The Association Between Eye Disease and Incidence of Dementia: Systematic Review and Meta-Analysis

Author: Feng J, Huang C, Liang L, Li C, Wang X, Ma J, Guan X, Jiang B, Huang S, Qin P. 

Geographical coverage: Europe, Asia and North America

Sector: Eye disease

Sub-sector: Risk of dementia

Equity focus: Not reported

Study population: Adults affected by four common eye diseases (age-related macular degeneration [AMD], glaucoma, diabetic retinopathy [DR] and cataract)

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: Dementia affects more than 55 million people worldwide and its prevalence is rising. Identifying modifiable risk factors is therefore essential. Visual impairment has been proposed as one such factor because the retina and the brain share microvascular characteristics. AMD, glaucoma, DR and cataract are leading causes of vision loss in older adults (e.g. cataract affected over 78 million people in 2020). These ocular conditions may mirror neuro-degenerative changes and have been linked to cognitive decline, yet previous studies—particularly for cataract and glaucoma—have yielded inconsistent results. A clearer understanding of these associations could inform early interventions for dementia prevention.

Objective:

To examine whether the four common eye diseases (AMD, glaucoma, DR and cataract) are associated with an increased risk of all-cause dementia and of specific dementia subtypes—Alzheimer’s disease (AD) and vascular dementia (VaD).

Main findings:

The review included 25 cohort studies encompassing 11 410 709 participants from Europe (8 studies), Asia (11) and North America (6). Follow-up exceeded five years in most studies, all of which adjusted for age and sex; methodological quality was rated high in 10 studies and moderate in 15. Pooled risk ratios (RR) indicated a higher incidence of all-cause dementia with AMD (RR = 1.29; 95 % CI 1.13 – 1.48), glaucoma (RR = 1.16; 95 % CI 1.03 – 1.32), DR (RR = 1.40; 95 % CI 1.21 – 1.63) and cataract (RR = 1.23; 95 % CI 1.09 – 1.40). Similar associations emerged for AD: AMD (RR = 1.27; 95 % CI 1.06 – 1.52), glaucoma (RR = 1.18; 95 % CI 1.02 – 1.38), DR (RR = 1.21; 95 % CI 1.04 – 1.41) and cataract (RR = 1.22; 95 % CI 1.07 – 1.38). No eye disease showed a significant link with VaD. Sub-group analyses for DR and all-cause dementia corroborated the primary results. Meta-regression suggested that geographical region partly explained heterogeneity for several associations (e.g. AMD with all-cause dementia and with AD; glaucoma with dementia and with AD).

Methodology:

Searches of PubMed, Embase and Web of Science up to 25 August 2022 identified English-language cohort studies evaluating the eye diseases and dementia outcomes. Reference lists of eligible papers and relevant reviews were also screened. Two reviewers independently performed screening, data extraction and quality appraisal using the Newcastle–Ottawa Scale; disagreements were resolved by discussion or third-reviewer adjudication. Random-effects meta-analysis generated pooled estimates; heterogeneity was assessed with the I² statistic, and publication bias with funnel plots and Egger’s test. Sub-group and meta-regression analyses explored potential modifiers and sources of heterogeneity.

Applicability/external validity:

Although the evidence base is large, results are tempered by heterogeneity across studies, regional imbalances (few data from low-income settings) and possible misclassification arising from self-reported ocular disease or medical-record dementia diagnoses. Consequently, findings may not generalise to under-represented regions or populations with different risk profiles. Future prospective studies in diverse cohorts with robust diagnostic ascertainment and longer follow-up are warranted.

Geographic focus:

The included studies originated from Europe, Asia and North America; no geographical restrictions were imposed by the reviewers.

Summary of quality assessment:

Overall confidence in the review’s conclusions is medium. The search strategy was comprehensive, eligibility criteria explicit, and assessments of study quality and data extraction were conducted independently in duplicate. Study characteristics were well documented and appropriate meta-analytic methods employed, with heterogeneity explored. Limitations include restriction to English-language publications and the absence of an excluded-study list.