Objective: Graves' disease is the commonest cause of hyperthyroidism in populations with sufficient dietary iodine intake. Anti-thyroid drugs (ATD) are often used as the initial treatment for Graves' hyperthyroidism, however there is a paucity of data relating the dose of ATD therapy to the effect on thyroid hormone levels, increasing the risk of both over- and under-treatment. We aimed to determine the pharmacodynamic response to the ATD carbimazole.

Design: Retrospective cohort study.

Methods: Participants were patients (n = 441) diagnosed with Graves' disease at Imperial College Healthcare NHS Trust between 2009 and 2018. The main outcome measure was change in thyroid hormone levels in response to ATD.

Results: Baseline thyroid hormone levels were positively associated with TSH receptor antibody titres (P < 0.0001). Baseline free triiodothyronine (fT3) were linearly related to free thyroxine (fT4) levels in the hyperthyroid state (fT3 = fT4*0.97–11), and fell proportionately with carbimazole. The percentage falls in fT4 and fT3 per day were associated with carbimazole dose (P < 0.0001). The magnitude of fall in thyroid hormones after the same dose of carbimazole was lower during follow up than at the initiation visit. The fall in thyroid hormone levels approximated to a linear response if assessed at least 3 weeks after commencement of carbimazole. Following withdrawal of antithyroid drug treatment, the risk of relapse was greater in patients with higher initial fT4, initial TSH receptor antibody titre, males, smokers, and British Caucasian ethnicity.

Conclusion: We identify a dose-response relationship for fall in thyroid hormones in response to carbimazole to aid in the selection of dose for Graves' hyperthyroidism.