NCO-15PREDICTION OF SUBJECTIVE MEMORY ABILITY BEFORE RESECTION OF HIGH GRADE GLIOMA

Abstract

OBJECTIVE: A significant relationship between subjective reports of cognitive function and objective neuropsychological test data is often absent in medical populations. Sociodemographic and clinical predictors of subjective memory ability are largely unknown in glioma patients. METHODS: Patients (n = 146) with high-grade glioma completed neuropsychological assessment preoperatively, including the Hopkins Verbal Learning Test-Revised (HVLT-R), the Functional Assessment of Cancer Therapy-Brain (FACT-Br), and the Beck Depression Inventory-Second Edition (BDI-II). Linear regression analyses were conducted with sociodemographic (age and sex), clinical (tumor location and lateralization), and neuropsychological variables (HVLT-R and BDI-II) as predictors of subjective memory ability (“I can remember new things”) from the FACT-Br. Three linear regression analyses were conducted to examine differences between the HVLT-R subtests (Immediate Recall, Delayed Recall, Delayed Recognition) in predicting subjective memory ability. RESULTS: FACT-Br subjective memory ability was significantly correlated with HVLT-R Immediate Recall, r(144) = .31, p < .001, Delayed Recall, r(136) = .26, p < .01, and Delayed Recognition, r(136) = .18, p < .05. Better subjective memory ability was predicted by a model (p < .001, R2 = .21) that included younger patient age (β = -.20; p = .018), better HVLT-R Immediate Recall (β = .19; p = .018), and less depressive symptomatology (β = -.24; p = .002). Separate models including HVLT-R Delayed Recall (p < .001, R2 = .20) and Delayed Recognition (p < .001, R2 = .20) did not find these subtests to be predictive of subjective memory ability; however, younger patient age (β = -.24; p = .006; and β = -.26; p = .002, respectively) and less depressive symptoms (β = -.25; p = .002; and β = -.26; p = .001, respectively) were predictive of better subjective memory ability. CONCLUSION: Subjective memory ability and objective memory test performance were modestly correlated in newly diagnosed patients with high-grade glioma. Multivariate analysis highlighted the additional importance of patient age and depressive symptoms in predicting subjective memory ability. Results underscore the importance of considering sociodemographic and clinical variables when patients self-report memory difficulties.