Division of Endocrinology, Diabetes and Bone Diseases, Mount Sinai School of Medicine, New York, New York 10029
Postpartum thyroiditis is a common thyroid disorder that presents during the first postpartum year. It is the occurrence of either transient hyperthyroidism, transient hypothyroidism, or transient hyperthyroidism followed by transient hypothyroidism (Fig. 1). Most, but not all, women are euthyroid 1 yr postpartum.
|
Postpartum thyroiditis is an exacerbation of an underlying autoimmune thyroiditis, aggravated by the immunological rebound that follows the partial immunosuppression of pregnancy (1, 2, 3). Women who express human leukocyte antigen haplotypes DR-3, DR-4, and DR-5 have an increased risk for postpartum thyroiditis. In women who develop postpartum thyroiditis, CD4+/CD8+ and CD4 + 2H4+ ratios are elevated throughout pregnancy and the postpartum (4). Histologically, thyroid aspirates reveal either a lymphocytic infiltrate or diffuse destruction, changes similar to those seen in both Hashimotos thyroiditis (2, 3, 5) and painless sporadic silent thyroiditis (6). In essence, the immunological rebound that follows the end of pregnancy precipitates the clinical expression of Hashimotos disease, which before pregnancy was clinically silent (2).
Prevalence
The prevalence of postpartum thyroiditis varies widely. Although geographic differences account for some of the variance, a critical factor has been methodological differences in study design. Not only has every prevalence study used a different methodology, but few studies performed their initial screen during pregnancy. Consequently, some of the women diagnosed with postpartum thyroiditis may have had preexisting Hashimotos disease. In calculating the prevalence rate of postpartum thyroiditis, only the 13 studies that met the following criteria were included: 1) follow-up extended until a minimum of 5 months postpartum, and 2) patients were evaluated on at least two distinct time points postpartum. All 13 studies were prospective, with screening for postpartum thyroiditis performed at predetermined time periods, dictated by study protocol.
The prevalence of postpartum thyroiditis ranges between 1.1% and 16.7% (1, 3, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15), with a mean prevalence rate of 7.2%. In individuals with type 1 diabetes mellitus, the prevalence rate triples (16, 17). Specifically, in the New York metropolitan area, the general prevalence of postpartum thyroiditis of 8.8% increased to 25% in women with type 1 diabetes (16). The increased prevalence of postpartum thyroiditis in women with type 1 diabetes mellitus is not surprising, because both are autoimmune disorders. Furthermore, prior studies have demonstrated an increased prevalence of thyroid antibodies in individuals with type 1 diabetes.
Postmiscarriage thyroiditis also occurs, but no prevalence data are available (18, 19). Furthermore, as in type 1 diabetes, I postulate that women with other autoimmune diseases, such as systemic lupus erythematosus, rheumatoid arthritis, and scleroderma would have an increased prevalence of postpartum thyroiditis. However, no studies to date have evaluated the prevalence of postpartum thyroiditis in these disorders.
The recurrence rate in postpartum thyroiditis is high. Seventy percent of women with a prior episode of postpartum thyroiditis develop a recurrence in the subsequent pregnancy. Antithyroid peroxidase-positive women, who did not develop postpartum thyroiditis during an initial pregnancy, have a 25% chance of having postpartum thyroiditis after the next pregnancy. In contrast, women who had neither antithyroid peroxidase nor postpartum thyroiditis during an initial pregnancy, do not develop postpartum thyroiditis in future pregnancies (20).
Presentation
A total of 371 episodes of postpartum thyroiditis are described in the 13 prevalence studies. The most common presentation, found in 43% of the cases, was hypothyroidism without preceding hyperthyroidism. Hyperthyroidism, by itself, was present in 32% of the reported cases. Least common, comprising 25% of the episodes, was hyperthyroidism, followed by hypothyroidism.
