significant differences between euthyroid and SH subjects (14, 37, 40, 41, 42, 43, 47), or lower TC in SH subjects (13). Two striking themes emerge from these observational surveys: 1) stage b SH subjects manifest much milder degrees of dyslipidemia than those with stage c (12, 14, 39, 41, 43, 47, 48), with only two (12, 39) of these seven reports showing significantly elevated TC or LDL in stage b SH compared with euthyroid individuals; 2) all three studies that separately considered men with SH reported an elevation of 1% or less of TC in stage b SH compared with euthyroid controls (12, 37, 43). This result, in view of the much lower prevalence of SH in men, suggests that many of these subjects are euthyroid outliers.

Consider the findings of the large-scale epidemiological surveys with respect to thyroid dysfunction and lipid levels. The initial Whickham study observed that lipid levels were not associated with TSH elevations after age adjustment (37). The follow-up Whickham survey found no association between elevated serum TSH and increased risk of ischemic heart disease or dyslipidemia (23). A large cross-sectional survey of 3410 elderly subjects in Maryland noted a significantly elevated LDL cholesterol in subjects with SH, but only for those with TSH higher than 10 mU/liter (48); no increased frequency of diagnosed atherosclerotic diseases was found in this entire cohort of SH subjects. A report from Rotterdam noted that SH subjects actually had lower TC than controls, but manifested increased atherosclerotic vascular disease (13), suggesting that factors other than TC contribute to the increased risk of atherosclerosis, although LDL was not explicitly determined. From the conflicting data in the randomized community surveys of Whickham (23, 37), Maryland (48), and Rotterdam (13), one may question 1) whether SH is associated with dyslipidemia, 2) whether treatment of SH may reverse dyslipidemia, and 3) whether L-thyroxine therapy may prevent development of atherosclerosis if dyslipidemia is not the predominant mechanism causing presumed atherosclerotic disease in SH.

More than 20 intervention studies have assessed the effect of L-thyroxine therapy in subjects with SH (Table 4Go) (18, 19, 20, 21, 22, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63) to address the questions posed above. Unfortunately, the vast majority of these trials were not controlled and many involved patients with iatrogenic SH (overtreated overt hyperthyroidism or undertreated overt hypothyroidism), which is not an appropriate comparison group for community SH subjects, 90% of whom have no history of thyroid disease or thyroid medication use (12).


View this table:
[in this window]
[in a new window]
 
Table 4. Response of lipid levels to L-thyroxine therapy in SH

 
Despite the statistically significant improvement in serum lipid levels reported by half of these intervention trials (22, 53, 54, 55, 56, 58, 59, 60, 62, 63), the only four studies to stratify TSH levels unanimously noted nonsignificant effects in those with TSH less than 10 mU/liter (21, 22, 58, 60). Of the studies that did not stratify by TSH levels, the average TSH value exceeded 10 mU/liter in 10 reports, suggesting that they recruited patients with markedly abnormal pituitary-thyroid axes, who are not representative of the majority of those with SH. A closer examination of Table 4Go also reveals that although most published studies demonstrate modest decrements in TC and LDL levels after L-thyroxine therapy, four studies reported a dramatic decline in high-density lipoprotein (HDL) level to the extent that the TC/HDL, an indicator of cardiovascular risk, actually worsened after treatment (50, 52, 57, 58)!

What can we conclude from these small, variably defined, predominantly nonblinded, nonplacebo-controlled studies that yield conflicting results? The observation that individuals with stage b SH do not respond significantly to L-thyroxine therapy suggests that the majority of SH subjects ought not to be treated on the grounds of reversing dyslipidemia. Moreover, because the majority of subjects (>80%) in these trials were women, the evidence to use L-thyroxine therapy for hyperlipidemia in men with SH is tenuous.

Two meta-analyses that separately reviewed studies assessing the efficacy of L-thyroxine therapy on lipoproteins in SH subjects noted a mean decrease in serum TC of 0.4 mmol/liter and 0.2 mmol/liter, respectively (64, 65), although the former analysis included two studies (39, 43) that treated patients with overt hypothyroidism. The latter meta-analysis was able to attain adequate power to achieve overall statistical significance, but the stratification of study criteria found that two subsets of intervention trials did not achieve significance in the lipid response to L-thyroxine therapy: 1) studies that merited higher scores of internal validity; and 2) studies that enrolled subjects with baseline TC less than 240 mg/dl and who did not have inadequately treated overt hypothyroidism (65). These observations suggest that 1) more rigorously controlled intervention studies are needed; 2) community SH subjects, who have an overall mean TC of 224 mg/dl (12), and the vast majority of whom do not have undertreated overt hypothyroidism, would not benefit from L-thyroxine; and 3) patients with undertreated overt hypothyroidism manifest more severe metabolic dysfunction than community SH subjects and require separate therapeutic strategies.

