Longitudinal population-based study of progression rates for borderline thyroid-stimulating hormone abnormalities

Current guidelines recommend thyroid-stimulating hormone (TSH) for initial biochemical evaluation of thyroid function, with borderline TSH abnormalities a common finding. The likelihood of a borderline TSH progressing to overt abnormality is not well characterised at the population level.

To determine risk factors and likelihood for progression of borderline TSH to overt abnormality.

Population-based retrospective longitudinal data-linkage study for TSH tests performed in Tasmania (1996-2013). Kaplan-Meier methodology was used to summarise conversion time for overt TSH elevation (≥10 mU/L) and overt suppression (≤0.1 mU/L) in patients whose initial TSH was borderline elevated (BeTSH; 4.0-9.99 mU/L) and borderline suppressed (BsTSH; 0.10-0.39 mU/L) respectively. Main outcome measures are the progression from borderline to overt TSH abnormality.

A total of 1 296 060 TSH tests was performed on 367 917 patients. Of these, 14 507 (3.9%) patients had BeTSH on initial assessment; mean age 51.4 ± 21.8 years and median TSH of 5.0 mU/L (interquartile range (IQR) 4.4, 5.2). Patients aged ≥80 years were most likely to progress (hazard ratio (HR) = 2.09 compared with age <20 years reference group (95% confidence interval (CI): 1.64, 2.68)). Patients aged 20-39 years had the second-highest rate of progression (HR = 1.49 (95% CI: 1.18, 1.88)). Seven thousand, eight hundred and eighty-three (2.14%) patients had BsTSH; mean age 50.7 ± 22.1 years and median TSH 0.30 mU/L (IQR 0.22, 0.35). Patients aged 60-79 years had the highest rate of progression to overt TSH suppression (HR = 2.47 compared with age <20 years reference group (95% CI: 1.88, 3.22)).

Follow-up intervals for patients with borderline TSH abnormalities should take into account patient age as a progression risk factor.