Confirmatory factor analysis of the WMS-III in patients with temporal lobe epilepsy

Machine learning as a new paradigm for characterizing localization and lateralization of neuropsychological test data in temporal lobe epilepsy

In this study, we employed a kernel support vector machine to predict epilepsy localization and lateralization for patients with a diagnosis of epilepsy (n = 228). We assessed the accuracy to which indices of verbal memory, visual memory, verbal fluency, and naming would localize and lateralize seizure focus in comparison to standard electroencephalogram (EEG). Classification accuracy was defined as models that produced the least cross-validated error (CVϵ). In addition, we assessed whether the inclusion of norm-based standard scores, demographics, and emotional functioning data would reduce CVϵ. Finally, we obtained class probabilities (i.e., the probability of a particular classification for each case) and produced receiver operating characteristic (ROC) curves for the primary analyses. We obtained the least error assessing localization data with the Gaussian radial basis kernel function (RBF; support vectors = 157, CVϵ = 0.22). There was no overlap between the localization and lateralization models, such that the poorest localization model (the hyperbolic tangent kernel function; support vectors = 91, CVϵ = 0.36) outperformed the strongest lateralization model (RBF; support vectors = 201, CVϵ = 0.39). Contrary to our hypothesis, the addition of norm, demographics, and emotional functioning data did not improve the accuracy of the models. Receiver operating characteristic curves suggested clinical utility in classifying epilepsy lateralization and localization using neuropsychological indicators, albeit with better discrimination for localizing determinations. This study adds to the existing literature by employing an analytic technique with inherent advantages in generalizability when compared to traditional single-sample, not cross-validated models. In the future, class probabilities extracted from these and similar analyses could supplement neuropsychological practice by offering a quantitative guide to clinical judgements.

Spatial memory for asymmetrical dot locations predicts lateralization among patients with presurgical mesial temporal lobe epilepsy

This study examined the ability of an asymmetrical dot location memory test (Brown Location Test, BLT) and two verbal memory tests (Verbal Selective Reminding Test (VSRT) and California Verbal Learning Test, Second Edition (CVLT-II)) to correctly lateralize left (LTLE) or right (RTLE) mesial temporal lobe epilepsy that was confirmed with video-EEG. Subjects consisted of 16 patients with medically refractory RTLE and 13 patients with medically refractory LTLE who were left hemisphere language dominant. Positive predictive values for lateralizing TLE correctly were 87.5% for the BLT, 72.7% for the VSRT, and 80% for the CVLT-II. Binary logistic regression indicated that the BLT alone correctly classified 76.9% of patients with left temporal lobe epilepsy and 87.5% of patients with right temporal lobe epilepsy. Inclusion of the verbal memory tests improved this to 92.3% of patients with left temporal lobe epilepsy and 100% correct classification of patients with right temporal lobe epilepsy. Though of a limited sample size, this study suggests that the BLT alone provides strong laterality information which improves with the addition of verbal memory tests.

Confirmatory Factor Analysis of the Dutch Version of the Wechsler Memory Scale-Fourth Edition (WMS-IV-NL)

The latent factor structure of the Dutch version of the Wechsler Memory Scale-Fourth Edition (WMS-IV-NL) was examined with a series of confirmatory factor analyses. As part of the Dutch standardization, 1,188 healthy participants completed the WMS-IV-NL. Four models were tested for the Adult Battery (16-69 years; N = 699), and two models were tested for the Older Adult Battery (65-90 years; N = 489). Results corroborated the presence of three WMS-IV-NL factors in the Adult Battery consisting of Auditory Memory, Visual Memory, and Visual Working Memory. A two-factor model (consisting of Auditory Memory and Visual Memory) provided the best fit for the data of the Older Adult Battery. These findings provide evidence for the structural validity of the WMS-IV-NL, and further support the psychometric integrity of the WMS-IV.

