INTRODUCTION — Although thyroid disease is common in the general population, there appears to be an increased frequency of subclinical and, less frequently, overt hypothyroidism in association with many autoimmune rheumatic diseases. The associations of thyroid and connective tissue diseases and data concerning the presence of antinuclear antibodies in patients with thyroid disease are reviewed here. The individual thyroid and connective tissue diseases are reviewed in detail in the individual topics.
FREQUENCY OF ASSOCIATION — The frequency of thyroid disease, particularly chronic autoimmune thyroiditis (Hashimoto's thyroiditis), may be increased in patients with connective tissue diseases. Similarly, the frequency of connective tissue disease in patients with autoimmune thyroiditis may also be increased. In a study of 3069 patients with autoimmune thyroiditis, there was a significantly increased prevalence of rheumatoid arthritis (2.4 percent), polymyalgia rheumatica (1.4 percent), Sjogren's syndrome (0.9 percent), psoriatic arthritis (0.6 percent), systemic sclerosis (0.5 percent), and systemic lupus erythematosus (0.8 percent) in patients with thyroiditis compared with controls [1].
However, it is important to remember that chronic autoimmune thyroiditis, manifested by positive tests for antithyroid antibodies, is common in the general population (occurring in 10 to 20 percent of women and 1 to 2 percent of men) and hypothyroidism (subclinical or overt) is only slightly less common [2]. In addition, both hypothyroidism and connective tissue diseases often cause muscle and joint aches, pains, and stiffness, and common treatments for connective tissue diseases and the illnesses themselves may affect thyroid function or thyroid function tests.
Given the high frequency of chronic autoimmune thyroiditis and hypothyroidism in patients with no other problems, the literature describing the occurrence of thyroid disease in patients with a connective tissue disease should be viewed with considerable caution. Most of the studies were performed in specialty clinics, so the likelihood of selection bias is high. Diagnostic criteria for thyroid disease may also vary between studies. Furthermore, there is only limited evidence that the concurrence in the same patient of thyroid disease and connective tissue disease alters the clinical manifestations or natural history of either disorder. Examples are scleroderma, in which fibrosis of the thyroid gland can occur and cause hypothyroidism independently of chronic autoimmune thyroiditis, and complete congenital heart block in children of hypothyroid, anti-Ro-positive mothers.
Another part of this relationship is that the drugs used to treat connective tissue diseases and the underlying illness can affect thyroid function. (See "Drug interactions with thyroid hormones" and "Thyroid function in nonthyroidal illness".)
●Glucocorticoids inhibit thyroid-stimulating hormone (TSH) secretion and slightly reduce serum thyroid hormone concentrations.
●Aspirin and some nonsteroidal anti-inflammatory drugs lower serum total thyroid hormone concentrations by interfering with thyroid hormone binding to its carrier proteins; serum free thyroid hormone concentrations do not change.
●Any major illness, including connective tissue diseases, can lower serum thyroid hormone and TSH concentrations.
THYROID DISORDERS IN CONNECTIVE TISSUE DISEASES
Polymyalgia rheumatica/giant cell arteritis — The data concerning the prevalence of hypothyroidism in patients with polymyalgia rheumatica and giant cell arteritis (GCA) have been conflicting. Some older reports have described an increased prevalence [3,4], while others have questioned the association [5,6]. More modern day studies suggest an increased frequency of thyroid disease in patients with GCA, particularly in women. As examples:
●In a large health care database study of 5663 patients with GCA, compared with 23,308 matched controls, hypothyroidism was significantly increased in patients with GCA (18.2 versus 6.91 percent in controls [7].
●In a population-based cohort study of 761 Swedish patients (571 women) with biopsy-proven GCA, there was an increased risk of thyroid disease in women with GCA, but not men [8].
Polymyalgia rheumatica/GCA may be more common in patients with autoimmune thyroid disease. One series found a prevalence of 2.8 percent in 250 patients with autoimmune thyroid disease, rising to 9.3 percent in women over the age of 60 years; in comparison, there were no cases of polymyalgia rheumatica/GCA among 150 control patients [9].
Fibromyalgia — Nonspecific rheumatic symptoms such as polyarthralgia, myalgia, and fibromyalgia appear more common in people with Hashimoto's thyroiditis. In a study comparing 500 patients with Hashimoto's thyroiditis with 310 controls, 20 percent of the patients had nonspecific rheumatic symptoms compared with 6.8 percent of the controls. The presence of symptoms did not appear to be related to thyroid dysfunction [10].
A blunted response of serum TSH to thyrotropin-releasing hormone (TRH) has been described in a small number of patients with fibromyalgia [11,12]. The importance of this finding is unclear. It could be due to an increase in cortisol secretion, due either to the fibromyalgia itself or to the depression that often accompanies it. (See "Pathogenesis of fibromyalgia".)
