Abstract
Abstract: :
Purpose:Retinal arteriovenous ratio (AVR) is a predictor for various vascular outcomes, yet few studies focus specifically on its venous component. This study investigates factors associated with retinal venular diameter in an older population. Methods: The Blue Mountains Eye Study examined 3654 Australian residents aged 49 years or older in two postcode regions in 1992–4. Retinal venule diameters were measured from digitized images in 3368 participants and summarised as central retinal venular equivalents. Factors associated with venule caliber were assessed both continuously and dichotomously for widest vs other and narrowest vs other quintiles. Results: Mean venule diameter was 224.4 microns (224.2–225.6) overall, narrowing significantly with age (by 3 microns per decade, p for trend <0.0001). A final multivariable model including age, gender, fibrinogen, smoking, systolic blood pressure (BP), body mass index (BMI), hematocrit and HDL cholesterol was developed. After multivariable adjustment, venules were significantly narrower in men than women (mean 223 vs 226.5 microns) and narrowed with increasing systolic blood pressure (by 0.8 microns per 10 mmHg, p<0.0001), and hdl cholesterol level (by 1.9 microns per mmol/L, p=0.04). Retinal venules were significantly wider in current smokers (mean 232 vs 223) and widened with increasing fibrinogen level (by 1.6 microns per mg/dL, p<0.0001), BMI (by 0.3 microns per kg/m2, p<0.001) and hematocrit (by 11 microns per 0.1 unit, p<0.0001). A trend toward wider venules in participants with impaired fasting glucose was not statistically significant after age–gender or multivariable adjustment (p=0.1). There was no obvious trend to suggest association between mean venule caliber and creatinine, cholesterol, alcohol use, or diabetes. Male gender, and increasing age and systolic BP were significantly associated with higher odds of having the narrowest quintile of venule caliber, and lower odds of the widest. Current smoking and increasing haemoglobin levels were associated with lower odds of having the narrowest quintile of venule caliber, and higher odds of the widest. These findings were essentially unchanged when participants with diabetes were excluded from the analysis. Conclusions: This study identifies associations with retinal venule caliber which may influence AVR, so may need inclusion in predictive modelling. Venule and arteriolar diameters may reflect different systemic phenomena, warranting separate attention.
Keywords: clinical (human) or epidemiologic studies: risk factor assessment • retina