Bone disease with hyperthyroidism and thyroid hormone therapy

INTRODUCTION — The adverse effects of hyperthyroidism on the skeleton were known before the advent of satisfactory treatment for hyperthyroidism. One of the first reports of hyperthyroid bone disease was in 1891 when von Recklinghausen described the "worm-eaten" appearance of the long bones of a young woman who died from hyperthyroidism [1]. With the introduction of antithyroid drugs and radioiodine in the 1940s, clinically apparent hyperthyroid bone disease became less common. However, bone density measurements demonstrate that bone loss is common in patients with overt hyperthyroidism and to a lesser extent in those with subclinical hyperthyroidism, whether caused by endogenous hyperthyroidism or excessive doses of thyroid hormone(s) (thyroxine [T4] or triiodothyronine [T3]).

This topic will review bone disease in patients with overt and subclinical hyperthyroidism. The clinical manifestations, diagnosis, and treatment of overt and subclinical hyperthyroidism are reviewed separately. (See "Overview of the clinical manifestations of hyperthyroidism in adults" and "Diagnosis of hyperthyroidism" and "Subclinical hyperthyroidism in nonpregnant adults" and "Graves' hyperthyroidism in nonpregnant adults: Overview of treatment".)

PATHOGENESIS — Thyroid hormone directly stimulates bone resorption in organ culture [2]. This action is mediated by nuclear triiodothyronine (T3) receptors, predominantly thyroid receptor (TR)-alpha-1, which has been found in rat and human osteoblast cell lines [3-5] and in osteoclasts derived from an osteoclastoma [5]. Thyroid hormone may affect bone calcium metabolism either by a direct action on osteoclasts or by acting on osteoblasts, which in turn mediate osteoclastic bone resorption [6]. Thyroid-stimulating hormone (TSH) may also have a direct effect on bone formation and bone resorption, mediated via the TSH receptor on osteoblast and osteoclast precursors [7,8]; this putative effect is, however, controversial, since bone loss appeared independent of TSH levels in the experiments in mice with a loss-of-function TSH receptor [9].

Increased serum interleukin-6 (IL-6) concentrations in hyperthyroid patients may also play a role in thyroid hormone-stimulated bone loss [10]. IL-6 stimulates osteoclast production and may be an effector of the action of parathyroid hormone (PTH) on bone. (See "Hypercalcemia of malignancy: Mechanisms", section on 'Multiple myeloma'.)

OVERT HYPERTHYROIDISM — Overt hyperthyroidism is associated with accelerated bone remodeling, reduced bone density, osteoporosis, and an increase in fracture rate. The bone density changes may or may not be reversible with therapy. These changes in bone metabolism are associated with negative calcium balance, hypercalciuria, and rarely, hypercalcemia.

Bone density — Bone loss is a uniform feature of overt hyperthyroidism [11]. The extent of the reduction in bone density in most studies of hyperthyroid patients ranges from 10 to 20 percent. The extent of the reversibility of bone loss with therapy is unclear. Studies that have looked at changes in bone density after treatment of hyperthyroidism have yielded variable results. As examples:

●A meta-analysis of studies demonstrated improvement in bone mineral density (BMD) with treatment of hyperthyroidism [11]. Within the first year after treatment, BMD was below age-matched controls (Z-score below 0). However, BMD was similar to age-matched controls one to four years after treatment of hyperthyroidism. A subsequent study also showed complete recovery [12].

●In another study, bone density was measured using dual-energy x-ray absorptiometry (DXA) technology at baseline and 12 months after definitive management of hyperthyroidism in 50 young patients (mean age 29.4 years) with hyperthyroidism and either bone pain or an elevation in serum alkaline phosphatase [13]. At baseline, 92 percent had low bone density. One year after treatment, bone density improved by 6 percent. However, bone density remained below aged-matched controls. Several other studies have also shown incomplete recovery [14-16].

●In a cross-sectional, population-based study of 5778 women without and 252 women with a self-reported history of hyperthyroidism, bone density was measured at the ultradistal forearm using DXA technology. Women with a history of hyperthyroidism had a higher prevalence of osteoporosis (adjusted odds ratio [OR] 1.5, 95% CI 1.1-2.0) [17].

