| Background/Aims: Diabetes mellitus is the leading reason of end stage renal disease. The number of patients with normal albuminuria and renal function decline is growing. However, the patients with DKD of them still haven’t been able to get proper clinical diagnosis according to the existing DKD clinical standards of NKF-K/DOQI guideline.This study was to investigate whether glomerular filtration rate (GFR) could be the clue for clinical diagnosis of DKD in type2diabetes mellitus (T2DM), and whether the combination of DR and GFR decline could be a supplement standard for NKF standard, by analyzing the correlation between GFR and DR, and by evaluating the efficiency of supplement standard with Receiver Operating Characteristic curve (ROC curve) and the evaluation criteria of DKD likelihood in NKF guide. We explored the clinic-valuable cut-off point of GFR decline, by comparing the extra-diagnostic application situations of supplement standard among different cut-off points. We further evaluated the clinical application value of the supplement standard, by analyzing whether the patients diagnosed by supplement standard or combination standard accorded with the characteristics of progression in DKD. The ultimate purpose of this study was to try to provide a new idea for the clinical diagnosis of normoalbuminuric DKD in T2DMMethods:656patients with T2DM in the Shanghai downtown and568patients with T2DM in Huashan Hospital were included in the cross-sectional study. The information including the fundus examination, albuminuria creatinine ratio (ACR), serum creatinine (sCr), height, weight, waist circumference, systolic blood pressure(SBP), diastolic blood pressure(DBP), glycated hemoglobin(HbAlc), fasting plasma glucose(FBG),2-hour postprandial blood glucose(2hPBG), fasting insulin (Fins), total cholesterol (TC), triglycerides(TG), low-density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol(HDL-C), blood urea nitrogen(BUN), uric acid(UA), and the social characteristics of patients were collected completely.392community patients with complete information in the cross-sectional study accepted the return visits2years later, and their sCr were measured again.The clinical characteristics were compared between the community patients and the inpatients. The analyses below were performed respectively in the community patients and the inpatients.(1)The percentage of DR, normal albuminuria or chronic kidney disease (CKD) was calculated.(2)The differences of clinical characteristics were compared between the patients with DKD and those with non-DKD (NDKD) when NKF standard was used for diagnosis.(3)The differences of ACR and GFR were compared between the patients with DR and those with non-diabetic retinopathy (NDR).(4)GFR and proportion of DR were compared between the normoalbuminuric patients and the albuminuric patients.(5)The proportion of DR and the proportion of albuminuria were compared among different levels of GFR.(6) The pairwise correlation coefficients of ACR, DR and GFR were calculated.(7)The efficiency of GFR decline to predict DR was evaluated by ROC curve analysis. The cut-off point corresponding to the maximum of Youden’s index was explored. The sensitivity and specificity were discussed among different cut-off points of GFR decline.(8) Within the patients with DR, the efficiency of supplement standard to diagnose DKD was evaluated by ROC curve analysis. The cut-off point corresponding to the maximum of Youden’s index was explored. The sensitivity and specificity were discussed among different cut-off points of GFR decline.(9)The proportions of patients with unlikely DKD among the patients diagnosed by supplement standard were analyzed among different cut-off points of GFR decline, using the evaluation criteria of DKD likelihood in NKF guide.(10) The number of patients extra-diagnosed by combined standard was accounted and the differences of clinical characteristics were compared between the extra-diagnosed patients and the patients diagnosed by NKF standard, respectively among different cut-off points of GFR decline. In the patients of the return visit, the baseline prognostic factors of GFR progressive decline in T2DM were analyzed by Logistic stepwise regression model. DKD was diagnosed respectively by NKF standard, supplement standard and combined standard. And the situations of GFR changes within these2years were compared between the patients with DKD and NDKD among these three standards. The baseline prognostic factors of GFR progressive decline in DKD were analyzed by Logistic stepwise regression model under these three standards. The progression situations of GFR and baseline clinical characteristics were compared between the patients extra-diagnosed by the supplement standard and the patients diagnosed by NKF standard. The combined standard meant someone met NKF standard or the supplement standard.Results:(1) The age, gender, duration of diabetes, smoking, body mass index (BMI), HbAlc, proportion of DR and proportion of albuminuria were significantly different between the community patients and the inpatients with T2DM.(2)Of the community patients with T2DM,23.02%were with DR,55.6%were with normal albuminuria,12.37%were with CKD stages3-5when GFR was calculated by the CKD-EPI equation. Among the patients with CKD stages3-5,26.32%were normoalbuminuric.Of the inpatients with T2DM,34.5%were with DR,66.2%were with normal albuminuria,12.68%were with CKD stages3-5when GFR was calculated by the CKD-EPI equation. Among the patients with CKD stages3-5,31.94%were normoalbuminuric.Among the community patients, smoking, BUN and ACR of the subjects with DKD were significantly higher than the subjects with NDKD when using the NKF standard.