Abstract
Purpose: :
Variations in retinal vessel diameters measured from fundus photographs (static diameters) have been shown to be risk markers for diabetic retinopathy (DR). We investigated the association of both static and dynamic retinal vessel diameter (induced by flicker-light stimulation) with DR.
Methods: :
We recruited 224 persons with diabetes (85 with type 1 and 139 type 2 diabetes) from the International Diabetes Institute, Melbourne, Australia. Static retinal vessel diameter was measured from digital fundus photographs using a computer-assisted method, and summarized as central retinal artery or vein equivalent (CRAE or CRVE). Dynamic retinal vessel diameter, defined as percentage increase in vessel diameter after flicker-light stimulation relative to baseline diameter, was measured using Dynamic Vessel Analyzer (IMEDOS, Germany). The presence of DR was graded from fundus photographs.
Results: :
After adjusting for age, gender, diabetes duration, hemoglobin A1c, systolic blood pressure, cholesterol, triglycerides and current smoking, CRVE and dynamic venodilation simultaneously, wider static retinal venular diameter and reduced dynamic response to flicker-induced venodilation were independently associated with the presence of DR: odds ratio (OR) 1.98 (95% confidence interval [CI] 1.05-3.73, comparing highest vs. lowest tertile of static venular diameter; and OR 2.62 (95% CI 1.45-4.73) comparing lowest vs. highest tertile of flicker-induced venodilation. There was an additive effect of static and dynamic retinal venular measures on the association: patients in the highest tertile of static venular diameter and lowest tertile of dynamic venodilation had a seven-fold risk of DR (OR 7.77, 95% CI 2.84-21.2), compared to those with the lowest tertile of static venular diameter and highest tertile of dynamic venodilation. We did not find significant associations of static or dynamic arteriolar diameter with DR.
Conclusions: :
Larger static retinal venular diameter and reduced dynamic flicker-induced venodilation are independently associated with DR. Retinal vessel image analyses may allow further understanding of the underlying pathologic processes related to DR.
Keywords: diabetic retinopathy • retina • imaging/image analysis: clinical