The mediating role of inflammation in the association between coagulation parameters and diabetic kidney diseases in patients with type 2 diabetes mellitus: a muti-center cross-sectional retrospective study in Southwest China from 2013 to 2022 Authors: Bei-Hui Ren, Ling-Zhi Xing, Bai-Yu-Lu Chen, Guo-Jia Liao, Li-Yuan Chen, Ying-Dan Zhang, Jie Fan, Ling Chen Published: 2025 ABSTRACT PurposeThis study aims to investigate the association between coagulation parameters and diabetic kidney diseases (DKD) with a specific focus on evaluating the potential mediating effects of inflammation.MethodsA total of 10,523 patients from 6 different centers were enrolled in our study. The association between coagulation parameters and DKD was evaluated using the logistic regression model and subgroup analysis. The predictive ability of coagulation parameters for DKD was assessed through a series of investigations including receiver operating characteristic curve (ROC) analysis and decision curve analysis (DCA). Additionally, mediation analysis was performed to examine the potential role of different inflammatory factors on the association between coagulation parameters and DKD.ResultsMultivariate logistic analysis revealed that fibrinogen (FIB) (OR = 1.086, 95%CI = 1.035–1.039, p-value < 0.001) and prothrombin activity (PTA) (OR = 1.015, 95%CI = 1.011–1.019, p-value < 0.001) were positively associated with DKD. Conversely, prothrombin time (PT) was negatively associated with DKD when converted into a categorical variable (Tertile 3: OR = 0.701, 95%CI = 0.594–0.827, p-value < 0.001). Moreover, PT, PTA, and FIB were examined for their predictive ability and PT performed best in the prediction of DKD (AUC: 0.583, 95%CI: 0.569–0.598). Mediation analysis demonstrated that neutrophils attenuated 8.27% of the total effects in the DKD-PTA relationship and mediated 3.74% of the relationship between DKD and PT. Moreover, multiple inflammation parameters including eosinophils and albumin were also found to mediate the association between DKD and coagulation parameters.ConclusionThese findings suggest that coagulation parameters are cost-effective screening tools for early DKD detection. Inflammation parameters are identified to mediate the association between coagulation parameters and DKD while the internal mechanism requires further exploration.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12882-025-04698-9. Background Diabetic kidney disease (DKD), characterized by albuminuria and progressive renal failure, is a common complication of long-term diabetes mellitus. Researchers indicate the incidence of DKD has been significantly rising in recent years, with 30%-40% of diabetic patients developing this condition. Moreover, DKD is one of the leading causes of end-stage kidney disease, cardiovascular and all-cause mortality [1–3]. Thus, it is essential to identify the risk factors of DKD for the early detection, diagnosis, and management of DKD. Multiple processes contribute to the progression of DKD including glomerular hyperfiltration, progressive albuminuria, declining, and eventually end-stage renal diseases. These processes are driven by hydrodynamic, metabolic, and immunological changes caused by long-term diabetes [2–4]. Although the development of DKD results from the combination of multiple factors and mechanisms, a considerable number of studies suggest that the pathophysiological mechanism of DKD is closely relevant to the inflammation caused by elevated glucose in diabetic patients [5]. It has been reported that a series of pro-inflammatory factors will be produced in hyperglycemia status [6]. These factors will give rise to the recruitment of cytokines and inflammatory cells [7]. These substances will then infiltrate the kidney and produce inflammatory mediators which can contribute to the damage of renal cells and the progression of inflammation [7–9]. Thus, inflammation plays a vital role in the development and progression of DKD. Coagulation has been demonstrated to be activated in diabetic patients due to the vascular damage caused by chronic hyperglycemia [10]. Emerging evidence suggests that coagulation may cause inflammation by interacting with the complement system and recruiting inflammatory cells [11, 12]. Previous studies have revealed the association between some coagulation parameters and DKD. For instance, fibrinogen (FIB) and its ratio with albumin (ALB) have been associated with DKD [13, 14]. Moreover, D-dimer has been identified as a key factor in DKD progression [15]. However, other coagulation parameters, such as prothrombin time (PT) and prothrombin activity (PTA) have received relatively less attention or been observed to be less significant by some studies. Additionally, we hypothesize that inflammation may mediate the association between coagulation parameters and DKD while no studies have investigated this. Thus, our study intended to examine the relationships between various coagulation parameters and DKD, and further explore the potential mediating role of inflammation in the association between coagulation parameters and DKD, expecting to find latent therapeutic strategies for DKD. Methods Study population We primarily selected 12,102 patients during the 2013–2022 period with Type 2 diabetes from Big Data Intelligence Platform which was jointly built by Chongqing