Another small study showed an increased prevalence of antithyroid antibodies in patients with fibromyalgia compared with controls and similar to the prevalence seen in patients with rheumatoid arthritis [13]. Sicca symptoms seemed more frequent in these patients.
Systemic lupus erythematosus — Approximately 15 to 30 percent of patients with systemic lupus erythematosus (SLE) have antithyroid antibodies (most often measured as antithyroid peroxidase [anti-TPO; microsomal] antibodies [14-20]), a slightly higher percentage than is found in normal subjects of the same age [14]. The antibody titers may fluctuate, and clinical thyroid disease is associated with the persistent presence of these antibodies [16]. In terms of thyroid disease, the following observations have been noted:
●The prevalence of overt thyroid disease appears to be between 3 and 19 percent [14,16-23]. In a meta-analysis of 10 studies totaling 10,500 patients with SLE patients and 44,170 healthy controls, subclinical hypothyroidism and overt hypothyroidism were more common in patients with SLE (odds ratio [OR] 5.67 and 2.93, respectively), but the prevalence of subclinical and overt hyperthyroidism was not significantly different between patients with SLE and controls [24].
In one study, thyroid dysfunction was not associated with the duration of SLE [19]. In other studies, older age, smoking, overlapping connective tissue diseases (Sjogren's, rheumatoid arthritis, dermatomyositis, and polymyositis), and anticyclic citrullinated peptide (CCP) antibody positivity were associated with autoimmune thyroid disease [20,25].
●Isolated low triiodothyronine (T3) has been observed in 4 to 8 percent of patients with SLE [18,19].
●In the study of 153 SLE patients, papillary thyroid cancer was found in five of the SLE patients compared with one of the 459 controls, with a relative risk (RR) of 14.5 [18]. The presence of thyroid cancer did not appear to be related to the treatment of SLE, but thyroid autoimmunity may be a predisposing condition for thyroid cancer in this group of patients. An uncontrolled ultrasound study of 55 lupus patients found a prevalence of thyroid nodules in 27 percent. This finding was associated with prior azathioprine treatment. None were malignant [26].
●In a study of 3286 people with Graves' disease (2791 cases) or Hashimoto's thyroiditis (495 cases), there was a significantly increased RR of SLE in both diseases, with the RR in women being over 10-fold [27]. There was also an increased risk of SLE noted in the men with Graves' disease.
●Part of the association between SLE and thyroid disease may be due to genetic factors. In a study of 378 multiplex SLE families, 35 also had autoimmune thyroid disease [28]. A genetic linkage effect in the region of D5S1462 on the chromosome 5q14.3-15 was demonstrated between these two related autoimmune conditions. In a study of families with more than one member affected by SLE, family members with SLE and secondary Sjögren's syndrome had a higher prevalence of autoimmune thyroid disease than SLE family members without secondary Sjögren's syndrome (30 versus 13 percent) [29].
●An increased risk of preterm delivery and other pregnancy complications has been noted in pregnant women with SLE and thyroid disease [30,31]. In a retrospective study of 63 pregnant women with SLE, 37.9 percent of the cohort had thyroid disease diagnosed before, during, or immediately after pregnancy [30]. Preterm delivery occurred in 67 percent of the women who had thyroid disease, compared with 18 percent of the women with SLE who remained free of thyroid disease. Thyroid antibodies did not predict preterm delivery in this group of women, but in another study, the presence of connective tissue disease and thyroid autoantibodies was independently associated with increased risks of pregnancy complications (miscarriage, fetal growth restriction, preeclampsia, preterm delivery) [31]. (See "Pregnancy in women with systemic lupus erythematosus", section on 'Preterm birth'.)
Sjögren's syndrome — Estimates of the prevalence of thyroid disease in patients with Sjögren's syndrome vary widely, ranging from 10 to 70 percent. Some of the variance may be accounted for by different criteria for thyroid disease (eg, abnormalities in thyroid structure, hormonal function, or presence of autoantibodies). Observations of a higher prevalence of thyroid disease in this population may also represent confounding factors as demographic features that favor Sjögren's syndrome also increase the risk of thyroid disease (eg, female sex and advancing age). However, a study of 114 patients with primary Sjögren's syndrome showed 14 percent to have hypothyroidism, while thyrotoxicosis had occurred in 1.8 percent of the same cohort [32] (see "Clinical manifestations of Sjögren's syndrome: Extraglandular disease"). Autoimmune disease, including autoimmune thyroid disease, has been shown to cluster in first-degree relatives of patients with Sjögren's syndrome [33].