Should treatment of SH be considered with the intent to normalize Lp(a) and homocysteine, two cardiovascular risk factors that increase in overt hypothyroidism, and both of which potentially may be invoked to explain the increased atherosclerosis seen in SH subjects in the Rotterdam study (13) who had lower TC than controls? Once again, current evidence does not favor the use of L-thyroxine in SH to treat either of these metabolic abnormalities. One study actually showed nonsignificantly increased Lp(a) levels in SH subjects after L-thyroxine treatment (59), whereas the study that solely enrolled stage b SH subjects noted no change (21). The only study that has reported a significant decrement in Lp(a) after L-thyroxine treatment of SH involved patients with an average baseline TSH of 22 mU/liter (62). There is even conflict about the ability of L-thyroxine treatment to change Lp(a) levels in patients with overt hypothyroidism (66). With respect to normalizing homocysteine levels, the only two published studies apply to L-thyroxine treatment of overt hypothyroidism (not SH) and are dichotomous in conclusions (67, 68). Although both noted significant lowering of fasting homocysteine after restoration of euthyroidism, the larger study with a control group reported that L-thyroxine alone was unable to lower postmethionine homocysteine levels, concluding that the hyperhomocysteinemia in overt hypothyroidism could not be fully reversed by L-thyroxine therapy (68).

Treatment of cardiovascular, neuromuscular, and neuropsychiatric dysfunction in SH

Features of neuromuscular dysfunction attributable to SH include myocardial dyscontractility, sensory and motor neurological impairment, and skeletal muscle dysfunction (15, 17). However, the only cross-sectional study that stratified individuals by TSH levels noted significantly altered ankle reflex times and myoglobin levels only in those patients with TSH higher than 12 mU/liter (41). Some, certainly not all, studies have determined statistically significant alterations in subjects with SH involving such indices of questionable biological significance as the stapedial reflex (69) and systolic time interval (STI) (70). Only two placebo-controlled studies have reported on L-thyroxine intervention to improve neuromuscular dysfunction, and they were in conflict about the presence of significant beneficial changes in STI after L-thyroxine (18, 19). A few noncontrolled studies showed significant improvement in neuromuscular and cardiovascular parameters after L-thyroxine treatment of SH (51, 52, 71, 72). Three of these studies enrolled patients with mean baseline TSH levels higher than 15 mU/liter (51, 52, 71). Interestingly, another study that administered triiodothyronine to euthyroid controls and SH subjects noted improved STI in both groups (73); yet another study showed significantly improved left ventricular systolic and diastolic function in T4-treated SH subjects, with some posttreatment parameters exceeding those of euthyroid controls (74). These latter results raise again the possibility that L-thyroxine may beneficially affect even euthyroid controls and emphasize the absolute need for using placebo controls and blinding participants in any studies that assess the efficacy of L-thyroxine therapy for SH. Furthermore, several other studies (not mentioned here) have correlated the diagnosis of SH with various unique measures of physiological abnormalities, but the reversal of such abnormalities with L-thyroxine therapy have not yet been demonstrated.

An abundant psychiatric literature has attempted to establish a relationship between subclinical thyroid dysfunction and affective as well as psychotic disorders (75, 76). Many such reports are hindered by not controlling for the effects of age, gender, inpatient hospitalization, and use of lithium on the prevalence of elevated TSH levels (75). Additionally, the mere association between SH and psychiatric disorders should not lead to the conclusion that SH brings about these associated disorders (76). The three large epidemiological studies that examined cognitive and affective scores in SH subjects found few (47, 77) or no (78) significant differences overall, and none in those with TSH between 5 and 10 mU/liter (47, 77). Two placebo-controlled studies assessing the effects of L-thyroxine therapy on psychometric measures noted that less than 25% of subjects responded in one study (19), and that statistically significant improvement occurred in only one parameter in the other study (20). Two nonplacebo-controlled studies noted significant reversal of only a subset of neurobehavioral and cognitive baseline abnormalities in SH subjects receiving L-thyroxine (79, 80). Given the paucity of data regarding the potential benefit of L-thyroxine for SH subjects with psychiatric disturbances, treatment is empirical and must be weighed against the risks of lifelong therapy.

Treatment of SH to prevent progression

Proponents of the view that most SH subjects should be treated with L-thyroxine advocate early initiation of therapy to prevent the future morbidity associated with development of overt hypothyroidism. The Whickham survey estimated that the progression of SH to overt hypothyroidism occurs in 2–5% of SH subjects per year, an increased risk affecting those with thyroid autoantibodies (9). Although the progression rate to overt thyroid failure in SH is significantly higher than in euthyroid controls, it is nonetheless necessarily modest in scale, given that the prevalence of "preclinical" thyroid autoimmunity as well as SH are each more than 5–10 times the magnitude of overt hypothyroidism. In addition, the clinical course of autoimmune thyroiditis (the most common single cause of SH) is marked by shifting biochemical parameters, alternating serological positivity, and fluctuating L-thyroxine dosages needed to maintain euthyroidism.

