Treatment of primary hypothyroidism in adults

INTRODUCTION — In most patients, hypothyroidism is a permanent condition requiring lifelong treatment. Therapy consists of thyroid hormone replacement, unless the hypothyroidism is transient (as after painless thyroiditis or subacute thyroiditis) or reversible (due to a drug that can be discontinued). (See "Disorders that cause hypothyroidism".)

The goal of therapy is restoration of the euthyroid state, which can be readily accomplished in almost all patients by oral administration of synthetic thyroxine (T4, levothyroxine). Appropriate treatment should reverse all the clinical manifestations of hypothyroidism.

This topic will review the major issues that must be addressed in the treatment of adults with overt primary hypothyroidism. The treatment of subclinical and central hypothyroidism, as well as hypothyroidism during pregnancy and in children (congenital or acquired), is discussed separately.

●(See "Subclinical hypothyroidism in nonpregnant adults".)

●(See "Central hypothyroidism".)

●(See "Hypothyroidism during pregnancy: Clinical manifestations, diagnosis, and treatment".)

●(See "Treatment and prognosis of congenital hypothyroidism".)

●(See "Acquired hypothyroidism in childhood and adolescence".)

APPROACH TO TREATMENT — Our approach described below is largely consistent with the American Thyroid Association (ATA) Guidelines for the Treatment of Hypothyroidism [1].

Defining hypothyroidism

●Overt primary hypothyroidism is characterized biochemically by a high serum thyroid-stimulating hormone (TSH) concentration and a low serum free thyroxine (T4) concentration. The clinical manifestations are highly variable, depending upon the age at onset and the duration and severity of thyroid hormone deficiency. (See "Clinical manifestations of hypothyroidism" and "Diagnosis of and screening for hypothyroidism in nonpregnant adults".)

●Subclinical hypothyroidism is characterized biochemically by a high serum TSH concentration and a normal serum free T4 concentration. Most patients are asymptomatic. The diagnosis and management of subclinical hypothyroidism are reviewed separately. (See "Subclinical hypothyroidism in nonpregnant adults".)

●Central hypothyroidism caused by hypothalamic or pituitary disease is characterized by a low serum T4 concentration and a serum TSH concentration that is not appropriately elevated. TSH may be low, normal, or even slightly elevated (up to approximately 10 mU/L). The clinical manifestations of central hypothyroidism are similar to but sometimes milder than those of primary hypothyroidism. The diagnosis and management of central hypothyroidism are reviewed separately. (See "Central hypothyroidism".)

All patients with overt primary (or central) hypothyroidism require treatment (regardless of symptoms), unless the hypothyroidism is transient (as after painless thyroiditis or subacute thyroiditis) or reversible (due to a drug that can be discontinued). (See "Disorders that cause hypothyroidism", section on 'Transient hypothyroidism'.)

Thyroid hormone should not be prescribed to biochemically euthyroid individuals with nonspecific symptoms (eg, fatigue, weight gain, depression). Thyroid hormone has been administered to euthyroid patients with several clinical problems in whom it was hoped that outcome would be improved. These include T3 therapy in patients undergoing coronary artery bypass graft surgery [2] and in those with refractory depression in an attempt to enhance the response to antidepressant drug therapy. A meta-analysis on the efficacy of thyroid hormone therapy in depressed patients was equivocal [3]. Although there was evidence for overall benefit, the analysis limited to randomized, double-blind trials revealed no benefit of synthetic T3. In addition, T3 did not help in an open-label study of patients with severe depression [4]. Finally, T4 therapy in euthyroid patients with "hypothyroid symptoms" was no more effective than placebo in ameliorating symptoms [5].

Goals of therapy — The goals of therapy are to:

●Ameliorate symptoms

●Normalize serum TSH secretion

●Reduce goiter size (if present)

●Avoid overtreatment (iatrogenic thyrotoxicosis)

We aim to keep serum TSH within the normal reference range (approximately 0.5 to 5.0 mU/L). However, it is important to note that there is an age-related shift towards higher TSH concentrations in older patients (eg, ≥70 years), with an upper limit of normal of approximately 7.5 mU/L in 80 year olds. Among patients with goiter, approximately 50 percent will have some decrease in goiter size, which lags behind the fall in TSH secretion [6,7].

There is considerable controversy as to the appropriate upper limit of normal for serum TSH [8]. Most laboratories have used values of approximately 4.5 to 5.0 mU/L. However, others have argued that the upper limit of normal of the euthyroid reference range should be reduced to 2.5 mU/L because 95 percent of rigorously screened, young, euthyroid volunteers have serum values between 0.4 and 2.5 mU/L [9]. In contrast, others have reported that age-adjusted upper limits of normal for TSH should be higher than 4.5 to 5 mU/L, especially for patients over age 70 years [10]. Until there are data demonstrating an adverse biologic impact for serum TSH values between 2.5 and 5.0 mU/L, the wisdom of labeling such patients as hypothyroid is questionable [11]. This topic is reviewed in detail separately. (See "Laboratory assessment of thyroid function", section on 'Serum TSH'.)

The approach to patients with persistent symptoms of hypothyroidism despite adequate treatment with levothyroxine (eg, a normal serum TSH level) is reviewed below. (See 'Persistent symptoms on T4 monotherapy (normal or minimally elevated TSH)' below.)

STANDARD REPLACEMENT THERAPY — Hypothyroidism is corrected with synthetic thyroxine (T4, levothyroxine). T4 is a prohormone with very little intrinsic activity. It is deiodinated in peripheral tissues to form T3, the active thyroid hormone. This deiodination process accounts for approximately 80 percent of the total daily production of T3 in normal subjects. Approximately 70 to 80 percent of a dose of T4 is absorbed and, because the plasma half-life of T4 is long (seven days), once-daily treatment results in nearly constant serum T4 and triiodothyronine (T3) concentrations when a steady state is reached [12].

T4 formulations — T4 is available in tablet, soft gel, and liquid formulations [13]. Either a generic or a brand-name formulation of T4 is acceptable. The brand-name and most generic formulations are available in color-coded tablets at small increments in hormone content to allow precise titration of the dose according to the serum concentration of TSH.

●Tablet versus soft gel capsule or liquid – Most patients are treated with a T4 tablet. While in one study, the pharmacokinetics of the soft gel capsule were similar to tablets in healthy individuals [14], in other studies, the soft gel capsule was less dependent upon gastric pH than a branded tablet [15-17]. Thus, the soft gel capsule or liquid is an option for patients with suspected poor absorption of the standard solid tablet [18], especially in the presence of atrophic gastritis. It may also be better absorbed after bariatric surgery [19]. However, another, less costly option is to increase the dose of a generic T4 tablet with monitoring of TSH.

●Generic versus brand name – Either a generic or a brand-name formulation of T4 is acceptable [20].

There has been considerable controversy about the bioequivalence of the various T4 formulations. In the past, variation in the T4 content of brand-name and generic formulations led many experts to prefer a specific manufacturer [21]. In 1997, a study of two brand-name and two generic formulations of T4, using US Food and Drug Administration (FDA)-recommended methodology for determining bioequivalence, reported that all four were equivalent [22]. The methodology used to determine bioequivalence in the study is considered by some to be flawed since endogenous T4 concentrations were not taken into account [23].

