Age-related macular degeneration in people with diabetes mellitus: prevalence, incidence, and determinants

Romdhoniyyah, Dewi ORCID: 0000-0003-3330-4847 (2022) Age-related macular degeneration in people with diabetes mellitus: prevalence, incidence, and determinants. PhD thesis, University of Liverpool.

Text 201305967_October2022.pdf - Author Accepted Manuscript Access to this file is embargoed until 1 August 2026. Download (5MB)

Abstract

Title: Age-related macular degeneration in people with diabetes mellitus: prevalence, incidence, and determinants. Background: There were limited epidemiology data of age-related macular degeneration (AMD) prevalence and incidence in people with diabetes and its determinants, including the effect of DR and diabetes medication, to the incidence and progression of AMD. Fundus photographic image from DR screening is reported to be reliable to detect AMD. In the UK, the universal photographic screening programme for DR in diabetes provided sufficient data of medical record to understand the risk and protective factors that affect the development and progression of the disease whilst also provided an opportunity to identify other retinal pathologies such as AMD. Methods: Of 10,336 people aged ≥50 years who attended the Liverpool Diabetic Eye Screening Programme (LDESP) in 2011 and were enrolled in the Individualised Screening for Diabetic Retinopathy (ISDR) study, I selected randomly 2,600 cohort who had gradable retinal photographs (N=2589) to grade for AMD and used these data to measure the prevalence of AMD in diabetes. Excluding 466 cohort who did not attend the LDESP in 5-year or ungradeable/unavailable images of both eyes, retinal photographs of 2123 cohort were graded for AMD to study the five-year incidence and progression. Retinal photographic grading was performed following the modified Age-Related Eye Disease Study (AREDS) grading system for each eye. Twenty percent of all the images that were used in this study were subjected to quality grading reassurance that was completed separately by a senior AMD grader at Liverpool Ophthalmic Reading Centre. A multivariable logistic regression model was used to assess the association of multiple factors, including anti diabetes medication metformin and AMD incidence and progression. Multiple imputation (MI) applied in multivariable logistic regression was used to handle missing data. Results: The age-standardised prevalence of early, intermediate, and late AMD was 13.8%, 4.4%, and 0.9%, respectively. Prevalence of any AMD were consistently increased with increasing age (crude OR 1.06, 95%CI 1.05-1.07, p<0.001). These results persisted in multivariable logistic regression of complete record analysis and MI to missing data (both OR 1.05, 95%CI 1.05-1.07, p<0.001). The presence of any DR was associated with reduced risk of any AMD in multivariable logistic regression (complete record: OR 0.70, 95% CI 0.51-0.97, p=0.03; MI: 0.78, 0.61-0.99, p=0.04). The age-standardised incidence of early, intermediate, early-intermediate, and late AMD was 11.6%, 10.3%, 16.4%, and 2.9%, respectively. The progression from early/intermediate to late AMD was 8.4%. In multivariable logistic regression analysis, increasing age was associated with increased odds of incidence of early-intermediate (OR 1.03; 95%CI: 1.01-1.05, p=0.002), intermediate (1.05, 1.03-1.07, p<0.001), late (1.19, 1.14-1.24, p<0.001), and any AMD (1.03, 1.02-1.05, p<0.001). Likewise, increasing age was related to progression to late AMD (1.14, 1.09-1.19, p<0.001). No significant evidence of association was identified between the presence of any DR or diabetes control at baseline with AMD incidence and progression in final model of multivariable logistic regression (p0.05). Metformin use lowered the odds of intermediate AMD incidence consistently in multivariable logistic regression model (complete record: OR 0.63, 95%CI 0.43-0.92, p=0.02; MI: 0.66, 0.48-0.91, p=0.01). The interobserver agreement kappa statistic for AMD severity level was more than 0.7 in both eyes of the baseline and follow-up images. Conclusions: I found that detecting and grading AMD during routine DR screening is feasible and reliable, indicated by good interobserver agreement in my study. Inconsistent comparison of prevalence, incidence, and progression of AMD with selected reference studies was observed. Due to differences in follow-up period, age ranges, grading categories, demographic, and methodologies between studies, cautious interpretation is warranted. Inconsistent association between presence of any DR and any AMD in cross-sectional and follow-up analysis were observed in my study. Ultimately, there is no straight-forward explanation for these conflicting results. Metformin use suggested to be protective to incidence of intermediate stage of AMD and warrant further observational studies in other population settings as foundational knowledge for potential randomised controlled clinical trial.

Item Type:	Thesis (PhD)
Divisions:	Faculty of Health and Life Sciences
Depositing User:	Symplectic Admin
Date Deposited:	29 Aug 2023 10:14
Last Modified:	29 Aug 2023 10:14
DOI:	10.17638/03170560
Supervisors:	Beare, Nicholas Harding, Simon Finana, Marta Cheyne, Christopher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3170560