Serum TSH concentration — First-generation thyroid-stimulating hormone (TSH) radioimmunoassays (RIAs) had detection limits of approximately 1 mU/L. Since the normal range for serum TSH is approximately 0.4 to 5.0 mU/L, these assays were useful for the diagnosis of primary hypothyroidism (in which serum TSH concentrations are appropriately elevated) but were not sufficiently sensitive to distinguish between normal serum TSH concentrations and the low serum TSH concentrations present in most patients with hyperthyroidism.
Second-generation TSH immunometric assays have detection limits of approximately 0.1 mU/L. Since this detection limit is just below the normal range for TSH of approximately 0.4 to 5.0 mU/L, these assays can be used as screening tests to distinguish hyperthyroidism from euthyroidism and hypothyroidism. However, since the range of subnormal TSH measurement is very limited, values near or at the detection limit do not distinguish the degree of hyperthyroidism, and poor quality control in many laboratories can lead to erroneous values.
Third-generation TSH chemiluminometric assays, currently in wide use, have detection limits of approximately 0.01 mU/L. They can therefore provide detectable TSH measurements even in mild hyperthyroidism. Because of the considerably lower detection limit, even with poor quality control, serum TSH values in patients with overt hyperthyroidism are easily distinguished from those in euthyroid patients. In order to reliably detect values of serum TSH in the hyperthyroid range, one needs a third-generation assay with a functional sensitivity of at least ≤0.05 mU/L. As an example, second-generation methods will often give values well over 0.1 mU/L when the true value is 0.03 mU/L, a typical hyperthyroid value, and can thus lead to a false negative test. Third-generation values <0.1 mU/L are, however, generally reliable.
Presently there is considerable controversy as to the appropriate upper limit of normal for serum TSH. Most laboratories have used values of approximately 4.5 to 5.0 mU/L. A monograph published by the National Academy of Clinical Biochemistry argues 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 euthyroid volunteers have serum values between 0.4 and 2.5 mU/L. However, a population study from Germany, which excluded patients with a positive family history, goiter, nodules, or positive antithyroid peroxidase (TPO) antibodies, found a normal reference range of 0.3 to 3.63 mU/L. The use of 2.5 mU/L as the upper limit of normal for serum TSH will increase substantially the number of patients in the United States diagnosed with subclinical hypothyroidism. Presently, controversy exists as to whether patients with serum TSH values between 5 and 10 mU/L require treatment. Until there are data demonstrating an adverse biologic significance for serum TSH values between 2.5 and 5.0 mU/L, the wisdom of labeling such patients as hypothyroid is questionable.
Age-based normal ranges for TSH are important, as illustrated by an analysis of 16,533 individuals in the National Health and Nutrition Examination Survey III (NHANES III). In this report, there was an age-related shift towards higher TSH concentrations in older patients, which persisted when those with positive antithyroid antibodies were excluded. For example, the 97.5 centile for TSH in subjects aged 20 to 29 years or over age 80 was 3.56 and 7.49 mU/L, respectively. Seventy percent of subjects in the older group with a TSH greater than 4.5 mU/L were within the normal range for their age. Similar age-associated increases in serum TSH concentrations were found in other prospective cohort studies.
