Effects of aging on the neural networks underlying visuomotor adaptation

Abstract

Visuomotor adaptation is the ability to correct movements in response to an error and becomes increasingly important during the aging process when assistive devices may become necessary. Neuroimaging has observed age-related atrophy in several brain regions implicated with visuomotor adaptation, and the integrity of white matter connections between these regions is also known to decline. The present research is a three-part study that assessed age-related changes in brain function and structure during healthy aging. Functional magnetic resonance images (fMRI) and diffusion tensor images (DTI) were acquired from 37 neurologically healthy, right handed participants from three age groups: young (N = 12, age range = 22-39), middle-aged (N = 13, age range = 41-58), and older (N=12, age range = 65-80). During two fMRI scans, participants completed a novel moving-target task in which visual feedback of a cursor on a computer screen was manipulated in order to elicit a visuomotor response. In the first paper of this thesis, the behavioural and neural differences between the first fMRI scan (early adaptation trials) and second scan (late adaptation trials) were analyzed for the young age group. This study aimed to outline the visuomotor transformations associated with a novel task that incorporates more natural behaviours. Visuomotor behaviour was examined through measurements of the mean cursor velocity and number of reversals in cursor direction. Although the interaction between scan number and cursor type was found to be insignificant for cursor velocity along with an insignificant main effect of scan number, a repeated-measures analysis of variance (ANOVA) found a significant main effect of cursor type. Pairwise comparisons revealed significant differences between normal and x-flip cursor conditions, and x- and y-flip conditions. Following an insignificant interaction between scan and condition, a significant main effect of cursor condition (normal, x-flip, or y-flip) for the number of reversals was determined. Pairwise comparisons revealed that participants made fewer reversals during the normal cursor condition when compared to either the x- or y-flip conditions. Areas of neural activity related to visuomotor transformations were subsequently identified through fMRI using contrast to examine regions of greater signal change for distorted vs. normal cursor conditions. A random-effects General Linear Model (GLM) was then performed to examine differences in neural activity between the two fMRI scans. Frontal regions and the postcentral and anterior cingulate gyri were activated for the early adaptation trials, while the superior temporal, inferior frontal and parahippocampal gyri, as well as the putamen and parietal regions were activated for the late trials.