Anti-saccade performance predicts executive function and brain structure in normal elders

Abstract

Objective: To assess the neuropsychological and anatomical correlates of anti-saccade (AS) task performance in normal elders.

Background: The AS task correlates with neuropsychological measures of executive function and frontal lobe volume in neurological diseases, but has not been studied in a well-characterized normal elderly population. Because executive dysfunction can indicate an increased risk for cognitive decline in cognitively normal elders, we hypothesized that AS performance might be a sensitive test of age-related processes that impair cognition.

Method: The percentage of correct AS responses was evaluated in 48 normal elderly subjects and associated with neuropsychological test performance using linear regression analysis and gray matter volume measured on magnetic resonance imaging scans using voxel-based morphometry.

Results: The percentage of correct AS responses was associated with measures of executive function, including modified trails, design fluency, Stroop inhibition, abstraction, and backward digit span, and correlated with gray matter volume in 2 brain regions involved in inhibitory control: the left inferior frontal junction and the right supplementary eye field. The association of AS correct responses with neuropsychological measures of executive function was strongest in individuals with fewer years of education.

Conclusions: The AS task is sensitive to executive dysfunction and frontal lobe structural alterations in normal elders.

Figure 1. Eye movement paradigm and traces

The eye movement paradigm consisted of 1000 ms of central fixation followed by a 200 ms gap and an eccentric target for 1000 ms (illustrated as moving to the right) (A). Relative to target position (B), representative eye traces for a correct response (C), a self-corrected error response (D) and an uncorrected error response (E) are shown. In B-E, down indicates left and up indicates right; this particular trial indicates a target to the left.

Figure 2. ROIs used in VBM analysis

Regions of interest are visualized in axial views at z values of 35 mm (A), 45 mm (B), 55 mm (C), as well as a rendered whole brain image with a cutout at y value of 0 mm (D) on the CH2bet template. Brown = left DLPFC, red = left IFJ, light blue = left FEF, yellow = left SEF, orange = right SEF, green = right FEF, dark blue = right IFJ, tan = right DLPFC, purple = left PEF, pink = right PEF.

Figure 3. AS performance predicts executive function

Scatter plots demonstrate the relationship between the percentage of correct AS responses and tests of executive function: modified trails (A), design fluency (B) and the Stroop task (C). Circles indicate subjects with low education (unfilled) and high education (filled). Lines indicate the trend for all subjects (thick, solid), subjects in the low education group (thin, solid) and subjects in the high education group (dotted). * indicates p < 0.05; ** indicates p < 0.001.