Clinical utility of the Wechsler Memory Scale--Fourth Edition (WMS-IV) in predicting laterality of temporal lobe epilepsy among surgical candidates

This study evaluated the accuracy of the Wechsler Memory Scale--Fourth Edition (WMS-IV) in identifying functional cognitive deficits associated with seizure laterality in localization-related temporal lobe epilepsy (TLE) relative to a previously established measure, the Rey Auditory Verbal Learning Test (RAVLT). Emerging WMS-IV studies have highlighted psychometric improvements that may enhance its ability to identify lateralized memory deficits. Data from 57 patients with video-EEG-confirmed unilateral TLE who were administered the WMS-IV and RAVLT as part of a comprehensive presurgical neuropsychological evaluation for temporal resection were retrospectively reviewed. We examined the predictive accuracy of the WMS-IV not only in terms of verbal versus visual composite scores but also using individual subtests. A series of hierarchal logistic regression models were developed, including the RAVLT, WMS-IV delayed subtests (Logical Memory, Verbal Paired Associates, Designs, Visual Reproduction), and a WMS-IV verbal-visual memory difference score. Analyses showed that the RAVLT significantly predicted laterality with overall classification rates of 69.6% to 70.2%, whereas neither the individual WMS-IV subtests nor the verbal-visual memory difference score accounted for additional significant variance. Similar to previous versions of the WMS, findings cast doubt as to whether the WMS-IV offers significant incremental validity in discriminating seizure laterality in TLE beyond what can be obtained from the RAVLT.

Influence of anxiety on memory performance in temporal lobe epilepsy

This study examined the degree to which anxiety contributed to inconsistent material-specific memory difficulties among 243 patients with temporal lobe epilepsy from the Multisite Epilepsy Study. Visual memory performance on the Rey Complex Figure Test (RCFT) was poorer for those with high versus low levels of anxiety but was not found to be related to the TLE side. The verbal memory score on the California Verbal Learning Test (CVLT) was significantly lower for patients with left-sided TLE than for patients with right-sided TLE with low anxiety levels but equally impaired for those with high anxiety levels. These results suggest that we can place more confidence in the ability of verbal memory tests like the CVLT to lateralize to left-sided TLE for those with low anxiety levels, but that verbal memory will be less likely to produce lateralizing information for those with high anxiety levels. This suggests that more caution is needed when interpreting verbal memory tests for those with high anxiety levels. These results indicated that RCFT performance was significantly affected by anxiety and did not lateralize to either side, regardless of anxiety levels. This study adds to the existing literature which suggests that drawing-based visual memory tests do not lateralize among patients with TLE, regardless of anxiety levels.

The Evolution of the Wechsler Memory Scale: A Selective Review

In clinical use since 1940, the Wechsler Memory Scale was formally introduced to the psychological community in 1945 . By 1946, it ranked 90th out of the 100 most frequently used psychological tests. By 1969, it was the 19th most used psychological test and the 2nd most used test of memory. By 1982, it was the 12th most used test and the most used memory test-a popularity it continues to enjoy. The present article will briefly trace the origin of the Wechsler Memory Scale and examine its evolution across the revisions that appeared in 1987 , 1997 , and 2009 . Issues with norming and standardization, as well as reliability and validity, will be summarized. It is argued that the test continues to have several serious shortcomings, including a lack of anchoring in an explicit neuroanatomical theory of memory and an underlying factor structure that appears to have changed little despite changes in the manifest structure and content of the test.

History of neuropsychology through epilepsy eyes

In the 19th century, Hughlings Jackson relied on clinical history, seizure semiology, and the neurologic examination as methods for seizure localization to inform the first epilepsy surgeries. In the 20th century, psychological and neuropsychological tests were first employed as both diagnostic and prognostic measures. The contemporary practice of epilepsy evaluation and management includes neuropsychology as a critical component of epilepsy care and research, and epilepsy and neuropsychology have enjoyed a very special and synergistic relationship. This paper reviews how epilepsy has shaped the practice of neuropsychology as a clinical service by asking critical questions that only neuropsychologists were in a position to answer, and how clinical care of epilepsy patients has been significantly improved based on neuropsychology's unique contributions.