●In a study of 61 hyperthyroid women compared with matched controls, high-resolution peripheral quantitative computed tomography (HRpQCT) of the radius revealed higher trabecular and lower cortical areas, lower volumetric bone density, and reduced cortical strength, which was reversed with treatment [18].

The variability in these results is likely due to differences in bone densitometry techniques, site of measurement, patient-specific factors (eg, duration and severity of hyperthyroidism prior to treatment, form of hyperthyroidism treatment, vitamin D sufficiency), and length of follow-up [19,20]. Taken together, the above data suggest that bone loss is incompletely reversible in some patients, and therefore, a past history of hyperthyroidism should be considered during an assessment of clinical risk factors for osteoporosis. (See "Osteoporotic fracture risk assessment", section on 'Clinical risk factor assessment'.)

Histomorphometric studies of iliac crest bone biopsies show a greater deleterious effect of hyperthyroidism on cortical as opposed to trabecular bone [21,22]. In one report, hyperthyroid patients had only a 2.7 percent reduction in trabecular bone volume, but there was a 40 percent increase in osteoclast resorption surfaces in cortical bone and a 32 percent increase in cortical bone porosity [21]. There was no change in osteoid volume (ie, no osteomalacia).

Three-dimensional reconstructions of the remodeling sequence have shown how these changes occur. In the normal remodeling sequence, osteoclastic resorption and osteoblastic bone formation are synchronized. In overt hyperthyroidism, osteoclastic resorption is stimulated out of proportion to osteoblastic remineralization [23]. Both resorption and formation occur rapidly; the normal cycle duration of approximately 200 days is halved. In addition, more surfaces are resorbing and forming. But with each cycle, there is more bone resorbed than formed, and this is associated with loss of mineralized bone with each remodeling cycle. In contrast, cycle length approximates 700 days in hypothyroid patients and is associated with an increase in mineralized bone.

Fracture risk — Overt hyperthyroidism is a risk factor for fracture [11,24-26]. As examples:

●In a prospective cohort study of over 230,000 patients (9217 of whom had low thyroid-stimulating hormone [TSH] [<0.3 mU/L]) followed for a median 7.5 years, there was an increased incidence of major osteoporotic fracture (hip, clinical spine, humerus, forearm) in patients with low TSH (13.5 percent, compared with 6.9 percent in those with normal TSH) [26].

In a subgroup analysis of patients with overt hyperthyroidism, there was a significant increase in the risk of hip fracture (hazard ratio [HR] 1.14, 95% CI 1.01-1.29) but not major osteoporotic fracture.

●In a prospective cohort study of 686 White women over age 65 years followed for a mean of 3.7 years, women with serum TSH concentrations of 0.1 mU/L or less at baseline were at increased risk for both hip and vertebral fracture (relative risk [RR] 3.6 and 4.5, respectively) [27]. Exogenous thyroid hormone therapy was not a risk factor for fracture in women with normal serum TSH concentrations, but a history of hyperthyroidism was a risk factor for hip fracture, even after adjustment for serum TSH concentration and BMD. Serum thyroxine (T4) was not measured, so the proportion of women with overt and subclinical hyperthyroidism is not known.

Symptomatic bone disease — The majority of patients with hyperthyroidism and bone disease are asymptomatic unless fracture occurs. In one older study of 187 patients with hyperthyroidism, as an example, only 15 (8 percent) had symptoms, predominantly severe bone pain or fracture [21]. All were women and 80 percent were over the age of 50 years. Three-quarters had been hyperthyroid for less than a year. Radiographic studies demonstrated generalized osteoporosis with frequent vertebral compression fractures or hypertransparency of the spine in the absence of fracture.

Mineral metabolism — The increased calcium release into the circulation due to the increased bone resorption has effects on mineral metabolism, which leads to negative calcium balance in hyperthyroid patients [28].

●Hypercalcemia occurs in up to 8 percent of patients [29]. However, increases in the serum ionized calcium concentration are more common [30]. The hypercalcemia suppresses the secretion of parathyroid hormone (PTH), leading to hypercalciuria, which protects against hypercalcemia but leads to a negative calcium balance and is a potential cause of nephrolithiasis.

●Low serum PTH concentrations reduce the conversion of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D (calcitriol) [31]. The decline in calcitriol production is compounded by an increase in calcitriol metabolism induced by hyperthyroidism [32].