Although the use of brand-name T4 products might be preferred theoretically to avoid the issue of variable bioavailabilities when there is interchange of different generic manufacturers by the pharmacy, often this is not possible, because of cost considerations. In addition, it has yet to be shown that switching among various generic manufacturers is a clinical problem. In the United States, the manufacturer of the generic formulation is included on the label. (See 'Switching between T4 manufacturers' below.)

Dose and monitoring

Initial dose — The average full replacement dose of T4 in adults is approximately 1.6 mcg/kg body weight per day (112 mcg/day in a 70-kg adult), but the range of required doses is wide, varying from 50 to ≥200 mcg/day. T4 requirements correlate better with lean body mass than total body weight [24]. In one study, the average full replacement dose after thyroidectomy was 1.76 mcg/kg body weight for body mass index (BMI) <25 kg/m², 1.47 mcg/kg for BMI 25 to 29 kg/m², 1.42 mcg/kg for BMI 30 to 34 kg/m², 1.27 mcg/kg for BMI 35 to 39 kg/m², and 1.28 mcg/kg for BMI over 40 kg/m² [25].

The initial dose can be the full anticipated dose (approximately 1.6 mcg/kg/day) in young, healthy patients (algorithm 1). Older patients (eg, >60 years) or those with coronary heart disease, in whom the duration of hypothyroidism is unknown, should be started on a lower dose (25 to 50 mcg daily). If the duration of hypothyroidism is known to be short, eg, less than approximately two months, starting doses in older patients or in those with coronary heart disease can be two-thirds to three-quarters of the anticipated dose needed to achieve a euthyroid state. (See 'Older patients or those with coronary heart disease' below.)

In one prospective study of different starting doses of T4, 50 hypothyroid patients (mean age 47 years but some patients were in their 70s and 80s) were randomly assigned to receive a full starting dose of T4 (1.6 mcg/kg/day) or T4 25 mcg/day with dose adjustments every four weeks. Euthyroidism was achieved more rapidly in the full-dose group, but signs and symptoms of hypothyroidism and quality of life improved at a similar rate in the two groups. No adverse cardiac effects were seen in either group, but the subjects in the study, including the older patients, had been carefully screened to rule out cardiovascular disease prior to enrollment [26]. In another study of post-thyroidectomy patients, only 285 of 951 (30 percent) who received initial weight-based dosing of T4 met their TSH goal at the first postoperative assessment [27]. In a validation cohort, a T4 dose calculator based on weight, height, age, sex, and calcium supplementation and developed using machine learning methods was able to modestly increase the number of patients meeting their TSH goal to 43.2 percent (329 of 762 patients). Further validation of the calculator is needed.

Timing of dose

●T4 (tablets, gel capsules, or liquid) should be taken on an empty stomach with water, ideally 30 to 60 minutes before breakfast.

●T4 (tablets, gel capsules, or liquid) should not be taken with other meds that interfere with its absorption (eg, bile acid resins, calcium carbonate, ferrous sulfate). (See "Drug interactions with thyroid hormones", section on 'Drugs that affect gastrointestinal absorption of thyroid hormone'.)

●Some patients take their T4 at bedtime (at least two hours after their last meal, ideally longer).

Few patients are able to wait a full hour before eating breakfast. The proximity to food ingestion, rather than time of day, is the more critical parameter. A meta-analysis demonstrated no difference in effectiveness of morning versus bedtime dosing based on TSH measurements [28]. In some studies, serum TSH concentrations were lower and less variable with standard fasting administration of levothyroxine than with nonfasting administration (eg, mean serum TSH 1.06±1.23, 2.93±3.29, and 2.19±2.66 mU/L if taken one hour before breakfast, with breakfast, or at bedtime two hours after the last meal, respectively) [29,30]. In another small study, there was no difference in serum TSH levels after ingestion of liquid T4 at breakfast compared with the same dose 30 minutes prior to breakfast [31]. In some studies, consumption of coffee and tea within one hour of levothyroxine intake interfered with tablet absorption [32,33]. In two small studies, espresso coffee, in comparison with water, appeared to interfere with T4 absorption of levothyroxine tablets [32] but not soft gel capsules [34].

Initial monitoring and dose adjustments — Patients who are treated with T4 usually begin to improve symptomatically within two weeks, but complete recovery can take several months in those with severe hypothyroidism. Although symptoms may begin to resolve after two to three weeks, steady-state TSH concentrations are not achieved for at least six weeks. Serum thyroid hormone concentrations increase first and then TSH secretion begins to fall because of the negative feedback action of T4 on the pituitary and hypothalamus.

After initiation of T4 therapy:

●The patient with symptomatic improvement should be re-evaluated and serum TSH measured in four to six weeks (algorithm 1). If the TSH remains above the reference range, the dose of T4 can be increased by 12 to 25 mcg/day in older patients, or it can be increased by a higher dose in younger patients based on the degree to which the initial dose increased free T4 concentrations and reduced TSH concentrations. The patient will require a repeat TSH measurement in six weeks. (See 'Adjustment of maintenance dose' below.)

●The patient with persistent symptoms after two to three weeks should be reevaluated and a serum free T4 and TSH measured in three weeks. If the serum free T4 is below normal, the dose can be increased at three weeks without additional testing, but it should be recognized that serum T4 (and TSH) concentrations at this time are not steady-state values, and serum TSH levels may still be falling despite normal (or even high) serum T4 concentrations. Given the one-week plasma half-life of T4, it takes approximately six weeks (six half-lives) before a steady state is attained after therapy is initiated or the dose is changed.

This process of increasing the dose of T4 every three to six weeks (depending upon the patient's symptoms) should continue, based upon periodic measurements of serum TSH (and free T4 if steady-state conditions have not yet been achieved), until the high values of TSH in patients with primary hypothyroidism return to the reference range. (See 'Goals of therapy' above.)

The maintenance dose may vary according to the cause of hypothyroidism. In a study of patients receiving chronic T4 therapy who were clinically euthyroid and had serum free T4 index values within the upper half of the normal range and normal serum TSH concentration, 73 patients with hypothyroidism caused by chronic autoimmune thyroiditis or radioiodine therapy were receiving less T4 (118 mcg/day, 1.6 mcg/kg/day) than 36 patients with thyroid cancer after near-total thyroidectomy (152 mcg/day, 2.1 mcg/kg/day) [35]. In 36 patients with central hypothyroidism and similar serum free T4 index values, the T4 dose was higher (155 mcg/day, 1.9 mcg/kg/day). These results suggest that both normal amounts of TSH and the presence of residual thyroid tissue are determinants of T4 dose in patients with hypothyroidism. In general, doses >2 mcg/kg/day suggest T4 malabsorption or poor adherence to the medication regimen. (See 'Persistent elevation in TSH' below.)