One study on the undulating course of Hashimoto’s disease determined that 20% of patients diagnosed with overt hypothyroidism regressed to euthyroidism after termination of T4 therapy (81). Moreover, the treatment of SH subjects would require similar, if not more intensive, biochemical and clinical monitoring than would be needed for those subjects being expectantly followed; no study has even demonstrated that early L-thyroxine intervention improves the outcomes of subjects with SH. As will be discussed in the following section, the overzealous dosing of L-thyroxine therapy with insufficient biochemical monitoring puts a substantial number of patients at risk for iatrogenic disease associated with hyperthyroidism. Once again, the benefit to risk ratio of initiating lifelong L-thyroxine therapy in the typical asymptomatic individual with stage b SH, who has minimal evidence of end-organ dysfunction, must be seriously weighed against the real morbidity associated with suppression of TSH.

Danger of treatment: overreplacement L-thyroxine therapy in hypothyroidism

Abnormal TSH values are frequently found in patients treated with L-thyroxine. Between 10% and 33% of individuals on L-thyroxine therapy have TSH values less than normal (12, 82, 83, 84, 85, 86, 87), and approximately one third to one half of these TSH levels are less than 0.1 mU/liter (83, 87). Given that most patients on L-thyroxine therapy are older individuals, the prescription of thyroid hormone for poorly established indications compounds an increasing problem of polypharmacy and drug interactions in the elderly.

Does suppression of TSH by exogenous L-thyroxine lead to a decrease in bone mineral density in postmenopausal women, the subjects who are at greatest risk for both SH and osteoporosis? An extensive recent review of this subject including a number of cross-sectional and longitudinal studies was inconclusive, with 10 studies showing a significant negative effect of L-thyroxine on skeletal integrity and a nearly equal number of studies showing no effect (88). Data from the Framingham study have shown that elderly persons 60 yr of age or older with a low TSH level have a 3-fold higher risk of developing atrial fibrillation over a 10-yr period compared with those with a normal TSH level, although only 5.7% of the subjects were taking a thyroid hormone preparation at entry (89). Left ventricular hypertrophy may be another adverse consequence of a suppressed TSH (90). Three blinded placebo-controlled studies of SH also noted overt manifestations of L-thyroxine toxicity: anxiety and tachycardia in 2 of 20 subjects (19), angina and atrial fibrillation in 2 of 18 participants (20) (all four individuals needed to be removed from these trials), and significantly increased frequency of anxiety in the L-thyroxine intervention vs. the placebo group (21).

Of additional great concern is the monitoring and management by community physicians of patients on L-thyroxine therapy. In a broad community survey of 25,862 Colorado residents, Canaris et al. (12) noted that 92% of subjects had seen a physician in the preceding year, yet 40% of the patients on L-thyroxine therapy had an abnormal TSH level, including 22% with a low TSH. It is likely that the majority of such patients were being poorly managed; two studies noted that overreplacement of primary hypothyroidism was the most common cause of a suppressed TSH in patients on L-T4 who did not have pituitary disease (91, 92). In another report, only 44% of T4-treated patients with an elevated TSH had dosages increased by their physicians who were aware of the TSH findings; even more astounding was the fact that only 11% of the patients with low TSH levels on L-thyroxine replacement had their doses decreased (86)! In view of the studies documenting the unacceptably high prevalence of low TSH levels in T4-treated patients who are managed by community physicians, it would be irrational to advocate lifelong L-thyroxine therapy for SH subjects unless the benefits are compelling and clearly outweigh the risks.

Conclusions

After reviewing a considerable spectrum of primary data, we conclude that subjects with SH should seldomly be treated with L-thyroxine. The current univariate method of measuring thyroid function is inadequate for distinguishing euthyroid outliers from subjects with mild incipient thyroid failure. Epidemiological reports observe that subjects with TSH lower than 10 mU/liter comprise a clear majority of those with SH. Observational reports suggest that a number of these stage b subjects, especially men, may actually be euthyroid outliers. Cross-sectional and interventional studies demonstrate that stage b subjects have minimal metabolic and physiological abnormalities and would be unlikely to benefit from L-thyroxine therapy. The vast majority of such studies, unfortunately, are not appropriately designed to assess the benefit of L-thyroxine treatment in SH, because they include large numbers of patients with iatrogenic SH, large numbers of patients with TSH higher than 10 mU/liter, few numbers of men, and with very few exceptions do not have placebo arms. A number of studies do illustrate interesting physiological abnormalities associated with SH, but the reversibility of such dysfunction with L-thyroxine therapy remains to be proven. The benefit to risk ratio of treating subjects with SH also remains to be determined, given the lack of outcome data, and the considerable risk of TSH suppression in patients on L-thyroxine replacement. If it is true that most subjects with SH should not be treated, then the recommendations advocating systematic screening of individuals for thyroid disease, especially in men, can also be justifiably challenged (93, 94).

