Changes in thyroid function and evolution of subclinical thyroid disease in older men

Abstract

Objective: Prevalence of subclinical thyroid disease increases with age, but optimal detection and surveillance strategies remain unclear particularly for older men. We aimed to assess thyroid stimulating hormone (TSH) and free thyroxine (FT4) concentrations and their longitudinal changes, to determine the prevalence and incidence of subclinical thyroid dysfunction in older men.

Design, participants and measurements: Longitudinal study of 994 community-dwelling men aged ≥70 years without known or current thyroid disease, with TSH and FT4 concentrations assessed at baseline and follow-up (after 8.7 ± 0.9 years). Factors associated with incident subclinical thyroid dysfunction were examined by logistic regression and receiver operating characteristic analyses.

Results: At baseline, 85 men (8.6%) had subclinical hypothyroidism and 10 (1.0%) subclinical hyperthyroidism. Among 899 men euthyroid at baseline (mean age 75.0 ± 3.0 years), 713 (79.3%) remained euthyroid, 180 (20.0%) developed subclinical/overt hypothyroidism, and 6 (0.7%) subclinical/overt hyperthyroidism. Change in TSH correlated with baseline TSH (r = .16, p < .05). Change in FT4 correlated inversely with baseline FT4 (r = -0.35, p < .05). Only higher age and baseline TSH predicted progression from euthyroid to subclinical/overt hypothyroidism (fully-adjusted odds ratio [OR] per year=1.09, 95% confidence interval [CI] = 1.02-1.17, p = .006; per 2.7-fold increase in TSH OR = 65.4, CI = 31.9-134, p < .001). Baseline TSH concentration ≥2.34 mIU/L had 76% sensitivity and 77% specificity for predicting development of subclinical/overt hypothyroidism.

Conclusions: In older men TSH concentration increased over time, while FT4 concentration showed little change. Subclinical or overt hypothyroidism evolved in one fifth of initially euthyroid men, age and higher baseline TSH predicted this outcome. Increased surveillance for thyroid dysfunction may be justified in older men, especially those with high-normal TSH.

Keywords: ageing; free thyroxine; hypothyroidism; men; subclinical hyperthyroidism; subclinical hypothyroidism; thyroid-stimulating hormone.

Figure 1

Participant flow chart showing derivation of the study cohort. There were four waves of data collection, and during two of those (Waves 2 and 4) blood samples were collected. The analysis population for this study comprises men with blood samples collected (and hence available for analysis of thyroid‐stimulating hormone and free thyroxine concentrations) at both timepoints (Wave 2 and Wave 4), after excluding those with missing data or known or current thyroid disease.

Figure 2

Scatterplots and regression lines showing relationships between (A) baseline thyroid‐stimulating hormone (TSH) concentrations (x axis) and follow‐up TSH concentrations (y axis); (B) baseline TSH concentrations (x axis) and change in TSH concentrations (y axis); (C) baseline free thyroxine (FT4) concentrations (x axis) and follow‐up FT4 concentrations (y axis); (D) baseline FT4 concentrations (x axis) and change in FT4 concentrations (y axis); in men who were euthyroid at baseline and remained so at follow‐up (○) and those who were euthyroid at baseline and had developed subclinical hypothyroidism at follow‐up (●).

Figure 3

Probability of developing subclinical or overt hypothyroidism during follow‐up (y axis), according to baseline TSH concentrations (x axis), in men who were euthyroid at baseline. The fitted exponential function is Probability = 0.0158*exp(0.8798*TSH).