Medical University and Yidu Cloud Company based on electronic medical system of hospitals including The First Affiliated Hospital of Chongqing Medical University, The Second Affiliated Hospital of Chongqing Medical University, University-Town Hospital of Chongqing Medical University, The Affiliated Rehabilitation Hospital of Chongqing Medical University, Chongqing Medical University Affiliated Yongchuan Hospital, and The Affiliated Dazu’s Hospital of Chongqing Medical University. Inclusion criteria: (1) Patients who were diagnosed with Type 2 diabetes. (2) Adults over 18 years old. (3) Patients with complete clinical and laboratory parameters. Exclusion criteria: (1) Patients without a reliable and complete diagnosis of Type 2 diabetes. (2) Patients with missing data on coagulation parameters. Then, 10,523 patients were formally enrolled in our study. The study population flow chart was shown in Fig. 1. Fig. 1Flow chart of the study population Flow chart of the study population Data collection We included age, gender, drinking status, smoking status, systolic blood pressure (SBP), diastolic blood pressure (DBP), hemoglobin (HGB), albumin (ALB), red blood cell (RBC), white blood cell (WBC), eosinophil, neutrophil, monocyte, lymphocyte, platelet (PLT), glycated hemoglobin (HbA1c), serum creatinine (SCR), uric acid (UA), triglycerides (TyG), high-density lipoprotein cholesterol (HDL-C), low- density lipoprotein cholesterol (LDL-C). Moreover, coagulation parameters were selected including activated partial thromboplastin time (APTT), prothrombin time (PT), fibrinogen (FIB), prothrombin activity (PTA), and thrombin time (TT). Assessment of outcomes DKD was diagnosed by albumin-to-creatinine ratio (UACR) ≥ 30 mg/g and/or estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73m3, with eGFR calculated using the CKD-EPI 2021 creatinine-cystatin C equation (CKD-EPIscr-cys), this diagnostic criteria for DKD in our research was based on established clinical criteria, as defined in the authoritative literature on this subject [16]. Assessment of inflammation status A series of inflammatory parameters were used including single inflammatory indicators like ALB, PLT, WBC, eosinophil, neutrophil, monocyte and complex inflammatory parameters like neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte to monocyte ratio (LMR), neutrophil-to-platelet ratio (NPR), platelet-to-albumin ratio (PAR), systemic immune system (SII), and systemic inflammation response index (SIRI) to assess inflammation status. Formulas used to calculate parameters mentioned above are as follows: NLR = Neutrophil(109/L)/Lymphocyte(109/L),PLR = Platelet(109/L)/Lymphocyte(109/L),LMR = Lymphocyte(109/L)/Monocyte(109/L),NPR = Neutrophil(109/L)/Platelet(109/L),PAR = Platelet(109/L)/Albumin(g/L),SII = platelet (109/L) *neutrophil (109/L)/lymphocyte (109/L),SIRI=(Neutrophil (109/L) *Monocyte (109/L))/Lymphocyte (109/L). NLR = Neutrophil(109/L)/Lymphocyte(109/L), NLR = Neutrophil(109/L)/Lymphocyte(109/L), PLR = Platelet(109/L)/Lymphocyte(109/L), PLR = Platelet(109/L)/Lymphocyte(109/L), LMR = Lymphocyte(109/L)/Monocyte(109/L), LMR = Lymphocyte(109/L)/Monocyte(109/L), NPR = Neutrophil(109/L)/Platelet(109/L), NPR = Neutrophil(109/L)/Platelet(109/L), PAR = Platelet(109/L)/Albumin(g/L), PAR = Platelet(109/L)/Albumin(g/L), SII = platelet (109/L) *neutrophil (109/L)/lymphocyte (109/L), SII = platelet (109/L) *neutrophil (109/L)/lymphocyte (109/L), SIRI=(Neutrophil (109/L) *Monocyte (109/L))/Lymphocyte (109/L). SIRI=(Neutrophil (109/L) *Monocyte (109/L))/Lymphocyte (109/L). Statistical analysis Continuous variables with normal distributions are presented as mean ± standard deviation while the median with quartile (25th and 75th percentiles) was used to express non-normal distribution continuous variables. Categorical variables are expressed as percentages. Between-group comparisons are realized by t-tests, Wilcoxon rank-sum tests, or Chi-square tests. Univariate logistic regression was carried out to explore the relationship between different variables and DKD. Multivariate logistic regression was then conducted with the progressive adjustment for confounding factors assessing the relationship between coagulation parameters and DKD. Model 1 was unadjusted, Model 2 adjusted for age and sex, and Model 3 further adjusted for age, sex, SBP, HGB, ALB, HbA1c, SCR, UA, eosinophil, and neutrophil. Restricted cubic spline analysis (RCS) was used to explore the linear relationship between coagulation parameters and DKD. Both multivariate logistic regression and RCS were also conducted in subgroups stratified by age, sex, HGB, ALB, eosinophil, neutrophil, HbA1c, Cr, UA, SBP to assess subgroup-specific associations with DKD. The clinical predictive ability of coagulation parameters was evaluated using receiver operating characteristic curve (ROC) analysis, decision curve (DCA) analysis, and indexes including net reclassification index (NRI), integrated Discrimination Index (IDI), and area under the curve (AUC). Coagulation and relevant clinical parameters were used to establish a prediction model, and the contribution of each parameter was assessed by nomogram and calibration curve analysis. Additionally, the mediating role of inflammation parameters was explored using mediation analysis with the ‘mediation’ packages in R (4.4.1 version). All