Visual memory in patients after anterior right temporal lobectomy and adult normative data for the Brown Location Test

Several large meta-analytic studies have failed to support a consistent relationship between visual or "nonverbal" memory deficits and right mesial temporal lobe changes. The Brown Location Test (BLT), a recently developed dot location learning and memory test, uses a nonsymmetrical array and provides control over many of the confounding variables (e.g., verbal influence and drawing requirements) inherent in other measures of visual memory. In the present investigation, we evaluated the clinical utility of the BLT in patients who had undergone left or right anterior mesial temporal lobectomy. We also provide normative data of 298 healthy adults for standardized scores. Results revealed significantly worse performance on the BLT in the right as compared to the left lobectomy group and the healthy adult normative sample. The present findings support a role for the right anterior mesial temporal lobe in dot location learning and memory.

Validity of DSM-IV axis V global assessment of relational functioning scale: a multimethod assessment

We investigate the convergent validity of the DSM-IV Axis V Global Assessment of Relational Functioning Scale (GARF; American Psychiatric Association, 1994). This study included 79 patients at a university-based outpatient treatment clinic. We examined clinician-rated GARF and the relationship to self-reported (Inventory of Interpersonal Problems; IIP-C; Horowitz et al. 2000) and free response themes [Social Cognition and Object Relations Scale: SCORS; Hilsenroth, Stein & Pinsker, 2004; Westen, 1995] of interpersonal functioning. Clinician ratings of the GARF scale and SCORS variables were highly reliable and internally consistent. Convergent Validity among the GARF, SCORS, and IIP scores was calculated using a Principal Components Analysis and Confirmatory Factor Analysis (CFA). Results of the Principal Components Analysis revealed that the GARF, SCORS, and IIP scores converged on a single factor, although findings of the CFA did not fully confirm the 1 factor model originally proposed. Intercorrelations among the GARF, SCORS, and IIP variables were analyzed and a pattern of significant relationships was found between the GARF and SCORS variables. This study helps support the convergent validity GARF as a relational functioning measure and is one of the first investigations to examine this scale multidimensionally.

Wechsler Memory Scale-III Faces test performance in patients with mild cognitive impairment and mild Alzheimer's disease

Little is known about the sensitivity of the Wechsler Memory Scale-Third Edition (WMS-III) Faces subtest to memory impairment associated with mild cognitive impairment (MCI). In this study, Faces performance was examined in 24 MCI patients, 46 mild Alzheimer's disease (AD) patients, and 98 elderly controls. We hypothesized that participants with diagnoses of MCI or AD would be impaired relative to controls on Faces. Analyses showed that AD participants performed significantly worse than MCI and intact participants, although there were no significant differences between MCI and intact participants. Data suggest that brain areas specialized for face recognition memory may be less affected by MCI and mild AD than regions specialized for verbal memory.

A new measure of visual location learning and memory: development and psychometric properties for the Brown Location Test (BLT)

There are a variety of well-established neuropsychological tests that are helpful in identifying global and specific verbal memory deficits. In contrast, tests of visual memory have produced less consistent results likely due in part to confounding variables such as verbal encodability, administration difficulties, and insufficient differentiation of among types of visual memory. The Brown Location Test (BLT) was designed to specifically measure visual memory for location of identical objects (dots) and address limitations found in commonly employed visual memory tests. This paper describes the empirical basis for the BLT and reports the psychometric properties of the test. Results indicate good internal and alternate form reliabilities. Factor analysis of a brief test battery confirmed that BLT performance is generally independent of verbal memory and global intellectual abilities. BLT performance declined with age, but there was no association between performance and gender, education, or intellectual functioning. In view of the favorable psychometric properties observed during preliminary studies, additional normative and validation studies in healthy and patient populations are warranted.