Symptoms and diagnosis
Symptoms can occur during either phase of postpartum thyroiditis. The hyperthyroid symptoms are typically subtle. It is often diagnosed retrospectively, with the symptoms recognized at the time of diagnosis of hypothyroidism. The hypothyroid phase can also go unrecognized. Consequently, an undetermined percentage of women with postpartum thyroiditis remain undiagnosed. This is due to a woman experiencing minimal or no symptoms, the attribution of symptoms to the demands of child rearing, or a lack of recognition of the symptoms by the physician.
The hyperthyroid phase of postpartum thyroiditis occurs between 2 and 10 months postpartum. Most commonly, it presents at 3 months. Symptoms more common in women with hyperthyroid postpartum thyroiditis include palpitations, fatigue, heat intolerance, and irritability/nervousness (Refs. 1, 10, 21, 22 ; Table 1). The frequency of asymptomatic hyperthyroidism is 33% (10). Untreated, the hyperthyroidism resolves spontaneously within 23 months. This phase is diagnosed by the combination of a low serum TSH concentration in the presence of thyroid peroxidase antibodies, in women who are TSH receptor antibody-negative. Free T4 levels are typically elevated but may be normal.
|
|
Postpartum depression
Studies evaluating the association between postpartum thyroiditis and postpartum depression have yielded varying, and at times unexpected, results. Investigations have evaluated both the impact of hypothyroidism and the presence of thyroid antibodies (antithyroid peroxidase and/or antithyroglobulin) on the development of postpartum depression. Two studies reported an association between hypothyroidism and postpartum depression (21, 23), whereas two other reports discovered an increased rate of depression in euthyroid women who were thyroid antibody-positive postpartum (24, 25). One study found no correlation (26). There is no association between postpartum thyroiditis and postpartum psychosis (27).
Permanent hypothyroidism
Although most women with postpartum thyroiditis are euthyroid by 1 yr postpartum, long-term follow-up reveals an increased prevalence of permanent hypothyroidism. Prospective studies have shown a prevalence rate of hypothyroidism of 23% and 29% at 3.5 to 8.7 yr postpartum (28, 29). Progression to permanent hypothyroidism was more common in women who presented with higher TSH levels and higher titers of thyroid peroxidase antibodies in the hypothyroid phase of postpartum thyroiditis (12). The development of permanent hypothyroidism provides further evidence that postpartum thyroiditis is the clinical presentation of preexisting subclinical Hashimotos disease (28).
Treatment
Hyperthyroidism.
There have been no studies assessing the benefit of treating the hyperthyroid phase of postpartum thyroiditis. Subclinical hyperthyroidism, however, adversely affects quality of life as reflected by an increase in the Symptoms Rating Scale (30). Treatment of hyperthyroidism, when necessary, is based on symptom severity and should be a joint decision of patient and physician. Beta-blockers are given to alleviate palpitations, irritability, and nervousness. The morbidity associated with treatment are the side effects of beta-blockade. The downside of withholding treatment is allowing the woman to remain symptomatic. Antithyroid medicines (thioureas) are not a potential treatment alternative, because the hyperthyroidism is caused by a destructive thyroiditis resulting in release of preformed thyroid hormone.
Regarding treatment, my philosophy is that symptomatic thyrotoxic women should receive beta-blockers (Fig. 2). Propranolol is the treatment of choice, because it allows for easy titration to a dose that alleviates palpitations, irritability, and nervousness. Therapy is usually for less than 3 months and is adjusted on the basis of thyroid hormone levels and symptoms. The Food and Drug Administration classifies the use of propranolol during lactation as acceptable.
|
The percentage of patients treated with levothyroxine (L-T4) for the hypothyroid phase of postpartum thyroiditis in the 13 studies of postpartum thyroiditis varied from 1373%, with a mean of 34%. Because none of the 13 studies were treatment trials, the indications for initiating therapy varied. In a recent survey of internists and endocrinologists, indications for treating the hypothyroid phase of postpartum thyroiditis included symptomatic hypothyroidism and a serum TSH concentration over 10 µU/ml (31). The potential benefits of treating subclinical hypothyroidism include alleviation of hypothyroid symptoms including dry skin, cold intolerance, easy fatigability, and impaired cognitive function (32), as well as enhanced fertility in women with ovulatory dysfunction (33). Women who become pregnant while in the hypothyroid phase of postpartum thyroiditis require therapy because pregnant women with subclinical hypothyroidism have an increased miscarriage rate (34). Furthermore, the children of mothers with low normal T4 levels in the first trimester, or an elevated TSH early in the second trimester, have decreased intellectual function (35, 36). Consequently, it seems more than prudent to treat women who are hypothyroid in the first trimester of pregnancy.
