A digital measure of eye movements during reading sensitively captures oculomotor and speech dysfunction, early changes, and disease progression in ataxias.

Abstract

Objective: Sensitive behavioral measures are needed for clinical trials in ataxias and other neurodegenerative diseases. We hypothesized that quantitative analysis of eye movements during a natural multi-component task (passage reading) could produce a measure capable of capturing subclinical signs and disease progression.

Methods: Binocular gaze sampled at 1000 Hz was collected from 102 individuals with ataxia (including 36 spinocerebellar ataxias, 12 Friedreich's ataxia, and 5 multiple system atrophy) and 70 healthy controls. Longitudinal data were available for 26 participants with ataxia in the ongoing natural history study. The Reading Eye Abnormality Digital (READ) score was developed by training a regression model to aggregate saccade and fixation kinematics.

Results: Mean displacement of fixations, the number and frequency of saccades, and the proportion of regressive saccades were related to oculomotor dysfunction, speech dysfunction, and overall ataxia severity. The READ score was reliable (ICC=0.96, p<0.001) and correlated with Brief Ataxia Rating Scale total score (r=0.82, p<0.001), oculomotor (r=0.52, p<0.001) and speech (r=0.73, p<0.001) subscores, and patient reports of function. The READ score detected subclinical oculomotor (AUC=0.69, p=0.02) and speech (AUC=0.72, p<0.001) signs and disease progression (d=0.36, p=0.03). The Brief Ataxia Rating Scale was less sensitive to progression (d=0.27, p=0.08).

Interpretation: Digital measures of eye movements are a promising approach for sensitively measuring ataxia in clinical trials (including early-stage disease) and may have utility in other neurodegenerative diseases affecting speech or ocular control.