Abstracts

RATIONALE: The objective of this abstract is to indicate that memory impairment is relatively common in temporal lobe epilepsy (TLE) even after taking into account IQ. One method of characterizing an individual[ssquote]s memory is to calculate the discrepancy between IQ and a memory index (IQ minus memory). In the past, a single discrepancy cut-off score has been used in making the judgment of impaired vs. unimpaired memory. A recent publication (Hawkins & Tulsky, 2001) that analyzed data from a large sample of healthy controls, who were administered the latest editions of the Wechsler adult intelligence and memory scales (WAIS-III & WMS-III; Psychological Corporation, 1997), demonstrated that the discrepancy score method requires consideration of IQ level. A patient[ssquote]s IQ must be taken into account because the General Memory Index (GMI) among controls tends to be higher than full scale IQ (FSIQ) at lower IQ levels, whereas the converse is true at higher IQ levels. Thus, a particular FSIQ-GMI discrepancy score may be clinically significant at one IQ level, but not at another. It is well known that TLE is associated with high risk of memory impairment when the memory score is considered in isolation. We assessed frequency of memory impairment in TLE using the FSIQ minus GMI discrepancy method, while taking into account FSIQ level.
METHODS: This study analyzed Bozeman Epilepsy Consortium data for 77 adult TLE patients (38 right, 39 left TLE). The WAIS/WMS III standardization sample (N = 1,250) served as the control group (Hawkins & Tulsky, 2001). For patients, means were: age 35 ([plusminus]11), age of epilepsy onset 16 ([plusminus]12), FSIQ 88 ([plusminus]11), and GMI 85 ([plusminus]15). Thus, there was a mild deficit in both IQ and memory at the group level. The TLE sample was divided into 4 subgroups based on IQ (70-79; 80-89; 90-99; 100-109). Individuals whose FSIQ-GMI discrepancy score was [gte] 84th percentile of the standardization sample were considered to have memory impairment relative to IQ level. The FSIQ-GMI discrepancy score cut-off corresponding to the 84th percentile is as follows for each of the four FSIQ subgroups: 0, 6, 9, 12.
RESULTS: By the FSIQ-GMI discrepancy criteria, 16% of the WAIS/WMS III standardization sample had abnormal memory (i.e., an atypical GMI-FSIQ score) at each IQ level. In contrast, the frequency of discrepancy score-based memory dysfunction among the TLE patients ranged from 36 to 50% across the IQ subgroups, with 47% impaired overall. Frequency of memory impairment did not differ by laterality of seizure focus or age of epilepsy onset. Considering the GMI in isolation, 51% of the patients had impaired memory (GMI [lte] -1 SD from the control mean).
CONCLUSIONS: Classification of memory functioning can be based on a memory score alone. Alternatively, memory functioning can be judged in the context of IQ. However, it is not appropriate to use a single FSIQ-GMI discrepancy score cut-off for all patients. In this study, TLE patients were three times more likely than controls to demonstrate memory impairment when IQ stratified base rates of the FSIQ-GMI discrepancy score were considered. Thus, memory impairment is common, although not pervasive, in TLE even after accounting for the fact that intelligence is often lower than average in this patient population.
[Supported by: The NIH.]