Regional Brain Structure Mediates Age‐Related Decline in Eye Movement Control

Abstract

AbstractBackgroundEye movements (EM) are considered potential biomarkers of neurodegeneration because they are altered in patients with neurodegenerative disease, including Alzheimer’s disease (AD). Alterations in EM control have also been observed in old age, with particularly strong age effects on saccadic latency, inhibitory errors in the antisaccade task, and smooth pursuit accuracy. The goal of our study was to characterize the neural pathways of age‐related changes in these EM outcomes.MethodThe analysis was based on 8,318 participants (aged 30 to 95 years) from the population‐based Rhineland Study. EMs were recorded using video‐based infrared oculography at 1,000 Hz. Volumes and thickness measures of selected brain structures were obtained using FreeSurfer from 3 Tesla T1‐weighted MR images. Relations of brain structure with EM outcomes were quantified using multivariable linear regression models. Brain structure variables were also analyzed as potential mediators in the relation between age and EM outcomes using causal mediation analysis under a counterfactual framework.ResultIn women, higher volume of the right thalamus and right globus pallidus was associated with lower antisaccade latency, whereas higher cortical thickness in the anterior cingulate cortex was associated with higher antisaccade latency. Higher cortical thickness in right supramarginal gyrus and right paracentral lobule and sulcus was associated with fewer antisaccade errors in women. There were no significant associations between regional brain structure and antisaccade performance in men. Higher cortical thickness in the right calcarine sulcus was associated with higher smooth pursuit velocity gain in both men and women. Additionally, individual brain structures mediated up to 11% of the age‐related decline in EM performance.ConclusionOur results support that EMs indeed reflect neurodegenerative processes as changes in brain structure partly account for age‐related decline in oculomotor performance. Relevant brain regions include both brain regions vulnerable to AD pathology and brain regions that are typically spared, suggesting that both “normal” age‐related and pathological mechanisms may be causative for decline in EM control. Our results highlight the importance of taking into account normal age‐related changes when considering EMs as early biomarkers for AD. Further work is needed to better distinguish normal age‐related changes from pathological change.