Thyroid function changes with advancing age, however there is little good quality data available that characterises the extent of thyroid disease in the elderly, or the parameters of normal thyroid function in the oldest old. Reference ranges for serum TSH and free thyroid hormones based on unselected younger adults are currently applied uniformly to older individuals.
The Newcastle 85+ study has collected health and lifestyle information from a cohort of 85-year olds, who are representative of the population of their age. This includes people living independently, in sheltered housing, residential and nursing care. Serum TSH, FT4, FT3 and thyroid peroxidase (TPO) antibodies were measured in batches. Out of the 764 eighty five year olds, 95 (12.4%) were taking levothyroxine, a further 64 people (8.3%) had known thyroid disease or were taking medication that could affect thyroid function; therefore these were excluded. Analysis was therefore confined to the remaining 605 people free of thyroid disease (disease-free).
Using our standard laboratory reference ranges (TSH 0.34.7 mU/l), 545 (90.1%) were classified as euthyroid, with 47 (7.8%) having subclinical hypothyroidism. Subclinical hyperthyroidism was found in 8 (1.3%) of the cohort, with overt hyperthyroidism and overt hypothyroidism being found in 1 (0.2%) and 4 (0.7%), respectively. 83 (13.7%) of the disease-free cohort were positive for TPO antibodies (>61 kU/l). In the disease-free population, the median TSH was 2.0 mU/l; with a reference range (2.5th to 97.5th centile) from 0.41 to 6.04 mU/l. If patients with positive TPO antibodies were excluded, the median TSH reduced to 1.94 mU/l; reference range 0.44 to 5.67 mU/l.
These unique data show that TSH levels are subtly increased in many healthy older people without overt thyroid disease. This demonstrates the need for age-appropriate reference ranges for serum TSH, with clear implications for the diagnosis and management of subclinical hypothyroidism.