With these caveats noted, we would generally limit use of L-thyroxine treatment to subjects with SH who possess a combination of the following compelling factors: TSH levels more than 10 mU/liter on repeated measurements, clear symptoms or signs (e.g. goiter) associated with thyroid failure, convincing family history of thyroid disease, state of pregnancy, strong habit of tobacco use (95), or severe hyperlipidemia not previously diagnosed. We do recognize that there exist other clinical scenarios (15), much less common than the ones noted above, and not within the scope of this discussion, in which the treatment of SH with L-thyroxine may be prudent; we leave open these possibilities in order that clinicians should consider treatment of SH on a case-by-case basis. In general, we would advocate obtaining FT4 and confirmatory TSH levels in any subject who is newly diagnosed with SH, and would recommend monitoring individuals expectantly if none of the compelling reasons for treatment are applicable.

Our position of seldomly treating SH is supported by a United Kingdom consensus panel, who advocates routine treatment of SH when the TSH exceeds 10 mU/liter (96), which is the only consistent predictor of response to L-thyroxine therapy. We propose that future studies be conducted to separately consider male subjects and to better assess the individuals who comprise the majority of cases with SH (i.e. those found on population screening to have TSH <10 mU/liter and without history of overt thyroid dysfunction). We hope that large-scale controlled intervention and outcome studies will be done to assess the potential benefit of L-thyroxine treatment in such SH subjects. We expect also that future recommendations regarding screening for thyroid disease in the community will be influenced by the results of such studies.

Acknowledgments

Received October 16, 2000. Accepted February 9, 2001.

Address all correspondence and requests for reprints to: Lawrence M. Crapo, M.D., Ph.D., Division of Endocrinology, Santa Clara Valley Medical Center, 751 South Bascom Avenue, San Jose, California 95128.

Footnotes

Abbreviations: FT4, Free T4; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SH, subclinical hypothyroidism; STI, systolic time interval; TC, total cholesterol.

