Serum Thyrotropin Levels Across Older Age Groups in Geriatric Outpatients Without Documented or Clinically Evident Thyroid Disease: A Retrospective Cross-Sectional Study

Background/Objectives: Reported age patterns in serum thyroid-stimulating hormone (TSH) vary across populations and assays. We examined age and sex associations with TSH levels in clinically selected geriatric outpatients in Ankara, Türkiye. Methods: This retrospective cross-sectional study analyzed records from 1 October to 31 December 2022. After clinical, medication, and laboratory exclusions, 206 participants aged ≥65 years remained. Linear regression of log-transformed TSH included age per five years and sex, with HC3 robust standard errors. Sensitivity models added comorbidities and medication count and applied the institution-specific TSH and fT4 reference intervals. Exploratory quantile regression used bootstrap confidence intervals (CIs) and formal comparisons of percentile-specific coefficients. Results: Mean age was 73.9 ± 6.5 years; 122 participants (59.2%) were women. Median TSH was 1.80 mIU/L (interquartile range, 1.25–2.61) and did not differ across age groups (p = 0.151). The primary age estimate was −6.6% per five years (95% CI, −13.1% to +0.5%; p = 0.066) and was unchanged after comorbidity adjustment (−6.7%; p = 0.077), but attenuated when analysis was restricted to the institutional TSH interval (−3.2%; p = 0.274). Women had 25.5% higher TSH in the primary model (95% CI, +1.3% to +55.5%; p = 0.037). An exploratory 10th-percentile age estimate (−19.6%; 95% bootstrap CI, −37.4% to −6.6%) was not supported by a joint test of percentile differences (p = 0.13). Conclusions: No statistically significant age association was demonstrated in this selected cohort. The primary analysis indicated higher TSH in women, although precision varied across restricted analyses. Isolated TSH results in older outpatients should be confirmed and interpreted in a clinical context; larger prospective studies are needed.