Session Title: Free Paper Session 3: AMD I
Session Date/Time: Thursday 07/09/2017 | 08:30-10:00
Paper Time: 09:48
Venue: Room 118
First Author: : M.Raimundo PORTUGAL
Co Author(s): : M. Cachulo I. Lains C. Lobo S. Nunes J. Cunha-Vaz R. Silva
To evaluate the prevalence of frequent retinal diseases in a large populational based cohort following digital fundus imaging grading.
Association for Innovation and Biomedical Research on Light and Image (AIBILI), Coimbra, Portugal.
Secondary analysis of the Coimbra Eye Study cohort, a cross-sectional population-based study including all patients aged ≥ 55 years of two different Portuguese primary health-care units. Responders underwent a full ophthalmological examination and digital fundus imaging. In addition to the prevalence of early and late age-related macular degeneration, AMD, in this secondary analysis we calculated the prevalence of diabetic retinopathy, pathological myopia, previous venous occlusion, epiretinal membrane and pigmentary retinopathy. Early and late AMD was defined per the International Age-Related Macular Epidemiological Study Group Classification. Point estimates of proportions and 95% confidence intervals (CI) are reported, overall and per age category, and associations were evaluated by logistic regression models (univariate and multivariate).
We included 5966 subjects from two sites (contributing with 3021 and 2975 subjects, respectively) with gradable digital fundus photography, 43.6% male and 99.5% Caucasian, with a mean age of 68.2 ± 8.6 years. Overall prevalence of early and late AMD in this cohort was 13.01% (95%CI 12.18–13.88) and 0.98% (95%CI 0.76–1.27), respectively. Neovascular AMD and geographic atrophy were seen in 0.51% (95%CI 0.36–0.73) and 0.58% (95%CI 0.42–0.81), respectively. Diabetes (type 1 or 2) was present in 22.2%, while diabetic retinopathy was observed in 14.9% of these, corresponding to an overall prevalence of 3.30% (95%CI 2.86–3.79) in the full cohort. In a multivariate regression model, presence of diabetic retinopathy was predicted by age (OR 1.21 per 10-year increase, p=0.028), body mass index (OR 1.28 per 5 kg/m2 increase, p=0.002), hypertension (OR 2.22, p<0.001) and dyslipidemia (OR 1.48, p=0.013). Evidence of venous occlusion was detected in 0.40% (95%CI 0.27–0.59) of cases; hypertension was the only significantly associated predictor (OR 3.95, p=0.012). Epiretinal membrane and pathological myopia were seen in 0.72% (95%CI 0.53–0.96) and 0.69% (95%CI 0.50–0.94), respectively. Pigmentary retinopathy was documented in 0.77% (95%CI 0.56–1.02) of cases.
Prevalence of retinal diseases in our study seems to follow other large-scale studies with similar methodologies. However, a higher than expected prevalence of pigmentary retinopathy was found, which may be explained by a cluster bias and a high level of consanguinity in the cohort sites. Predictors of diabetic retinopathy and retinal venous occlusion highlight the striking importance of cardiovascular risk factors. Large population-based epidemiological studies on retinal diseases are critical for health planning and may contribute to better designed screening initiatives and optimized resource allocation.