Rey Complex Figure Test Performance After Traumatic Brain Injury

The influence of factors such as a cerebral injury characteristics, education, perceptual organization skills, and speed of information processing on performance on the Rey Complex Figure Test & Recognition Trial (RCF) was examined by means of hierarchical regression analyses in 100 consecutively referred persons with traumatic brain injury at a median of 2.5 months post-injury. Patients with premorbid (e.g., psychiatric history) or comorbid (e.g. financial compensation seeking) complicating factors were excluded. Perceptual organization skills and the presence of a diffuse intracranial lesion but not education or speed of processing were statistically significant predictors of the variance in RCF variables. A large proportion of the sample improved by at least a standard deviation from independent delayed recall to multiple-choice recognition, and this was mediated by perceptual organization skills but not by injury parameters. It is concluded that performance on the RCF after traumatic brain injury is affected relatively more by perceptual organization skills than by injury severity characteristics.

Is Basic Memory Structure Invariant Across Epilepsy Patient Subgroups?

It is well known that epilepsy patients often exhibit material-specific deficits in memory for verbal versus nonverbal material. However, it is also apparent that such deficits are not always discernible, and that numerous factors may potentially moderate the degree to which fractionation between verbal and nonverbal material is evident. Using confirmatory factor analysis, we compared the relative fit of one-factor (general) and two-factor (material-specific) models of memory in data from 330 patients with intractable seizure disorder. Data from verbal and nonverbal selective reminding tests, as well as Logical Memory and Visual Reproduction (immediate and delayed recall indices), were used in analyses. The one-factor model fit the data poorly in the full sample, whereas the two-factor model fit the data significantly better. We then assessed the invariance of these two models using multiple-group modeling across subsamples of patients with left versus right sided seizure focus, earlier versus later age of seizure onset, lower versus higher Full Scale IQ, fewer versus more years of education, younger versus older patients, and male versus female patients. In all cases, the two-factor model comprised of verbal and nonverbal memory factors fit the data better than a one-factor (general) model. These findings indicate robust consistency of verbal and nonverbal memory constructs in the epilepsy population, which remain viable in this patient group despite considerable heterogeneity in other respects.

Memory Impairment Following Right Hemisphere Stroke: A Comparative Meta-Analytic and Narrative Review

We examined the evidence for widely held clinical beliefs about memory impairment following right hemisphere stroke (RHS), conducting both narrative and meta-analytic reviews of the literature [MEDLINE (1966-January 2003), PsycINFO (1974-January 2003), and CINAHL (1982-December 2002)]. We sought to determine whether RHS patients experience more problems with non-verbal memory than non-stroke controls (NSCs) and left hemisphere stroke (LHS) patients. Secondarily, we sought to determine whether RHS patients experience more problems with verbal memory than NSCs and fewer verbal memory problems than LHS patients. We also examined the effect of type of memory assessment (recall versus recognition) on reported findings. As regards non-verbal memory, narrative and meta-analytic reviews found that RHS patients had deficits relative to NSCs, on tests of both recall and recognition. The evidence for RHS non-verbal memory deficits relative to LHS was mixed in the narrative review, whereas the meta-analysis found RHS deficits on non-verbal recognition tests, but no difference between RHS and LHS patients on non-verbal recall tests. Deficits on recognition tests imply problems with early encoding of material or possibly its storage. Regarding verbal memory, the narrative review found that RHS patients performed more poorly than NSCs in about half of all studies. The meta-analytic review confirmed poorer RHS performance on tests of verbal recall, but none of the studies that compared RHS and NSCs on verbal recognition could be included in this type of review. The narrative review found mixed evidence as regards the performance of RHS and LHS patients on verbal memory tests, but the meta-analysis pointed to RHS superiority for both verbal recall and recognition. The relative strengths of both types of review are discussed.

