PURPOSE: The aim of the study was to evaluate serum thyroid hormone balance in children receiving long-term therapy with carbamazepine (CBZ), valproate (VPA), and phenobarbital (PB). METHODS: We determined serum levels of triiodothyronine (T3), thyroxine (T4), free thyroxine (FT4), thyroxine-binding globulin (TBG), and thyroid-stimulating hormone (TSH) in 148 healthy children and 141 children with epilepsy who had been receiving CBZ (61 patients), VPA (51 patients), or PB (29 patients) for 12-161 months. In view of TSH values, three categories of subclinical hypothyroidism were considered: I, TSH greater than the control-group mean + 2 SD (4.37 mIU/L in our study) and <6 mIU/L; II, TSH between 6 and 12 mIU/L; and III, TSH >12 mIU/L. RESULTS: In all treated groups, mean T4 and FT4 levels were lower than in the control group, whereas the CBZ- and VPA-treated children additionally showed reduced mean T3 and TBG levels and increased mean TSH levels. In the group receiving CBZ, 8.2% had TSH values higher than the normal-range maximum, by comparison with only 3.6% of healthy children. The increase in TSH levels was particularly marked in VPA-treated children, accounting for 26% of patients with subclinical hypothyroidism. CONCLUSIONS: Our results, in contrast to previous reports, suggest that CBZ and particularly VPA may induce subclinical hypothyroidism. This suggests a need for careful monitoring of TSH levels in children receiving CBZ or VPA.
PURPOSE: The aim of the study was to evaluate serum thyroid hormone balance in children receiving long-term therapy with carbamazepine (CBZ), valproate (VPA), and phenobarbital (PB). METHODS: We determined serum levels of triiodothyronine (T3), thyroxine (T4), free thyroxine (FT4), thyroxine-binding globulin (TBG), and thyroid-stimulating hormone (TSH) in 148 healthy children and 141 children with epilepsy who had been receiving CBZ (61 patients), VPA (51 patients), or PB (29 patients) for 12-161 months. In view of TSH values, three categories of subclinical hypothyroidism were considered: I, TSH greater than the control-group mean + 2 SD (4.37 mIU/L in our study) and <6 mIU/L; II, TSH between 6 and 12 mIU/L; and III, TSH >12 mIU/L. RESULTS: In all treated groups, mean T4 and FT4 levels were lower than in the control group, whereas the CBZ- and VPA-treated children additionally showed reduced mean T3 and TBG levels and increased mean TSH levels. In the group receiving CBZ, 8.2% had TSH values higher than the normal-range maximum, by comparison with only 3.6% of healthy children. The increase in TSH levels was particularly marked in VPA-treated children, accounting for 26% of patients with subclinical hypothyroidism. CONCLUSIONS: Our results, in contrast to previous reports, suggest that CBZ and particularly VPA may induce subclinical hypothyroidism. This suggests a need for careful monitoring of TSH levels in children receiving CBZ or VPA.