Abstract
Purpose: :
To explore associations of retinal arteriolar and venular caliber with diabetic retinopathy (DR) in the Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications Study (EDIC).
Methods: :
In the DCCT "primary prevention" cohort (no DR at entry), vessel diameters were measured in one eye per person with IVAN software (Nicola Ferrier, UW-Madison Engineering) at DCCT baseline (BL), 1 yr, and end-of-study (EOS, 4-9 yrs) in 473 persons; and at EDIC BL (DCCT EOS), 4 yrs, and 10 yrs in 479 persons. By visit, individual vessels were summarized into "central retinal arteriolar equivalent" (CRAE) and "central retinal venular equivalent" (CRVE) via revised Parr-Hubbard formulae. Multivariate logistic regressions examined odds ratios (OR) per caliber unit of earlier CRAE and CRVE changes for later DR progression (3+ steps, ETDRS person scale) and proliferative (PDR) onset. Models adjusted for age, gender, duration, mean arterial pressure (MABP), HbA1c, and DCCT treatment as covariates.
Results: :
Within DCCT overall, neither BL caliber nor 1 yr change predicted DR progression by EOS. However among adolescents, progressors had more dilated CRVE at BL (237 vs. 232µm; OR 1.05/µm [95% CI 1.01-1.11], p=0.04), and further CRVE dilation (vs. narrowing) from BL to 1 yr (+2 vs. -2 µm; OR 1.12/µm [1.03-1.22], p=0.01). Within EDIC, for DR progression by 10 yrs, progressors had more dilated CRVE at BL (231 vs. 227µm; OR 1.015/µm [1.002-1.028], p=0.02), and from BL to 4 yrs, less CRVE narrowing (p=0.02) and greater CRAE narrowing (p=0.03). These EDIC effects appeared limited to the former DCCT conventional treatment group, and to the subset with MABP <88 mm Hg at DCCT BL. For PDR onset by 10 yrs, progressors had further CRVE dilation (vs. narrowing) from BL to 4 yrs (+4 vs. -2µm; OR 1.07/µm [1.03-1.12], p<0.001), and greater CRAE narrowing from BL to 4 yrs (-5 vs. -2µm; OR 0.91/µm [0.86-0.96], p<0.001).
Conclusions: :
In DCCT, retinal vessel measurements did not predict DR development/progression, as seen previously in some epidemiologic studies - except in adolescents. However, in EDIC, venular dilation (expected) and arteriolar narrowing (novel) did predict DR progression. Concentration of the linkage between EDIC vessel change and DR progression in the DCCT conventional treatment group may represent long-term sequelae of not instituting strict metabolic control early.
Clinical Trial: :
http://www.clinicaltrials.gov NCT00360815
Keywords: diabetic retinopathy • blood supply • imaging/image analysis: clinical