Antisaccade task reflects cortical involvement in mild cognitive impairment

Abstract

Objective: The aims of this study were to examine executive dysfunction using an antisaccade (AS) task in normal elderly (NE) and patients with mild cognitive impairment (MCI) and Alzheimer disease (AD) as well as to evaluate the relationship between AS performance and cortical thinning within AD-associated regions.

Methods: We recorded eye movements in 182 subjects (NE: 118; MCI: 36; AD: 28) during an AS task. We also performed neuropsychological measures of executive function for comparison. Brain MRI scans were collected on most subjects, and cortical thickness was determined in 9 regions known to exhibit atrophy in AD dementia ("AD signature"). We investigated the relationships between AS and neuropsychological performance, as well as possible correlations between AS performance and cortical thickness.

Results: AS performance in MCI resembled that in NE; subjects with AD were impaired relative to both MCI and NE. In all subjects, AS performance correlated with neuropsychological measures of executive function, even after controlling for disease severity. In the subjects with MCI but not in NE, cortical thickness in frontoparietal AD signature regions correlated with AS performance.

Conclusions: The AS task is a useful measure of executive function across the AD spectrum. In MCI, AS performance may reflect disease burden within cortical brain regions involved in oculomotor control; however, AS impairments in NE may have etiologies other than incipient AD.

Figures

Figure 1. Oculomotor experiments

(A) Trial timing schematic. (B) Example responses. Bottom panel shows the eye movement response in a self-corrected error antisaccade trial; uncorrected error responses appear similar to prosaccades.

Figure 2. Oculomotor group performance

(A) Latency for horizontal prosaccades and (B) correct antisaccade responses, and incorrect (error) antisaccade responses for all subject groups. Brackets indicate significant between-group differences: *p < 0.01; **p < 0.001. (C) Antisaccade performance by group. Left panel is percentage correct, right is ASTotal (percentage of total trials performed correctly or self-corrected). Brackets indicate between-group differences: *p < 0.02; **p < 0.001. Individual data points have been distributed horizontally for visual clarity. Thick horizontal lines indicate median values; upper and lower boxes show the 25th and 75th percentiles. AD = Alzheimer disease; MCI = mild cognitive impairment; NE = normal elderly.