Nonverbal memory tests revisited: Neuroanatomical correlates and differential influence of biasing cognitive functions.

The detection of right temporal lobe dysfunction with nonverbal memory tests has remained difficult in the past. Reasons for this might be the potential influence of other biasing cognitive functions such as executive functions or the verbalisability of nonverbal material. The aim of this study was to investigate three classic nonverbal memory tests by identifying their neuroanatomical correlates with lesion-symptom mapping (LSM) and by probing their independence from verbal encoding abilities and executive functions. In a cohort of 119 patients with first-time cerebrovascular accident, memory performance was assessed in the Nonverbal Learning and Memory Test for Routes (NLMTR), the Rey Complex Figure Test (RCFT), and the Visual Design Learning Test (VDLT). Calculating multivariate LSM, we identified crucial brain structures for these three nonverbal memory tests. Behavioural analyses were performed to assess the impact of executive functions and verbal encoding abilities with regression analyses and likelihood-ratio tests. LSM revealed for the RCFT mainly right-hemispheric frontal, insular, subcortical, and white matter structures and for the NLMTR right-hemispheric temporal (hippocampus), insular, subcortical, and white matter structures. The VDLT did not reach significance in LSM analyses. Behavioural results showed that amongst the three nonverbal memory tests the impact of executive functions was most pronounced for RCFT, and the impact of verbal encoding abilities was most important in VDLT. Likelihood-ratio tests confirmed that only for NLMTR did the goodness of fit not significantly improve by adding executive functions or verbal encoding abilities. These results suggest that amongst the three nonverbal memory tests the NLMTR, as a spatial navigation test, could serve as the most suitable marker of right-hemispheric temporal lobe functioning, with the right hippocampus being involved only in this test. In addition, the behavioural results propose that only NLMTR seems mostly unaffected by executive functions and verbal encoding abilities.

Lesion-symptom mapping corroborates lateralization of verbal and nonverbal memory processes and identifies distributed brain networks responsible for memory dysfunction.

Memory disorders are a common consequence of cerebrovascular accident (CVA). However, uncertainties remain about the exact anatomical correlates of memory impairment and the material-specific lateralization of memory function in the brain. We used lesion-symptom mapping (LSM) in patients with first-time CVA to identify which brain structures are pivotal for verbal and nonverbal memory and to re-examine whether verbal and nonverbal memory functions are lateralized processes in the brain. The cognitive performance of a relatively large cohort of 114 patients in five classic episodic memory tests was analysed with factor analysis. Two factors were extracted that distinguished the verbal and nonverbal components of these memory tests, and their scores were subsequently tested for anatomical correlates by combining univariate and multivariate LSM. LSM analysis revealed for the verbal factor exclusively left-hemispheric insular, subcortical and adjacent white matter regions and for the nonverbal factor exclusively right-hemispheric temporal, occipital, insular, subcortical and adjacent white matter structures. These results corroborate the long-standing hypothesis of a material-specific lateralization of memory function in the brain and confirm a robust association between right temporal lobe lesions and nonverbal memory dysfunction. The right-hemispheric correlates for the nonverbal aspects of episodic memory include not only classic memory structures in the medial temporal lobe but also a more distributed network that includes cortical and subcortical structures also known for implicit memory processes.