Development of demographic norms for four new WAIS-III/WMS-III indexes

Following the publication of the third edition Wechsler scales (i.e., WAIS-III and WMS-III), demographically corrected norms were made available in the form of a computerized scoring program (i.e., WAIS-III/WMS-III/WIAT-II Scoring Assistant). These norms correct for age, gender, ethnicity, and education. Since then, four new indexes have been developed: the WAIS-III General Ability Index, the WMS-III Delayed Memory Index, and the two alternate Immediate and Delayed Memory Indexes. The purpose of this study was to develop demographically corrected norms for the four new indexes using the standardization sample and education oversample from the WAIS-III and WMS-III. These norms were developed using the same methodology as the demographically corrected norms made available in the WAIS-III/WMS-III/WIAT-II Scoring Assistant.

Temporal lobe epilepsy and the selective reminding test: the conventional 30-minute delay suffices

Conventional memory assessment may fail to identify memory dysfunction characterized by intact recall for a relatively brief period but rapid forgetting thereafter. This study assessed learning and retention after 30-min and 24-hr delays on auditory and visual selective reminding tests (SRTs) in right (n=20) and left (n=22) temporal lobe epilepsy (TLE) patients and controls (n=49). The left TLE group performed significantly worse than controls on all 3 trials of both tests. The right TLE group differed from the controls on all 3 visual SRT trials and on learning for the auditory SRT. There were no between-groups differences in rate of information lost at the 30-min versus the 24-hr delay. At the individual level, there was no difference in the percentage of patients versus controls who demonstrated isolated memory impairment at the 24-hr delay. Accelerated forgetting over 24 hr is uncommon in TLE patients.

Confirmatory factor analysis of the behavioral dyscontrol scale in a mixed clinical sample

The Behavioral Dyscontrol Scale (BDS) is a brief measure of executive functioning originally designed for use primarily with older adults. Previous research suggests that factors derived from exploratory factor analysis of the BDS have clinical utility and are theoretically meaningful; however, the factor structure has not been previously submitted to confirmatory analysis, and there are several potential methodological problems with the previous exploratory factor analysis. In this study, we conducted a confirmatory factor analysis of the BDS. Findings were largely consistent with the original exploratory analysis (Grigsby, Kaye, & Robbins, 1992) in suggesting a three-factor model, although the three-factor model was a better fit if factors were allowed to correlate with one another. The findings of this study lend support to the current practice of using factor scores in clinical and research applications.

Verbal mediation and memory for novel figural designs: a dual interference study

To the extent that all types of visual stimuli can be verbalized to some degree, verbal mediation is intrinsic in so-called "visual" memory processing. This impurity complicates the interpretation of visual memory performance, particularly in certain neurologically impaired populations (e.g., aphasia). The purpose of this study was to investigate the relative contributions of verbal mediation to recognition memory for visual stimuli that vary with respect to their amenability to being verbalized. In Experiment 1, subjects attempted to verbally describe novel figural designs during presentation and then identify them in a subsequent recognition memory test. Verbalizing these designs facilitated memory. Stimuli that were found to be easiest or most difficult to verbalize at the group level were retained for the second study. In Experiment 2, subjects evidenced superior recognition memory for the relatively easy to verbalize items. This advantage was attenuated in subjects who performed a concurrent verbal interference task during encoding, but not in those who performed an analogous visual interference task. These findings provide evidence that impoverished verbal mediation disproportionately impedes memory for visual material that is relatively easy to verbalize. Implications for the clinical assessment of visual memory are discussed.