●Low serum calcitriol concentrations diminish intestinal calcium (and phosphorous) absorption, resulting in fecal calcium loss. Malabsorption of calcium may be aggravated by steatorrhea and increased gut motility [33].

Biochemical markers of bone and mineral metabolism are also affected. The serum concentrations of alkaline phosphatase, osteocalcin, osteoprotegerin [34], and fibroblast growth factor-23 (FGF-23) [35] are increased in overt hyperthyroidism; they may remain high for months after treatment, presumably due to a persistent increase in osteoblastic activity [28,36,37]. Urinary excretion of bone resorption markers (collagen-derived pyridinium cross-links and urinary N-telopeptide of type 1 collagen) is increased and falls to normal after treatment [28,38].

Treatment and prevention — Overt hyperthyroidism should be treated in all patients. In a case-control study from Denmark, the increased fracture risk with hyperthyroidism was reduced to a nonsignificant risk in those patients who were treated with an antithyroid drug [39]. (See "Graves' hyperthyroidism in nonpregnant adults: Overview of treatment" and "Treatment of toxic adenoma and toxic multinodular goiter" and "Exogenous hyperthyroidism", section on 'Treatment'.)

In addition to treating the hyperthyroidism, patients should be advised to have adequate amounts of calcium and vitamin D in their diets, typically 1000 to 1200 mg of calcium (total of diet and supplement) and 800 to 1000 international units of vitamin D (see "Calcium and vitamin D supplementation in osteoporosis", section on 'Optimal intake'). Calcium and vitamin D supplementation may be especially important in the setting of overt hyperthyroidism, since some data suggest that post-thyroidectomy hypocalcemia may be due in part to a pathophysiology similar to that seen in "hungry bone syndrome" following surgical correction of hyperparathyroidism [40]. (See "Hungry bone syndrome following parathyroidectomy in patients with end-stage kidney disease".)

Some patients may benefit from bisphosphonates. In two studies, there was greater improvement in bone density after resolution of the hyperthyroidism when hyperthyroid women were treated with both alendronate and methimazole versus methimazole alone [41,42]. Fracture rate was not reported.

The decision to treat with bisphosphonates should be based upon fracture risk assessment. Fracture risk is determined by a combination of BMD and clinical risk factors. Patients at highest risk for fracture are the most likely to benefit from bisphosphonate therapy. Fracture risk assessment is reviewed in detail separately. (See "Overview of the management of low bone mass and osteoporosis in postmenopausal women", section on 'Patient selection' and "Treatment of osteoporosis in men", section on 'Patient selection'.)

SUBCLINICAL HYPERTHYROIDISM — Patients with subclinical hyperthyroidism have normal serum concentrations of free thyroxine (T4) and triiodothyronine (T3) but subnormal concentrations of thyroid-stimulating hormone (TSH). (See "Subclinical hyperthyroidism in nonpregnant adults".)

Endogenous or exogenous subclinical hyperthyroidism is associated with reduced bone density, particularly in cortical-rich bone in postmenopausal women. The risk of fracture appears to be related to the degree of TSH suppression and to specific patient factors (eg, older age) that confer an increased risk of osteoporotic fracture. (See 'Fracture risk' below.)

The causes of subclinical hyperthyroidism are the same as the causes of overt hyperthyroidism, and like overt hyperthyroidism, subclinical hyperthyroidism can be persistent or transient (table 1). Common causes of subclinical hyperthyroidism include autonomously functioning thyroid adenomas and multinodular goiters (endogenous subclinical hyperthyroidism), Graves' disease (endogenous subclinical hyperthyroidism), or excessive thyroid hormone therapy (exogenous subclinical hyperthyroidism).

Endogenous subclinical hyperthyroidism: Nodular goiter and Graves' disease — Any form of hyperthyroidism can be subclinical, but this disorder most commonly occurs in older patients with a multinodular goiter or, less often, mild Graves' disease.

Bone density — Although symptomatic bone disease is not a feature of subclinical hyperthyroidism, observational data strongly suggest that this disorder has an adverse effect upon bone density and is a risk factor for osteoporosis [43].

