Validating the usefulness of the NSI validity-10 with the MMPI-2-RF.

Purpose/Aim of the Study: The Neurobehavioral Symptom Inventory (NSI) is a 22-item self-report measure created to quantify the somatosensory, cognitive, and affective symptoms of Post-concussive Syndrome. Developers of the NSI used a subset of 10 items, the Validty-10, to measure symptom overreporting. We compared the Validity-10 versus the remaining NSI items (i.e., the Remaining-12) for how accurately they detect symptom exaggeration on the Minnesota Multiphasic Personality Inventory Second Edition - Restructured Form (MMPI-2-RF).Materials and Methods: We used a sample of 45 veterans evaluated in a Polytrauma/TBI Clinic of a Midwest VA Healthcare System who completed the NSI and MMPI-2-RF.Results: The Vaidity-10, Remaining-12, and Total Score all strongly correlated with mean of the MMPI-2-RF validity scales (r = .65, .67, and .70, respectively), illustrating equivalency among the various NSI scores. Groups were created based on significant T score elevation on any MMPI-2-RF validity scale (i.e. F-r > 119, or Fp-r, F-s, FBS, or RBS > 99). ROC analyses demonstrated that areas under the curve were equivalent for NSI Total Score (.84), Validity-10 (.81), and Remaining-12 (.81) in detecting overreporting.Conclusions: These findings do not support the notion that the Validity-10 has unique utility as an embedded symptom validity scale and highlights the likelihood that NSI Total Score can also serve this function.

Detecting mild cognitive impairment remotely with the modified memory impairment screen by telephone.

The original Memory Impairment Screen by Telephone (MIST) was designed to identify individuals with dementia but was relatively ineffective for identification of less severe impairment observed in mild cognitive impairment (MCI). We expanded the original MIST to create a modified instrument (mMIST) with greater sensitivity to less severe memory impairment. Older men and women with subjective cognitive decline were assessed by phone with the mMIST and subsequently classified independently with MCI or non-pathological cognitive decline. Participants with MCI produced lower scores on the mMIST than did participants without MCI, 10.8 ± 2.7 vs 13.3 ± 1.3, t = 5.68, p < 0.001, and performance on the mMIST predicted performances on the California Verbal Learning Test (CVLT), Verbal Paired Associate Learning Test (VPAL), Montreal Cognitive Assessment (MoCA) total score, and MoCA memory index score, p < 0.001. Receiver operating characteristic (ROC) analyses identified the optimal cut score on the mMIST to distinguish participants with and without MCI with Sensitivity = 73.1%, Specificity = 79.1%, and AUC = 0.79. Predictive values for distinguishing the amnestic form of MCI (aMCI) from non-amnestic MCI were Sensitivity = 81.8%, Specificity = 30%, and AUC = 0.82. These findings indicate that the mMIST is a valid screening instrument for identifying MCI. It can be administered remotely at low cost and low participant burden. Also, the mMIST has potential utility for remote cross-sectional and longitudinal evaluation in research and clinical contexts. Further investigation is indicated to corroborate its utility for assessment of aging patients and research participants.

Cross-validation of incidental learning tasks from the WAIS-IV as a measure of memory.

Although incidental learning (IL) routinely occurs in everyday life, it is infrequently assessed during neuropsychological evaluations. This study aimed to further examine the concurrent validity of IL measures based on the Vocabulary and Similarities subtests from the Wechsler Adult Intelligence Test-Fourth Edition (WAIS-IV). Participants included 43 Veterans referred for outpatient neuropsychological testing. Performances on the IL procedures correlated strongly with performances on the Repeatable Battery for the Assessment of Neuropsychological Status Immediate and Delayed Recall Indices (r = .48 to r = .78). These results indicate that the IL procedures from selected WAIS-IV subtests provided an efficient and valid measure of memory. In particular, the task based on the Similarities subtest provided exceptionally high value as a screen for memory problems. These IL procedures, which require minimal additional administration time, capitalize on the semantic encoding that is inherent in completing the Vocabulary and Similarities subtests, and offer a complementary approach to standard memory assessment.

Incremental utility of an extended stroop task for identifying subtle differences in cognitive performance among healthy older adults.

The Stroop Color and Word Test is a test of processing speed, response inhibition, and executive functioning (EF). This project examined whether extending the Stroop Color-Word trial beyond the standard time limit could more accurately assess performance on EF measures. Cognitively healthy older individuals (n = 198) enrolled in a study of cardiovascular health completed the Stroop as part of a neuropsychological battery. Two scores were computed for the Color-Word trial: the number of items completed within the first 45 seconds (traditional Color-Word score) and the speed of page completion beyond the first 45 seconds (Stroop-Extended score). Criterion measures included the Trail-Making Test Part B (TMT-B), Digit Span Backward, Symbol Digit Modalities Test, Short Category Test, and measures of verbal fluency. Results from hierarchical linear regression analyses indicated that the extended Stroop score accounted for small but statistically significant variance in TMT-B (additional 2.6%) and Digit Span Backwards (additional 2.6%) beyond the standard Color-Word score. These findings suggest that extending the Stroop Color-Word trial beyond the first 45 seconds provides a limited increase in predictive power within a healthy sample with restricted range of performance. The extended Stroop requires additional examination in heterogeneous samples, including clinical populations, to determine its predictive utility.

Development of a scale of executive functioning for the RBANS.

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) is a cognitive battery that contains scales of several cognitive abilities, but no scale in the instrument is exclusively dedicated to executive functioning. Although the subtests allow for observation of executive-type errors, each error is of fairly low base rate, and healthy and clinical normative data are lacking on the frequency of these types of errors, making their significance difficult to interpret in isolation. The aim of this project was to create an RBANS executive errors scale (RBANS EE) with items comprised of qualitatively dysexecutive errors committed throughout the test. Participants included Veterans referred for outpatient neuropsychological testing. Items were initially selected based on theoretical literature and were retained based on item-total correlations. The RBANS EE (a percentage calculated by dividing the number of dysexecutive errors by the total number of responses) was moderately related to each of seven established measures of executive functioning and was strongly predictive of dichotomous classification of executive impairment. Thus, the scale had solid concurrent validity, justifying its use as a supplementary scale. The RBANS EE requires no additional administration time and can provide a quantified measure of otherwise unmeasured aspects of executive functioning.

Validity of a verbal incidental learning measure from the WAIS-IV in older adults.

Incidental memory may reflect a form of learning in everyday life, although it is not consistently evaluated during standard neuropsychological evaluations. Further validation of a recently created measure of verbal Incidental Learning (IL) from the Wechsler Adult Intelligence Scale-IV is necessary to understand the utility of such a measure in clinical settings. Sixty-eight adults aged 50 to 89 were recruited from a Cognitive Disorders Clinic while receiving a standard neuropsychological assessment, along with two additional measures of IL. IL-Total Score was significantly correlated with immediate and delayed memory trials from standard neuropsychological tests (rs = .43 to .73, ps < .001, ds = 0.94-2.14), with worse IL performance being associated with lower memory abilities. Participants with probable Alzheimer's disease performed worse on the IL-Total Score than participants with Mild Cognitive Impairment, t(39.997) = 5.46, p < .001, d = 1.13. Given the strong relationships between this IL task and traditional memory measures in our sample, and the discrimination of IL-Total Score performance among diagnostic groups despite its short administration time, this IL task may play a role as a measure of memory in brief cognitive evaluations.