References

  1. Tunbridge WMG, Evered DC, Hall R, et al. 1977 The spectrum of thyroid disease in a community: the Whickham survey. Clin Endocrinol 7:481–493[Medline]
  2. Nystrom E, Bengtsson C, Lindquist O, Noppa H, Lindstedt G, Lundberg PA 1981 Thyroid disease and high concentration of serum thyrotropin in a population sample of women. Acta Med Scand 210:39–46[Medline]
  3. Falkenberg M, Kagedal B, Norr A 1983 Screening of an elderly female population for hypo- and hyperthyroidism by use of a thyroid hormone panel. Acta Med Scand 214:361–365[Medline]
  4. Sawin CT, Castelli WP, Hershman JM, McNamara P, Bacharach P 1985 The aging thyroid. Thyroid deficiency in the Framingham Study. Arch Intern Med 145:1386–1388[Abstract]
  5. Okamura K, Nakashima T, Ueda K, Inoue K, Omae T, Fujishima M 1987 Thyroid disorders in the general population of Hisayama Japan, with special reference to prevalence and sex differences. Int J Epidemiol 16:545–549[Abstract]
  6. Parle JV, Franklyn JA, Cross KW, Jones SC, Sheppard MC 1991 Prevalence and follow-up of abnormal thyrotropin (TSH) concentrations in the elderly in the United Kingdom. Clin Endocrinol 34:77–83[Medline]
  7. Konno N, Yuri K, Taguchi H, et al. 1993 Screening for thyroid diseases in an iodine sufficient area with sensitive thyrotropin assays, and serum thyroid autoantibody and urinary iodide determinations. Clin Endocrinol 38:273–281[Medline]
  8. Konno N, Makita H, Yuri K, Iizuka N, Kawasaki K 1994 Association between dietary iodine intake and prevalence of subclinical hypothyroidism in the central regions of Japan. J Clin Endocrinol Metab 78:393–397[Abstract]
  9. Vanderpump MPJ, Tunbridge WMG, French JM, et al. 1995 The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol 43:55–68[Medline]
  10. Danese D, Arduino G, Andreoli M, Pontecorvi A 1997 Screening for subclinical hypothyroidism in a flight personnel population. Thyroid 7:S-61
  11. Knudsen N, Jorgensen T, Rasmussen S, Christiansen E, Perrild H 1999 The prevalence of thyroid dysfunction in a population with borderline iodine deficiency. Clin Endocrinol 51:361–367[CrossRef][Medline]
  12. Canaris GJ, Manowitz NR, Mayor G, Ridgway EC 2000 The Colorado thyroid disease prevalence study. Arch Intern Med 160:526–534[Abstract/Free Full Text]
  13. Hak AE, Pols HAP, Visser TJ, Drexhage HA, Hofman A, Witteman JCM 2000 Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women: the Rotterdam Study. Ann Intern Med 132:270–278[Abstract/Free Full Text]
  14. Pirich C, Mullner M, Sinzinger H 2000 Prevalence and relevance of thyroid dysfunction in 1922 cholesterol screening participants. J Clin Epidemiol 53:623–629[CrossRef][Medline]
  15. Arem R, Escalante D 1996 Subclinical hypothyroidism: epidemiology, diagnosis, and significance. Adv Intern Med 41:213–250[Medline]
  16. Cooper DS 1998 Subclinical thyroid disease: a clinician’s perspective. Ann Intern Med 129:135–138[Free Full Text]
  17. Ayala AR, Danese MD, Ladenson PW 2000 When to treat mild hypothyroidism. Endocrinol Metab Clin North Am 29:399–415[Medline]
  18. Cooper DS, Halpern R, Wood LC, Levin AA, Ridgway EC 1984 L-thyroxine therapy in subclinical hypothyroidism. A double-blind, placebo-controlled trial. Ann Intern Med 101:18–24[Medline]
  19. Nystrom E, Caidahl K, Fager G, Wikkelso C, Lundberg PA, Lindstedt G 1988 A double-blind cross-over 12-month study of L-thyroxine treatment of women with "subclinical" hypothyroidism. Clin Endocrinol 29:63–75[Medline]
  20. Jaeschke R, Guyatt G, Gerstein H, et al. 1996 Does treatment with L-thyroxine influence health status in middle-aged and older adults with subclinical hypothyroidism? J Gen Intern Med 11:744–749[Medline]
  21. Kong WM, Sheikh M, Lumb P, Naoumova RP, Freedman DB, Finer N A randomized controlled trial of thyroxine treatment in mild subclinical hypothyroidism. Proceedings of the 82nd Annual Meeting of The Endocrine Society, Toronto, Canada, 2000; 2466:597
  22. Meier C, Roth CB, Huber G, Guglielmetti M, Huber P, Staub JJ Clinical and metabolic effects of thyroxine replacement in patients with mild thyroid failure: results from a double-blind placebo-controlled study. Proceedings of the 82nd Annual Meeting of The Endocrine Society, Toronto, Canada, 2000; 2372:573
  23. Vanderpump MPJ, Tunbridge WMG, French JM, et al. 1996 The development of ischemic heart disease in relation to autoimmune thyroid disease in a 20-year follow-up study of an English community. Thyroid 6:155–160[Medline]
  24. Evered DC, Ormston BJ, Smith PA, Hall R, Bird T 1973 Grades of hypothyroidism. Br Med J 1:657–662[Medline]