Adjustment of maintenance dose — After identification of the proper maintenance dose, the patient should be examined and serum TSH measured once yearly or more often if there is an abnormal result or a change in the patient's status (algorithm 1). The dose of T4 need not be altered in patients who are asymptomatic and clinically euthyroid if their serum TSH concentration is normal or only slightly above (or below) the reference range, assuming there are no new factors that might increase T4 requirements (table 1). TSH values may be slightly high (or low) because of laboratory error or normal circadian fluctuations in TSH secretion, so a slightly high (or low) value should be confirmed with repeat measurement before the dose is changed. Repeat TSH is often normal, and further dose adjustment is usually not required. However, if the initial abnormal TSH is not unexpected and likely due to the presence of a factor known to alter T4 dosing (table 1), the TSH does not need to be repeated before changing the dose. The serum TSH should be remeasured six to eight weeks after any change in dose. (See 'Goals of therapy' above.)

●Increasing the dose – Increases in dose may be required in the following settings:

•Pregnancy, and if increased, the dose should be reduced to the prepregnancy maintenance dose postpartum. (See "Hypothyroidism during pregnancy: Clinical manifestations, diagnosis, and treatment", section on 'Preexisting treated hypothyroidism'.)

•Weight gain of more than 10 percent of body weight.

•Diminished thyroid hormone absorption (eg, initiation of drugs that interfere with T4 absorption (table 1), development of gastrointestinal disorders that impair gastric acid secretion or cause malabsorption [eg, uncontrolled celiac disease]). (See "Diagnosis of celiac disease in adults".)

•Increased thyroid hormone excretion (nephrotic syndrome). (See "Endocrine dysfunction in the nephrotic syndrome".)

•Increased rate of thyroid hormone metabolism (therapy with rifampin, carbamazepine, phenytoin, or phenobarbital). (See "Drug interactions with thyroid hormones", section on 'Drugs that affect thyroid hormone metabolism or clearance'.)

When drugs that affect the absorption of T4 (table 1) are initiated for coexisting medical conditions, serum TSH should be measured four to six weeks later to confirm that the T4 dose is still adequate. The dose should be increased if the serum TSH value is high. Medications that interfere with T4 absorption should be taken several hours after the T4 dose. (See "Drug interactions with thyroid hormones", section on 'Drugs that affect gastrointestinal absorption of thyroid hormone'.)

If the TSH is slightly elevated (eg, 5 to 10 mU/L), a small increase of 12 to 25 mcg/day is usually sufficient. If the TSH is ≥10 mU/L, a larger dose increase (eg, 25 to 50 mcg/day) is usually necessary. Because of the seven-day half-life of levothyroxine, another method of changing the dose without the need for a new prescription is to recommend increasing the dose by 15 percent by adding one tablet a week. For example, if a patient is taking 100 mcg/d (1 tablet a day), if they were to take 8 tablets a week (eg, 1 tablet a day for 6 days and 2 tablets on 1 day), this is equivalent to 114 mcg a day ([(100 x 8)/7 = 114]).

Free T4 measurements can help determine an appropriate dose increase when TSH is very high (eg, ≥20 mU/L) since the magnitude of TSH elevation in hypothyroid patients is quite variable. For example, if steady-state conditions exist and the TSH exceeds 20, and the free T4 measurement is only half the mean value of the normal reference range (for example, a free T4 of 0.7 ng/dL in an assay with a normal range of 0.9 to 1.9 ng/dL), the dose could initially be doubled to target a mid-normal free T4 (a free T4 of 1.4 ng/dL in the example given). If the free T4 measurement is in the lower third of the normal range (for example, a free T4 1.2 ng/dL in the example above), a 20 percent increase in dose would target a mid-normal free T4. Further dose adjustments may be needed. The normal range for free T4 is assay dependent, and the target free T4 level may be higher than the average level, for example, in patients with thyroid cancer.

●Decreasing the dose – Decreases in dose may be required in the following settings [36,37]:

•Normal aging

•Weight loss of roughly more than 10 percent of body weight

•Initiation of androgen therapy

If the TSH is slightly below normal (eg, 0.05 to 0.3 mU/L), a small dose reduction of 12 to 25 mcg/day is usually sufficient. An alternative is to reduce the dose by 15 percent by omitting one pill a week. Lower TSH values may require larger dose reductions. For TSH values below 0.05 mU/L (below 0.1 mU/L in a second-generation assay), measurement of free T4 can help determine the estimated dose reduction. For example, using the free T4 assay described above, if the free T4 is 2.8 ng/dL, which is twice the average free T4, the necessary dose reduction may be as high as 50 percent if the target is a mid-normal free T4. However, in young patients with longstanding overtreatment, a stepwise reduction in dose over three to four months might be better tolerated. When TSH is suppressed to <0.05 mU/L, it will occasionally take longer than eight weeks for steady-state levels to be re-established.

Persistent elevation in TSH — Occasionally a patient will insist they are taking T4, but TSH will be quite high. Often these patients are taking larger doses of T4 than expected. This finding may be due to poor compliance or to poor absorption of T4.

Patients with autoimmune gastritis have higher T4 requirements (especially tablet formulations). In one study, the T4 dose was 17 percent higher in patients with parietal cell antibodies [38]. A similar effect can be seen in patients with occult celiac disease [39]. In this setting, free T4 levels are typically low or low-normal. In contrast, poorly compliant patients may have free T4 concentrations that are low, normal, or high, depending on how much T4 they have taken and when they have taken it (some poorly compliant patients take extra T4 in the days leading up to the appointment with their clinician and may have a high-normal or even an elevated free T4 with a TSH that hasn't had time to fall into the normal or subnormal range).

In patients with persistently elevated TSH despite what appears to be an adequate dose of T4, it should be confirmed that T4 is taken daily on an empty stomach with water, ideally an hour before breakfast, and that medications that interfere with T4 absorption (table 2) are taken several hours after the T4 dose.

If the TSH remains elevated, and noncompliance is not acknowledged, adequate T4 absorption can be assessed by a T4 absorption test [40]. Patients are administered their weight-based weekly oral dose of T4 (eg, 1.6 mcg/kg body weight times 7), and free T4 is measured at baseline and at two hours. In one study, the average normal increase in free T4 at 120 minutes was 54 percent [41]. Values well below this suggest malabsorption, whereas values similar to this suggest poor compliance. Only 1 of 16 tests done at one institution over a four-year period documented malabsorption [42]. (See 'Poorly compliant patients' below.)

In patients with celiac disease, a gluten-free diet improves T4 absorption [39]. Other options for patients with poor T4 absorption include increasing the dose of T4 tablets or switching to a soft gel or liquid preparation. In one study, patients who appeared to be resistant to levothyroxine administration did not absorb T4 tablets well but absorbed T4 tablets after they were pulverized [43]. (See 'T4 formulations' above.)