On the other hand, symptoms of Sjögren's syndrome may occur with increased frequency in patients with autoimmune thyroiditis. As an example, 32 percent of 19 patients with chronic autoimmune thyroiditis had keratoconjunctivitis sicca and xerostomia [34].
In a study analyzing the risk of cancer in a cohort of 1300 patients with Sjogren's syndrome, there was an increased risk of hematological malignancy (standard incidence ratio [SIR] 11.02), and of thyroid cancer (SIR 5.17) [35].
Rheumatoid arthritis — In a retrospective cohort study of 1035 patients with rheumatoid arthritis, the prevalence of thyroid disease was 11.8 percent and the presence of thyroid disease was associated with a poorer initial response to treatment of the rheumatoid arthritis [36]. Another smaller study by the same authors in newly diagnosed patients with rheumatoid arthritis found that the presence of thyroid disorders was 15.7 percent and significantly associated with female sex, antinuclear antibody (ANA) positivity, high levels of anti-CCP antibodies, and a poorer response to treatment in the first 4 months [37]. The mechanism of poor response may be due to the association of thyroid disease with high levels of anti-CCP antibodies.
In one study from Norway, the frequency of antithyroid antibodies was increased in patients with rheumatoid arthritis [38]. In contrast, a study from the United States showed a similar prevalence of antithyroid antibodies in 1290 patients with rheumatoid arthritis and 1236 controls (15.6 versus 15.9 percent) [39]. In a large, population-based study of 1998 cases and 2252 controls, T4 (levothyroxine) replacement was associated with a twofold risk of both anti-citrullinated peptide antibody (ACPA)-positive and negative rheumatoid arthritis [40]. For ACPA-positive rheumatoid arthritis, the risk associated with the combination of T4 substitution and the shared epitope alleles HLA-DRB1 was higher than that associated with either alone.
Psoriatic arthritis — In a large cohort study from Taiwan, both psoriasis and psoriatic arthritis were associated with an increased incidence of thyroid disease (hyperthyroidism, Graves' disease, hypothyroidism, thyroiditis, and Hashimoto's thyroiditis) compared with controls [41]. In other studies, antithyroid antibodies and subclinical hypothyroidism were found more frequently in women with psoriatic arthritis than controls [42]. Men with psoriatic arthritis had an increased prevalence of antithyroid antibodies alone. Subclinical hypothyroidism occurred in 25 percent of women with psoriatic arthritis, compared with 8 percent of controls and 12 percent of women with rheumatoid arthritis, and appeared to be associated with polyarticular involvement and longer disease duration.
In a prospective study of 97 euthyroid patients with psoriatic arthritis compared with 97 matched controls, there was an increased incidence of clinical and subclinical hypothyroidism over a median follow-up of 74 months in patients with psoriatic arthritis, particularly in women [43]. At baseline, there was a higher level of both antithyroid peroxidase and antithyroglobulin antibodies in the patients with psoriatic arthritis.
Ankylosing spondylitis — A study of 80 patients with ankylosing spondylitis found 10 percent had Hashimoto's thyroiditis, defined as positive thyroid autoantibodies plus at least one of the following features (diffuse goiter, abnormal thyroid function tests, parenchymal heterogeneity on ultrasound), compared with 1.3 percent of healthy controls. As only one ankylosing spondylitis patient had subclinical hypothyroidism, with the rest being euthyroid, the clinical significance is uncertain [44].
Mixed connective tissue disease — There are limited data on the frequency of thyroid disease in mixed connective tissue disease. One report of 22 patients with this disorder found antithyroid antibodies in 23 percent and hypothyroidism in 16 percent [45]. A larger study of 159 patients found a similar prevalence of antithyroid antibodies, present in 21 percent of patients, all of whom required T4 treatment [46].
Scleroderma — Hypothyroidism occurs more frequently in women with scleroderma than in control populations [47,48]. As mentioned above, fibrosis of the thyroid gland can directly cause hypothyroidism in patients with scleroderma and appears to be responsible for most cases [48-50]. This was illustrated in a study of 202 patients with scleroderma compared with population-based controls [48]. Clinical (4 versus 0.3 percent) and subclinical (17 versus 6 percent) hypothyroidism was more common in female scleroderma subjects than controls. Anti-TPO antibodies were not significantly different between hypothyroid and non-hypothyroid scleroderma subjects.
However, in a prospective study of 179 patients with scleroderma, baseline TSH was higher and anti-TPO antibodies were more likely to be present in scleroderma patients than controls, as was a small thyroid volume and hypoechoic pattern on ultrasound [51]. After a mean follow-up of 73 months, new thyroid dysfunction (defined as clinical or subclinical hypothyroidism or hyperthyroidism) occurred in 23 of 179 patients compared with 7 of 179 in the control group [51]. An increased risk of hyperthyroidism has also been reported in other studies [52].