  25. Tsay JY, Chen IW, Maxon HR, Heminger L 1979 A statistical method for determining normal ranges from laboratory data including values below the minimum detectable value. Clin Chem 25:2011–2014[Abstract/Free Full Text]
  26. Szabolcs I, Ploenes C, Beyer M, Bernard W, Herrmann J 1991 Reference intervals for serum thyrotropin: dependence on the population investigated. Exp Clin Endocrinol 98:23–31[Medline]
  27. Sawin CT 1991 Thyroid dysfunction in older persons. Adv Intern Med 37:223–248
  28. Nagayama I, Yamamoto K, Saito K, Kuzuya T, Saito T 1993 Subject-based reference values in thyroid function tests. Endocr J 40:557–562[Medline]
  29. Meier CA, Maisey MN, Lowry A, Muller J, Smith MA 1993 Interindividual differences in the pituitary-thyroid axis influence the interpretation of thyroid function tests. Clin Endocrinol 39:101–107[Medline]
  30. Brown CA, Hennessey JV 1989 Clinical significance of mildly elevated thyrotropin levels with normal thyroxine levels. South Med J 82:681–685[Medline]
  31. Spencer CA, Schwarzbein D, Guttler RB, LoPresti JS, Nicoloff JT 1993 Thyrotropin (TSH)-releasing hormone stimulation test responses employing third and fourth generation TSH assays. J Clin Endocrinol Metab 76:494–498[Abstract]
  32. Mojiminiyi OA, Rege V, Bolodeoku J, Wilcox AH, Barron JL 1995 Thyroid autoantibodies and the response to thyrotropin releasing hormone in patients with subclinical hypothyroidism. J Clin Pathol 48:463–465[Abstract]
  33. Faughnan M, Lepage R, Fugere P, Bissonnette F, Brossard JH, D’Amour P 1995 Screening for thyroid disease at the menopausal clinic. Clin Invest Med 18:11–18[Medline]
  34. Kabadi UM 1993 Subclinical hypothyroidism. Natural course of the syndrome during a prolonged follow-up study. Arch Intern Med 153:957–961[Abstract]
  35. Carr D, McLeod DT, Parry G, Thornes HM 1988 Fine adjustment of thyroxine replacement dosage: comparison of the thyrotropin releasing hormone test using a sensitive thyrotropin assay with measurement of free thyroid hormones and clinical assessment. Clin Endocrinol 28:325–333[Medline]
  36. Becker C 1985 Hypothyroidism and atherosclerotic heart disease: pathogenesis, medical management, and the role of coronary artery bypass surgery. Endocr Rev 6:432–440[Medline]
  37. Tunbridge WMG, Evered DC, Hall R, et al. 1977 Lipid profiles and cardiovascular disease in the Whickham area with particular reference to thyroid failure. Clin Endocrinol 7:495–508[Medline]
  38. Althaus BU, Staub JJ, Ryff-De Leche A, Oberhansli A, Stahelin HB 1988 LDL/HDL-changes in subclinical hypothyroidism: possible risk factors for coronary heart disease. Clin Endocrinol 28:157–163[Medline]
  39. Elder J, McLelland A, O’Reilly DS, Packard CJ, Series JJ, Shepherd J 1990 The relationship between serum cholesterol and serum thyrotropin, thyroxine and tri-iodothyronine concentrations in suspected hypothyroidism. Ann Clin Biochem 27:110–113[Medline]
  40. Parle JV, Franklyn JA, Cross KW, Jones SR, Sheppard MC 1992 Circulating lipids and minor abnormalities of thyroid function. Clin Endocrinol 37:411–414[Medline]
  41. Staub JJ, Althaus BU, Engler H, et al. 1992 Spectrum of subclinical and overt hypothyroidism: effect on thyrotropin, prolactin, and thyroid reserve, and metabolic impact on peripheral target tissues. Am J Med 92:631–642[Medline]
  42. Geul KW, van Sluisveld IL, Grobbee DE, et al. 1993 The importance of thyroid microsomal antibodies in the development of elevated serum TSH in middle-aged women: associations with serum lipids. Clin Endocrinol 39:275–280[Medline]
  43. Johnston J, McLelland A, O’Reilly DS 1993 The relationship between serum cholesterol and serum thyroid hormones in male patients with suspected hypothyroidism. Ann Clin Biochem 30:256–259[Medline]
  44. Valdemarsson S, Hansson P, Hedner P, Nilsson-Ehle P 1983 Relations between thyroid function, hepatic and lipoprotein lipase activities, and plasma lipoprotein concentrations. Acta. Endocrinol 104:50–56[Abstract]
  45. Kung AW, Pang RW, Janus ED 1995 Elevated serum lipoprotein(a) in subclinical hypothyroidism. Clin Endocrinol 43:445–449[Medline]
  46. Bauer DC, Ettinger B, Browner WS 1998 Thyroid function and serum lipids in older women: a population-based study. Am J Med 104:546–551[CrossRef][Medline]
  47. Lindeman RD, Schade DS, LaRue A, et al. 1999 Subclinical hypothyroidism in a biethnic, urban community. J Am Geriatr Soc 47:703–709[Medline]
  48. Ladenson PW, Wilson MC, Gardin J, et al. 1994 Relationship of subclinical hypothyroidism to cardiovascular risk factors and disease in an elderly population. Thyroid 4:S18
  49. Nilsson G, Nordlander S, Levin K 1976 Studies on subclinical hypothyroidism with special reference to the serum lipid pattern. Acta Med Scand 200:63–67[Medline]