The following studies illustrate the range of findings in women with subclinical hyperthyroidism due to nodular goiter or Graves' disease:

●In one report, women with nodular goiter and subclinical hyperthyroidism had decreased forearm bone density, which correlated inversely with serum free T4 values (although in the normal range) [44]. In another report, postmenopausal (but not premenopausal) women with nodular goiter and subclinical hyperthyroidism had reduced bone density in the radius and femoral neck but not lumbar spine [45].

●Postmenopausal women with subclinical hyperthyroidism treated with methimazole had higher distal forearm bone density as compared with untreated women [46].

●Postmenopausal women with subclinical hyperthyroidism treated with radioiodine and followed for two years did not lose bone from the spine or the hip, whereas untreated women lost bone at both sites (4.5 and 2.0 percent, respectively [p<0.02]) [47].

●In an analysis of 284 patients with subclinical hyperthyroidism from six prospective, population-based cohorts that included 5458 individuals, the annual loss of femoral neck bone density was -0.59 percent greater among those with TSH <0.1 mU/L compared with those who were euthyroid [48].

Fracture risk — Endogenous subclinical hyperthyroidism increases the risk of fracture. In a meta-analysis of 13 prospective cohort studies, compared with euthyroidism, endogenous subclinical hyperthyroidism was associated with an increased risk of hip fracture (6.1 versus 4.4 percent; hazard ratio [HR] 1.52, 95% CI 1.19-1.93), any fracture (13.9 versus 8.4 percent; HR 1.42, 95% CI 1.16-1.74), and clinical spine fracture (2.3 versus 1.2 percent; HR 1.74, 95% CI 1.01-2.99) [49]. In this analysis, lower TSH levels (<0.10 mU/L) were associated with higher fracture rates. In a prospective cohort study, for each six months in which the TSH was below 0.3 mU/L, hip fracture risk increased by 1.07 (95% CI 1.04-1.10) [26].

Subsequent studies similarly reported an association between endogenous subclinical hyperthyroidism and fracture risk:

●In a Swedish subcohort of the Osteoporotic Fractures in Men (MrOS) study (mean age 75), the 25 participants with subclinical hyperthyroidism (without levothyroxine treatment) had an increased risk for vertebral fracture compared with euthyroid individuals (31.8 versus 9.2 percent, adjusted HR 3.12, 95% CI 1.37-7.12) [50].

●In a cohort study of 10,946 individuals who experienced 3556 fractures with a median follow-up of 21 years, those with subclinical hyperthyroidism (TSH <0.56 mIU/L) had a hazard ratio for fracture of 1.34 (95% CI 1.09-1.65) [51].

Mineral metabolism — Among patients with Graves' hyperthyroidism taking an anti-thyroid drug, those with persistent subclinical hyperthyroidism (ie, the dose of antithyroid drug is insufficient to normalize serum TSH) have higher serum bone alkaline phosphatase concentrations and urinary pyridinoline excretion than those who are euthyroid [52].

Treatment — We typically treat the underlying cause of subclinical hyperthyroidism in patients at risk for skeletal complications, such as postmenopausal women with low bone density. There are no studies evaluating the long-term benefits of correcting subclinical hyperthyroidism, particularly studies with clinically important endpoints such as fracture. In two nonrandomized studies, postmenopausal women with nodular goiter and subclinical hyperthyroidism treated with antithyroid drugs or radioiodine for two years had higher bone density than similar women who were not treated [46,47]. The management of endogenous subclinical hyperthyroidism is reviewed in more detail separately. (See "Subclinical hyperthyroidism in nonpregnant adults", section on 'Endogenous subclinical hyperthyroidism'.)

In addition to treating the underlying thyroid disease, lifestyle measures (adequate calcium and vitamin D, exercise, smoking cessation, counseling on fall prevention, and avoidance of heavy alcohol use) should be adopted universally to reduce bone loss in postmenopausal women. (See "Overview of the management of low bone mass and osteoporosis in postmenopausal women", section on 'Lifestyle measures to reduce bone loss'.)

Exogenous thyroid hormone therapy — Thyroid hormone (T4 [levothyroxine]) is administered in replacement doses for the treatment of hypothyroidism and in higher doses (suppression therapy) to patients with thyroid cancer. (See "Treatment of primary hypothyroidism in adults", section on 'Dose and monitoring' and "Differentiated thyroid cancer: Overview of management", section on 'Thyroid hormone suppression'.)