Switching between T4 manufacturers — If possible, we suggest that patients remain on the same T4 formulation. However, switching between synthetic thyroxine (T4, levothyroxine) manufacturers is typically not a clinical problem. Most large pharmacy chains have been using the same generic manufacturer for years. If a switch from one manufacturer to another is made by the pharmacy and the patient is concerned regarding equivalent efficacy of the formulations, or if maintaining the serum TSH within a narrow range is important (eg, thyroid cancer treatment), we measure a serum TSH six weeks after the change to document that the serum TSH is still within the therapeutic target.

Studies evaluating a switch to a different T4 manufacturer show that the majority of individuals maintain a TSH within the normal range, particularly when the T4 dose is ≤100 mcg daily. As examples:

●In a comparative effectiveness study evaluating changes in TSH levels after switching or not switching between generic levothyroxine products in 2780 propensity-matched patient pairs, the proportion of individuals with TSH levels within the normal range was similar in nonswitchers and switchers (82.7 and 84.5 percent, respectively, risk difference -0.018, 95% CI -0.038 to 0.002) [44].

●In a registry study from the Netherlands during an unexpected shortage of one branded T4 product, the proportion of individuals with subsequent TSH levels within the normal range was slightly higher in 6438 nonswitchers than 1204 switchers to another branded or generic T4 manufacturer (81 versus 76 percent, respectively, when the dose was ≤100 mcg daily) [45]. Fewer patients had normal TSH levels if their dose exceeded 100 mcg/day (76 versus 37 percent in nonswitchers and switchers, respectively).

Long-term outcomes — Successful treatment reverses all the symptoms and signs of hypothyroidism, although some neuromuscular and psychiatric symptoms may not disappear for several months. Long-term treatment of hypothyroidism is not associated with impaired cognitive function or depressed mood in some studies [46], but other studies have documented a persistent defect in psychological well-being that was not corrected with adequate amounts of T4 [47]. Limited evidence also suggests no increase in all-cause mortality among patients with treated hypothyroidism [48,49].

In contrast, infants with congenital hypothyroidism in whom treatment is inadequate or delayed for several months may have permanent brain damage, even if they are adequately treated several months later. (See "Treatment and prognosis of congenital hypothyroidism".)

Adverse effects — Adverse effects of T4 replacement are rare if the correct dose is given. Over-replacement and under-replacement of thyroid hormone are associated with increased cardiovascular risk [50-52]. (See 'Adjustment of maintenance dose' above.)

Over-replacement, reduced TSH — Over-replacement with thyroid hormone should especially be discouraged. Over-replacement causes subclinical hyperthyroidism (normal serum T4 and T3 and low serum TSH concentrations) or overt hyperthyroidism. Atrial fibrillation is the main risk of subclinical hyperthyroidism, and it occurs three times more often in older patients with serum TSH concentrations <0.1 mU/L than in normal individuals (figure 1) [53]. Furthermore, in a retrospective study of 705,307 patients on levothyroxine replacement followed for an average of four years, risk of cardiovascular mortality was increased with exogenous hyperthyroidism (TSH <0.1 mU/L) compared with euthyroidism (hazard ratio [HR] 1.39, 95% CI 1.32-1.47) [50]. In another retrospective study of 663,809 patients on levothyroxine replacement, risk of stroke was increased with exogenous hyperthyroidism (TSH <0.1 mU/L) compared with euthyroidism (odds ratio [OR] 1.33, 95% CI 1.24-1.43) [51].

Patients with iatrogenic subclinical hyperthyroidism, particularly postmenopausal women, may also have accelerated bone loss. It is therefore important to educate patients about the potential adverse effects of overtreatment with T4. (See "Bone disease with hyperthyroidism and thyroid hormone therapy".)

The risks associated with over-replacement of thyroid hormone are greatest in those with the most suppressed TSH concentrations. This was illustrated by the findings from a cohort study of 17,684 patients taking T4 replacement therapy. Patients with TSH concentrations between 0.04 and 0.4 mU/L were not at risk for arrhythmias or fractures compared with those with a TSH in the normal reference range. However, patients with more severe iatrogenic thyrotoxicosis (TSH <0.03 mU/L) had a significantly increased risk of arrhythmia (HR 1.6) and fractures (HR 2.0) [54].

Under-replacement, elevated TSH — Cardiovascular morbidity and mortality are also increased in patients with persistent hypothyroidism due to inadequate treatment with thyroid hormone. In the retrospective study of 705,307 patients on levothyroxine replacement, risk of cardiovascular mortality was increased in people with exogenous hypothyroidism (TSH >20 mU/L) compared with euthyroidism (HR 2.67, 95% CI 2.55-2.80) [50]. In the other retrospective study of 663,809 patients on levothyroxine replacement, risk of stroke was increased with exogenous hypothyroidism (TSH >5.5 mU/L) compared with euthyroidism (OR 1.29, 95% CI 1.26-1.33) [51].

Allergy to dye or excipients — Rare patients have an allergy to the dye or excipients (filler) in the tablets. For dye sensitivities, multiples of the white 50 mcg tablets can be given. For allergies to excipients (except gelatin), the soft gel capsule (which contains T4 as a liquid) can be given. The liquid formulation contains glycerol and water.

PERSISTENT SYMPTOMS ON T4 MONOTHERAPY (NORMAL OR MINIMALLY ELEVATED TSH) — The majority of patients feel well on properly dosed T4 monotherapy. However, some hypothyroid patients remain symptomatic despite T4 replacement and normal serum TSH concentrations [55]. In a large, community-based questionnaire study of patients taking T4 who had normal serum TSH concentrations, 9 to 13 percent more patients had impaired psychologic well-being compared with normal subjects [47]. Because many symptoms of hypothyroidism are nonspecific, the possibility of an inadequate current T4 dose should be verified:

●Measure serum TSH

●Evaluate for alternative causes of the symptoms

TSH in upper half of or above reference range — If a patient has possible hypothyroid symptoms and the serum TSH is confirmed by repeat measurement to be at the upper limits or minimally above the reference range, it is reasonable to increase the dose and to aim for a serum TSH value in the lower half of the reference range. However, it is important to note that there is an age-related shift towards higher TSH concentrations in older patients, with an upper limit of normal of approximately 7.5 mU/L in 80 year olds. (See "Laboratory assessment of thyroid function", section on 'Serum TSH'.)

Additionally, it is likely that improved symptoms with higher doses are due to the expectation of feeling better on a higher dose of levothyroxine, rather than a true physiologic benefit. In one study of 697 patients on thyroid hormone replacement, psychological well-being assessed by the General Health Questionnaire (GHQ)-12 correlated positively with serum free T4 and negatively with serum TSH for TSH values between 0.3 to 4.0 mU/L [56]. More specifically, psychological well-being was better in patients with lower serum TSH concentrations. However, in a blinded, placebo-controlled, crossover trial, patients could not distinguish between their usual T4 dose and doses that were 25 to 50 mcg/day higher, ie, they could not distinguish between TSH values that averaged 2.8 mU/L from those that averaged 0.3 mU/L [57], and in a second double-blinded trial, patients had no preference after varying doses of levothyroxine that resulted in average TSH levels between 1.85 and 9.49 mU/L [58].