  50. Lithell H, Boberg J, Hellsing K, et al. 1981 Serum lipoprotein and apolipoprotein concentrations and tissue lipoprotein-lipase activity in overt and subclinical hypothyroidism: the effect of substitution therapy. Eur J Clin Invest 11:3–10[Medline]
  51. Ridgway EC, Cooper DS, Walker H, Rodbard D, Maloof F 1981 Peripheral responses to thyroid hormone before and after L-thyroxine therapy in patients with subclinical hypothyroidism. J Clin Endocrinol Metab 53:1238–1242[Abstract]
  52. Bell GM, Todd WT, Forfar JC, et al. 1985 End-organ responses to thyroxine therapy in subclinical hypothyroidism. Clin Endocrinol 22:83–89[Medline]
  53. Powell JT, Wiseman SA, Carter G, Alaghband Zadeh J, Fowler PB, Greenhalgh RM 1989 Can thyroxine half the progression of peripheral arterial disease? Eur J Vasc Surg 3:85–87[Medline]
  54. Arem R, Patsch W 1990 Lipoprotein and apolipoprotein levels in subclinical hypothyroidism. Effect of levothyroxine therapy. Arch Intern Med 150:2097–2100[Abstract]
  55. Caron P, Calazel C, Parra HJ, Hoff M, Louvet JP 1990 Decreased HDL cholesterol in subclinical hypothyroidism: the effect of L-thyroxine therapy. Clin Endocrinol 33:519–523[Medline]
  56. Bogner U, Arntz HR, Peters H, Schleusener H 1993 Subclinical hypothyroidism and hyperlipoproteinaemia: indiscriminate L-thyroxine treatment not justified. Acta Endocrinol 128:202–206[Medline]
  57. Franklyn JA, Daykin J, Betteridge J, et al. 1993 Thyroxine replacement therapy and circulating lipid concentrations. Clin Endocrinol 38:453–459[Medline]
  58. Miura S, Iitaka M, Yoshimura H, et al. 1994 Disturbed lipid metabolism in patients with subclinical hypothyroidism: effect of L-thyroxine therapy. Intern Med 33:413–417[Medline]
  59. Arem R, Escalante DA, Arem N, Morrisett JD, Patsch W 1995 Effect of L-thyroxine therapy on lipoprotein fractions in overt and subclinical hypothyroidism, with special reference to lipoprotein(a). Metabolism 44:1559–1563[Medline]
  60. Diekman T, Lansberg PJ, Kastelein JJ, Wiersinga WM 1995 Prevalence and correction of hypothyroidism in a large cohort of patients referred for dyslipidemia. Arch Intern Med 155:1490–1495[Abstract]
  61. Romaldini JH, Biancalana MM, Figueiredo DI, Farah CS, Mathias PC 1996 Effect of L-thyroxine administration on antithyroid antibody levels, lipid profile, and thyroid volume in patients with Hashimoto’s thyroiditis. Thyroid 6:183–188[Medline]
  62. Yildirimkaya M, Ozata M, Yilmaz K, Kilinc C, Gundogan MA, Kutluay T 1996 Lipoprotein(a) concentration in subclinical hypothyroidism before and after levo-thyroxine therapy. Endocr J 43:731–736[Medline]
  63. Paoli M, Bellabarba G, Velazquez E, et al. 1998 Sex steroids, lipids, and lipoprotein cholesterols in women with subclinical and overt hypothyroidism before and after L-thyroxine therapy. Clin Chim Acta 275:81–91[CrossRef][Medline]
  64. Tanis BC, Westendorp RGJ, Smelt AHM 1996 Effect of thyroid substitution on hypercholesterolaemia in patients with subclinical hypothyroidism: a reanalysis of intervention studies. Clin Endocrinol 44:643–649[Medline]
  65. Danese MD, Ladenson PW, Meinert CL, Powe NR 2000 Effect of thyroxine therapy on serum lipoproteins in patients with mild thyroid failure: a quantitative review of the literature. J Clin Endocrinol Metab 85:2993–3001[Abstract/Free Full Text]
  66. Pazos F, Alvarez JJ, Rubies-Prat J, Varela C, Lasuncion MA 1995 Long term thyroid replacement therapy and levels of lipoprotein(a) and other lipoproteins. J Clin Endocrinol Metab 80:562–566[Abstract]
  67. Hussein WI, Green R, Jacobsen DW, Faiman C 1999 Normalization of hyperhomocysteinemia with L-thyroxine in hypothyroidism. Ann Intern Med 131:348–351[Abstract/Free Full Text]
  68. Catargi B, Parrot-Roulaud F, Cochet C, Ducassou D, Roger P, Tabarin A 1999 Homocysteine, hypothyroidism, and effect of thyroid hormone replacement. Thyroid 9:1163–1166[Medline]
  69. Goulis DG, Tsimpiris N, Delaroudis S, et al. 1998 Stapedial reflex: a biological index found to be abnormal in clinical and subclinical hypothyroidism. Thyroid 8:583–587[Medline]
  70. Tseng KH, Walfish PG, Persaud JA, Gilbert BW 1989 Concurrent aortic and mitral valve echocardiography permits measurement of systolic time intervals as an index of peripheral tissue thyroid functional status. J Clin Endocrinol Metab 69:633–638[Abstract]
  71. Arem R, Rokey R, Kiefe C, Escalante DA, Rodriguez A 1996 Cardiac systolic and diastolic function at rest and exercise in subclinical hypothyroidism: effect of thyroid hormone therapy. Thyroid 6:397–402[Medline]
  72. Monzani F, Caraccio N, Del Guerra P, Casolaro A, Ferrannini E 1999 Neuromuscular symptoms and dysfunction in subclinical hypothyroid patients: beneficial effect of L-T4 replacement therapy. Clin Endocrinol 51:237–242[CrossRef][Medline]
