Association between over- and under-replacement with thyroid hormone and incident heart failure

CONTEXT: Whether over- or under-replacement with thyroid hormone is associated with incident heart failure remains unclear. OBJECTIVE: Elucidate the relationship between thyroid hormone over- and under-replacement and incident heart failure. DESIGN: Retrospective cohort study of 641 504 adults who initiated thyroid hormone treatment between 2004 and 2017 at the Veterans Health Administration. Two cohorts of adults aged ≥18 years treated with thyroid hormone were separately studied: 638 323 adults with at least 2 outpatient thyroid stimulating hormone (TSH) measurements, and 338 249 adults with at least 2 outpatient free thyroxine (FT4) measurements, between thyroid hormone initiation and incident heart failure or study conclusion. Associations between time-varying TSH and FT4 exposure and incident heart failure as the outcome were analyzed. RESULTS: Overall, 564 152/641 504 (87.9%) patients were men. Median age was 66 years. Incident heart failure occurred in 81 286/641 504 (12.7%) patients. Following adjustment for age, sex, and cardiovascular risk factors (eg, smoking, hypertension), under-replacement (eg, TSH > 20 mIU/L, adjusted odds ratio [AOR], 1.93, 95% CI, 1.84-2.02; FT4 < 0.7 ng/dL, AOR 1.08; 95% CI, 1.04-1.12) and over-replacement (eg, TSH < 0.1 mIU/L, AOR, 1.06; 95% CI, 1.01-1.12; FT4 > 1.9 ng/dL, AOR, 1.23; 95% CI, 1.14-1.32) were associated with increased incident heart failure compared to euthyroidism. Incident heart failure risk was cumulative over time with both thyroid hormone over- and under-replacement, but association was strongest with under-replacement (eg, TSH > 5.5 mIU/L associated with a 5.8-fold increase in incident heart failure over 5 years). CONCLUSION: Our findings suggest that thyroid hormone treatment intensity may be a modifiable risk factor for heart failure.