Blighe, Kevin; Gurudas, Sarega; Lee, Ying; Sivaprasad, Sobha; (2020) Diabetic retinopathy environment-wide association study (EWAS) in NHANES 2005-8. medRXiv: Cold Spring Harbor, NY, USA.

Preview

Text Blighe_2020.09.20.20198218v1.full.pdf Download (797kB) | Preview

Abstract

<h4>SUMMARY</h4> <h4>Background</h4> Several circulating biomarkers are reported to be associated with diabetic retinopathy (DR). However, their relative contributions to DR compared to known risk factors, such as hyperglycemia, hypertension, and hyperlipidemia, remain unclear. In this data driven study, we used novel models to evaluate the associations of over 400 laboratory parameters with DR. <h4>Methods</h4> We performed an environment-wide association study (EWAS) of laboratory parameters available in National Health and Nutrition Examination Survey (NHANES) 2007-8 in individuals with diabetes with DR as the outcome (test set). We employed independent variable (‘feature’) selection approaches, including parallelized univariate regression modeling, Principal Component Analysis (PCA), penalized regression, and RandomForest™. These models were replicated in NHANES 2005-6 (replication set). <h4>Findings</h4> The test and replication set consisted of 1025 and 637 individuals with available DR status and laboratory data respectively. Glycohemoglobin (HbA1c) was the strongest risk factor for DR. Our PCA-based approach produced a model that incorporated 18 principal components (PCs) that had AUC 0.796 (95% CI 0.761-0.832), while penalized regression identified a 9-feature model with 78.51% accuracy and AUC 0.74 (95% CI 0.72-0.77). RandomForest™ identified a 31-feature model with 78.4% accuracy and AUC 0.71 (95% CI 0.65-0.77). On grouping the selected variables in our RandomForest™, hyperglycemia alone achieved AUC 0.72 (95% CI 0.68-0.76). The AUC increased to 0.84 (95% CI 0.78-0.9) when the model also included hypertension, hypercholesterolemia, hematocrit, renal and liver function tests. <h4>Interpretation</h4> All models showed that the contributions of established risk factors of DR especially hyperglycemia outweigh other laboratory parameters available in NHANES. <h4>RESEARCH IN CONTEXT</h4> What is already known about this subject? ▪ There are >500 publications that report associations of candidate circulating biomarkers with diabetic retinopathy (DR). ▪ Although hyperglycemia, hypertension, and hyperlipidemia are established risk factors, they do not always explain the variance of this complication in people with diabetes; DR also shares risk factors with other diabetes complications including markers of renal and cardiovascular disease. ▪ ‘Holistic’ studies that quantify risk across all of these parameters combined are lacking. What is the key question? ▪ It is unclear whether risk models for DR may be improved by adding some of these reported biomarkers - there is an unmet need to systematically evaluate as many circulating biomarkers as possible to help rank their associations with DR. What are the new findings? ▪ We show that hyperglycemia is the strongest risk factor across all models. ▪ We stratified the rest of the highest ranked parameters into groups related to diabetes control, renal and liver function, and hematocrit changes. How might this impact on clinical practice in the foreseeable future? ▪ The importance of focusing on parameters beyond hyperglycemia control to reduce risk of progression from diabetes to DR is emphasized.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item