statistical analyses were carried out using SPSS (version 27.0), GraphPad Prism (version 9.5.1), and R software (version 4.4.1). P-value < 0.05 was considered to be significant. Result The baseline characteristics of T2DM patients In this study, we included 10,523 patients with T2DM (Table 1). Compared to the non-DKD group, patients in DKD group were more likely to be males and younger. Besides, they also tended to have higher SBP, RBC, eosinophil, neutrophil, monocyte, glucose, HbA1c, Cr, UA, LDL, FIB, PTA, NLR, PAR, SIRI, SII and lower HGB, ALB, calcium, TyG, HDL, PT, LMR. Furthermore, we also analyzed the distribution of different levels of coagulation parameters between the DKD and non-DKD groups. It has been found that DKD patients were more likely to have high levels of PTA and FIB along with low levels of PT (Supplementary Fig. 1). Table 1Baseline demographic and clinical characteristics in type 2 diabetesTotal (n = 10523)Non-DKD group (n = 8645)DKD group (n = 1878)p-valueAge, years72 (62, 80)72.00 (63, 81)70.00 (59, 79)< 0.001Female (%)5031 (47.81)4192 (48.49)839 (44.68)0.003Drinking, n(%)2703 (25.69)2195 (25.39)508 (27.05)0.143Smoking, n(%)3461 (32.89)2814 (32.55)647 (34.45)0.118SBP, mmHg135.00 (122.00, 150.00)135.00 (122.00, 150.00)137.00 (124.25, 152.00)< 0.001DBP, mmHg80.00 (70.00, 87.00)80.00 (70.00, 87.00)80.00 (70.00, 87.00)0.951HGB, gL134.00 (120.00, 148.00)135.00 (121.00, 148.00)131.00 (113.00, 145.00)< 0.001ALB, gL40.70 (37.20, 44.10)41.20 (37.80, 44.50)38.60 (35.00, 41.90)< 0.001RBC, 109/L4.14 (2.39, 4.63)4.13 (0.73, 4.62)4.20 (3.51, 4.69)< 0.001WBC, 109/L6.34 (5.06, 8.11)6.34 (5.06, 8.12)6.42 (5.11, 8.08)0.166Eosinophil, 109/L0.09 (0.04, 0.16)0.08 (0.04, 0.15)0.10 (0.05, 0.17)< 0.001Neutrophil, 109/L4.03(2.90,5.63)4.14(3.08,5.63)4.01(2.86,5.63)< 0.001Monocyte, 109/L0.34 (0.24, 0.46)0.34 (0.24, 0.46)0.34 (0.26, 0.46)0.001Lymphocytes, 109/L1.46 (1.05, 1.94)1.46 (1.05, 1.95)1.46 (1.05, 1.92)0.471PLT, 109/L178.00 (140.00, 222.00)178.00 (140.00, 221.00)180.00 (141.00, 224.00)0.171Glucose, mmol/L8.12 (6.06, 11.95)7.97 (6.03, 11.67)8.95 (6.22, 13.75)< 0.001HbA1c, %7.60 (6.60, 9.80)7.51 (6.60, 9.70)8.10 (6.80, 10.34)< 0.001SCR, umol/L68.00 (54.80, 86.70)67.00 (54.10, 83.30)76.15 (58.80, 118.27)< 0.001UA, umol/L327.30 (263.75, 405.00)323.40 (260.90, 396.90)350.85 (281.40, 436.22)< 0.001Calcium, mmol/L2.26 (2.15, 2.35)2.26 (2.16, 2.35)2.25 (2.14, 2.34)0.018TyG, mmol/L1.60 (1.11, 2.39)1.59 (1.09, 2.36)1.66 (1.19, 2.54)< 0.001HDL-C, mmol/L1.04 (0.87, 1.27)1.05 (0.88, 1.28)1.01 (0.85, 1.23)< 0.001LDL-C, mmol/L2.52 (1.91, 3.22)2.50 (1.90, 3.19)2.60 (1.96, 3.34)< 0.001APTT, s32.60 (29.90, 35.80)32.60 (29.90, 35.90)32.50 (29.80, 35.60)0.103PT, s12.30 (11.10, 13.20)12.40 (11.20, 13.20)11.90 (10.80, 12.90)< 0.001FIB, g/L3.29 (2.76, 4.01)3.26 (2.75, 3.97)3.39 (2.82, 4.17)< 0.001PTA, %103.00 (92.00, 113.00)102.00 (92.00, 112.00)107.00 (96.00, 117.00)< 0.001TT, s16.20 (14.60, 17.50)16.20 (14.60, 17.50)16.00 (14.60, 17.30)0.031NLR2.62 (1.70, 4.44)2.59 (1.67, 4.41)2.74 (1.84, 4.63)< 0.001PLR119.02 (87.26, 170.72)118.59 (86.78, 170.00)121.71 (88.93, 176.08)0.077LMR4.44 (2.85, 6.65)4.49 (2.86, 6.81)4.25 (2.77, 6.13)< 0.001NPR0.02 (0.02, 0.03)0.02 (0.02, 0.03)0.02 (0.02, 0.03)0.001PAR4.33 (3.27, 5.52)4.28 (3.24, 5.41)4.61 (3.45, 6.04)< 0.001SII456.75 (268.95, 822.96)451.17 (262.05, 813.97)484.38 (297.55, 873.31)< 0.001SIRI0.90 (0.49, 1.78)0.89 (0.48, 1.77)0.96 (0.55, 1.86)< 0.001Abbreviations: SBP, Systolic blood pressure; DBP, Diastolic blood pressure; HGB, Hemoglobin; ALB, Albumin; RBC, Red blood cell; WBC, White blood cell; PLT, Platelet; HbA1c, Glycated hemoglobin; SCR, Serum Creatinine; UA, Uric acid; TyG, Triglycerides; HDL-C, High density lipoprotein cholesterol; LDL-C, Low density lipoprotein cholesterol; APTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time; NLR, Neutrophil-to-lymphocyte ratio; PLR, Platelet-to-lymphocyte ratio; LMR, Lymphocyte-to-monocyte ratio; NPR, Neutrophil-to-platelet ratio; PAR, Platelet-to-albumin ratio; SII, Systemic immune-inflammation index; SIRI, Systemic inflammation response index Table 1Baseline demographic and clinical characteristics in type 2 diabetesTotal (n = 10523)Non-DKD group (n = 8645)DKD group (n = 1878)p-valueAge, years72 (62, 80)72.00 (63, 81)70.00 (59, 79)< 0.001Female (%)5031 (47.81)4192 (48.49)839 (44.68)0.003Drinking, n(%)2703 (25.69)2195 (25.39)508 (27.05)0.143Smoking, n(%)3461 (32.89)2814 (32.55)647 (34.45)0.118SBP, mmHg135.00 (122.00, 150.00)135.00 (122.00, 150.00)137.00 (124.25, 152.00)< 0.001DBP, mmHg80.00 (70.00, 87.00)80.00 (70.00, 87.00)80.00 (70.00, 87.00)0.951HGB, gL134.00 (120.00, 148.00)135.00 (121.00, 148.00)131.00 (113.00, 145.00)< 0.001ALB, gL40.70 (37.20, 44.10)41.20 (37.80, 44.50)38.60 (35.00, 41.90)< 0.001RBC, 109/L4.14 (2.39, 4.63)4.13 (0.73, 4.62)4.20 (3.51, 4.69)< 0.001WBC, 109/L6.34 (5.06, 8.11)6.34 (5.06, 8.12)6.42 (5.11, 8.08)0.166Eosinophil, 109/L0.09 (0.04, 0.16)0.08 (0.04, 0.15)0.10 (0.05, 0.17)< 0.001Neutrophil, 109/L4.03(2.90,5.63)4.14(3.08,5.63)4.01(2.86,5.63)< 0.001Monocyte, 109/L0.34 (0.24, 0.46)0.34 (0.24, 0.46)0.34 (0.26, 0.46)0.001Lymphocytes, 109/L1.46 (1.05, 1.94)1.46 (1.05, 1.95)1.46 (1.05, 1.92)0.471PLT, 109/L178.00 (140.00, 222.00)178.00 (140.00, 221.00)180.00 (141.00, 224.00)0.171Glucose, mmol/L8.12 (6.06, 11.95)7.97 (6.03, 11.67)8.95 (6.22, 13.75)< 0.001HbA1c, %7.60 (6.60, 9.80)7.51 (6.60, 9.70)8.10 (6.80, 10.34)< 0.001SCR, umol/L68.00 (54.80, 86.70)67.00 (54.10, 83.30)76.15 (58.80, 118.27)< 0.001UA, umol/L327.30 (263.75, 405.00)323.40 (260.90, 396.90)350.85 (281.40, 436.22)< 0.001Calcium, mmol/L2.26 (2.15, 2.35)2.26 (2.16, 2.35)2.25 (2.14, 2.34)0.018TyG, mmol/L1.60 (1.11, 2.39)1.59 (1.09, 2.36)1.66 (1.19, 2.54)< 0.001HDL-C, mmol/L1.04 (0.87, 1.27)1.05 (0.88, 1.28)1.01 (0.85, 1.23)< 0.001LDL-C, mmol/L2.52 (1.91, 3.22)2.50 (1.90, 3.19)2.60 (1.96, 3.34)< 0.001APTT, s32.60 (29.90, 35.80)32.60 (29.90, 35.90)32.50 (29.80, 35.60)0.103PT, s12.30 (11.10, 13.20)12.40 (11.20, 13.20)11.90 (10.80, 12.90)< 0.001FIB, g/L3.29 (2.76, 4.01)3.26 (2.75, 3.97)3.39 (2.82, 4.17)< 0.001PTA, %103.00 (92.00, 113.00)102.00 (92.00, 112.00)107.00 (96.00, 117.00)< 0.001TT, s16.20 (14.60, 17.50)16.20 (14.60, 17.50)16.00 (14.60, 17.30)0.031NLR2.62 (1.70, 4.44)2.59 (1.67, 4.41)2.74 (1.84, 4.63)< 0.001PLR119.02 (87.26, 170.72)118.59 (86.78, 170.00)121.71 (88.93, 176.08)0.077LMR4.44 (2.85, 6.65)4.49 (2.86, 6.81)4.25 (2.77, 6.13)< 0.001NPR0.02 (0.02, 0.03)0.02 (0.02, 0.03)0.02 (0.02, 0.03)0.001PAR4.33 (3.27, 5.52)4.28 (3.24, 5.41)4.61 (3.45, 6.04)< 0.001SII456.75 (268.95, 822.96)451.17 (262.05, 813.97)484.38 (297.55, 873.31)< 0.001SIRI0.90 (0.49, 1.78)0.89 (0.48, 1.77)0.96 (0.55, 1.86)< 0.001Abbreviations: SBP, Systolic blood pressure; DBP, Diastolic blood pressure; HGB, Hemoglobin; ALB, Albumin; RBC, Red blood cell; WBC, White blood cell; PLT, Platelet; HbA1c, Glycated hemoglobin; SCR, Serum Creatinine; UA, Uric acid; TyG, Triglycerides; HDL-C, High density lipoprotein cholesterol; LDL-C, Low density lipoprotein cholesterol; APTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time; NLR, Neutrophil-to-lymphocyte ratio; PLR, Platelet-to-lymphocyte ratio; LMR, Lymphocyte-to-monocyte ratio; NPR, Neutrophil-to-platelet ratio; PAR, Platelet-to-albumin ratio; SII, Systemic immune-inflammation index; SIRI, Systemic inflammation response index Baseline demographic and clinical characteristics in type 2 diabetes Abbreviations: SBP, Systolic blood pressure; DBP, Diastolic blood pressure; HGB, Hemoglobin; ALB, Albumin; RBC, Red blood cell; WBC, White blood cell; PLT, Platelet; HbA1c, Glycated hemoglobin; SCR, Serum Creatinine; UA, Uric acid; TyG, Triglycerides; HDL-C, High density lipoprotein cholesterol; LDL-C, Low density lipoprotein cholesterol; APTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time; NLR, Neutrophil-to-lymphocyte ratio; PLR, Platelet-to-lymphocyte ratio; LMR, Lymphocyte-to-monocyte ratio; NPR, Neutrophil-to-platelet ratio; PAR, Platelet-to-albumin ratio; SII, Systemic immune-inflammation index; SIRI, Systemic inflammation response index The association between coagulation parameters and DKD To investigate the association between coagulation parameters and DKD, univariate logistic regression was carried out. It was found that PT was negatively related to DKD (OR = 0.885, 95%CI = 0.857–0.914, p-value < 0.001) while FIB (OR = 1.106, 95%CI = 1.064–1.149, p-value < 0.001) and PTA (OR = 1.015, 95%CI = 1.012–1.018, p-value < 0.001) was positively associated with DKD (Supplementary Table 1). Then, a multivariate logistic analysis was used to identify the independent factors affecting DKD. After the full adjustment of all the interfering factors, the association between PT and DKD was insignificant when PT was included in the model as a continuous factor (OR = 0.986, 95%CI = 0.948–1.025, p-value = 0.475), but it was observed that higher PT was less likely related to DKD when PT was turned into categorical factor according to tertiles (Tertile 3: OR = 0.701, 95%CI = 0.594–0.827, p-value < 0.001). In addition, FIB and PTA were still associated with DKD (FIB: OR = 1.086, 95%CI = 1.039–1.035, p-value < 0.001 PTA: OR = 1.015, 95%CI = 1.011–1.019, p-value < 0.001) when they were included in the model as continuous variables. Besides, higher FIB (Tertile 3: OR = 1.193, 95%CI = 1.044–1.364, p-value = 0.009) and PTA (Tertile 3: OR = 1.766, 95%CI = 1.524–2.047, p-value < 0.001) were found to be more likely associated with DKD when FIB and PTA were included in the model as categorical factors (Table 2). Table 2Multivariate logistic regression analysis of coagulation parameters with diabetic kidney disease in patients with type 2 diabetesModel 1Model 2Model 3OR(95%CI)p-valueOR(95%CI)p-valueOR(95%CI)p-valueAPTT(continuous)0.991(0.982–0.999)0.0350.993(0.984–1.001)0.0920.999(0.990–1.008)0.832APTT(categorical)Tertile 1Reference0.54Reference0.859Reference0.508Tertile 20.973(0.862-1.100)0.6650.974(0.862-1.100)0.671.012(0.889–1.152)0.858Tertile 30.933(0.826–1.055)0.2710.96890.855–1.096)0.6051.082(0.936–1.251)0.287PT(continuous)0.885(0.857–0.914)< 0.0010.902(0.874–0.932)< 0.0010.986(0.948–1.025)0.475PT(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference< 0.001Tertile 20.636(0.565–0.717)< 0.0010.661(0.586–0.747)< 0.0010.715(0.627–0.814)< 0.001Tertile 30.532(0.471–0.602)< 0.0010.6567(0.498–0.645)< 0.0010.701(0.594–0.827)< 0.001PTA(continuous)1.015(1.012–1.018)< 0.0011.014(1.011–1.017)< 0.0011.015(1.011–1.019)< 0.001PTA(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference< 0.001Tertile 21.241(1.087–1.418)0.0011.229(1.075–1.404)0.0031.275(1.101–1.477)< 0.001Tertile 31.836(1.625–2.074)< 0.0011.780(1.573–2.014)< 0.0011.766(1.524–2.047)< 0.001FIB(continuous)1.106(1.064–1.149)< 0.0011.123(1.080–1.167)< 