Association of urinary proteome scores with early markers of complications and all-cause mortality in individuals with prediabetes

Abstract

Preliminary work and aims of the project: Proteomic classifiers derived from urinary proteome profiles offer the potential for non-invasive personalised risk stratification and have been associated with cardiovascular, hepatic and renal outcomes. The project aims to systematically characterise multidimensional urinary proteomics scores, namely CKD273, HF1 and HF2, CAD160, 193-GTM, LivFib-50, Cov50, and Col156, as non-invasive novel markers for vascular diabetic complications in prediabetes and type 2 diabetes (T2D). Description of results: A nested case-control study within the EPIC-DZD sample was established consisting of 110 cases (47 with prediabetes and 63 with T2D) that died during follow-up and 110 controls matched 1:1 for pre-/diabetes status, age, sex, and time of urine sampling. Urinary proteomics were measured and the pre-defined proteomic scores were derived. First preliminary results suggest that Col156, Cov50, HF1, HF2, LivFib-50 and CKD273 are cross-sectionally associated with early diagnostic markers of nephropathy after multivariable adjustment. Col156 And CAD160 were associated with elevated advanced glycation end products (AGE)-load quantified via skin autofluorescence. HF2 and CKD273 were associated with early markers of macrovascular complications, namely peripheral artery disease and vascular stiffness. Prospective analyses in the EPIC-DZD sample showed that HF2 was associated with overall mortality after multivariable adjustment and HF2 and CKD273 were informative for prediction of total mortality beyond established standard markers. Results slightly deviated when stratifying for diabetes-state, but were overall comparable. Potential applications and outlook: HF2 and CKD273 were identified as potential early markers for total mortality in individuals with prediabetes and T2D. They improved the prediction of mortality beyond established risk markers and could qualify for non-invasive personalised risk stratification.