Use of antiepileptic drugs and risk of hypothyroidism

Abstract

This study aimed to investigate the risk of clinically significant hypothyroidism among all the currently available antiepileptic drugs (AED). The Taiwan National Health Insurance Research Database (NHIRD) from 2004 to 2010 was analyzed using a prescription sequence symmetry analysis, and thyroxine treatment was used as a proxy to identify a hypothyroidism event. A cohort of patients who have been treated with both AED and thyroxine was selected, and the chronological order of AED and thyroxine use constituted the basis of the prescription sequence symmetry analysis. A causal relationship was suspected if there was a significantly higher proportion of patients who initiated thyroxine after AED than those who initiated thyroxine before AED. The ratio of the two proportions was described as a sequence ratio. To benchmark the effect size of AEDs on thyroid function, amiodarone was selected as the reference indicator. A total of 1,878,189 AED users was found in the database, with 16,200 of them also used thyroxine. The adjusted sequence ratio of thyroxine use after each AED was 1.75 (99% confidence interval, 1.58-1.94) for phenytoin, 1.34 (1.20-1.49) for valproate, 1.25 (1.15-1.36) for phenobarbital, 1.21 (1.08-1.34) for carbamazepine, and 1.22 (1.03-1.46) for oxcarbazepine. The risk of hypothyroidism from phenytoin use within a shorter time frame was similar that associated with amiodarone use. No association was shown in most of the new generation AEDs. The results indicated an increased risk of hypothyroidism among patients using AEDs, especially phenytoin, valproate, phenobarbital, carbamazepine, and oxcarbazepine. The findings also provided strong grounds for further investigations on acute thyroid adverse effect induced by phenytoin.