Purchase this article with an account.
Alasdair Warwick, Robert Luben, Tjebo Heeren, Abraham Olvera Barrios, Sharon Yu Lin Chua, Paul J Foster, Adnan Tufail, Catherine A Egan, Anthony P Khawaja; Diabetic retinopathy prevalence and associations in a large population-based cohort: a three-way data linkage study. Invest. Ophthalmol. Vis. Sci. 2020;61(7):3833.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Diabetic retinopathy (DR) is a leading cause of blindness in the working age population. It remains unclear why some patients with diabetes develop sight-threatening diabetic retinopathy (STDR) while others do not. Here, we describe the EPIC-DR study methods and present associations with DR. The overall aim of EPIC-DR is to create a resource for examining the genetic and lifestyle determinants of DR via data linkage.
EPIC-Norfolk is a large observational cohort study which collected detailed phenotypic and genetic data on 26,152 adult participants. An eye examination was performed in only 8,623 subjects and the present EPIC-DR study aimed to extend this outcome ascertainment to the whole cohort. Study data from the EPIC-Norfolk cohort was linked with DR gradings from the UK National Diabetic Eye Screening Programme (DESP) and Norfolk and Norwich University Hospital Medisoft ophthalmic electronic health records. The prevalence of DR and STDR in 2018-2019 was ascertained for subjects with DESP data. Associations between presence of DR and putative risk factors were assessed using logistic regression. STDR was defined as moderate-severe non-proliferative DR, proliferative DR or maculopathy, as described by the DESP DR grading system. For participants with differing DR severities between the two eyes, the more severe grade was selected.
1,613 EPIC-Norfolk participants had available DESP data, representing 6.2% (95% confidence interval (CI) 5.9-6.5) of the entire cohort. Of these, 1,611 participants had DESP data in 2018-2019 for analysis. The mean age was 79 years (standard deviation (SD) 6.9) and 51% were men. The prevalence of any DR and STDR was 26.4% (95% CI 24.3-28.6) and 1.9% (95% CI 1.2-2.5) respectively. 42 subjects had both Medisoft and DESP data available, 2 of whom were receiving anti-VEGF therapy for diabetic macular oedema. Presence of any DR was associated with higher body mass index (BMI) (odds ratio (OR) 1.03 per kg/m2, 95% CI 1.00-1.06; P=0.03) and older age (OR 1.02 per year, 95% CI 1.01-1.04; P<0.01).
The prevalence of any DR and STDR among diabetics in the EPIC-Norfolk cohort are lower than for the general UK population (~30% and ~3% respectively). BMI and age were associated with DR presence. Future work will explore associations between lifestyle and genetic factors with the presence and/or development of STDR.
This is a 2020 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only