TSH in mid to lower half of reference range — Persistent symptoms of hypothyroidism despite a frankly normal serum TSH level may be due to inadequacy of T4 to physiologically restore tissue thyroid hormone levels to normal or to factors unrelated to hypothyroidism, such as inflammation in other tissues from autoimmune disease [59]. In this setting, we do not increase the dose of T4.

The question of whether thyroid peroxidase (TPO) antibodies themselves, or associated immunologic modulators, cause inflammation in other tissues was evaluated in a trial from Norway, in which 150 patients with Hashimoto's thyroiditis and persistent symptoms (eg, fatigue, joint and muscle tenderness, dry mouth and eyes) despite normal TSH and free T4 were randomly assigned to total thyroidectomy or medical management (with optimally titrated thyroid hormone replacement in both groups) [60,61]. After 18 months, there was a greater decrease in TPO antibody levels and a greater improvement in the primary outcome (SF-36 general health score) in the patients who received surgery (+26 versus -3 points in the medical management group). The absence of a sham surgical group limits the validity of this study, however it is unlikely that the study can be repeated with an appropriate control group, and this provocative finding is likely to remain controversial. Thyroidectomy has surgical complications (in this trial, infection [4.1 percent], hypocalcemia [4.1 percent], and recurrent laryngeal nerve palsy [5.5 percent]) and is currently not a recommended treatment for biochemically euthyroid patients with autoimmune thyroiditis who have persistent nonspecific symptoms.

Is there a role for combination T4 and T3 therapy? — For most patients with persistent symptoms of hypothyroidism and a normal TSH, we do not prescribe combination T4-T3 therapy. Commercially available preparations either do not reflect normal T4:T3 ratios, and/or they are short acting, difficult to dose and titrate, and may have potential adverse effects (see 'Adverse effects' below). Our approach is in agreement with the 2014 American Thyroid Association (ATA) guidelines, which found insufficient evidence to support the routine use of a combination of T4 and T3 therapy in patients unhappy with T4 monotherapy [1].

However, for selected patients with persistent symptoms and a normal TSH on levothyroxine monotherapy, a trial of T4-T3 combination therapy may have some benefit. In patients with residual thyroid function, such as many patients with hypothyroidism due to Hashimoto's thyroiditis, measuring a serum T3 level may help facilitate the decision to trial combined T4-T3 therapy.

●Patient selection for combined T4 and T3 therapy – A therapeutic trial using doses of T4 and T3 that attempt to mimic normal physiology (ratio T4:T3 of 13:1 to 16:1) while maintaining a normal TSH is most reasonable in the following patients:

•Patients with no residual endogenous T3 production (eg, after thyroidectomy or ablative therapy with radioiodine), and have not felt well since the procedure.

•Patients with hypothyroidism due to Hashimoto's thyroiditis who do not feel well and who have serum T3 near or below the lower end of the reference range.

●Patients in whom combined T4 and T3 is not likely to be effective or should be avoided

•Older patients and patients with underlying cardiovascular disease in whom excessive T3 levels might precipitate an arrhythmia or other adverse cardiovascular events. (See 'Over-replacement, reduced TSH' above and 'Adverse effects' below.)

•Patients who have previously felt well on T4 monotherapy but now feel poorly, and patients with mild hypothyroidism who have persistent endogenous T3 production and who are taking low doses of T4. Combined therapy is not likely to improve symptoms in these patients.

•Pregnancy – An important caveat for women of childbearing age using combined T4 and T3 treatment is that fetal neurogenesis is primarily dependent upon maternal free T4 concentrations until week 16 to 18 of gestation [62]. Regimens containing excessive T3 cause hypothyroxinemia, which has been associated with impaired neurologic development. For example, patients taking desiccated thyroid extract in the trial noted above [63] had a mean free T4 of 0.85 ng/dL (normal 0.89 to 1.76 ng/dL).

There is controversy as to whether T4 replacement alone can mimic normal physiology. T4 is deiodinated in peripheral tissues to form T3, the active thyroid hormone. The prohormone nature of T4 is an advantage over other thyroid hormone formulations because the patient's own physiologic mechanisms control the production of active hormone. In some [12,64], but not all [65,66], studies, mean serum T3 concentrations were within the normal range in hypothyroid patients receiving adequate T4 therapy. In one study, serum T3 levels were more uniformly restored to preoperative levels when T4 doses were high enough to suppress the serum TSH (≤0.3 mU/L) [65]. In another study, 15 percent of athyreotic patients taking T4 monotherapy had serum T3 levels below the reference range for individuals with intact thyroid glands [66]. These studies raise the question of whether some hypothyroid patients might benefit from substitution of some T3 for T4. However, in a prospective study of recently athyreotic patients receiving T4 therapy to normalize serum TSH concentrations (TSH level of ≤4.6 mU/L), serum T3 concentrations on treatment were, in most (but not all) cases, comparable with the patients' preoperative T3 values [64].

Multiple randomized trials have evaluated combination T4-T3 therapy, and almost all showed that combination therapy does not appear to be superior to T4 monotherapy for the management of hypothyroid symptoms [67-78]. The normal ratio of T4:T3 secretion by the thyroid gland is approximately 13:1 to 16:1 (mcg T4 to mcg T3) [59]. The majority of the randomized, controlled trials used excessive and nonphysiologic amounts of T3 when assessing combination therapy. In addition, a slow-release T3 preparation, which may avoid supraphysiologic peaks in serum T3 concentrations, is not yet commercially available [79]. A combination T4-slow release T3 preparation may better replicate physiologic T4-T3 production. Well-designed, blinded studies are still needed to address this ongoing controversy [80].

Efficacy — In a systematic review of nine randomized trials, only one trial reported beneficial effects of combination T4-T3 therapy on mood, quality of life, and psychometric performance when compared with T4 therapy alone [81]. Subsequent meta-analyses did not report a significant benefit with combined therapy compared with T4 monotherapy [82,83]. As examples:

●In a meta-analysis of 11 published randomized trials (1216 patients), there was no benefit (fatigue, bodily pain, anxiety, depression, quality of life) of combined therapy [82].

●In a meta-analysis of seven blinded randomized trials (348 patients), there was no difference in prevalence rate for patient preference of combined therapy over T4 monotherapy (46.2 percent, 95% CI 40.2-52.4 percent) [83].

In some [76], but not all [63,74], trials in the meta-analyses, patients were given overzealous doses of T3, resulting in mild hyperthyroidism. As examples:

●In one trial comparing T4 alone with T4 and T3 in a molar ratio of 10:1 or 5:1, there were similar improvements in mood, fatigue, psychological symptoms, or neurocognitive testing [76]. However, patients preferred combined therapy to T4 alone. Forty-four percent of the patients who preferred combined therapy had TSH values less than 0.11 mU/L. In addition, those patients taking T4:T3 in a molar ratio of 5:1 had a 1.8 kg weight loss, which correlated with a preference for the combined treatment, and 54 percent of these patients had subnormal serum TSH concentrations.

●In one small, well-designed study, women taking 100 mcg of T4 were changed to 75 mcg T4 and 5 mcg T3 (ratio 15:1) [74]; while no benefit was found using standardized questionnaires, patients preferred combination therapy over monotherapy despite a higher TSH within the normal range in the group receiving combination therapy.