0.0011.086(1.035–1.039)< 0.001FIB(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference0.004Tertile 21.109(0.899–1.156)0.7671.078(0.950–1.225)0.2450.973(0.853–1.109)0.677Tertile 31.309(1.159–1.478)< 0.0011.408(1.244–1.593)< 0.0011.193(1.044–1.364)0.009TT(continuous)1.000(0.994–1.007)0.9191.002(0.996–1.008)0.5311.006(1.000-1.013)0.048TT(categorical)Tertile 1Reference0.004Reference0.007Reference0.27Tertile 21.067(0.944–1.205)0.2991.146(1.012–1.297)0.0321.099(0.959–1.259)0.176Tertile 30.871(0.770–0.985)0.0280.947(0.835–1.075)0.41.007(0.865–1.172)0.931Note: OR, odds ratios; 95%CI, 95% confidence intervalModel 1 adjusted for noneModel 2 adjusted for age and sexModel 3 adjusted for age, sex, systolic blood pressure, hemoglobin, albumin, hemoglobin A1c, creatinine, uric acid, eosinophil, neutrophilAPTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time Table 2Multivariate logistic regression analysis of coagulation parameters with diabetic kidney disease in patients with type 2 diabetesModel 1Model 2Model 3OR(95%CI)p-valueOR(95%CI)p-valueOR(95%CI)p-valueAPTT(continuous)0.991(0.982–0.999)0.0350.993(0.984–1.001)0.0920.999(0.990–1.008)0.832APTT(categorical)Tertile 1Reference0.54Reference0.859Reference0.508Tertile 20.973(0.862-1.100)0.6650.974(0.862-1.100)0.671.012(0.889–1.152)0.858Tertile 30.933(0.826–1.055)0.2710.96890.855–1.096)0.6051.082(0.936–1.251)0.287PT(continuous)0.885(0.857–0.914)< 0.0010.902(0.874–0.932)< 0.0010.986(0.948–1.025)0.475PT(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference< 0.001Tertile 20.636(0.565–0.717)< 0.0010.661(0.586–0.747)< 0.0010.715(0.627–0.814)< 0.001Tertile 30.532(0.471–0.602)< 0.0010.6567(0.498–0.645)< 0.0010.701(0.594–0.827)< 0.001PTA(continuous)1.015(1.012–1.018)< 0.0011.014(1.011–1.017)< 0.0011.015(1.011–1.019)< 0.001PTA(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference< 0.001Tertile 21.241(1.087–1.418)0.0011.229(1.075–1.404)0.0031.275(1.101–1.477)< 0.001Tertile 31.836(1.625–2.074)< 0.0011.780(1.573–2.014)< 0.0011.766(1.524–2.047)< 0.001FIB(continuous)1.106(1.064–1.149)< 0.0011.123(1.080–1.167)< 0.0011.086(1.035–1.039)< 0.001FIB(categorical)Tertile 1Reference< 0.001Reference< 0.001Reference0.004Tertile 21.109(0.899–1.156)0.7671.078(0.950–1.225)0.2450.973(0.853–1.109)0.677Tertile 31.309(1.159–1.478)< 0.0011.408(1.244–1.593)< 0.0011.193(1.044–1.364)0.009TT(continuous)1.000(0.994–1.007)0.9191.002(0.996–1.008)0.5311.006(1.000-1.013)0.048TT(categorical)Tertile 1Reference0.004Reference0.007Reference0.27Tertile 21.067(0.944–1.205)0.2991.146(1.012–1.297)0.0321.099(0.959–1.259)0.176Tertile 30.871(0.770–0.985)0.0280.947(0.835–1.075)0.41.007(0.865–1.172)0.931Note: OR, odds ratios; 95%CI, 95% confidence intervalModel 1 adjusted for noneModel 2 adjusted for age and sexModel 3 adjusted for age, sex, systolic blood pressure, hemoglobin, albumin, hemoglobin A1c, creatinine, uric acid, eosinophil, neutrophilAPTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time Multivariate logistic regression analysis of coagulation parameters with diabetic kidney disease in patients with type 2 diabetes Note: OR, odds ratios; 95%CI, 95% confidence interval Model 1 adjusted for none Model 2 adjusted for age and sex Model 3 adjusted for age, sex, systolic blood pressure, hemoglobin, albumin, hemoglobin A1c, creatinine, uric acid, eosinophil, neutrophil APTT, Activated partial thromboplatin time; PT, Prothrombin time; FIB, Fibrinogen; PTA, Prothrombin activity; TT, Thrombin time Sensitivity analysis Then, statistically significant coagulation parameters including PT, PTA, FIB were incorporated in sensitivity analysis to further explore the stability of results (Figs. 2 and 3). There were no differences within groups in the association of PT, PTA with DKD. Conversely, the relationship between FIB and DKD can be insignificant in subgroups with higher HGB, ALB, UA, SBP or lower levels of SCR (Fig. 2). After adjusting age, sex, HGB, ALB, Eosinophil, Neutrophil, HbA1c, SCR, UA, SBP, and APTT, restricted cubic spline (RCS) analysis was used to explore the potential nonlinearity in the association between coagulation parameters and DKD. It was observed that there was a nonlinear relationship between PT and DKD as well as in subgroup RCS analysis. While PTA had a linear association with DKD except in subgroups where SBP < 135. In addition, FIB had a linear relationship with DKD although this linear relationship was insignificant when Cr ≥ 68 or SBP < 135 (Fig. 3). Fig. 2Subgroup analysis of the association between coagulation parameters and DKD. Note: OR, odds ratios; 95%CI, 95% confidence interval Subgroup analysis of the association between coagulation parameters and DKD. Note: OR, odds ratios; 95%CI, 95% confidence interval Fig. 3Restricted cubic spline analysis of the association between coagulation parameters and DKD (A) Association of PT with DKD. (B-E) Association of PT with diabetic kidney disease stratified by age, HGB, Cr and SBP. (F) Association of PTA with diabetic kidney disease. (G-J) Association of PTA with diabetic kidney disease stratified by age, HGB, SCR and SBP. (K) Association of FIB with diabetic kidney disease. (L-O) Association of FIB with diabetic kidney disease stratified by age, HGB, SCR and SBP. PT. Prothrombin time; PTA, Prothrombin activity; FIB, Fibrinogen; HGB, Hemoglobin; ALB, Albumin; HbA1c, Glycated hemoglobin; SCR, Serum