Color vision in epileptic adolescents treated with valproate and carbamazepine

OBJECTIVE The aims of our study were to evaluate whether deficits in color vision exist in epileptic adolescents, to study if monotherapy with valproic acid (VPA) and carbamazepine (CBZ) can affect color vision, and to determine the possible relationship between abnormal color vision tests and AEDs dosage and their serum concentrations. PATIENTS We examined 45 epileptic patients before the beginning of therapy and after 1 year of VPA or CBZ monotherapy and 40 sex- and age-matched healthy controls. METHODS Color vision was evaluated with Farnsworth Munsell 100 (FM100) hue test and achromatic and short-wavelength automated perimetry (SWAP). STATISTICAL ANALYSIS To evaluate intergroup differences we used ANOVA with Scheffe's post hoc test, when appropriate. Repeated measures ANOVA was used to evaluate the intragroup modifications of total error score (TES) and perimetric threshold during the follow-up. Pearson's correlation test was performed to correlate chromatic sense and perimetric data and AEDs dosage and serum concentrations. RESULTS Before the beginning of therapy, there were no differences in central color vision and SWAP between controls and epileptic patients. After 1 year, patients treated with VPA or CBZ showed a deficit in FM100 hue test and SWAP parameters while no significant deficit was found in achromatic perimetry. In particular, with the FM100 hue test a higher number of errors was found in both groups of patients (CBZ patients: 166.00 +/- 27.72 TES; VPA patients: 151.19 +/- 44.09, P < 0.001) in comparison with controls (controls: 109.29 +/- 24.73) and baseline values (CBZ patients: 110.65 +/- 22.9; VPA patients 107.43 +/- 21.70). With SWAP patients of both groups showed significant variation of foveal threshold (controls: 21.07 +/- 2.01 dB; CBZ patients: 19.35 +/- 1.32, P < 0.001; VPA patients: 18.88 +/- 1.89, P < 0.001), full-field mean threshold perimetric sensitivity (controls: 18.50 +/- 1.24 dB; CBZ patients: 16.60 +/- 1.47, P < 0.001; VPA patients: 16.23 +/- 1.55, P < 0.001) and mean threshold perimetric sensitivity of the three evaluated subareas of the visual field (area 1 controls: 21.01 +/- 1.15; CBZ patients: 19.45 +/- 1.74, P = 0.001; VPA patients: 18.25 +/- 1.61, P < 0.001; area 2 controls: 18.40 +/- 1.43; CBZ patients: 16.07 +/- 1.58, P +/- 0.001; VPA patients: 16.13 +/- 1.46, P = 0.001; area 3 controls: 17.20 +/- 1.49; CBZ patients: 14.28 +/- 1.51, P < 0.001; VPA patients: 14.31 +/- 2.90, P = 0.001). CONCLUSIONS Our study demonstrates that treatment with VPA or CBZ can affect significantly both central and paracentral color vision after a short treatment period.