Training effects in oculography in multiple sclerosis patients: a one-year longitudinal study

Abstract Purpose Longitudinal quantitative data on eye movement abnormalities in people with multiple sclerosis (PwMS) are scarce, and the impact of repeated oculomotor testing on these measures is unknown. We aimed to use infrared oculography to characterise one-year changes in oculomotor metrics in PwMS and to determine whether frequent exposure to eye movement tasks influences these measurements. Methods At baseline, 3- and 12 months 86 PwMS and 22 healthy controls (HCs, baseline, and 3 months only) were prospectively assessed using infrared oculography (DEMoNS protocol), including fixation, pro-saccades, express-saccades, anti-saccades, and double-step saccades. Participants received separate smartphone-based eye-tracking tasks, daily in the first month and weekly up to 12 months. Results At baseline PwMS showed increased 15° latency (+ 18ms, P = 0.029) and versional dysconjugacy (index + 0.08, P = 0.028) on pro-saccades compared to HCs. At 12 months PwMS showed increased performance in anti-saccades (correct responses + 9%, P < 0.001) and double-step saccades (correct responses + 8%, P < 0.001; error − 0.39°, P < 0.001). Changes in double-step saccade accuracy correlated with disability (std. β = 0.28, P = 0.02). Changes in anti-saccade performance correlated with cognitive processing speed (std. β = 0.23, P = 0.049) and verbal memory (std. β = 0.30, P = 0.01). Conclusions Frequent oculomotor tasks may induce training effects in complex oculography tasks, such as anti-saccades and double-step saccades. Simple oculomotor tasks, such as prosaccades and parameters related to internuclear ophthalmoplegia, seem to be relatively unaffected by training effects. These effects should be considered in MS trials using oculomotor outcomes.