●In a double-blind, crossover trial comparing T4 and desiccated thyroid extract (T4:T3 ratio 4:1), there were no differences in symptoms and neurocognitive measurements between the two groups, but 49 percent of the patients preferred thyroid extract over T4 (19 percent preferred T4 and 33 percent had no preference), and those who preferred thyroid extract had lost on average 1.8 kg during the study [63]. In this trial, thyroid medications were adjusted to maintain a TSH level between 0.5 and 3.0 mU/L (mean achieved TSH levels were 1.30 and 1.67 mU/L for T4 and desiccated thyroid extract, respectively).

Whether a combination of T4 and T3 is beneficial in a subset of hypothyroid patients has also been studied, but the results are conflicting:

●One analysis suggested patients with a polymorphism in the type 2 deiodinase, which converts T4 to T3, might benefit from combination therapy [84]. Sixteen percent of the population studied had the CC genotype of the rs225014 polymorphism in the deiodinase 2 gene (DIO2); these patients had worse baseline quality-of-life scores and showed greater improvement after T4-T3 therapy compared with T4 alone. In another study, there was a clear preference for T4-T3 therapy in patients with a combination of polymorphisms (DIO2 and MCT10) [85].

●However, a smaller study was unable to show a difference in response to combined T4-T3 therapy based on DIO2 genotype [86]. In addition, a large population-based study showed lower health-related quality-of-life scores in levothyroxine-treated hypothyroid patients compared with controls, but this did not differ among individuals who did or did not have the rs225014 polymorphism [87].

●Another study of 75 patients randomly assigned to T4 alone, combined T4 and T3 therapy, or thyroid extract in a crossover design found no difference among the three groups in scores on the thyroid symptom or general health questionnaires or patient preference [88]. There was also no difference among patients with autoimmune versus non-autoimmune hypothyroidism, or those who had the Thr92Ala polymorphism in the type 2 deiodinase. However, the 20 most symptomatic patients at baseline had a strong preference for both combined T4 and T3 or thyroid extract over T4 alone, and this was associated with improved thyroid and general health scores and improvement in the Beck Depression Inventory and Visual Memory Index.

Dosing and available formulations — There are commercial thyroid hormone preparations containing T3 alone and T4 alone (table 3).

●We do not use combined T4 and T3 therapy when the ratio of T4 to T3 is not physiologic (ie, when T3 doses are excessive). The normal ratio of T4:T3 is 13:1 to 16:1.

●We do not use desiccated thyroid extract, which has a T4:T3 ratio of 4:1. (See 'Converting from desiccated thyroid extract to T4' below.)

Although for most patients with hypothyroidism we do not suggest treatment with T3, if combined therapy is given, it should be given as separate pills, and the doses of T4 and T3 should mimic normal physiology as closely as possible. T3 alone is available as 5 and 25 mcg tablets, so available doses utilizing half tablets are 2.5, 5, 7.5, 10, and 12.5 mcg. When possible, the dose of T3 should be divided into morning and afternoon doses. One can calculate an optimal dose by considering that T3 is three to four times more potent metabolically than T4 and aiming for a T4:T3 ratio of approximately 13:1 to 16:1 [59]. The table is offered as a guide for the conversion of T4 monotherapy to combined T4 and T3 therapy (table 4). This approach is largely consistent with the European Thyroid Association (ETA) and joint British Thyroid Association/Society for Endocrinology guidelines on the use of combination therapy, published with the intent of enhancing its safety [59,89].

Adverse effects — Commercially available combination T4-T3 products contain excessive and nonphysiologic amounts of T3, exacerbating the risk of over-replacement (see 'Over-replacement, reduced TSH' above). As described above, over-replacement of thyroid hormone causes subclinical hyperthyroidism or even overt hyperthyroidism. The risks associated with over-replacement of thyroid hormone are greatest in those with the most suppressed TSH concentrations and in older adults.

In a retrospective study of 1434 T3 users (monotherapy or combined therapy) compared with 3908 T4 users, the risk of heart failure and stroke was higher in T3 users (3.13 versus 1.9 per 1000 person-years for heart failure, incidence rate ratio [IRR] 1.664, 95% CI 1.002-2.764 and 3.4 versus 1.9 per 1000 person-years for stroke, IRR 1.757, 95% CI 1.073-2.877), but the risk was not statistically higher for atrial fibrillation or osteoporosis [90]. (See "Subclinical hyperthyroidism in nonpregnant adults", section on 'Potential consequences of subclinical hyperthyroidism'.)

Monitoring combined therapy — In patients taking combined therapy (in a ratio of 13:1 to 16:1), we typically monitor TSH six weeks after initiating therapy. TSH levels in steady-state conditions will reflect adequacy of therapy. Dose adjustments should be individualized based on the degree that TSH diverges from the normal reference range and on specific patient characteristics (eg, body mass index [BMI], age). The following examples provide an approximation of how the dose may be adjusted:

●If the TSH is slightly elevated (eg, approximately 5 to 10 mU/L), increase T4 initially by 12 mcg. If TSH is significantly above the reference range (eg, ≥10 mU/L), the doses of both T4 and T3 may need to be adjusted, maintaining the appropriate ratio of 13:1 to 16:1.

●If the TSH is normal, continue the same dose.

●If the TSH is below the lower limit of normal but detectable, and the ratio is closer to 13:1, one should reduce the T3 dose by 2.5 mcg; however, in a younger patient with a ratio closer to 16:1, reducing the T4 dose by 12 mcg instead may also be appropriate. If the TSH is fully suppressed, the doses of both T4 and T3 may need to be adjusted, maintaining the appropriate ratio of 13:1 to 16:1.

Serum free T4 can be useful, especially in non-steady-state conditions, in patients receiving combined T4 and T3 therapy at a physiologic ratio of 13:1 to 16:1 but may be misleading (due to low values) in patients receiving combined therapy where the ratio is low. For example, over one-half of patients receiving desiccated thyroid extract (T4:T3 ratio 4:1) will have subnormal T4 concentrations despite normal serum TSH [63].

We do not typically monitor T3 levels. Patients treated with currently available T3-containing formulations have wide fluctuations in serum T3 concentrations throughout the day due to its rapid gastrointestinal absorption and its relatively short half-life in the circulation (approximately one day). In fact, T3 serum levels may be elevated three to four hours after the last dose (eg, at noon if the dose is taken at 8 AM) and low if T3 is measured before the next dose. Thus, T3 measurements primarily reflect the interval since the dose was administered and are not used for monitoring.

Converting from desiccated thyroid extract to T4 — For patients who are taking desiccated thyroid, we prefer to switch them to T4.

