Studies that track cognitive change in patients with epilepsy over time are among the most important data documenting or refuting mental deterioration. Most longitudinal studies of this type have provided proof that intellectual decline is indeed progressive.17-19 However, some such studies showed no deterioration, or even improvement, possibly related to therapy with antiepileptic drugs.20,21
The most useful studies are those with prospective designs that provide cognitive and neuropsychological data before the onset of epilepsy. They are uncommon. Studies that report percentages of patients undergoing decline rather than just test-score means are also infrequent.
A prospective study of 35 adults with epilepsy was performed to determine the nature of changes in cognitive or neuropsychological performance over a 10-year period.22 Although some changes in subtests were noted, no overall changes in the WAIS-R or the Neuropsychological Battery for Epilepsy were observed over the 10 years of the trial. The mean duration of epilepsy before the initial evaluation was 20 years, however, and substantial, unobserved cognitive decline may have occurred during this interval.
Bourgeois et al.23 evaluated children with epilepsy within 2 weeks of seizure onset and then over a mean time of 4 years. Unaffected siblings were used as controls. The mean IQ of the affected children did not change appreciably over time, but persistent decreases of 10 points or more occurred in 8 of the 72 patients:
|IQ Pattern over Time||No. of Patients |
(n = 72)
|IQ at First Testing |
(mean ▒ SD)
|IQ at Last Testing|
(mean ▒ SD)
|Consistent decrease = |
|8 (11.1%)||108.1 ▒ 19.5||88.0 ▒ 19.5*|
|Consistent increase = |
|12 (16.7%)||93.7 ▒ 24.2||108.8 ▒ 23.5*|
|Fluctuation: Difference of =10 points |
between highest and lowest scores
but no permanent increase or decrease
|29 (40.3%)||99.3 ▒ 19.0||100.8 ▒ 19.3|
|No change: difference between highest and lowest score <10 points||23 (31.9%)||100.3 ▒ 19.8||101.6 ▒ 19.8|
*Significantly different from IQ on first testing (p <.0001).
The children with consistent decreases were more likely to have toxic drug levels, epilepsy that was difficult to control, and earlier onset of seizures. Of these three factors, drug toxicity and early age of onset were the best predictors of poor prognosis. These results suggest that progressive cognitive deterioration may occur in a percentage of children with epilepsy and that complete seizure control should not be sought at the expense of overall cognitive outcome.
A prospective trial by Ellenberg and colleagues29 tested children at ages 4 and 7 years. Those who experienced seizure onset between these two ages did not show significant decreases in mean IQ scores as compared to normal controls. This finding challenges the notion of intellectual deterioration in epilepsy (see below). However, IQ was measured using two different scales: the Stanford-Binet Intelligence Scale at age 4 and the Wechsler Intelligence Scale for Children at age 7. Also, children with epilepsy were compared to their unaffected siblings. The IQ was slightly (but not significantly) higher in the unaffected group. More variation in the IQ of the children with seizures was noted, however, and they had a higher prevalence of mental retardation. The authors did not believe that their findings indicated intellectual deterioration, because mental retardation was present in the group with epilepsy before the onset of seizures and probably reflected preexisting neurologic damage rather than changes caused by seizures.
Other studies have attempted to demonstrate that most cognitive change takes place early in the course of epilepsy. Meinardi et al.24 described a cascadic model of deterioration, in which irreversible decline occurs shortly after the onset of seizures. In subsequent years, intellectual function plateaus, and no changes may be noted on cognitive testing. This model could explain previous longitudinal studies in which epilepsy is associated with a baseline of impaired intellectual function but no further deterioration.23,25-27
To detect and control for the early cognitive decline that they hypothesized, Neyens and colleagues28 chose patients with a maximum 3-year duration of epilepsy as subjects for a longitudinal study. These patients were matched to controls and underwent cognitive assessment every 6 months for 1.5 years. At the onset of the study, children with epilepsy demonstrated a lower level of intelligence as compared to the control group. Follow-up testing revealed that most subjects experienced a gain in IQ score; this gain was greatest for normal controls. The children with the longest duration of epilepsy (who had presumably already undergone their greatest period of cognitive decline) had smaller gains in IQ scores than the controls. Those who had epilepsy for no more than 2 years showed an even more modest gain in IQ. The group with the most recent onset of epilepsy showed a relative decline in performance. These results were consistent with an early deterioration followed by relatively little change later in the course.
Seidenberg et al.27 conducted a longitudinal study in which they administered the Wechsler Adult Intelligence Scale (WAIS) to patients with epilepsy on two occasions 18 months apart. The patients were divided into two groups:
Both groups had higher scores on the second test administration, but the increase in the group with unimproved seizures was smaller and was present in fewer tests. These findings support some intellectual deterioration in the more severely affected group. The practice effect, a phenomenon that describes improved scores on subsequent test administration resulting from familiarity, may also explain improvement of IQ scores in patients with uncontrolled epilepsy.
Studies of IQ changes in children Table: Adapted from BFD Bourgeois, AL Prensky, HS Palkes, et al. Intelligence in epilepsy: A prospective study in children. Ann Neurol 1983;14:438ĘC444.
Reviewed and revised May 2004 by Steven C. Schachter, MD, epilepsy.com Editorial Board.
© 2014 Epilepsy.com. All rights reserved.