A-168: CVLT-II Process Scores Differentiate Mild Cognitive Impairment (MCI) From Normal Aging but Do Not Improve MCI Subtype Classification Beyond Delayed Recall

Abstract

Objective: To evaluate the clinical utility of selected CVLT-II process scores (intrusions, semantic clustering, and learning slope) for identifying mild cognitive impairment (MCI) in a memory clinic sample. We examined whether these indices provide incremental diagnostic value beyond long delay free recall (LDFR) and whether they distinguish amnestic from non-amnestic MCI relative to cognitively normal (WNL) individuals. Method/Data Selection: Participants included 151 memory clinic patients (M age = 74.09, SD = 6.31; 46.4% female) including 87 WNL, 32 amnestic MCI, and 32 non-amnestic MCI. CVLT-II standard scores (LDFR, intrusions, semantic clustering, and learning slope) were analyzed using MANOVA and hierarchical regression to assess group differences and determine whether process scores predicted LDFR and improved diagnostic classification. Results/Data Synthesis: Diagnostic groups differed on combined CVLT-II process scores, F(6, 294)=4.06, p<.001, partial η2=.076. WNL demonstrated greater semantic clustering (ps< .004), learning slope (p=.034), and fewer intrusion errors (p=.020) than both MCI groups; amnestic and non-amnestic MCI did not differ. Process scores predicted LDFR (ΔR2 = .438, p<.001), but did not improve amnestic versus non-amnestic MCI classification beyond LDFR, χ2 change(3)=4.37, p=.224. Conclusions: CVLT-II process scores were sensitive to early learning and memory inefficiencies, differentiating MCI from normal cognition. However, they did not distinguish amnestic from non-amnestic MCI and did not provide incremental classification beyond LDFR. These findings suggest that, despite their theoretical and descriptive value, process scores may offer limited clinical utility for MCI subtype differentiation, reinforcing LDFR as the most robust indicator of memory impairment.