Mayo's older African American normative studies: confirmatory factor analysis of a core battery

Covariance structure analyses of a core neuropsychological test battery consisting of the Wechsler Adult Intelligence Scale-Revised, Wechsler Memory Scale-Revised, and Auditory Verbal Learning Test have previously identified a 5-factor model in a sample of cognitively normal White volunteers from Mayo's Older Americans Normative Studies (MOANS). The present study sought to replicate this factor structure in a sample of 289 cognitively normal, community-dwelling African American elders from Mayo's Older African Americans Normative Studies (MOAANS). The original 5-factor model was tested against 2 alternative 4-factor models and a 6-factor model generated on a substantive basis. Confirmatory factor analysis supported the construct validity of this core battery in older African Americans by replicating the original 5-factor model of Verbal Comprehension, Perceptual Organization, Attention/Concentration, Learning, and Retention as viable in the present sample.

Significant discrepancies between immediate and delayed WMS-III indices are rare in temporal lobe epilepsy patients

Initial factor analysis of Wechsler Memory Scale-Third Edition (WMS-III) standardization data resulted in a five-factor model that included separate immediate (IMM) and delayed (DEL) memory factors for auditory (AUD) and visual (VIS) memory. However, recent factor analyses that revealed three factors--AUD, VIS, and working memory--were discovered to be more accurate. Continued use of separate WMS-III IMM and DEL indices has been recommended because future studies with clinical groups might support this distinction. Data from this investigation of 88 temporal lobe epilepsy (TLE) patients suggested separate IMM and DEL WMS-III indices are not necessary in this population. Results were as follows: 1) There was no significant difference between the IMM and DEL indices for AUD or VIS memory. 2) The percentage of individuals with a significant difference between the IMM and DEL indices (IMM minus DEL >11 or <-12) was similar for our TLE group and the standardization sample. 3) The mean percent retention standard score (SS) was at the average level for the four memory subtests, and for each subtest only about 10% of the TLE patients had impaired scores. 4) Hippocampal volumes correlated significantly with both IMM and DEL indices. These data suggest that combining the appropriate WMS-III IMM and DEL memory subtest scores to form AUD and VIS memory indices is appropriate for TLE patients.

WMS-III performance in epilepsy patients following temporal lobectomy

We examined performances on the Wechsler Memory Scale-3rd Edition (WMS-III) among patients who underwent temporal lobectomy for the control of medically intractable epilepsy. There were 51 right (RTL) and 56 left (LTL) temporal lobectomy patients. All patients were left hemisphere speech-dominant. The LTL and RTL patients were comparable in terms of general demographic, epilepsy, and intellectual/attention factors. Multivariate analyses revealed a significant crossover interaction (p <.001), with the RTL group scoring significantly lower on the visual than auditory indexes while the LTL group scored significantly lower on the auditory than visual memory indexes. Within-group pairwise analyses revealed statistically significant auditory versus visual index score comparisons (all p <.001) for both surgical groups. Discriminant analysis (p <.001) identified Verbal Paired Associates I, Faces I, and Family Pictures II to significantly discriminate RTL and LTL patients, with an overall correct classification rate of 81.3%. Our findings suggest that the WMS-III is sensitive to modality-specific memory performance associated with unilateral temporal lobectomy.

Neuropsychological aspects of epilepsy surgery

Neuropsychological evaluation of a patient's cognitive capabilities before and after epilepsy surgery is essential in elective epilepsy surgery. On the one hand, neuropsychology provides accessory information regarding the localization and lateralization of epilepsy-associated cognitive impairment; on the other hand, it is a useful tool for quality and outcome control of epilepsy surgery which helps to make surgery more effective and safe. Evaluation of the adequacy of the brain tissues to be resected and of the patient's mental reserve capacities allows for a prediction of the postoperative cognitive development. Successful surgery can stop mental decline due to chronic epilepsy and it can reverse this negative trend by release of functions and capacities that were secondarily affected before surgery. However, surgery bears the risk of additional impairments which, in interaction with normal or even pathological processes of mental aging, may accelerate cognitive decline at an older age. From a neuropsychological point of view, early recognition of pharmacoresistance is important along with early and complete seizure control with maximal sparing of functional tissues.