Interrelationship between Serum FGF21, GDF15, and Microalbuminuria as Predictive Biomarkers for Early Detection of Diabetic Retinopathy in Type 2 Diabetes Mellitus.

Abstract

Diabetic retinopathy (DR) is a leading cause of preventable visual loss in type 2 diabetes mellitus (T2DM), yet current screening largely detects disease after microvascular damage has occurred. This review synthesizes emerging evidence that three measurable biomarkers - fibroblast growth factor-21 (FGF21), growth differentiation factor-15 (GDF15), and microalbuminuria -capture complementary dimensions of early DR pathobiology. FGF21 reflects hepatometabolic stress and adaptive signaling via the β-Klotho-FGFR1 axis; GDF15 mirrors mitochondrial/integratedstress-response activation with systemic effects through the glial cell-derived neurotrophic factor family receptor-α-like (GFRAL) REarranged during Transfection (RET); and microalbuminuria indicates systemic endothelial dysfunction and increased microvascular permeability. We narratively integrate cross-sectional, cohort, and meta-analytic studies in human T2DM populations (excluding type 1 diabetes and animal experiments) and map these signals onto a unified metabolic-renal-retinal framework. Across studies, higher circulating FGF21 and GDF15 associate with DR presence and severity, while microalbuminuria correlates with DR grade and predicts progression independent of glycemic control. We propose a tri-biomarker model in which chronic lipotoxicity and oxidative stress elevate FGF21/GDF15, drive endothelial injury detectable as microalbuminuria, and together forecast early retinal microangiopathy. Translational implications include risk stratification before funduscopic changes, therapy monitoring for metabolic/mitochondrial targets, and development of multimarker algorithms alongside OCTA imaging. To our knowledge, this is the first review to integrate FGF21, GDF15, and microalbuminuria as a coherent predictive axis for early DR in T2DM, highlighting priorities for longitudinal validation and population-specific cut-offs.