creatinine; UA, Uric acid; SBP, Systolic blood pressure. Note: OR, odds ratios; 95%CI, 95% confidence interval. PT, Prothrombin time; PTA, Prothrombin activity; FIB, Fibrinogen; HGB, Hemoglobin; ALB, Albumin; HbA1c, Glycated hemoglobin; SCR, Serum creatinine; UA, Uric acid; SBP, Systolic blood pressure Restricted cubic spline analysis of the association between coagulation parameters and DKD (A) Association of PT with DKD. (B-E) Association of PT with diabetic kidney disease stratified by age, HGB, Cr and SBP. (F) Association of PTA with diabetic kidney disease. (G-J) Association of PTA with diabetic kidney disease stratified by age, HGB, SCR and SBP. (K) Association of FIB with diabetic kidney disease. (L-O) Association of FIB with diabetic kidney disease stratified by age, HGB, SCR and SBP. PT. Prothrombin time; PTA, Prothrombin activity; FIB, Fibrinogen; HGB, Hemoglobin; ALB, Albumin; HbA1c, Glycated hemoglobin; SCR, Serum creatinine; UA, Uric acid; SBP, Systolic blood pressure. Note: OR, odds ratios; 95%CI, 95% confidence interval. PT, Prothrombin time; PTA, Prothrombin activity; FIB, Fibrinogen; HGB, Hemoglobin; ALB, Albumin; HbA1c, Glycated hemoglobin; SCR, Serum creatinine; UA, Uric acid; SBP, Systolic blood pressure The prediction ability of coagulation parameters in DKD To further explore the prediction ability of coagulation parameters with DKD, a series of analyses was conducted. According to the ROC analysis, PT had slightly higher AUC (0.583, 95%CI: 0.569–0.598, Reference) than PTA (0.578, 95%CI: 0.564–0.592, p-value = 0.433) and FIB (0.537, 95%CI: 0.523–0.552, p-value < 0.001) (Fig. 4; Table 3). Then, NRI and IDI were calculated and compared to assess the prediction ability of PT, PTA, and FIB. Although the differences of NRI and IDI between PT, PTA, and FIB were slight, compared with PT, the prediction ability of PTA was worse (NRI=-0.0030, 95%CI=-0.0074-0.0015, p-value:=0.194 IDI=-0.0024, 95%CI=-0.0039–0.0080, p-value = 0.003), while FIB performed better in the prediction of DKD (NRI=-0.0012, 95%CI=-0.0027-0.0004, p-value:=0.138 IDI = 0.0054, 95%CI = 0.0034–0.0073, p-value < 0.001) (Table 3). After that, we established a clinical prediction model including coagulation parameters and other parameters related to DKD. Moreover, a nomogram was plotted to assess the contribution of different parameters to the prediction of DKD. According to Fig. 5(A), PTA contributed the most in the model for the prediction of DKD compared to PT and FIB. Besides, we also used the calibration curve to assess the prediction ability of this clinical prediction model (Fig. 5(B)). Table 3Comparisons of NRI, IDI, AUC of PT, PTA, and FIBModelsNRI(95%CI)p-valueIDI(95%CI)AUC(95%CI)p-valueDKDPT Reference Reference 0.583(0.569–0.598) Reference PTA-0.0030(-0.0074-0.0015)0.194-0.0024(-0.0039-0.0080)0.0030.578(0.564–0.592)0.433FIB-0.0012(-0.0027-0.0004)0.1380.0054(0.0034–0.0073)< 0.0010.537(0.523–0.552)< 0.001 Table 3Comparisons of NRI, IDI, AUC of PT, PTA, and FIBModelsNRI(95%CI)p-valueIDI(95%CI)AUC(95%CI)p-valueDKDPT Reference Reference 0.583(0.569–0.598) Reference PTA-0.0030(-0.0074-0.0015)0.194-0.0024(-0.0039-0.0080)0.0030.578(0.564–0.592)0.433FIB-0.0012(-0.0027-0.0004)0.1380.0054(0.0034–0.0073)< 0.0010.537(0.523–0.552)< 0.001 Comparisons of NRI, IDI, AUC of PT, PTA, and FIB Fig. 4Prediction ability of PT, PTA, FIB with DKD. (A) Receiver operating characteristic curve (ROC) for predicting the occurrence of diabetic kidney disease with hemostatic parameters. (B) Decision curve analysis (DCA) for the assessment of the predicting value of hemostatic parameters to diabetic kidney disease in the clinic Prediction ability of PT, PTA, FIB with DKD. (A) Receiver operating characteristic curve (ROC) for predicting the occurrence of diabetic kidney disease with hemostatic parameters. (B) Decision curve analysis (DCA) for the assessment of the predicting value of hemostatic parameters to diabetic kidney disease in the clinic Fig. 5Clinical prediction model assessment. (A) Nomogram estimating the risk of getting diabetic kidney disease. (B) Calibration curve analysis assessing the predicting ability of nomogram Clinical prediction model assessment. (A) Nomogram estimating the risk of getting diabetic kidney disease. (B) Calibration curve analysis assessing the predicting ability of nomogram The mediation role of inflammation parameters to the relationship between coagulation parameters and DKD For the further study of the relationship between coagulation parameters and DKD, a mediation analysis was taken. Initially, the associations between coagulation parameters and DKD were examined. PT was observed to be weakly positively associated with albumin (r = 0.001158, p-value = 0.0370) and neutrophil (r = 0.05904, p-value < 0.0001), and negatively correlated with eosinophil (r=-1.024, p-value < 0.0001). Besides, the associations between PTA and albumin (r = 0.01316, p-value < 0.0007) was weakly positive, the correlation between PTA and eosinophil (r = 8.608, p-value < 0.0001) was positive while association between PTA and neutrophil was negative (r=-0.9870, p-value < 0.0001). Moreover, there was also a positive association between FIB and neutrophil (r = 0.1115, p-value < 0.0001) while FIB was weakly negatively associated to eosinophil (r=-0.3290, p-value < 0.0001) and LMR (r=-0.0006938, p-value = 