By US Food and Drug Administration (FDA) mandate, 1 grain of desiccated thyroid extract (60 mg) should contain approximately 38 mcg T4 and 9 mcg T3. Since T3 is approximately four times as metabolically active as T4, 9 mcg T3 is equivalent to approximately 36 mcg T4; therefore, 1 grain would be equivalent to approximately 74 mcg T4. However, in a randomized trial comparing T4 with desiccated thyroid extract, 1 grain (60 mg) of extract was equivalent to 88 mcg of T4 [63]. Both of these conversions result in lower amounts of T4 than traditionally recommended; the United States Pharmacopeia Drug Information suggests 1 grain (60 mg) is roughly equivalent to 100 mcg of T4 [63]. Therefore, many clinicians use a simple conversion factor of 1 grain = 100 mcg T4. For a patient who is taking 1.5 grains (90 mg) of desiccated thyroid, an equivalent T4 dose ranges from 112 to 150 mcg. Some clinicians prefer to begin with the higher dose (150 mcg) and slowly taper the dose if the TSH measured six weeks after switching remains below the reference range.

Converting from combined T4 and T3 therapy to T4 monotherapy — For patients converting from T4-T3 combination therapy back to T4 monotherapy, the equivalent dose of T4 can be calculated by considering that T3 is three to four times more potent metabolically than T4. As an example, for a patient who is taking 12.5 mcg T3 and 50 mcg T4, the equivalent dose of T4 is approximately 100 mcg (50 mcg T4 plus four times the dose of T3).

SPECIAL TREATMENT SITUATIONS — There are several situations in which therapy should be more conservative or the dose may need modification:

Older patients or those with coronary heart disease — Older patients (>60 years), patients with coexisting cardiopulmonary problems, or patients with a history of coronary heart disease, should initially be treated with 25 to 50 mcg T4 (levothyroxine)/day (algorithm 1). Patients with coronary artery disease without other cardiopulmonary problems who have had recent successful interventions to treat ischemia (eg, coronary artery bypass grafting [CABG] or coronary artery stenting) can initially receive up to 80 percent of their weight-based dose (1.6 mcg/kg/day).

The dose can be increased by 12 to 25 mcg/day every three to six weeks until replacement is complete, as determined by a normal serum TSH concentration or an increase in dose results in cardiac symptoms, in which case something less than full replacement may have to be accepted. It is important to note that there is an age-related shift towards higher TSH concentrations in older patients, with an upper limit of normal of approximately 7.5 mU/L in 80 year olds. (See 'Goals of therapy' above.)

Thyroid hormone increases myocardial oxygen demand, which is associated with a small risk of inducing cardiac arrhythmias, angina pectoris, or myocardial infarction in older patients. A 1961 report remains the largest and best study of the effects of beginning thyroid hormone on chest pain in patients with hypothyroidism [91,92]. Among 1503 hypothyroid patients, the following findings were noted:

●Fifty-five had angina before thyroid hormone replacement therapy. During therapy, 21 improved, 25 had no change, and 9 had more angina.

●Thirty-five patients developed new angina during therapy, 6 during the first month, 6 during the first year, and 23 after one year.

Thus, angina may improve with T4 treatment, and it does not often first appear during T4 replacement therapy.

Many older patients receiving thyroid hormone replacement are over- or undertreated, as illustrated by a community survey that identified 339 individuals over age 65 years taking thyroid hormone [93]. Forty-one percent of patients had a subnormal TSH, 16 percent had a high TSH, and only 43 percent were euthyroid [93]. Patients with low body weight were more likely to have a subnormal TSH, while those with diabetes were at risk for having low and high serum TSH. (See 'Goals of therapy' above.)

The clinical manifestations and consequences of hypothyroidism (subclinical and overt) in older adults are discussed in detail elsewhere. (See "Clinical manifestations of hypothyroidism" and "Subclinical hypothyroidism in nonpregnant adults", section on 'Consequences of subclinical hypothyroidism'.)

Pregnancy — Women need more thyroid hormone during pregnancy and, unlike normal women, those with hypothyroidism are unable to increase thyroidal T4 and T3 secretion. Approximately 75 to 85 percent of women with preexisting hypothyroidism need a higher dose of T4 during pregnancy to maintain normal TSH secretion. The increase in T4 requirements occurs as early as the fifth week of gestation and plateaus by week 16 to 20. The treatment of hypothyroidism during pregnancy is reviewed separately. (See "Hypothyroidism during pregnancy: Clinical manifestations, diagnosis, and treatment", section on 'Preexisting treated hypothyroidism'.)

Euthyroid women with TPO antibodies who become pregnant are also discussed separately. (See "Overview of thyroid disease and pregnancy", section on 'Thyroid peroxidase antibodies in euthyroid women'.)

Estrogen therapy — In women receiving T4 therapy, estrogens increase serum thyroxine-binding globulin (TBG) concentrations, as they do in normal women, and may increase the need for T4. In a study of postmenopausal women (25 women with hypothyroidism and 11 normal women) treated with 0.625 mg conjugated estrogens daily for 48 weeks, serum T4 and TBG concentrations increased in both the hypothyroid and normal women. Serum free T4 and TSH concentrations did not change in the normal women but decreased and increased, respectively, in the hypothyroid women [94]. Among the latter, seven women had serum TSH concentrations >7 mU/L and were given more T4. These data suggest that serum TSH should be measured approximately 6 to 12 weeks after starting estrogen therapy in postmenopausal women receiving T4 therapy to determine if an increase in T4 dose is needed. Whether younger hypothyroid women receiving oral contraceptives require dose adjustments is uncertain. Such patients may require dose adjustments, especially when oral contraceptives are initiated because of hypoestrogenic states.

Surgical patients — Patients receiving chronic T4 therapy who undergo surgery and are unable to eat for several days need not be given T4 parenterally. If oral intake cannot be resumed in five to seven days, then T4 should be given intravenously. The dose should be approximately 70 to 80 percent of the patient's usual oral dose because that is approximately the fraction of oral T4 that is absorbed [12,95]. We typically give 80 percent.

Several studies have investigated the safety of general anesthesia and surgery in patients with untreated or inadequately treated hypothyroidism. This topic is reviewed in detail elsewhere. (See "Nonthyroid surgery in the patient with thyroid disease", section on 'Hypothyroidism'.)

Poorly compliant patients — Some patients do not take their T4 regularly and do not respond to efforts to improve compliance. These patients may be given their total weekly dose of T4 once per week [96]. The efficacy of this approach was evaluated in a crossover trial of 12 patients [97]. The mean serum TSH concentration one week after a single weekly dose was slightly higher than when the usual dose was given daily (6.6 versus 3.9 mU/L), but the raised value returned to normal one day after the next weekly dose. There was no difference in symptoms between daily or weekly dosing. Weekly dosing should probably not be used in patients with coronary heart disease.

Thyroid cancer — Patients who have had a thyroidectomy for thyroid cancer, with or without additional treatment with radioiodine (I-131), need to take T4 not only for treatment of hypothyroidism but also to prevent recurrence of their thyroid cancer, especially those with higher risk disease. (See "Differentiated thyroid cancer: Overview of management", section on 'Thyroid hormone suppression'.)