In a study of 138 patients with scleroderma, anti-TPO antibodies were increased in patients with limited scleroderma, whereas the levels of these antibodies in patients with diffuse scleroderma were similar to those in healthy controls [53]. An ultrasound study of thyroid volume in a series of 106 Indian patients with systemic sclerosis (94 percent diffuse) showed thyroid volumes consistent with atrophy in 57.5 percent, although the prevalence of subclinical and overt hypothyroidism was low (8.5 and 1.9 percent, respectively) [54]. These findings support the hypothesis that hypothyroidism in limited scleroderma may reflect thyroid autoimmunity, while fibrosis may be the mechanism of hypothyroidism in diffuse scleroderma.
In an Italian series of 327 patients with systemic sclerosis (52 percent diffuse), there was an increased prevalence of anti-TPO and antithyroglobulin antibodies in comparison with controls [55]. There was a significant increase in papillary thyroid cancer in patients with systemic sclerosis. All the patients with systemic sclerosis and papillary thyroid cancer had evidence of thyroid autoimmunity.
Antinuclear antibodies in patients with thyroid disease — Antinuclear antibodies have been detected in the serum of patients with both Graves' disease and chronic autoimmune thyroiditis in some, but not all, studies [56-58]. In a study including 22 patients with Graves' hyperthyroidism and 28 with chronic autoimmune thyroiditis, 13 (26 percent) had antinuclear antibodies and 17 (34 percent) had anti-single-stranded DNA antibodies (the percentages were approximately the same in each patient group) [58]. No patient had antibodies against double-stranded DNA, extractable nuclear antigens (anti-Ro/SSA, anti-La/SSB, anti-Sm, and anti-RNP), mitochondria, or rheumatoid factor. A study in a Middle Eastern population followed 61 patients (2 with Graves' disease, 59 with Hashimoto's thyroiditis) in comparison with 61 healthy controls [59]. At baseline, patients with thyroid disease had significantly higher levels of ANA, anti-Ro, anti-La, anti-CCP, and anti-double-stranded DNA antibodies and rheumatoid factor compared with controls. After two years of follow-up, 14.8 percent of patients had developed rheumatoid arthritis, 3.3 percent Sjogren's syndrome, and 6.6 percent SLE. No information was given on the frequency of these diseases in the control group.
Maternal anti-Ro antibodies are associated with complete congenital heart block in infants. A study showed that maternal hypothyroidism was more common in women with anti-Ro antibodies and was associated with an increased risk of a child with complete congenital heart block compared with anti-Ro positive women with normal thyroid function (56 versus 13 percent) [60].
SUMMARY
●Although thyroid disease is common in the general population, there appears to be an increased frequency of subclinical and, less frequently, overt hypothyroidism in association with many autoimmune rheumatic diseases. The frequency of hyperthyroidism in these patients is less than hypothyroidism and appears to be similar or only slightly increased compared to that seen in the general population. It is reasonable to assess thyroid function tests in most patients on presentation and periodically thereafter. (See 'Thyroid disorders in connective tissue diseases' above.)
●There is limited evidence that the concurrence in the same patient of thyroid disease and connective tissue disease alters the clinical manifestations or natural history of either disorder. In patients with scleroderma, fibrosis of the thyroid gland can occur and cause hypothyroidism independently of chronic autoimmune thyroiditis. In addition, an increased risk of preterm delivery has been noted in pregnant women with systemic lupus erythematosus (SLE) and thyroid disease, and the presence of anti-Ro antibodies in hypothyroid mothers is associated with complete congenital heart block in infants. (See 'Scleroderma' above and 'Systemic lupus erythematosus' above and 'Antinuclear antibodies in patients with thyroid disease' above.)
●An association between polymyalgia rheumatica and thyroid disease remains controversial. There appears to be an increased frequency of thyroid disease in patients with giant cell arteritis (GCA), particularly in women. (See 'Polymyalgia rheumatica/giant cell arteritis' above.)
●Although there is no clear association between fibromyalgia and thyroid dysfunction, the symptoms of fibromyalgia and hypothyroidism can be similar. Thus, it is important for fibromyalgia patients to have routine thyroid function tests as part of their initial clinical screen. (See 'Fibromyalgia' above.)
●Pregnant women with anti-Ro antibodies should be screened for hypothyroidism as there is some evidence to suggest screening may identify those at a greater risk of having a child with congenital complete heart block. (See 'Antinuclear antibodies in patients with thyroid disease' above.)
