Abstract
Purpose :
Cross-sectional studies have demonstrated that axial myopia is associated with a decreased risk of having any diabetic retinopathy (DR). However, long-term data are lacking and, in this study, we examined the longitudinal association of refractive error (RE) and its associated determinants (axial length [AL] and corneal curvature [CC]) with the incidence and progression of DR.
Methods :
From the Singapore Epidemiology of Eye Diseases Study, a population-based cohort, we included 1024 eyes of 554 individuals with diabetes and gradable retinal photographs (mean age [SD]: 57.6 [7.9], 49.4% female) at both baseline and follow-up examinations (mean [SD] follow-up duration: 6.2 (0.9) years). RE was calculated as sphere plus half negative cylinder, while AL and CC were assessed using optical biometry. DR, graded from 2-field retinal photographs using the modified Airlie House classification system, was categorized as none, minimal, mild, moderate and vision-threatening (VTDR). Incident DR was defined as having no DR at baseline and any DR at follow-up; incident VTDR as no VTDR at baseline but present at follow-up; and progression as an increase in ≥ one step in severity at follow-up from at least minimal DR at baseline. Eye-specific data and general estimating equation models to account for the correlation between both eyes were utilized to determine the relationships between refractive error, AL and CC; and incident DR, VTDR and DR progression, adjusted for traditional DR risk factors.
Results :
At follow-up, 101 of 764 (13.2%) and 17 of 981 (1.7%) eyes had incident DR and VTDR, respectively. From 217 eyes with DR at baseline, 60 (27.7%) experienced progression. In multivariable models, a longer AL was associated with lower incidence of DR (per mm increase, odds ratio, 95% confidence interval: 0.44, 0.25 to 0.80), while a flatter CC (per mm increase, 7.44, 1.46 to 38.0) was instead associated with increased risk. In contrast, no associations were found between RE and the outcomes assessed.
Conclusions :
Our finding that a longer AL is protective of incident DR confirms previous cross-sectional observations, although the role of a flatter CC as a risk factor for DR incidence is novel. Further studies are needed to better understand the mechanisms underlying these relationships, so as to inform future novel prophylactic strategies to prevent DR development and progression.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.