Spatial navigation in the real world is a separate cognitive domain

Editor's Notes

• #1: Ladies and gentlemen, good morning. It is a great pleasure for me to be given a chance to present the results of our study aiming to uncover the relationship between spatial navigation and various cognitive domains in cognitively normal elderly and patients with amnestic mild cognitive impairment.
• #2: I has consulted for Pfizer and I hold shares of Polyhymnia-TS Ltd., otherwise I have nothing to disclose.
• #3: What do we know about spatial navigation: It depends on the brain structures that are affected early in the course of AD – on the hippocampus and the posterior parietal cortex. Previous studies found out that spatial navigation impairment is present already in patients with amnestic mild cognitive impairment. Further we know that the same brain regions that form a navigation network are essential also for correct functioning of other cognitive domains, but the case-report studies reported an isolated spatial navigation impairment with preservation of other cognitive domains and vice versa. Even though some studies indicated this possible independence of spatial navigation from other cognitive functions, it is not fully clear whether they in fact overlap, which of them overlap and, if they do, to what extent.
• #4: Based on the results of previous studies, we hypothesized that spatial navigation in a real-space human analogue of the Morris Water Maze would be a separate cognitive domain. Specifically, using a sample of cognitively normal older adults and patients with amnestic MCI, we tested how the two basic types of spatial navigation – world-centered (allocentric) and body-centered (egocentric) are associated with six well-established cognitive domains above and beyond variations in age, gender and education.
• #5: 53 cognitively normal elderly and 55 amnestic MCI patients meeting Petersen‘s criteria were recruited in this cross sectional study. They underwent a clinical and laboratory evaluation, a detailed neuropsychological and spatial navigation testing in the real-space human analogue of the MWM and an MRI brain scan with subsequent automated MRI volumetry.
• #6: Neuropsychological battery was chosen to cover all 6 cognitive domains including verbal memory, non-verbal memory, visuospacial functions, executive functions, attention and working memory and language. Where possible the each domain was examined by two neuropsychological tests.
• #7: Brain MR images were obtained on a 1.5T scanner. 170 contiguous coronal brain images were acquired using T1-weighted 3-dimensional Fast Field Echo sequence with slice thickness 1.0 mm with no gap and isometric voxel size 1x1x1.25 . Automated MRI volumetry of right and left hippocampus and total brain volume was performed using FreeSurfer algorithm. (The hippocampal volumes in each individual were normalized using the total intracranial volume to control for variation in head size.)
• #8: For spatial navigation testing we used the real-space human analogue of the MWM - a circular arena approximately 3 m in diameter and 3 m high, where the task was to locate an invisible goal in the arena using a start postition and 2 orientation cues for navigation. The subjects moved to the goal with a long pole indicating their position.
• #9: Depending on the subtask, the start position or orientation cues could be used for navigation to the invisible goal. Therefore, we were able to test separately two basic types of navigation – allocentric (world-centered, hippocampus-dependent) and egocentric (self-centered, parietal cortex-dependent).
• #10: There were less females among the aMCI patients, who were further less educated and had a lower MMSE score and than the control group.
• #11: According to the Petersen‘s criteria the aMCI patients were impaired on verbal and non-verbal memory tests. The aMCI group was further impaired on executive functions and language compared to the control group.
• #12: Further aMCI patients performed worse on allocentric and egocentric navigation and had lower right and left hippocampal volumes than the control group. The difference in total brain volume was not significant.
• #13: Initial correlation analyses revealed that among demographic variables only age was low to moderately correlated with spatial navigation performance, especially allocentric navigation, in aMCI and control groups.
• #14: First, to to support our previous findings that spatial navigation has physiological underpinning in the right hippocampus, we tested whether spatial navigation, especially allocentric navigation, would be associated with right hippocampal volume. In these main analyses we used multivariate linear regression to model the relationship between a basic covariate model consisting of age, gender and education plus a variable of interest in this case it was a right hippocampal volume and spatial navigation performance in this case it was allocentric navigation. Thus change in model fit expressed as r-squared was assessed by adding right hippocampal volume to a model already adjusted for gender, age and education. (This implicates that this regression line is composed of covariate variables and a right hippocampal volume.) Using this approach we found an association between right hippocampal volume and allocentric navigation performance in the aMCI group. Where the right hippocampal volume explained additional 26% of variance in the allocentric navigation performance. There was no association between right hippocampal volume and allocentric navigation in the control group.
• #15: Among cognitive domains, there was an association between executive functions and allocentric navigation performance in the aMCI group. Where the executive functions explained additional 11% of variance in the allocentric navigation performance. There was no association between executive functions and allocentric navigation in the control group.
• #16: Further, there was an borderline association between visuospacial functions and allocentric navigation performance in the aMCI group. There was no association between visuospacial functions and allocentric navigation in the control group. Associations between other cognitive functions and allocentric navigation were not significant.
• #17: Finally, there was an association between verbal memory and egocentric navigation performance in the aMCI group. Where the verbal memory explained additional 9% of variance in the egocentric navigation performance. There was no association between verbal memory and egocentric navigation in the control group. Associations between other cognitive functions and egocentric navigation were not significant.
• #18: Although these two types of navigation are supported by different brain structures, they have strong interconnections and cooperate closely with each other. We tested this interrelation – ie whether egocentric navigation would be associated with allocentric navigation. There was an association between egocentric and allocentric navigation performance in the aMCI group. Where the egocentric navigation explained additional 43% of variance in the allocentric navigation performance. There was also an association between egocentric and allocentric navigation performance in control group. Where the egocentric navigation explained additional 12% of variance in the allocentric navigation performance.
• #19: Our results suggested that in amnestic MCI, allocentric navigation is associated very low with executive and marginally with visuospatial performance. Egocentric navigation is associated very low with verbal memory performance. We did not find associations between spatial navigation and other cognitive functions in aMCI. Further the results supported our previous findings that allocentric navigation is associated with right hippocampal volume in aMCI. Finally or results suggested that spatial navigation is associated with age in aMCI.
• #20: Our results suggested that in cognitively normal elderly, spatial navigation is not associated neither with cognitive performance nor with right hippocampal volume, but there is an association between allocentric navigation and age.
• #21: Finally our results suggested that both types of navigation are partially associated with each other, especially in amnestic MCI.
• #22: Spatial navigation appears to be a separate cognitive entity that is impaired early in the course of AD, where hippocampal atrophy plays a major role. It seems that spatial navigation testing yields an additional information next to well established cognitive domains like memory, executive and visuospatial functions. Spatial navigation testing may thus be useful in assessment of people at risk of AD at the stage of MCI and may be therefore used in a wider clinical setting.
• #23: Finally I would like to thank my colleagues who patricipated at this study. Thank you for your attention.
• #24: Further we also found an association between left hippocampal volume and allocentric navigation performance in the aMCI group. Where the left hippocampal volume explained additional 12% of variance in the allocentric navigation performance. There was no association between left hippocampal volume and allocentric navigation in the control group.
• #25: Finally we did not find any association between total brain volume and allocentric navigation performance in the aMCI and control groups. We also did not find any association between right and left hippocampi and total brain volume and egocentric navigation performance in the aMCI and control groups.