The diagnosis of subclinical hypothyroidism--serum TSH levels above and T(4) levels within the laboratory reference ranges--depends critically on the upper limit of the TSH reference interval. Calls have been made to lower the current upper TSH reference limit of 4.0 mU/l to 2.5 mU/l to exclude patients with occult hypothyroidism. However, data from population studies do not indicate that the distribution of TSH is altered owing to inclusion of such individuals. The opposite suggestion has also been put forward; the TSH upper reference limit is often too low, especially in the elderly, in women and in white individuals, which may lead to unnecessary or even harmful therapy. Studies in elderly individuals have shown that although aging may be associated with increased TSH levels, paradoxically, overt hypothyroidism in this population may be associated with a less robust TSH response than in young individuals. This Review highlights the interindividual and intraindividual variability of TSH levels and discusses the current controversy that surrounds the appropriateness of reference ranges defined on the basis of age, race, sex and amount of iodine intake. Moreover, the current evidence on lowering or increasing the upper limit of the TSH reference interval is reviewed and the need to individualize levothyroxine treatment in patients with elevated TSH levels is discussed.
© 2011 Macmillan Publishers Limited. All rights reserved