Keywords: Diabetic Retinopathy, Mixed Models, Neurodegeneration Corresponding author: Diana Tavares dstavares@aibili.pt Supplementary material: Available online: Link Conflict of interest: D.T., M.H.M, I.P.M., T.S, A.R.S and C.L. declare no conflict of interests. J.C.-V. reports grants from Carl Zeiss Meditec and is consultant for Alimera Sciences, Allergan, Bayer, Gene Signal, Novartis, Pfizer, Precision Ocular Ltd., Roche, Sanofi-Aventis, Vifor Pharma, and Carl Zeiss Meditec. Clinical study registration number: NCT03010397 First published: 22JUN2021 EXTENDED ABSTRACT © 2020 The Authors. This is an open access article distributed under CC BY license, whis license allows reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use (https://creativecommons.org/licenses/by/4.0/). Open Access Publication J. Stat. Health Decis. 2021;3(1):36-37 | https://doi.org/10.34624/jshd.v3i1.24805 36 A10 Progression of Ganglion Cell-Inner Plexiform layer thickness in the initial stages of diabetic retinopathy in type 2 diabetic patients: a 5-year longitudinal study Diana Tavares 1 , Maria H. Madeira 1,2,3 , Inês P. Marques 1,2,3 , Torcato Santos 1 , Ana Rita Santos 1,2,3,4 , Conceição Lobo 1,2,3,5 , José Cunha-Vaz 1,2,3 1 AIBILI, Association for Innovation and Biomedical Research on Light and Image, 3000-548 Coimbra, Portugal 2 University of Coimbra, Coimbra Institute for Clinical and Biomedical Research (iCBR), Faculty of Medicine, 3000-548 Coimbra, Portugal; 3 University of Coimbra, Center for Innovative Biomedicine and Biotechnology (CIBB), 3000-548 Coimbra, Portugal 4 Department of Orthoptics, School of Health, Polytechnic of Porto, 4002-072 Porto, Portugal; 5 Department of Ophthalmology, Centro Hospitalar e Universitário de Coimbra (CHUC) 3000-075 Coimbra, Portugal. Introduction Diabetic Retinopathy (DR) is a frequent complication of Diabetes Mellitus (DM) and the main cause of vision loss in the working population in western countries. Diabetic Retinopathy has always been con- sidered a microvascular disease, but it has been suggested that neurodegeneration is also associated with this complex pathology[1], although there is evidence indicating that the neurodegenerative process may progress independently[2]. To evaluate this potential association, we have examined the progression of neurodegeneration over a 5-year period of follow-up (considering thinning of ganglion cell + inner plexi- form retinal layers (GCL+IPL) in individuals with type 2 diabetes (T2D) and nonproliferative DR) and explored whether it is associated with microaneurysm turnover (MAT), disease level at baseline and sever- ity progression. Methods This study was designed as a 5-year prospective, longitudinal study (ClinicalTrials.gov identifier: NCT03010397), to evaluate disease progression in T2D individuals. 212 T2D individuals with mild non-proliferative diabetic retinopathy (NPDR, ETDRS (Early Treat- ment Diabetic Retinopathy Study) level 20 or 35)[3] were included in the study, of which 145 completed the 5-year follow-up, with ophthalmological examinations performed at baseline and annually (one eye per patient). GCL+IPL average thickness was evaluated by optical coherence tomography (OCT). Classi- fication in ETDRS levels assesses severity of DR and was performed by grading of 7-fields color fundus photography. Severity progression was determined as step changes between levels of ETDRS at baseline and at the 5-year follow-up and was classified as improvement/maintenance or worsening. Microan- eurysm turnover (MAT) was evaluated using the RetMarkerDR considering a cut-off of 6, identified previously to characterize microvascular disease activity [4]. Linear mixed-effects models with a random intercept for the patient were used to assess the progression of GCL+IPLthickness over time (6 annual visits), adjusting for age and gender. Separate models were used to analyse the effect of ETDRS severity progression, ETDRS at baseline, MAT≥6 and the interaction of these parameters with time on GCL+IPL thickness. Statistical analysis was performed with Stata 16.1 (StataCorp. 2019. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC.Stata Corp. LP, College Station, TX, USA), and a P-value ≤0.05 was considered statistically significant. Results The analysis included 144 subjects, as one individual was excluded due to inconsistencies in the auto- matically collected OCT data. The effect estimates of time, adjusted for age and gender, on GCL+IPL average thickness are reported in table 1 and expected progression of GCL+IPL thickness over the 6 visits is depicted in figure 1. GCL+IPL thickness showed an estimated decrease of 1.40µm over the 5-year period in relation to the first visit, when other variables remain constant, adjusted for age (p-value: <0.001) and gender (p-value: 0.015). When evaluating the effect of the variables of interest, by adding them to the model, the effect of ETDRS step change was not statistically significant (p-value: 0.332), but the interaction of ETDRS level progression with time showed a significant effect on GCL+IPL average thickness (p-value: 0.008) (figure 2). Noteworthy, the effect of time on ETDRS step change groups was significant on those individuals