2007

A Practical Approach to the Treatment of Subclinical Hypothyroidism

Points to Remember

• Subclinical hypothyroidism or mild thyroid failure is a common problem, with a prevalence of 4% to 8.5% in the adult population.
• Before diagnosing subclinical hypothyroidism, other causes of elevated serum thyroid-stimulating hormone (TSH) should be excluded.
• Many clinicians initiate thyroxine replacement therapy for all patients with TSH higher than 10 mIU/L, even if the free thyroxine level is within the normal laboratory range. However, this approach is controversial, as is whether patients with serum TSH levels between 5 and 10 mIU/L should be treated.
• To date, most studies have not shown an association of subclinical hypothyroidism with cardiac events and cardiovascular mortality.

The Challenge

Subclinical hypothyroidism occurs when the serum thyroid-stimulating hormone (TSH) level rises above the upper limit of normal (ULN) despite a normal serum free thyroxine (FT4) concentration. Subclinical hypothyroidism or mild thyroid failure is a common problem, with a prevalence of 4% to 8.5% in the adult population. The prevalence of subclinical hypothyroidism increases with advancing age and is higher in women.

Because serum TSH has a log-linear relationship with circulating thyroid hormone levels (eg, a 2-fold change in FT4 produces a 100-fold change in TSH), it is the key test for the diagnosis of subclinical hypothyroidism. Before diagnosing subclinical hypothyroidism, other causes of elevated serum TSH should be excluded. These include recovery from nonthyroidal illness, assay variability, heterophil antibodies, central hypothyroidism with biologically inactive TSH, and thyroid hormone resistance. However, the most common cause of elevated serum TSH is autoimmune thyroid disease.

Figure. Serum TSH levels in 123,958 patients aged 50 years or older seen at Mayo Clinic, 1995-1997.

Enlarge

What Is the Upper Limit of Normal for TSH?

The ULN for serum TSH is the subject of hot debate. The reference range used by Mayo Medical Laboratories is 0.3 to 5.0 mIU/L. However, data that support a move to lower the ULN of TSH to 3.0 mIU/L and possibly to 2.5 mIU/L have been published. These lower ULN cut-offs are obtained if individuals at risk of thyroid disease are excluded from the reference range population.

Should All Patients With Subclinical Hypothyroidism Be Treated With Thyroid Hormone Replacement?

There is consensus for initiating thyroxine replacement therapy in patients with TSH levels higher than 10 mIU/L, even if FT4 is within the normal laboratory range. However, this treatment approach is somewhat controversial, as is whether patients with serum TSH levels between 5 and 10 mIU/L should be treated. The argument in favor of replacement therapy is based on numerous proposed consequences of untreated subclinical hypothyroidism: progression to clinical hypothyroidism, subtle systemic symptoms of hypothyroidism, lipid abnormalities, adverse cardiac end points, cardiac dysfunction, adverse fetal effects and pregnancy outcomes, possible contribution to infertility, neuromuscular dysfunction, psychiatric dysfunction, and cognitive dysfunction.

Several investigators have demonstrated subtle cardiovascular dysfunction in patients with subclinical hypothyroidism, but the clinical significance is questionable. To date, studies have not shown an association of subclinical hypothyroidism with cardiac events and cardiovascular mortality.

Table 1. At a Glance: Controversy Over Limits

Because of the large number of individuals potentially affected, there is a need to resolve the controversy over how to treat subclinical hypothyroidism.

• Argument for lowering the ULN: The strongest argument for lowering the ULN of TSH is the higher rate of positive antithyroid antibodies (reflecting underlying autoimmune thyroid disease) for individuals with TSH concentrations between 3 and 5 mIU/L and the higher rate of progression to clinical thyroid disease in this subgroup.
• Argument against lowering the ULN: The argument against lowering the ULN for serum TSH is that 22 million to 28 million additional individuals in the United States would be considered hypothyroid if the ULN of the TSH range were decreased to 2.5 to 3.0 mIU/L. Mayo Clinic data show that lowering the ULN of the TSH reference range to 3.0 mIU/L results in a 3-fold increase in the diagnosis of hypothyroidism in patients without a history of thyroid disease. Yet there is no evidence that intervention at these levels of TSH is beneficial.

Note of caution: Some evidence shows that lowering patients' serum TSH to the proposed new normal range by adjustment of the thyroxine dose does not improve their well-being or relieve their nonspecific complaints. For patients with TSH levels between 3 and 5 mIU/L, follow-up and possible measurement of thyroperoxidase antibody may be considered.

Table 2. Clinical Approach Until More Evidence Is Available

Until guidance from carefully designed randomized trials becomes available, a practical approach is
needed. Mayo Clinic physicians generally follow these guidelines:

• Individuals with serum TSH levels between 5 and 10 mIU/L should be treated selectively.
• Thyroxine replacement therapy should be reserved for patients who have goiter, women who are anticipating pregnancy or are pregnant, or patients with depression or bipolar disorder.
• Patient preference, clinical circumstance, age, presence of symptoms of hypothyroidism, thyroperoxidase antibody positivity, and level of and progression of TSH over time should also be considered.
• Subclinical hypothyroidism associated with autoimmune thyroiditis of children and adolescents should be treated.
• Mayo data show that patients with serum TSH levels higher than 8 mIU/L have a high likelihood of progression to TSH above 10 mIU/L in 4 years and may be considered for thyroxine replacement
  therapy.
• Improvement in serum lipid levels with thyroxine replacement therapy is more likely for patients who have baseline TSH levels higher than 10 mIU/L. If hyperlipidemia is encountered in a patient with a serum TSH between 5 and 10 mIU/L, specific lipid-directed therapy or lifestyle changes are needed.