In a large-scale study of more than 1 million Medicare beneficiaries, investigators reported that atrial fibrillation was independently associated with an increased risk of retinal stroke.
The study, published in JAMA Network Open, examined the relationship between atrial fibrillation (AF) and retinal stroke using a retrospective cohort design. The investigators analyzed data from 545,072 patients with AF and an equal number of matched controls, drawn from Medicare claims between 2000 and 2020. The patients with AF experienced retinal stroke at a rate of 0.55 per 1,000 person-years compared with 0.50 per 1,000 person-years in matched controls without AF. After adjusting for covariates, AF was associated with a 14% higher risk of retinal stroke (adjusted hazard ratio [HR] = 1.14, 95% confidence interval [CI] = 1.02–1.28).
Secondary analyses revealed similar associations for related endpoints. AF was linked to retinal stroke when considered in any diagnostic position (adjusted HR = 1.12, 95% CI = 1.02–1.23) and to any retinal artery occlusion, including branch occlusions (adjusted HR = 1.21, 95% CI = 1.16–1.26). Cumulative incidence graphs further illustrated the persistent increased risk of retinal stroke among patients with AF over time.
AF was more strongly associated with cerebral ischemic stroke, with an adjusted HR of 1.73 (95% CI = 1.69–1.76). This stronger relationship underscored the broader vascular risks posed by AF.
The investigators employed statistical methods to strengthen the study's validity. Propensity score matching and overlap weighting were used to minimize baseline differences between the AF and control cohorts, ensuring robust and comparable groups. The comprehensive dataset included claims from office-based care, hospital outpatient clinics, emergency departments, and inpatient facilities, enhancing the reliability of retinal stroke diagnoses.
Despite these strengths, the investigators noted potential miscoding of retinal stroke in outpatient settings, where it may be confused with other disorders such as nonarteritic ischemic optic neuropathy. This limitation, along with the study’s focus on Medicare fee-for-service beneficiaries aged 66 and older, may have affected the generalizability of the findings.
The study’s negative control endpoints provided additional context. AF was not associated with central retinal vein occlusion (adjusted HR = 1.00, 95% CI = 0.78–1.27), validating the specificity of the primary findings. However, associations were observed with urinary tract infections (adjusted HR = 1.15, 95% CI = 1.13–1.16), cataracts (adjusted HR = 1.15, 95% CI = 1.12–1.17), and humeral fractures (adjusted HR = 1.12, 95% CI = 1.06–1.18), possibly reflecting increased health care interactions in patients with AF.
The study discussed potential relevance for cardiac monitoring in patients with cryptogenic retinal stroke but emphasized the need for further research before clinical recommendations can be established.
The research was funded by the National Institutes of Health (K23HL161426 and K23HL161426-S1).
Conflict of interest disclosures can be found in the study.