Context

Randomized controlled trials (RCTs) found no cardioprotective effects of levothyroxine therapy in older adults with subclinical hypothyroidism.

Objective

To assess levothyroxine effects on cardiometabolic biomarkers, which may serve as more sensitive treatment indicators.

Design

Post hoc analysis using (baseline and 12-month) data from two double-blind randomised controlled trials in older adults (≥ 65 years) with subclinical hypothyroidism.

Main outcome measure(s)

Cardiometabolic biomarkers included seven clinically relevant lipid measures (apolipoprotein B (ApoB), total cholesterol (Total-C), non-high-density lipoprotein cholesterol (non-HDL-C), remnant cholesterol (RC), low-density lipoprotein cholesterol (LDL-C), HDL-C, and triglycerides (TG)) and 167 standardised metabolomic measures from nuclear magnetic resonance. Analyses were additionally stratified by baseline TSH levels.

Results

Among 286 included participants (48% women; median age 75 [70, 82] years; median baseline TSH 6.44 [5.36, 7.81] mIU/L), 142 were randomized to levothyroxine. Overall, levothyroxine showed no effects on ApoB (-0.03 [95% CI: -0.07, 0.00] g/L), Total-C (-0.17 [-0.34, 0.00] mmol/L), non-HDL-C (-0.15 [-0.31, 0.00] mmol/L), RC (-0.09 [-0.16, -0.01] mmol/L), LDL-C (-0.07 [-0.15, 0.02] mmol/L), and TG (-0.07 [-0.15, 0.01] mmol/L). In participants with baseline TSH ≥10 mIU/L (n=27), potentially beneficial changes (P-values < 0.05, but not significant after multiple-testing correction) were observed for all clinically relevant lipids except HDL-C, as well as for ApoB-containing lipoproteins, VLDL size and fatty acids.

Conclusion

In older adults with subclinical hypothyroidism, levothyroxine treatment showed no effects on cardiometabolic biomarkers, although potentially favourable changes in lipids and lipoproteins were observed for individuals with baseline TSH ≥ 10 mIU/L.

© The Author(s) 2026. Published by Oxford University Press on behalf of the Endocrine Society.