The treatment of hypothyroid postpartum thyroiditis remains controversial. When, and if, to initiate treatment is an unresolved issue. Most clinicians would treat postpartum women with symptomatic hypothyroidism. However, the importance of treating asymptomatic women with an elevated TSH is unclear. There is also no consensus regarding duration of therapy. In hypothyroidism, the options are to wean the L-T4 612 months after initiation, or maintain L-T4 therapy until the woman has completed her family.
In hypothyroid women, my philosophy is to initiate L-T4 therapy in symptomatic women, women attempting pregnancy, or if the TSH level exceeds 10 µU/ml. The initial dose of L-T4 depends on the TSH level, but typically is 50 µg. Therapeutic adjustments are made subsequently based on symptoms and the TSH level. Once instituted, I maintain treatment until 1 yr after a woman completes her family. During the intervening pregnancies, careful monitoring of TSH levels should be performed while the woman is on L-T4, and adjustments should be made as needed. One year after the final pregnancy, L-T4 should be decreased by 50%, followed 6 wk later by a TSH measurement. If euthyroid, thyroid replacement is discontinued, and a repeat TSH is obtained in 6 wk. In women successfully weaned from L-T4, an annual serum TSH should be obtained to monitor for long-term hypothyroidism (Fig. 2).
Screening
Controversy surrounds whether or not to screen for postpartum thyroiditis and what the optimal strategy should be (37). During pregnancy, screening for thyroid peroxidase antibodies identifies women with an 11-fold increased risk of developing postpartum thyroiditis. TSH measurements in the postpartum will identify women who have developed postpartum thyroiditis. There are almost 4 million births in the United States annually. Screening all women has important health care, work force, fiscal, and psychological ramifications. A decision to screen must evaluate the clinical impact of postpartum thyroiditis and take into account the recent literature on 1) the effect of subclinical hypothyroidism during pregnancy on both pregnancy loss and the intellectual development of the unborn child, and 2) the correlation between thyroid antibodies and spontaneous miscarriage.
A decision regarding screening must evaluate the following issues: 1) Is postpartum thyroiditis prevalent enough to warrant screening? 2) Are there important morbidities associated with postpartum thyroiditis? 3) Can the complications of postpartum thyroiditis be prevented? and 4) Is there an economical, widely available screening method that has high sensitivity and specificity?
The answers to the first three questions are straightforward. Postpartum thyroiditis is a common disease with important clinical consequences that could be treated with L-T4. These include the symptoms the mother experiences (1) 10, 21, 22 , the negative impact of hypothyroidism on fertility (33), the effect of hypothyroxinemia (first trimester) and/or elevated TSH levels (early second trimester) on the intellectual development of the unborn child in a subsequent pregnancy (35, 36), the increased miscarriage rate in women with subclinical hypothyroidism (34), and the possibility of hypothyroidism being undetected for years.
Antithyroid peroxidase antibody is the best available screening tool for postpartum thyroiditis. It is widely available, economical, and reproducible. Studies evaluating thyroid peroxidase as a screening tool have revealed a sensitivity of 0.460.89, with a specificity of 0.910.98. The positive predictive value has varied from 0.400.78 (38). The wide variation in results reflects the different laboratory assays used and the decrease in thyroid peroxidase that occurs during pregnancy, often to undetectable levels, followed by a postpartum rebound.
