Purchase this article with an account.
G. Burlutsky, T. Y. Wong, G. Liew, N. Cheung, P. Mitchell, J. J. Wang, Blue Mountains Eye Study; Threshold Effect of Retinal Vessel Calibre on Diabetes and Retinopathy Lesions: The Blue Mountains Eye Study. Invest. Ophthalmol. Vis. Sci. 2008;49(13):2113. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We aimed to test whether a threshold effect exists in the relationships between retinal vessel calibre and (1) the presence of diabetes and (2) isolated retinopathy lesions in persons without diabetes.
The Blue Mountains Eye Study baseline data included 3654 persons aged 49+ years. Vessel calibre was measured from retinal photographs using standardised computer-assisted techniques to calculate central retinal arteriolar (CRAE) and venular (CRVE) equivalents, representing mean arteriolar and venular calibre of each eye, respectively. Right eye data of 3536 subjects were used to assess associations with diabetes prevalence, and eye-specific data of 3106 subjects without diabetes to assess associations with isolated retinopathy lesions. Change-point model was programmed1 using a macro function. Logistic regression models adjusting for age, gender, systolic BP, BMI and smoking were used to estimate odds ratios (OR) and 95% confidence intervals (CI) associated with change in vessel calibre.
A U-shaped, non-linear relationship best described the association between CRAE and diabetes prevalence. For CRAE >153.7 µm but <l88 µm (middle group), no association existed. Beyond these two limits, the likelihood of having diabetes increased compared to persons with CRAE in the middle group: CRAE ≤153.7 µm, OR 1.80, CI 0.96-3.36; CRAE ≥l88 µm, OR 1.73, CI 1.30-2.29. A similar non-linear association existed between CRVE and presence of isolated retinopathy lesions. For CRVE >176 µm but <260 µm (middle group) no association existed. Beyond these limits, the likelihood of having isolated retinopathy lesions increased compared to persons with CRVE in the middle group: CRVE ≤176 µm, OR 2.39, CI 1.03-5.54; CRVE ≥260 µm, OR 1.71, CI 1.05-2.78.
These findings suggest a ‘normal’ physiological range for both arteriolar and venular calibre in the population beyond which associations with high risk of diabetes or retinopathy lesions are evident. Future studies should consider possible threshold effects of retinal vessel calibre when investigating their associations with systemic vascular and metabolic conditions.1. Ulm, Statistics in Medicine 10:341-349, 1991
This PDF is available to Subscribers Only