Vision Loss May Double Delirium Risk, Major Analysis Finds

Delirium is one of the most feared complications in medicine: an abrupt, fluctuating disturbance of attention and awareness that can overtake a hospitalized patient within hours, leaving them disoriented, agitated, or unnaturally drowsy. It strikes older adults hardest, afflicts patients in intensive care units, on surgical wards, in emergency departments and in palliative care alike, and it carries a heavy price tag of longer hospital stays, accelerated cognitive decline, institutionalization and death. Now a sweeping systematic review and meta-analysis published in the Journal of General Internal Medicine has added a striking and, until recently, underappreciated entry to the roster of delirium risk factors: impaired vision. Pulling together evidence from 44 studies and more than 26,000 participants, an international team of ophthalmology and internal medicine researchers reports that adults with visual impairment face roughly twice the odds of experiencing delirium compared with those who see well, across a remarkably diverse range of clinical settings.

The research team, led by Abdelrahman Abu Osba of Dalhousie University in Halifax alongside collaborators from the University of Calgary, the University of Manitoba and Dalhousie’s department of ophthalmology, set out to answer a deceptively simple question with rigorous methods: does poor eyesight predict delirium? They systematically searched MEDLINE, Embase and the Cochrane Central Register of Controlled Trials from their inception through September 2025, and screened and extracted data with two independent reviewers, a design intended to guard against the subjective drift that can plague narrative reviews. The review was prospectively registered with PROSPERO, the international registry of systematic review protocols, before the data were examined.

The studies that made the cut spanned the full geography of acute care. Some followed older patients undergoing hip fracture surgery, where postoperative delirium is notoriously common. Others enrolled medical inpatients, intensive care unit patients, stroke survivors, oncology patients, emergency department visitors and elderly patients in sub-intensive and palliative care. Visual impairment was defined in various ways, from measured acuity deficits to documented eye disease or self-reported difficulty seeing, and delirium was identified through a patchwork of diagnostic criteria and screening instruments. That heterogeneity, the authors acknowledge, is one of the central challenges of interpreting the pooled result.

When the researchers ran their primary analysis, they were deliberately conservative. They restricted it to the 16 studies that reported multivariable-adjusted estimates, meaning statistical models that already accounted for other delirium risk factors such as age, dementia, illness severity and medications. Even after that adjustment, visual impairment remained firmly associated with delirium: the pooled odds ratio was 2.02, with a 95 percent confidence interval of 1.29 to 3.16. In plain terms, patients whose vision was impaired had about double the odds of delirium compared with similar patients whose vision was intact, a difference that would be hard to dismiss as statistical noise.

The secondary analysis, which pooled all available estimates including those unadjusted for confounders, produced an even larger association: an odds ratio of 2.40, with a confidence interval of 1.81 to 3.18. That the adjusted figure is somewhat smaller is exactly what one would expect if some of the apparent link in raw data flows through correlated factors, but the persistence of a doubled risk after adjustment is the analysis’s most consequential finding. Both pooled estimates, however, came with substantial statistical heterogeneity, with I-squared values above 82 percent, signaling that individual studies varied considerably in the magnitude of the association they observed, likely reflecting their different populations, settings, definitions and analytic choices.

One might have expected the association to be far stronger in some settings than others, for example in surgical wards where sensory inputs are already disrupted by anesthesia and unfamiliar environments, or in the intensive care unit where delirium rates can exceed 50 percent. Interestingly, meta-regression across clinical settings did not detect statistically significant differences in the strength of the association by setting. The signal, in other words, appears remarkably portable, rising in hospitals, ICUs, emergency rooms and palliative units alike, which strengthens the case that the relationship reflects a genuine phenomenon rather than an artifact of any single care environment.

Why would failing eyesight prime the brain for acute confusion? The authors and the broader literature they cite point to several plausible and not mutually exclusive mechanisms. Sensory deprivation is a long-recognized trigger of psychotic-like and confusional experiences, and vision supplies the brain with a dominant stream of information for orienting to reality. When that stream degrades, the brain loses grounding cues, and in a vulnerable, often older and medically ill patient, the result may be the disorganized perception and attention that define delirium. There is also the shared-pathology argument: the vascular and neurodegenerative processes that damage the eye and optic pathways frequently damage the brain as well, meaning poor vision may partly act as a marker of a fragile, delirium-prone nervous system rather than a cause of it. Reduced visual input may further diminish engagement, worsen sleep-wake disruption and limit a patient’s ability to use clocks, calendars, glasses and familiar faces, all staples of multicomponent delirium prevention programs such as the Hospital Elder Life Program.

The authors are careful, and refreshingly so, about what their evidence can and cannot say. Because all of the included studies are observational, the certainty of the evidence, evaluated with the GRADE framework, was rated low, with uncertainty stemming primarily from the substantial between-study heterogeneity. Risk of bias was assessed using the Quality in Prognosis Studies tool, and the analysis itself was conducted according to PRISMA 2020 reporting standards. The verdict the team offers is precise: visual impairment should be regarded as a prognostic factor for delirium, a flag that identifies higher-risk patients, rather than an established modifiable cause. Whether correcting vision, with new glasses, cataract surgery or other ophthalmic care, actually lowers delirium incidence was not evaluated by any of the included studies and remains an open question that will require testing in a causal framework, ideally randomized trials.

Even with that caveat, the clinical implications are difficult to ignore. Visual impairment is extraordinarily common in aging populations; prevalence studies using the National Health and Aging Trends Study data suggest that a large fraction of Americans over 71 have some form of vision loss, and projections indicate the burden will grow substantially in the coming decades. Delirium is equally common, affecting a large share of hospitalized older adults, yet it is frequently missed and its risk factors incompletely addressed. If a simple assessment of vision at admission, paired with low-cost interventions such as ensuring patients have their correct glasses, adequate lighting and orientation aids, could contribute to delirium prevention, the public health payoff could be enormous. This meta-analysis does not prove that such measures will work, but it elevates the eye from an organ routinely ignored in delirium risk models to one that clinicians, researchers and hospital designers can no longer afford to overlook. The next chapter, testing whether restoring sight restores clarity of mind, is now an urgent priority.

Subject of Research: The association between visual impairment and delirium risk in adults

Article Title: The Association Between Visual Impairment and Delirium: A Systematic Review and Meta-Analysis

Article References: Abu Osba, A., Bondok, M., Ibrahim, A., Ahmad, S., Al-Ani, A., Mishra, A., & Benard-Seguin, E. (2026). The Association Between Visual Impairment and Delirium: A Systematic Review and Meta-Analysis. Journal of General Internal Medicine. https://doi.org/10.1007/s11606-026-10804-w

Image Credits: AI Generated

DOI: 10.1007/s11606-026-10804-w

Keywords: delirium, visual impairment, systematic review, meta-analysis, older adults, ophthalmology, risk factors, hospitalized patients, sensory deprivation, geriatric medicine, delirium prevention, vision loss