Many patients treated with T4 have iatrogenic subclinical hyperthyroidism, and some have increased bone resorption and reduced bone density. Fracture risk appears to be related to the degree of TSH suppression and to specific patient factors (eg, older age) that confer an increased risk of osteoporotic fracture.

Bone density — Many cross-sectional studies and a smaller number of longitudinal studies have found that patients, particularly postmenopausal women, with exogenous subclinical hyperthyroidism can have the same reduction in bone density as occurs in patients with endogenous subclinical hyperthyroidism, and that careful adjustment of the dose of T4 can minimize this risk. However, a reduction in bone density with suppressive doses of thyroid hormone is not a uniform finding [53-60].

In a meta-analysis of 17 studies of patients with thyroid cancer taking suppressive doses of levothyroxine, postmenopausal women had a significant reduction in bone density of the lumbar spine and a nonsignificant trend in reduced bone density of the total hip, while premenopausal women had significant increases at both sites and bone density measurements in men were unchanged [61].

In premenopausal women taking T4, the annualized loss of femoral neck density is significantly correlated with dose, and bone density is not reduced if the T4 dose does not exceed approximately 2 mcg/kg body weight [62]. In contrast, postmenopausal women who are not on estrogen replacement and who are taking more than 1.6 mcg/kg body weight of T4 had significant bone loss in one study [63].

Bone geometry — In a small study of 25 pre- and 74 postmenopausal women receiving T4 suppressive therapy for thyroid cancer for more than three years, compared with 297 controls, postmenopausal patients with free T4 concentrations greater than 1.79 ng/dL had lower cross-sectional moment of inertia, cross-sectional area, section modulus, and thinner cortical thickness (but not lower bone mineral density [BMD]) at the femoral neck [64].

Fracture risk — In some [27,65], but not all [66,67], studies, patients with subclinical hyperthyroidism due to exogenous thyroid hormone therapy have an increased risk of fracture. The contradictory findings are likely due to differences in the patient populations studied and to the degree of TSH suppression. As examples:

●In a population-based study of 17,684 individuals taking T4 in Tayside, Scotland, there was no increase in osteoporotic fractures in the 3731 individuals whose TSH was low but detectable (between 0.04 and 0.4 mU/L), while those with undetectable TSH (below 0.03 mU/L) had a twofold increased risk [65].

●In another study of 686 women older than 65 years of age, women with a low TSH concentration (≤0.1 mU/L) had a nearly fourfold increased risk of hip and vertebral fracture compared with women who had normal serum TSH concentrations [27].

●In a study of 1180 patients taking T4, 59 percent had a suppressed serum TSH concentration (<0.05 mU/L) [66]. Over a five-year period, the overall fracture rate in the women over age 65 years was 2.5 percent in those with suppressed TSH values versus 0.9 percent in those with normal serum TSH values; this difference did not reach statistical significance.

●In the Cardiovascular Health Study of 4936 individuals over age 65 years, there was no association of subclinical hyperthyroidism and incident hip fracture [67].

●In a study of women on TSH-suppressive therapy for thyroid cancer, radiologic vertebral fractures were found in 44.6 percent of 83 women whose TSH target was <0.5 mU/L, compared with 4.3 percent of 46 women whose TSH target was >1 mU/L [68].

Thus, fracture risk appears to be related to the degree of TSH suppression and to specific patient factors (eg, older age) that confer an increased risk of osteoporotic fracture.

Mineral metabolism — Changes in several other measures of bone and mineral metabolism are also consistent with increased bone resorption in subclinical hyperthyroidism. As an example:

●Urinary excretion of bone collagen-derived pyridinium cross-links is increased in postmenopausal women [69].

●A negative correlation has been demonstrated between the serum osteocalcin and TSH concentrations [70].

●Serum carboxy-terminal-I-telopeptide (ICTP) concentrations are high more often than are serum osteocalcin concentrations in postmenopausal women taking suppressive doses of T4 [71].

●Serum ICTP, urine N-terminal telopeptide of type I collagen, and serum osteocalcin were elevated in estrogen-deficient postmenopausal women, but not premenopausal women, when T4 dose was carefully titrated to prevent overzealous TSH suppression in patients with thyroid cancer [72].