The effect of maternal subclinical hypothyroidism (either from unresolved postpartum thyroiditis or permanent hypothyroidism) on the fetus and the association of thyroid antibodies with miscarriage impacts on the screening debate. Low normal T4 levels in the first trimester are associated with a lower score on the Bayley Psychomotor Developmental Index (36), and early second semester hypothyroidism has been linked to a decreased IQ of seven points in the unborn child (35). Second semester hypothyroidism is associated with increased fetal death (34). Studies in the last decade have also revealed a doubling to quadrupling of the first trimester miscarriage rate in thyroid peroxidase women, irrespective of their thyroid hormonal status (39, 40).
A recently published article presented three opinions regarding screening for postpartum thyroiditis, ranging from universal screening, to no screening, to selective screening in high-risk patients (37). Universal screening was recommended based on the assumption that treatment of postpartum thyroiditis would improve maternal quality of life. The argument against screening cited the lack of consensus regarding an optimal screening strategy and the lack of data demonstrating that treatment is beneficial. The position for selective screening focused on the high prevalence of postpartum thyroiditis in select subgroups.
A cost analysis screening study for postpartum thyroiditis compared no screening, with first trimester thyroid peroxidase screening, to TSH screening at 6 wk postpartum (41). Women who tested positive for thyroid peroxidase had TSH levels performed at 6 wk and 34 months postpartum. Assuming a 5% prevalence of postpartum thyroiditis, the study concluded that screening for thyroid peroxidase at 6 wk was cost-effective at $60,000 per quality-adjusted life-years. This study does not, however, evaluate the impact of subclinical hypothyroidism that remains undetected in the first trimester of pregnancy.
Guidelines from professional organizations regarding screening vary and are evolving. Universal screening for postpartum thyroiditis is not recommended by any national organization. The American College of Obstetricians and Gynecologists does not recommend thyroid screening postpartum (42). The American Thyroid Association states that the decision regarding screening in women of childbearing age should be made jointly between each physician and patient (43), and the American Association of Clinical Endocrinologists recommends postpartum screening in women known to have high titers of thyroid peroxidase antibody (44).
Only the PostGravid Health Care group in Osaka, Japan, recommends that all women be screened (Nobuyuki Amino, personal communication). Specifically, the PostGravid Health Care group recommends that all women be evaluated for thyroid peroxidase antibodies in early pregnancy. Thyroid antibody-positive women are followed closely for the development of postpartum thyroiditis, with thyroid function tests performed at 3 and 6 months postpartum.
The complexities associated with universal screening for postpartum thyroiditis are well articulated in the Guide to Clinical Preventive Services (45). The Guide to Preventive Services concludes that although there is insufficient evidence to recommend screening all pregnant women, an argument could be made for universal screening based on both the prevalence of the disease and the possibility that symptoms could be overlooked in the postpartum period (45).
Given the potential benefit of treating hypothyroidism and the importance of achieving euthyroidism before and during the subsequent pregnancy, I recommend selective screening for women at high risk for postpartum thyroiditis. Specifically, women with 1) type 1 diabetes mellitus, 2) a prior episode of postpartum thyroiditis, 3) a history of being thyroid peroxidase antibody-positive, or 4) a prior miscarriage should be screened at 3 months postpartum. Women who develop postpartum depression, women with other autoimmune disorders, and women with a strong family history of autoimmune thyroid disease should also be screened. Screening should include an antithyroid peroxidase antibody titer and TSH level. Women who are euthyroid and antithyroid peroxidase antibody-negative require no further follow-up. Antithyroid peroxidase antibody-positive women should have a serum TSH performed at 6 and 9 months postpartum.
Acknowledgments
I am indebted to Dr. Terry F. Davies for his thoughtful critique of this manuscript.
Footnotes
Address all correspondence and requests for reprints to: Alex Stagnaro-Green, M.D., Associate Dean for Curriculum and Faculty Development, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, 185 South Orange Avenue, MSB C644, Newark, New Jersey 07103-2714.
Abbreviation: L-T4, Levothyroxine.
Received April 3, 2002.
Accepted May 31, 2002.
References