Renal function-stratified associations of serum 1,5-anhydroglucitol with glycaemic markers and its discriminative value for albuminuria In type 2 diabetes

Background: Serum 1,5-anhydroglucitol (1,5-AG) reflects recent hyperglycaemic exposure, although its concentration may be affected by renal filtration and reabsorption. This study analysed the relationships between 1,5-AG and conventional glycaemic markers at different levels of renal function and evaluated its value in identifying albuminuria. Methods: We retrospectively included 527 individuals who underwent glycaemic and renal function testing between October 2023 and February 2025, including 120 healthy controls, 120 patients with type 2 diabetes mellitus (T2DM) alone, and 287 patients with T2DM and renal impairment. We divided patients in the latter group by estimated glomerular filtration rate (eGFR) into the <span style="color: rgb(32, 33, 34); font-family: sans-serif; font-size: 16px;">≥</span>90, 60-89, 30-59, and &lt;30 mL/min/1.73 m2 groups, with 80, 80, 80, and 47 patients, respectively. The measured indicators included 1,5-AG, fasting plasma glucose (FPG), glycated haemoglobin A1c (HbA1c), glycated albumin (GA), serum creatinine (Scr), cystatin C (CysC), albumin (ALB), and the urinary albumin-to-creatinine ratio (UACR). Correlation analysis, multivariable linear regression, and interaction testing were used to evaluate factors associated with 1,5-AG; albuminuria was defined as UACR30 mg/g, and the independent and incremental discriminative value of 1,5-AG was evaluated using multivariable logistic regression, receiver operating characteristic (ROC) curves, the DeLong test, and validation with 1000 bootstrap resamples. Results: Among the 407 patients with T2DM, ln(1,5-AG) was negatively correlated with HbA1c, GA, and FPG (P&lt;0.001); multivariable analysis showed that HbA1c (b=-0.316, 95% confidence interval [CI]: -0.430--0.202, P&lt;0.001) and FPG (b=-0.127, 95% CI: -0.235--0.020, P=0.020) remained independently associated with ln(1,5-AG), whereas eGFR, UACR, Scr, and CysC showed no independent associations. The negative correlation between 1,5-AG and HbA1c did not reach statistical significance in the eGFR&lt;30 mL/min/1.73 m2 group, but the global interaction was not significant (P=0.263). Albuminuria was present in 270 patients, and ln(1,5-AG) was not independently associated with albuminuria (adjusted odds ratio [aOR]=1.069, 95% CI: 0.828-1.381, P=0.608). The area under the curve (AUC) for 1,5-AG alone was 0.530; after adding 1,5-AG to the model comprising age, sex, HbA1c, GA, and FPG, the AUC remained 0.782, with no incremental discriminative value (DeLong P=0.650). Conclusion: In patients with T2DM, 1,5-AG primarily reflects glycaemic exposure, but its interpretation becomes less reliable when renal function is severely reduced; 1,5-AG has no independent or incremental value for identifying albuminuria and should be regarded as an adjunctive glycaemic marker affected by renal function rather than a marker of renal injury