Table 2.
Cognitive measures
| Measure | LEV | LTG | CBZ | CTR | P |
|---|---|---|---|---|---|
| Letter fluency, mean (SD) |
n = 34, 12.85 (4.6)a |
n = 37, 14.16 (5.8)a |
n = 36, 10.28 (5.2) |
n = 36 17.75 (7.8) |
.000 |
| Category fluency, mean (SD) |
n = 36, 17.03 (5.8)a |
n = 37, 17.92 (6.1)a |
n = 36, 14.31 (5.8) |
n = 36, 19.42 (9.0) |
.016 |
| McKenna Graded Naming Test, median (range) |
n = 25, 16 (7‐27) |
n = 29, 15.5 (6‐26) |
n = 34, 14 (3‐25) |
– | .346 |
| National Adult Reading Test, median (range) |
n = 25, 98 (36‐119) |
n = 29, 100 (70‐125) |
n = 24, 97.5 (27‐117) |
– | .244 |
For letter and category fluency, analysis of variance was used to detect differences between groups at P < .05.CBZ, carbamazepine; CTR, healthy control subjects; LEV, levetiracetam; LTG, lamotrigine; SD, standard deviation.
a
Post hoc group comparisons (2‐tailed t test, P < .05) revealed that both LEV and LTG groups performed better on letter and category fluency than the CBZ group, and all 3 patient groups performed worse than healthy controls. For McKenna Graded Naming Test and National Adult Reading Test, healthy control data were not available. Nonparametric testing (Kruskal‐Wallis test) did not show any differences between the 3 different patient groups.