Cross-material memory prediction: Influence of preoperative visual memory on postoperative verbal memory change after left anterior temporal lobe resection in people with epilepsy.

OBJECTIVE: Left anterior temporal lobe resection (TLR) in people with epilepsy is associated with a high risk of postoperative verbal memory decline. Whereas substantial evidence supports the role of preoperative verbal memory in predicting postoperative verbal memory change, the potential contribution of preoperative visual memory has been rarely investigated. METHODS: We retrospectively analyzed data from 105 people with left TLR due to mesial temporal sclerosis and typical language dominance. Multiple linear and logistic regression analyses were conducted to predict the extent and risk of significant postoperative verbal memory decline based on preoperative verbal and visual memory performance, as well as clinical variables (age at surgery, age at onset, and preoperative focal to bilateral tonic-clonic seizures). In exploratory analyses, we examined the influence of dichotomized preoperative verbal and visual memory performance (impaired vs. unimpaired) on the extent and frequency of postoperative verbal memory decline, as well as the potential influence of general intellectual ability. RESULTS: Regression analyses showed that higher preoperative verbal memory performance was associated with more pronounced and more frequent postoperative verbal memory decline, whereas higher preoperative visual memory performance was associated with less severe and less frequent verbal memory decline. Clinical variables did not improve model prediction. Analyses with dichotomized preoperative verbal and visual memory performance confirmed the results of the regression analyses. General intellectual ability had an additional influence on postoperative verbal memory change, without changing the predictive value of preoperative verbal or visual memory performance. SIGNIFICANCE: These findings demonstrate a cross-material relationship between preoperative visual memory and postoperative verbal memory change following left TLR. Whereas preoperative verbal memory performance likely represents functional adequacy, preoperative visual memory appears to represent functional reserve. Therefore, our findings reconcile these two concepts of postoperative verbal memory change. Additionally, the influence of general intellectual ability suggests relevance of more general cognitive reserve.