0.0188) (Supplementary Fig. 2). Then, we also researched the relationship between inflammation parameters and DKD, according to Supplementary Table 2, ALB, eosinophil, neutrophil, and LMR were found to be associated with DKD. Afterward, the mediation analysis of ALB, eosinophil, neutrophil, and LMR to coagulation parameters was carried out (Fig. 6). ALB and LMR mediated 1.57% and 0.25% for the association between PT and DKD while eosinophil and neutrophil mediated − 3.07% and − 3.74% of this association. In addition, ALB, eosinophil, neutrophil, and LMR were also found to contribute − 1.8%, -8.27%, 2.05%, and − 0.15% of the association between PTA and DKD. Eosinophil and LMR were also mediated − 3.30% and 0.57% of the association between FIB and DKD. Fig. 6Mediation analysis of the role of inflammation parameters in the association between coagulation parameters and DKD. (A-D) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of PT with DKD; (E-H) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of PTA with DKD; (I-L) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of FIB with DKD Mediation analysis of the role of inflammation parameters in the association between coagulation parameters and DKD. (A-D) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of PT with DKD; (E-H) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of PTA with DKD; (I-L) Analysis of the mediation by ALB, eosinophil, neutrophil, LMR of the associations of FIB with DKD Discussion This observational study of 12,102 patients with type 2 diabetes from 2013 to 2022 in Chongqing is the first to systematically investigate the association between coagulation parameters and DKD, with a particular focus on the potential mediating role of inflammation. There were three key findings regarding the association between coagulation parameters and DKD. First, PT was negatively associated with the development of DKD, whereas PTA and FIB exhibited a positive association with the risk of DKD. Second, PT was demonstrated to be the most effective predictor of the incidence of DKD compared to other coagulation parameters. Third, mediation analysis revealed that ALB, eosinophil, neutrophil, and LMR were linked to coagulation parameters and played a mediating role in the association between coagulation parameters and DKD. In this study, PT was found to be negatively associated with DKD, and it further demonstrated a stronger predictive performance than the other coagulation parameters. Given that a shortened PT reflects a hypercoagulable state that may increase the risk of DKD progression, we propose two potential pathophysiological mechanisms to explain this association. First, hypercoagulability has been associated with elevated eNOS levels, which in turn accelerate DKD progression [17]. Second, hyperglycemia in patients with diabetes can damage endothelial cells, leading to the release of tissue factor (TF), which directly activates the extrinsic coagulation pathway [10]. Together, these mechanisms contribute to a sustained hypercoagulable state in patients with diabetes, a condition that is directly reflected by PT. The subsequent activation of the coagulation system may further injure renal cells and promote the development of kidney disease [11, 12]. As an inverse percentage derived from the PT ratio, PTA shows a predictive performance similar to PT. In addition, to the best of our knowledge, clinical trial evidence examining the associations between PTA or PT and DKD remains limited. Accordingly, our study helps to address this important gap in the current body of evidence. Conversely, FIB, another coagulation parameter positively associated with DKD, showed unstable results and relatively low predictive ability. Specifically, logistic regression analysis indicated that the association between Tertile 2 and Tertile 1 was not statistically significant, and subgroup analyses revealed significant differences in certain subgroups. This finding may be related to the glycation of FIB in diabetic patients under hyperglycemic status. Previous study reported that glycated FIB tends to aggregate together more easily and is more resistant to degradation [18]. As a result, the measured FIB levels in patients with DKD may not accurately reflect coagulation status, compromising the accuracy and reliability of this biomarker. However, it is noteworthy that although we have confirmed and compared the predictive value of coagulation parameters for DKD, the effect size was modest. This can be attributed to the fact that coagulation is a complex process involving multiple factors, and a single coagulation parameter can only reflect part of this process. Thus, combination of different coagulation parameters may comprehensively reflect the coagulation process and exhibit better predictive value than single coagulation parameters. In the mediation analysis, multiple inflammatory parameters including ALB, eosinophils, neutrophils, and LMR were found to effect the association between coagulation and DKD. Notably, although the association between specific inflammatory and coagulation parameters was relatively weak in our own research, their p-values are still significant, indicating a real but weak correlation between inflammation and coagulation. Consistent with previous research, the coagulation process has been shown to promote inflammation