  73. Ooi TC, Whitlock RM, Frengley PA, Ibbertson HK 1980 Systolic time intervals and ankle reflex time in patients with minimal serum TSH elevation: response to triiodothyronine therapy. Clin Endocrinol 13:621–627[Medline]
  74. Biondi B, Fazio S, Palmieri EA, et al. 1999 Left ventricular diastolic dysfunction in patients with subclinical hypothyroidism. J Clin Endocrinol Metab 84:2064–2067[Abstract/Free Full Text]
  75. Griffin JE 1990 Hypothyroidism in the elderly. Am J Med Sci 299:334–345[Medline]
  76. Haggerty Jr JJ, Garbutt JC, Evans DL, et al. 1990 Subclinical hypothyroidism: a review of neuropsychiatric aspects. Int J Psychiatry Med 20:193–208[Medline]
  77. Manciet G, Dartigues JF, Decamps A, et al. 1995 The PAQUID survey and correlates of subclinical hypothyroidism in elderly community residents in the southwest of France. Age Ageing 24:235–241[Abstract]
  78. Luboshitzky R, Oberman AS, Kaufman N, Reichman N, Flatau E 1996 Prevalence of cognitive dysfunction and hypothyroidism in an elderly community population. Isr J Med Sci 32:60–65[Medline]
  79. Monzani F, Del Guerra P, Caraccio N, et al. 1993 Subclinical hypothyroidism: neurobehavioral features and beneficial effect of L-thyroxine treatment. Clin Investig 71:367–71[Medline]
  80. Baldini IM, Vita A, Mauri MC, et al. 1997 Psychopathological and cognitive features in subclinical hypothyroidism. Prog Neuropsychopharmacol Biol Psychiatry 21:925–935[CrossRef][Medline]
  81. Takasu N, Komiya I, Asawa T, Nagasawa Y, Yamada T 1990 Test for recovery from hypothyroidism during thyroxine therapy in Hashimoto’s thyroiditis. Lancet 336:1084–1086[Medline]
  82. Ross DS, Daniels GH, Gouveia D 1990 The use and limitations of a chemiluminescent thyrotropin assay as a single thyroid function test in an out-patient endocrine clinic. J Clin Endocrinol Metab 71:764–769[Abstract]
  83. Parle JV, Franklyn JA, Cross KW, Jones SR, Sheppard MC 1993 Thyroxine prescription in the community: serum thyroid hormone stimulating level assays as an indicator of undertreatment or overtreatment. Br J Gen Pract 43:107–109[Medline]
  84. Bemben DA, Winn P, Hamm RM, Morgan L, Davis A, Barton E 1994 Thyroid disease in the elderly. Part 1. Prevalence of undiagnosed hypothyroidism. J Fam Pract 38:577–582[Medline]
  85. Nuovo J, Ellsworth A, Christensen DB, Reynolds R 1995 Excessive thyroid hormone replacement therapy. J Am Board Fam Pract 8:435–439[Medline]
  86. De Whalley P 1995 Do abnormal thyroid stimulating hormone level values result in treatment changes? A study of patients on thyroxine in one general practice. Br J Gen Pract 45:93–95[Medline]
  87. Watsky JG, Koeniger MA 1998 Prevalence of iatrogenic hyperthyroidism in a community hospital. J Am Board Fam Pract 11:175–179[Abstract]
  88. Greenspan SL, Greenspan FS 1999 The effect of thyroid hormone on skeletal integrity. Ann Intern Med 130:750–758[Abstract/Free Full Text]
  89. Sawin CT, Geller A, Wolf PA, et al. 1994 Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older persons. N Engl J Med 331:1249–1252[Abstract/Free Full Text]
  90. Biondi B, Fazio S, Carella C, et al. 1993 Cardiac effects of long term thyrotropin-suppressive therapy with levothyroxine. J Clin Endocrinol Metab 77:334–338[Abstract]
  91. Kasagi K, Takeuchi R, Misaki T, et al. 1997 Subclinical Graves’ disease as a cause of subnormal TSH levels in euthyroid subjects. J Endocrinol Invest 20:183–188[Medline]
  92. Figge J, Leinung M, Goodman AD, et al. 1994 The clinical evaluation of patients with subclinical hyperthyroidism and free triiodothyronine (free T3) toxicosis. Am J Med 96:229–234[Medline]
  93. Danese MD, Powe NR, Sawin CT, Ladenson PW 1996 Screening for mild thyroid failure at the periodic health examination: a decision and cost-effectiveness analysis. JAMA 276:285–292[Abstract]
  94. Ladenson PW, Singer PA, Ain KB, et al. 2000 American Thyroid Association guidelines for detection of thyroid dysfunction. Arch Intern Med 160:1573–1575[Abstract/Free Full Text]
  95. Muller B, Zulewski H, Huber P, Ratcliffe JG, Staub JJ 1995 Impaired action of thyroid hormone associated with smoking in women with hypothyroidism. N Engl J Med 333:964–969[Abstract/Free Full Text]
  96. Vanderpump MP, Ahlquist JA, Franklyn JA, Clayton RN 1996 Consensus statement for good practice and audit measures in the management of hypothyroidism and hyperthyroidism. The Research Unit of the Royal College of Physicians of London, the Endocrinology and Diabetes Committee of the Royal College of Physicians of London, and the Society for Endocrinology. BMJ 313:539–544[Free Full Text]