Prevention and treatment of reduced bone density — There are several measures that may prevent loss of bone density: titration of T4 therapy to lower doses, calcium supplementation, and inhibitors of bone resorption. (See "Subclinical hyperthyroidism in nonpregnant adults", section on 'Patients on T4 for the treatment of hypothyroidism' and "Subclinical hyperthyroidism in nonpregnant adults", section on 'Patients on suppressive T4 therapy'.)

Titration of T4 — In patients being treated for primary hypothyroidism, over-replacement of T4 should be avoided. A 1997 study demonstrated the benefit of titrating T4 dose to achieve a normal TSH. Both lumbar and femoral bone density increased, and serum osteocalcin and urinary excretion of bone collagen-derived pyridinium cross-links decreased, when the T4 dose was reduced in postmenopausal women whose initial serum TSH concentration was low [73]. (See "Treatment of primary hypothyroidism in adults", section on 'Goals of therapy' and "Treatment of primary hypothyroidism in adults", section on 'Over-replacement, reduced TSH'.)

For patients treated with thyroid hormone suppression therapy for thyroid cancer, the degree of thyroid hormone suppression should be based on the risk of disease recurrence. (See "Differentiated thyroid cancer: Overview of management", section on 'Thyroid hormone suppression'.)

Because of the adverse effects of suppressive doses of thyroid hormone on bone and the heart, the 2015 American Thyroid Association (ATA) management guidelines for thyroid nodules and differentiated thyroid cancer recommend against suppressive therapy for patients with thyroid nodules, and they suggest that patients with low-risk cancers be treated with T4 doses that target TSH values in the lower portion of the normal range [74].

Calcium supplementation — Adequate dietary calcium intake is essential to ameliorate the adverse effects of thyroid hormone on bone. In a study of 46 postmenopausal women taking suppressive doses of T4, those taking placebo had 5 to 8 percent reductions in bone density over a two-year period, while those given 1000 mg of calcium daily had no measurable bone loss [75]. (See "Calcium and vitamin D supplementation in osteoporosis".)

Inhibitors of bone resorption — Treatment with inhibitors of bone resorption may be useful in patients with continuing bone loss. In short-term studies in humans, pamidronate reduced thyroid hormone-mediated increases in measures of bone turnover [76,77]. In one study of patients with thyroid cancer taking suppressive doses of T4, bone density was not reduced as compared with normal subjects but increased with cyclic pamidronate therapy [77]. In another study, alendronate improved bone density in postmenopausal women with low BMD (T-score ≤-2.5), who were taking long-term thyroid hormone suppressive therapy for the treatment of thyroid cancer [78]. Alendronate was less effective in patients taking suppressive therapy for nine years than for those taking it three to six years. (See "Overview of the management of low bone mass and osteoporosis in postmenopausal women", section on 'Patient selection' and "Bisphosphonate therapy for the treatment of osteoporosis".)

Estrogen replacement therapy is protective when coadministered with thyroid hormone. In one study of 196 women taking thyroid hormone, significant reductions in bone density were found if the T4-equivalent dose was greater than 1.6 mcg/kg, but not at lower doses [79]. However, postmenopausal women who also were taking estrogen replacement therapy had no bone loss. The role of estrogen replacement therapy has been diminished since the Women's Health Initiative (WHI). This topic is discussed in detail elsewhere. (See "Menopausal hormone therapy: Benefits and risks".)

T4 REPLACEMENT THERAPY

Treatment of overt hypothyroidism — Bone loss would not be expected to occur when hypothyroidism is treated with oral T4 (levothyroxine) and the serum thyroid-stimulating hormone (TSH) concentration does not go below the reference range (ie, if subclinical hyperthyroidism is avoided). As examples:

●In a cross-sectional study, 50 women with primary or radioiodine-induced hypothyroidism receiving long-term therapy had no change in femoral neck or spine bone density [80].

●In a study of 44 children with congenital hypothyroidism treated and followed for an average of 8.5 years, bone mineral density (BMD) did not differ compared with that of age-matched normal subjects [81].

However, hypothyroidism is associated with an increase in bone density, and treatment with T4 may result in transient increased bone resorption and a decrease in bone density, particularly in women [63,82-84]. This finding has not been reported in men [85,86].