process through mechanisms such as platelet activation and aggregation, which lead to leukocyte recruitment and the release of cytokines and chemokines [19–23]. The resulting inflammatory response may further trigger the production of reactive oxygen species (ROS) and advanced glycation end products (AGEs), which in turn can initiate and amplify both renal and systemic inflammation [24–28]. Additionally, recent clinical trials have offered further evidence-based insights that support the key findings of our study. Wang et al. reported that immune-mediated inflammation can promote the progression of DKD, and that logSII is an independent predictor of DN (P < 0.05), with up to 23.41% (16.18%-30.64%) of DN cases attributable to increases in logSII levels [29]. Similarly, Li J. et al. found that SII, NLR, and PLR were highly valuable for diagnosing diabetic microvascular complications, as they were significantly associated with the risk of DN (ORs: 1.52, 1.71, and 1.60, respectively) and DR (ORs: 1.57, 1.79, and 1.55, respectively) [30]. Moreover, Liu W. et al. underscored the potential implications of utilizing SII and SIRI as cost-effective and readily available inflammatory indicators for monitoring DKD in primary care settings [31]. These findings suggest that inflammation may act as a crucial intermediary in the relationship between coagulation abnormalities and DKD progression, providing new insights into the underlying mechanisms of disease progression. Interestingly, neutrophils exhibited a suppressing effect to the association between PTA and DKD, which was contrary to our expectations. This phenomenon can be interpreted by the dual role of neutrophil extracellular traps (NETs). On one hand, they can induce glomerular endothelial dysfunction by activating the NLRP3 inflammasome and IL-1β signaling pathway [32]. On the other hand, their aggregated structure may serve to localize and control the inflammatory response, exhibiting an anti-inflammatory effect under certain conditions [33]. In contrast to the suppressing effect of neutrophil, eosinophil was found to significantly mediate the association between PT and DKD. This finding is supported by previous research that integrin which is commonly expressed on eosinophils can bind to vascular cell adhesion molecule 1 (VCAM-1)—a molecule implicated in glomerular filtration dysfunction and urinary albumin excretion [34, 35]. Further progression of these mechanisms may eventually lead to renal injury characteristic of DKD including glomerulosclerosis, ECM accumulation, and fibrosis can be viewed as a consequence of inflammatory processes initiated by coagulation and subsequently reflected within the kidney [36, 37]. Our study is a large multi-center study that includes more than 12,000 patients with diabetes from 2013 to 2022, which is its main strength. While another strength is that our study provides additional evidence supporting the associations of PT and PTA with DKD that had previously been overlooked, and compared their predictive abilities with that of FIB. Additionally, our study demonstrated the influence of inflammation on the association between coagulation parameters and DKD. The proposed coagulation–inflammation–DKD framework offers a new perspective on the potential pathophysiology of DKD, highlighting that the close link between inflammation and renal injury in patients with diabetes may be partly reflected through alterations in coagulation parameters, thereby providing useful insights for clinical prevention and diagnosis. One of the primary limitations of our study is its retrospective nature and the fact that the all the participants of our study came from southwestern China, which may introduce regional bias. In addition, due to various clinical constraints, several variables including underlying diseases, medication, and BMI were not included, which may lead to bias in clinical modeling. Among them, key cofounders including coronary diseases and anticoagulant use were of particular concern. Moreover, the inflammatory status of DKD patients could not be fully characterized, as inflammatory markers such as IL-6, TNF-α, and CRP were not available in our dataset. Another limitation of this research is that the coagulation parameters used in our study were all measured and calculated artificially, which may cause errors because of different operations. In addition, PT, PTA, APTT, and TT were parameters used to assess the coagulation function and can not show the specific change of substances in the clotting process. Future studies should take these limitations into consideration to further deepen understanding in this area. In summary, our study demonstrates the association between coagulation parameters and DKD, assesses their predictive ability, and explores the mechanisms mediating the relationship between coagulation parameters and DKD. These insights enrich the existing evidence regarding the clinical utility of coagulation parameters in predicting DKD, providing important references for early risk stratification and the development of intervention strategies in patients with diabetes. Moreover, identifying inflammation as a potential intermediary between coagulation and DKD offers new perspectives for exploring the pathophysiological mechanisms of DKD. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1