糖尿病腎症発症予測因子としての好中球・リンパ球比率の競合リスク分析

Diabetic nephropathy (DN) is a serious complication of diabetes mellitus. A high level of neutrophil-lymphocyte ratio (NLR) is an indicator of abnormal immune system activity which may serve as an effective potential inflammatory marker for identifying the risk of DN. This study aimed to investigate the relationship between neutrophil-lymphocyte ratio (NLR) and the incidence of DN in type 2 diabetes mellitus (T2DM) patients.

DN incidence was defined as the time from baseline diabetes diagnosis to first DN occurrence (KDIGO CKD criteria). NLR's effect and interactions were evaluated using covariate-adjusted competing risks regression (death as competing event). The optimal NLR cut-point for DN prediction was determined by ROC analysis. The Fine and Gray subdistribution hazard model assessed NLR's effect on DN incidence, with subdistribution hazard ratios (sHR) validated via bootstrap sampling. The final sample consisted of the records of 220 individuals (median age 64 years (IQR: 55-72)) with T2DM with complete covariates information which were available for incidence analysis with NLR.

Among 220 T2DM patients with complete covariates, 133 (60.45%) developed DN at 6 years, 20 (9.10%) were lost to competing events, and 67 remained DN-free. Median NLR was 2.4 (IQR: 1.8-3.3), positively correlating with urinary albumin-to-creatinine ratio and negatively with the estimated glomerular filtration rate (eGFR) (p<0.01). ROC analysis demonstrated diagnostic value for DN (AUC=0.772; 95% CI: 0.708, 0.836; p<0.01), with optimal cut-off at 3.02. NLR showed associations with DN in cause-specific (CSH=1.66; 95% CI: 1.13, 2.52) and FGR models (sHR=2.26; 95% CI: 1.72, 2.92). Bootstrap validation yielded consistent results (sHR= 2.36; 95% CI: 1.76, 3.02). Notably, NLR better predicts DN risk in older adults (>65 years) and those with well-controlled HbA1c (≤7.5%).

NLR shows promise for predicting DN incidence in Chinese patients, especially those >65 years or with good glycemic control.