In one study, 26 premenopausal hypothyroid women with Hashimoto's thyroiditis were treated with an average dose of 0.111 mg per day of T4 for an average of 7.5 years [87]. Serum TSH concentrations were normal throughout the study. The mean density of the femoral neck and trochanter were significantly lower than those of age-matched controls (-5.7 and -7 percent, respectively); lumbar spine BMD was similar in the two groups. This single study suggests that T4 replacement therapy may be sufficiently nonphysiologic that it could be associated with increased bone turnover; however, these data have yet to be confirmed.

Treatment of subclinical hypothyroidism — In randomized trials of T4 replacement in predominantly postmenopausal women with subclinical hypothyroidism, there was either no reduction [88] or a very small reduction (1.2 percent after 48 weeks) [89] in bone density with normalization of serum TSH concentrations. (See "Subclinical hypothyroidism in nonpregnant adults".)

THYROID HORMONE WITHIN THE NORMAL REFERENCE RANGE — Several, but not all, cohort studies have found that low-normal thyroid-stimulating hormone (TSH) and high-normal free thyroxine (T4) levels, within their respective normal reference ranges, correlate with low bone mineral density (BMD) [57] or fracture [90]. As examples:

●In a study from South Korea of 959 postmenopausal women, TSH within the normal range was positively correlated with lumbar spine and femoral neck bone density [58].

●In another prospective study of 2374 postmenopausal women, free T4 within the normal range was negatively correlated with total hip BMD and nonvertebral fractures, as well as an increased rate of bone loss at the hip [59].

●In a study of 2957 adults over age 45 from Taiwan, there was a negative correlation between free T4 and wrist BMD in women over age 50 years [60].

●In an analysis of 56,835 individuals from several prospective cohorts, the hazard ratio (HR) for hip fracture was 1.25 (95% CI 1.05-1.49) for TSH 0.45 to 0.99 mU/L and 1.12 (95% CI 0.94-1.33) for TSH 2.50 to 3.49 mU/L [90].

●In a study of 533 postmenopausal women, TSH was positively correlated with trabecular bone score [91]. Women who fractured during a five-year follow-up period had lower mean TSH levels than those who did not fracture (1.77 versus 2.05 mU/L).

SUMMARY AND RECOMMENDATIONS

●Overt hyperthyroidism – Overt hyperthyroidism is associated with accelerated bone remodeling, reduced bone mineral density (BMD), osteoporosis, and an increase in fracture rate. Hyperthyroidism should be treated in all patients. (See 'Overt hyperthyroidism' above and "Graves' hyperthyroidism in nonpregnant adults: Overview of treatment" and "Treatment of toxic adenoma and toxic multinodular goiter" and "Exogenous hyperthyroidism", section on 'Treatment'.)

●Subclinical hyperthyroidism – Endogenous or exogenous subclinical hyperthyroidism is associated with reduced bone density, particularly in cortical-rich bone in postmenopausal women. The risk of fracture appears to be related to the degree of thyroid-stimulating hormone (TSH) suppression and to specific patient factors (eg, older age) that confer an increased risk of osteoporotic fracture. (See 'Subclinical hyperthyroidism' above.)

•Endogenous – We typically treat the underlying cause of endogenous subclinical hyperthyroidism in patients at risk for skeletal complications, such as postmenopausal women with low bone density. (See 'Endogenous subclinical hyperthyroidism: Nodular goiter and Graves' disease' above and "Subclinical hyperthyroidism in nonpregnant adults", section on 'Management'.)

•Exogenous – Thyroid hormone (T4 [levothyroxine]) is administered in replacement doses for the treatment of hypothyroidism and in higher doses (suppression therapy) to patients with thyroid cancer. In patients being treated for primary hypothyroidism, over-replacement of T4 should be avoided. The risk of osteoporosis in postmenopausal women taking suppressive doses of T4 (levothyroxine) for thyroid cancer can be minimized by treatment with doses of T4 that yield serum TSH levels within the reference range in low-risk cancer patients and by institution of bisphosphonate therapy where indicated. (See 'Prevention and treatment of reduced bone density' above and "Differentiated thyroid cancer: Overview of management", section on 'Thyroid hormone suppression' and "Overview of the management of low bone mass and osteoporosis in postmenopausal women", section on 'Patient selection'.)

●Thyroid hormone replacement therapy – Appropriate replacement therapy for hypothyroidism with avoidance of TSH suppression probably does not adversely alter skeletal metabolism. (See 'T4 replacement therapy' above.)