Risk of incident atrial fibrillation in patients presenting with retinal artery or vein occlusion: a nationwide cohort study

Abstract Background The inter-relationships of atrial fibrillation (AF) to retinal vascular occlusions (whether retinal artery occlusion (RAO) or retinal venous occlusion (RVO)) remain unclear. It is unknown if a presentation of retinal artery or venous occlusions may indicate a new onset cardiac ar...

Full description

Bibliographic Details
Main Authors: Christine Benn Christiansen, Christian Torp-Pedersen, Jonas Bjerring Olesen, Gunnar Gislason, Morten Lamberts, Nicholas Carlson, Mathias Buron, Nikolai Juul, Gregory Y. H. Lip
Format: Article
Language:English
Published: BMC 2018-05-01
Series:BMC Cardiovascular Disorders
Subjects:
Online Access:http://link.springer.com/article/10.1186/s12872-018-0825-1
Description
Summary:Abstract Background The inter-relationships of atrial fibrillation (AF) to retinal vascular occlusions (whether retinal artery occlusion (RAO) or retinal venous occlusion (RVO)) remain unclear. It is unknown if a presentation of retinal artery or venous occlusions may indicate a new onset cardiac arrhythmia. To shed light on this association, we investigated the risk of new onset AF in patients with known RAO and RVO. Methods Patients with retinal occlusions from 1997 to 2011 were identified through Danish nationwide registries and matched 1:5 according to sex and age. Cumulative incidence and unadjusted rates of AF according to retinal vascular occlusions (i.e. RAO or RVO) were determined. Hazard ratios (HR) of AF according to retinal vascular occlusion were adjusted for hypertension, diabetes, vascular disease and prior stroke/systemic thromboembolism/transient ischemic attack. Results One thousand three hundred sixty-eight cases with retinal vascular occlusions were identified (median age 71.4 (inter quartile range (IQR); 61.2–79.8), 47.3% male). RAO constituted 706 cases (51.6%) and RVO 529 (38.7%). The rate of incident AF amongst all cases with retinal vascular occlusion was 1.74 per 100 person-years (95% confidence interval (CI), 1.47–2.06) compared to 1.22 (95% CI, 1.12–1.33) in the matched control group. The rate of AF in RAO was 2.01 (95% CI, 1.6–2.52) and 1.52 (1.15–2.01) in RVO. HRs of incident AF adjusted for cardiovascular comorbidities were 1.26 (95% CI; 1.04–1.53, p = 0.019) for any retinal vascular occlusion, 1.45 (95% CI; 1.10–1.89, p = 0.015) for RAO, and 1.02 (95% CI; 0.74–1.39, p = 0.920) for RVO. Conclusions A new diagnosis of retinal vascular occlusion in patients without prior AF was associated with increased risk of incident AF, particularly amongst patients with RAO. Awareness of AF in patients with retinal vascular occlusions is advised.
ISSN:1471-2261