Retinal vessel caliber and the long-term incidence of age-related cataract: The Blue Mountains eye study

Objective: To determine whether local nutritional or ischemic factors are involved in cataract pathogenesis, we aimed to assess whether narrowed retinal vessel caliber predicted the long-term incidence of age-related cataract, as shown in 1 previous report. Design: Population-based cohort study. Participants: The Blue Mountains Eye Study examined 3654 baseline participants (1992–1994), 2335 (75.1% of survivors) after 5 years, and 1952 (75.6% of survivors) after 10 years. Methods: Retinal vessel caliber was measured from baseline retinal photographs using computer-assisted techniques. Mean arteriolar and venular diameters of each eye were summarized as central retinal arterial (CRAE) and venular (CRVE) equivalents. Cataract was assessed from lens photographs of both eyes using the Wisconsin grading system. Main Outcome Measures: Nuclear cataract was defined as opacity > Wisconsin standard photograph 3, cortical cataract defined as opacity ≥5% of the lens area, and posterior subcapsular (PSC) cataract as any present. Eye-specific data were analyzed using generalized estimating equations. Odds ratios (OR) and 95% confidence intervals (CI) are reported. Results: After adjusting for age, gender, smoking, hypertension, diabetes, body mass index, and inhaled steroid use, reduced incidence of nuclear cataract was associated with the narrowest compared with the widest quintile of CRAE and CRVE (for CRAE: OR, 0.62; 95% CI 0.42–0.92; for CRVE: OR, 0.70; 95% CI, 0.47–1.05) but a higher incidence of PSC cataract (for CRAE: OR, 2.40; 95% CI, 1.34–4.29; for CRVE: OR, 3.17; 95% CI, 1.62–6.20) and cataract surgery (for CRAE: OR, 1.52; 95% CI, 1.06–2.17; for CRVE: OR, 1.58; 95% CI, 1.08–2.32). These associations were not maintained when both CRAE and CRVE were included simultaneously in the same models. Path analysis suggested that age was the most important contributor to nuclear cataract incidence, and CRAE and CRVE, as markers of a latent age-related variable, were only indirectly associated with its incidence (for CRAE: OR, 0.61; 95% CI, 0.41–0.91; for CRVE: OR, 0.62; 95% CI, 0.41–0.94). Conclusions: Retinal vessel narrowing predicted greater risk of long-term incidence of PSC cataract and cataract surgery, and was indirectly linked to a lower incidence of nuclear cataract. Retinal vessel narrowing could be a marker of age-related factors associated with risk of PSC and nuclear cataract