Genetics of diabetes-associated microvascular complications

“

TL;DR

This review synthesizes genetic studies of diabetic microvascular complications (nephropathy, retinopathy, neuropathy), highlighting that while hyperglycemia is a key driver, glucose-independent mechanisms and early-life programming also contribute. Genome-wide association studies have identified loci such as COL4A3 for nephropathy and GRB2 for retinopathy, but replication remains limited due to small sample sizes and inconsistent phenotyping. The findings underscore the need for better phenotyping, separate analyses by diabetes type, and integration of multi-omics to uncover causal pathways and enable personalized treatment.

Full summary

4,275 CHARS

**Background:** Diabetes is a leading cause of blindness, end-stage kidney failure, and limb amputation. Microvascular complications—retinopathy, nephropathy, and neuropathy—are traditionally attributed to chronic hyperglycemia, but individuals with similar HbA1c levels show variable progression, suggesting glucose-independent mechanisms. Genetic studies, including genome-wide association studies (GWAS), have aimed to uncover the heritable basis of these complications. This narrative review discusses progress in the genetics of diabetic microvascular complications, emphasizing the need to consider temporal risk factors (e.g., intrauterine programming, metabolic risk factors) and to conduct analyses separately in type 1 and type 2 diabetes.

**Methods:** The authors performed a narrative review of the literature, summarizing findings from large-scale GWAS, candidate gene studies, Mendelian randomization analyses, and whole-exome sequencing studies. Key cohorts mentioned include the DCCT, UK Biobank, CKDGen Consortium, DNCRI, EUROCONDOR, GoDARTS, and DOLORisk Consortium. The review covers studies in individuals with type 1 and type 2 diabetes, as well as general populations, and discusses methodological challenges such as sample size, phenotyping definitions, and confounding by metabolic risk factors.

**Key Results:** - **Nephropathy:** Familial clustering shows 83% concordance for diabetic kidney disease (DKD) in siblings of probands with nephropathy vs. 17% in siblings of probands without. GWAS have identified 16 loci for DKD in type 1 diabetes (e.g., protective variant in COL4A3, ~20% allele frequency, protective effect seen only with HbA1c >58 mmol/mol [7.5%]). In combined type 1/type 2 diabetes meta-analyses (n=27,000), novel loci include COL20A1, DCLK1, EIF4E, PTPRN–RESP18, INIP–SNX30, LSM14A, and MFF. The SLC47A1 variant (rs2252281) showed the strongest effect on eGFR (p=10^-75), CpG methylation (p=10^-39), and gene expression (p=10^-8), with co-localization probability 0.98. Mendelian randomization supported a causal effect of BMI on DKD but not of glycemic traits. - **Retinopathy:** Neurodegeneration precedes vascular changes: in the EUROCONDOR study, 58% of type 2 diabetes patients with no visible retinopathy had neurodegeneration, while only 33% had vascular alterations. Candidate gene studies have validated variants in VEGF and TCF7L2 for severe retinopathy. GWAS identified a signal near GRB2 (growth factor receptor bound protein 2) in combined White/Asian populations. In West African extreme-phenotype GWAS (64 cases, 227 controls), a locus in WDR72 was replicated in African Americans. - **Neuropathy:** Few GWAS exist; a UK Biobank study identified a signal in SLC25A3 (mitochondrial phosphate carrier) associated with neuropathic pain defined by prescription drug use. Candidate genes include SCN9A, GCH1, and KCNS1. - **Mechanisms:** Transcriptomic analysis in long-term type 1 diabetes non-progressors (no complications after >30 years) showed synergistic downregulation of mitochondrial OXPHOS and DNA repair genes, associated with higher insulin sensitivity and lower liver fat. Metabolomic profiling revealed lower pyruvate, thiamine monophosphate, and erythritol, and higher phenylalanine, glycine, and serine, suggesting shunting of glycolytic substrates toward pentose phosphate and one-carbon metabolism.

**Clinical Implications:** The review highlights that current clinical definitions of diabetic retinopathy rely solely on vascular lesions, ignoring early neurodegeneration. Similarly, nephropathy definitions based on eGFR/albuminuria may miss glucose-unrelated kidney dysfunction. The findings suggest that genetic risk profiling, combined with early detection of neurodegeneration (e.g., corneal confocal microscopy for neuropathy) and consideration of fetal programming (e.g., famine exposure increasing proliferative retinopathy risk), could improve risk stratification. The protective role of reduced mitochondrial ROS generation challenges the traditional view that oxidative stress uniformly promotes complications. Future research should integrate multi-omics, use prospective designs with repeated measures, and apply artificial intelligence to develop personalized treatment strategies.