Abstract
The upper limit of the thyroid-stimulating hormone (TSH) reference range is controversial. Historically, values have been based on 95% confidence intervals of log-transformed concentrations obtained from healthy individuals. A paucity of longitudinal studies examining the association between baseline TSH and development of hypothyroidism exists.
The Busselton Thyroid Study is based on the 1981 and 1994 cross-sectional health surveys of a white, iodine-replete population. Participants completed a health questionnaire, underwent physical examination and provided a fasting venous sample at each time point. Of the original 2108 participants, 1328 attended the 13 y follow-up survey. Archived samples stored securely at −70°C were assayed in 2001 for TSH, free thyroxine (FT4), thyroid peroxidase antibody (TPOAb) and thyroglobulin antibody (TgAb) on an Immulite 2000 chemiluminescent analyser. Hypothyroidism (TSH > 4.0 mU/L) and overt hypothyroidism (TSH > 10.0 mU/L) constituted the primary and secondary outcome measures. Participants with pre-existing thyroid disease, missing results, lithium and amiodarone therapy, pituitary disease and antibody interference were excluded. Predictors of hypothyroidism (age, gender, TSH, parity, body mass index, smoking status, TPOAb and TgAb) were examined.
Female gender and baseline TSH were the strongest predictors of hypothyroidism, while the predictive value of TPOAb and TgAb was attenuated by adjustment for age, gender and TSH. The optimal baseline TSH cut-off for predicting outcomes was 2.4 mU/L (sensitivity 76%, specificity 90%) for hypothyroidism and 2.6 mU/L (sensitivity 79%, specificity 90%) for overt hypothyroidism. A baseline TSH >2.5 mU/L yielded a positive predictive value (PPV) of 47% and negative predictive value (NPV) of 97% for hypothyroidism at follow-up, compared with a TSH >4 mU/L (PPV 84%, NPV 95%).
While the study demonstrated an increased risk at baseline TSH between 2.5 and 4.0 mU/L, most of these subjects did not develop hypothyroidism during the follow-up. The authors therefore do not support lowering the upper TSH limit to 2.5 mU/L, but recommend that women with TSH values between 2.5 and 4.0 mU/L and positive TPOAb should be monitored.