Eye Movements Expose Hidden Executive Decline in Late-Life Depression

Depression in older adults has long been one of the most elusive conditions in geriatric medicine. Unlike the memory lapses that flag dementia, the executive dysfunction that so often accompanies late-life depression tends to hide behind vague complaints of poor concentration, indecisiveness, and mental fatigue. Clinicians have few tools that can measure this impairment objectively, relying instead on lengthy neuropsychological batteries that older patients may find exhausting and that can be confounded by education, language, and motivation. Now a team of researchers in Taiwan has shown that a remarkably simple behavior—the flick of the eye from one point to another—may reveal exactly the cognitive fault lines that standard testing struggles to expose. The study, published in BMC Geriatrics, used video-based eye tracking to compare the saccadic eye movements of older adults with late-life depression against those of age-matched healthy controls, and the differences it found were both specific and telling.

The research focused on a task that cognitive neuroscientists have refined over decades: the interleaved pro-saccade and anti-saccade paradigm. In a pro-saccade trial, a participant simply looks toward a sudden visual target, an almost reflexive action governed by ancient circuitry linking the retina to the superior colliculus in the brainstem. In an anti-saccade trial, the instruction reverses: the participant must suppress the reflexive pull toward the target and instead look in the exact opposite direction. That act of suppression is anything but trivial. It depends on top-down executive control implemented by frontal-subcortical circuits—the same prefrontal networks that degrade in depression and that underpin planning, inhibition, and working memory. Because the reflexive and voluntary components of eye movement can be measured with millisecond precision, the task offers a window into executive function that is fast, non-invasive, and largely immune to the practice effects and cultural biases that plague paper-and-pencil tests.

The study enrolled 29 older adults diagnosed with late-life depression and 19 cognitively healthy controls of comparable age. Depression severity was quantified with the Geriatric Depression Scale-15, a widely used screening instrument, while global cognitive status was assessed with the Montreal Cognitive Assessment, or MoCA, a sensitive measure of mild cognitive impairment. Each participant then completed the interleaved saccade task while a high-speed video-based eye tracker recorded the trajectory, timing, and direction of every eye movement. The researchers extracted a rich set of measures: the rate of direction errors in anti-saccade trials, saccadic reaction times, the frequency of short-latency responses, the kinematic metrics of the movements themselves such as velocity and amplitude, and a less familiar quantity called voluntary override time, or VOT, which captures how long the executive system takes to override the reflexive response.

The headline finding was unambiguous. Older adults with late-life depression made significantly more direction errors on anti-saccade trials than the healthy controls did. In other words, when the task demanded that they suppress the automatic urge to glance at a flashing target, their eyes betrayed them more often. This is precisely the signature one would expect from weakened prefrontal inhibitory control, because the anti-saccade error rate is one of the most reliable behavioral indices of frontal-lobe integrity in the cognitive neuroscience literature. Importantly, the deficit was selective. The basic machinery of eye movement appeared intact: saccade metrics such as peak velocity and amplitude did not differ reliably between the groups, suggesting that the depression-related impairment lay not in the motor periphery or the brainstem reflex circuit itself but in the higher-order control system that decides whether the reflex should be allowed to run.

A second, subtler result added an intriguing twist. Short-latency pro-saccades—reflexive glances toward the target—were actually faster in the depressed group than in the controls. Rather than a general slowing, which clinicians often associate with depression in the form of psychomotor retardation, the eye-movement data revealed a pattern of disinhibition: the reflexive pathway fired more quickly and the voluntary brake applied less effectively. This combination of faster reflexes and weaker suppression points toward an imbalance in the competitive interaction between the superior colliculus, which generates saccade commands, and the frontal cortical networks that modulate them. In computational terms, the threshold for triggering a reflexive response appears lower while the top-down signal that raises that threshold appears weaker, a double hit that compounds executive failure.

Perhaps the most clinically significant finding emerged from correlations within the depressed group. MoCA scores, which index global cognitive performance, were significantly associated with both anti-saccade error rates and voluntary override time, meaning that patients whose eye movements showed the weakest executive control also scored lowest on the standard cognitive assessment. Yet depression severity, as measured by the GDS-15, showed no correlation with any of the oculomotor measures. This dissociation is important. It suggests that the saccade abnormalities track the cognitive dimension of late-life depression rather than the mood symptoms themselves, positioning the eye-tracking paradigm as a potential probe of the executive dysfunction that predicts poorer treatment outcomes, greater functional decline, and elevated risk of progression to dementia in this population.

The implications reach beyond the clinic. Late-life depression affects millions of older adults worldwide and is increasingly understood not merely as a mood disorder but as a condition with genuine neurobiological consequences, involving vascular injury, inflammation, and neurodegeneration in frontal-subcortical circuits. Executive dysfunction is one of its most disabling features, undermining medication adherence, financial decision-making, and the ability to benefit from psychotherapy. Yet because mood symptoms fluctuate and self-report is unreliable, the cognitive component often goes unrecognized until it has already compounded disability. A two-minute eye-tracking test that could be administered in a general practitioner’s office would offer a way to detect this component early, monitor it over time, and potentially identify which patients need closer neuropsychological follow-up.

The authors are careful to frame their results as preliminary. The sample sizes were modest, with 29 patients and 19 controls, and the study design was cross-sectional, capturing a single snapshot rather than tracking change over time. Larger longitudinal cohorts will be needed to establish whether saccade measures are reliable across repeated testing sessions, whether they predict clinical trajectories such as conversion to dementia, and whether they respond to treatment as depression remits. Questions also remain about how the measures perform in patients taking different classes of antidepressant medication, in those with coexisting medical illness, and in populations with different educational backgrounds. The researchers themselves state that further work is required to establish the reliability and clinical utility of saccade-based assessment in late-life depression.

Even so, the study adds to a growing movement in neuroscience and psychiatry toward behavioral measures that bypass self-report entirely. Eye movements occupy a privileged position in this effort because the circuits that control them are well mapped, their timing can be resolved to the millisecond, and the tasks that engage them are simple enough for frail or cognitively impaired patients to complete. The anti-saccade task in particular has a long pedigree as a probe of frontal function, having revealed inhibitory deficits in schizophrenia, Parkinson’s disease, attention deficit hyperactivity disorder, and normal aging. Extending it to late-life depression connects that literature to one of geriatric psychiatry’s most pressing clinical problems and demonstrates that a disorder of mood leaves measurable fingerprints on one of the body’s fastest and most automatic behaviors.

What makes the finding resonate is its elegance: the same neural machinery that lets a person resist an impulsive glance at a flashing light also lets them resist impulsive decisions, hold plans in mind, and steer attention where it needs to go. When that machinery falters in late-life depression, the failure shows up first where it can be measured most precisely—in the eyes. If larger studies confirm the reliability of these saccade-based markers, the humble eye movement could become a routine vital sign of executive health in aging, transforming how clinicians detect, track, and ultimately treat the cognitive shadow that follows depression into later life.

Subject of Research: Executive dysfunction in late-life depression assessed through saccadic eye-movement control

Article Title: Altered saccade control reveals executive dysfunction in late-life depression

Article References: Lee, Y.-T., Tai, Y.-H., Huang, L.-K., Chen, K.-Y., Chen, T.-L., Chen, C., & Wang, C.-A. (2026). Altered saccade control reveals executive dysfunction in late-life depression. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08427-1

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08427-1

Keywords: late-life depression, saccades, anti-saccade task, executive control, eye tracking, superior colliculus, geriatric psychiatry, MoCA, cognitive control, prefrontal cortex, voluntary override time, BMC Geriatrics