Notably, temporary treatment with T3 monotherapy is appropriate in patients with thyroid cancer who are to undergo radioiodine imaging and possible treatment. To shorten the period of hypothyroidism, the patient's T4 therapy is discontinued and T3 is substituted for three to four weeks until the T4 is cleared. (See "Differentiated thyroid cancer: Radioiodine treatment", section on 'Thyroid hormone withdrawal'.)

Myxedema coma — Myxedema coma is defined as severe hypothyroidism leading to decreased mental status, hypothermia, and other symptoms. It is a medical emergency with a high mortality rate. Fortunately, it is now a rare presentation of hypothyroidism, probably because of earlier diagnosis. The clinical presentation, diagnosis, and treatment of myxedema coma are reviewed separately. (See "Myxedema coma".)

Selenium deficiency — Selenium is required for deiodinase activity (the enzyme is a selenoprotein), and it has important effects on immune function. The effects of selenium deficiency on normal thyroid function are not well described. However, selenium deficiency has been shown to exacerbate both autoimmune thyroid disease and endemic cretinism [98]. Selenium supplementation in some studies reduces antithyroid peroxidase antibody levels, improves the ultrasound structure of the thyroid gland, and reduces the occurrence of postpartum thyroiditis in pregnant women with TPO antibodies, but a meta-analysis failed to show any improvement in thyroid function when selenium is given to hypothyroid individuals [99]. (See "Postpartum thyroiditis", section on 'Prevention'.)

When the diagnosis of hypothyroidism is uncertain — Some patients have been prescribed thyroid hormone for questionable indications (eg, obesity or hypercholesterolemia) or the diagnosis of hypothyroidism is uncertain. In such a patient, a high serum TSH concentration suggests that the patient is truly hypothyroid, and the T4 dose should be increased accordingly. If, however, the serum TSH values are normal or low, the dose of thyroid hormone can be reduced by one-half and serum TSH measured again in four to six weeks. If the value is normal, the dose can be reduced further or stopped. Most patients with hypothyroidism have symptoms and a high serum TSH concentration within one month after discontinuing therapy. In a meta-analysis of 1082 patients from 16 studies, 37 percent of putatively hypothyroid patients had their levothyroxine successfully discontinued [100].

Many of these patients are reluctant to discontinue their thyroid hormone, especially if they have taken it for many years. In the meta-analysis noted above, two-thirds of patients restarted levothyroxine, in most cases because of recurrent mildly elevated serum TSH levels [100]. In this case, the goal should be to provide an appropriate dose of T4 (adjusted to maintain a normal serum TSH concentration) to avoid the potential adverse cardiac and skeletal effects of overtreatment.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Hypothyroidism".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Hypothyroidism (underactive thyroid) (The Basics)")

●Beyond the Basics topics (see "Patient education: Hypothyroidism (underactive thyroid) (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Hypothyroidism definition – Overt primary hypothyroidism is characterized biochemically by a high serum thyroid-stimulating hormone (TSH) concentration and a low serum free thyroxine (T4) concentration. All patients with overt primary hypothyroidism require treatment (regardless of symptoms), unless the hypothyroidism is transient (as after painless thyroiditis or subacute thyroiditis) or reversible (due to a drug that can be discontinued). (See 'Defining hypothyroidism' above and "Disorders that cause hypothyroidism", section on 'Transient hypothyroidism'.)

●Goals of therapy – The goals of therapy are amelioration of symptoms, normalization of TSH secretion, reduction in size of goiter (if present), and avoidance of overtreatment (iatrogenic thyrotoxicosis). We aim to keep serum TSH within the normal reference range (approximately 0.5 to 5.0 mU/L). It is important to note that there is an age-related shift towards higher TSH concentrations in patients ≥70 years, with an upper limit of normal of approximately 7.5 mU/L in 80 year olds. (See 'Goals of therapy' above.)

●Standard thyroid hormone replacement – Hypothyroidism is corrected with synthetic thyroxine (T4, levothyroxine). Either a generic or a brand-name formulation is acceptable. (See 'Standard replacement therapy' above.)

●Initial dosing – The initial dose of T4 can be the full anticipated dose (1.6 mcg/kg/day) in young, healthy patients, but older patients (eg, >60 years) and those with coronary heart disease should be started on a lower dose (25 to 50 mcg daily) (algorithm 1). T4 should be taken on an empty stomach with water, ideally 30 to 60 minutes before breakfast. (See 'Initial dose' above and 'Timing of dose' above.)

●Initial monitoring and dose adjustments – After initiation of T4 therapy, the patient should be reevaluated and serum TSH should be measured in six weeks and the dose adjusted accordingly (algorithm 1). Symptoms may begin to resolve after two to three weeks, but steady-state TSH concentrations are not achieved for at least six weeks. After identification of the proper maintenance dose, the patient should be examined and serum TSH measured once yearly or more often if there is an abnormal result or a change in the patient's status (table 1). (See 'Initial monitoring and dose adjustments' above and 'Adjustment of maintenance dose' above.)

●Switching between T4 manufacturers – If possible, we suggest that patients remain on the same formulation of T4 (Grade 2C). However, switching from one manufacturer of T4 to another is usually not a clinical problem. If a switch from one manufacturer to another is made by the pharmacy and the patient is concerned regarding equivalent efficacy of the formulations, or if maintaining the serum TSH within a narrow range is important (eg, thyroid cancer treatment), we measure a serum TSH six weeks after changing to document that the serum TSH is still within the therapeutic target. (See 'T4 formulations' above.)

●Persistent symptoms on T4 monotherapy – The possibility of an inadequate current T4 dose should be verified by measuring serum TSH before the dose is increased. In addition, clinicians should evaluate for alternative causes of the symptoms. (See 'Persistent symptoms on T4 monotherapy (normal or minimally elevated TSH)' above.)

•TSH in upper half to just above reference range – If a patient has possible hypothyroid symptoms and the serum TSH is confirmed by repeat measurement to be at the upper limits or above the reference range, it is reasonable to increase the dose and to aim for a serum TSH value in the lower half of the normal range; however, improved symptoms with higher doses may be based on expectation of benefit rather than a true physiologic advantage. (See 'TSH in upper half of or above reference range' above.)

•TSH in mid to lower half of reference range – For most patients with persistent symptoms with normal TSH, we suggest not using combination T4-liothyronine (T3) therapy (Grade 2B). For selected patients, however, a trial of T4-T3 combination therapy may have some benefit. A therapeutic trial using separate pills of T4 and T3 is most reasonable in patients who have not felt well since thyroidectomy or ablative therapy with radioiodine, or in patients with hypothyroidism due to Hashimoto's thyroiditis who do not feel well and who have serum T3 near or below the lower end of the reference range. We discourage the use of combined therapy in older patients, patients with underlying cardiovascular disease in whom excessive T3 levels might precipitate an arrhythmia or other adverse cardiovascular events, and in pregnant women. (See 'TSH in mid to lower half of reference range' above and 'Is there a role for combination T4 and T3 therapy?' above.)

When T4-T3 therapy is used, the T4:T3 dose ratio should be approximately 13:1 to 16:1, and TSH should remain normal (table 4). (See 'Dosing and available formulations' above.)