(3) Among the community patients with T2DM, ACR and the estimated GFR were significantly different between the patients with DR and those with NDR. The estimated GFR and proportion of DR were significantly different between albuminuric and normoalbuminuric patients. Proportions of DR or albuminuria both increased when the levels of the estimated GFR decreased. The estimated GFR was negatively correlated with ACR (r=-0.203, p<0.001) and DR (r=-0.119, p<0.001).(4) Among the inpatients with T2DM, The estimated GFR was also negatively correlated with ACR (r=-0.387, p<0.001) and DR (r=-0.161, p<0.001).(5)Among the community patients with T2DM, area under curve (AUC) of ROC curve of estimated GFR decline to predict DR was0.581(95%CI0.530,0.634, p=0.003). The cut-off point corresponding to the maximum of Youden’s index was85.5. The cut-off point of30corresponded to the specificity of0.994and the sensitivity of0.007. The cut-off point of60corresponded to the specificity of0.951and the sensitivity of0.105.(6)Among the inpatients with T2DM, AUC of ROC curve of estimated GFR decline to predict DR was0.575(95%CI0.524,0.627, p=0.003). The cut-off point corresponding to the maximum of Youden’s index was84.5. The cut-off point of30corresponded to the specificity of0.995and the sensitivity of0.061. The cut-off point of60corresponded to the specificity of0.909and the sensitivity of0.194.(7) Among the community patients,98patients were with DKD diagnosed by NKF standard, of which90.82%were with DR. Among the patients with DR, AUC of ROC curve of supplement standard to diagnose DKD was0.635(95%CI0.548,0,723, p=0.01). The cut-off point corresponding to the maximum of Youden’s index was86.5. The cut-off point of30corresponded to the specificity of1.000and the sensitivity of0.011. The cut-off point of60corresponded to the specificity of0.984and the sensitivity of0.157.Among the inpatients,119patients were with DKD diagnosed by NKF standard, of which84.03%were with DR. Among the patients with DR, AUC of ROC curve of supplement standard to diagnose DKD was0.655(95%CI0.578,0.732, p<0.001). The cut-off point corresponding to the maximum of Youden’s index was76.5. The cut-off point of30corresponded to the specificity of1.000and the sensitivity of0.120. The cut-off point of60corresponded to the specificity of0.938and the sensitivity of0.310.According to the evaluation criteria of DKD likelihood in the NKF guide, among the community patients diagnosed by supplement standard, the proportion of patients with unlikely DKD was1.81%(1/55) on the cut-off point of85,6.67%(1/15) on the cut-off point of60, and0%(0/1) on the cut-off point of30. Among the community patients diagnosed by supplement standard, the proportion of patients with unlikely DKD was23%(13/56) on the cut-off point of75,34%(13/38) on the cut-off point of60, and60%(6/10) on the cut-off point of30.(8)Among the community patients,16patients could be extra-diagnosed by the combined standard on the cut-off point of85, with16.3%more than NKF standard. The extra-diagnosed patients had no differences of clinical characteristics from those diagnosed by NKF standard (p>0.01) except that they had normal albuminuria. One patient could be extra-diagnosed on the cut-off point of60, with1.02%more than NKF standard. No patients could be extra-diagnosed on the cut-off point of30.Among the inpatient,12patients could be extra-diagnosed by the combined standard on the cut-off point of75, with10.08%more than NKF standard. The extra-diagnosed patients had no differences from those diagnosed by NKF standard (p>0.01) except that they had no albuminuria.7patients could be extra-diagnosed on the cut-off point of60, with5.88%more than NKF standard. No patients could be extra-diagnosed on the cut-off point of30.(9) Of the community patients with T2DM, the baseline prognostic factors of estimated GFR progressive decline was age (OR=1.0695%CI1.02,1.11, p=0.01)(10)With these three diagnostic standards, the proportions of the patients with estimated GFR progressive decline were all significantly higher in those with DKD than without DKD.(NKF standard (25.42%vs12.31%, p=0.01), supplement standard (27.27%vs12.64%, p=0.012) and combined standard (23.81%vs11.69%, p<0.001)). Under all of these three diagnostic standards, the proportions of the patients without estimated GFR decline were found no differences between those with DKD and those without DKD (p>0.05).(11)The baseline prognostic factors of estimated GFR progressive decline in the patients with DKD were estimated GFR when DKD was diagnosed by NKF standard or the combined standard, and HbAlc when DKD was diagnosed by supplement standard. The consequences were accordant with the prognostic factors of progression in DKD reported in the other literatures.(12)25patients were extra-diagnosed by supplement standard in the followed-up community patients. They had no significant differences from those diagnosed by NKF standard, in the proportion of the subjects with estimated GFR progressive decline within2years and baseline clinical characteristics, except that they had no albuminuria.Conclusions:(1) Though the clinical characteristics were widely different between the community patients with T2DM and the inpatients with T2DM, the same inferences could be made below. GFR was negatively correlated with DR. Combination of DR and GFR decline as the supplement standard had the clinical value of diagnosis for DKD. The supplement standard for DKD could diagnose few extra patients but it had high specificity, when the cut-off points of GFR decline were set on30or60. The cut-off points corresponding to the maximum of Youden index in the community patients and the inpatients were both between60and90, on these points the combined standard could diagnose10%to20%more patients than NKF standard. The extra-diagnosed patients had no significant differences from those diagnosed by NKF standard except that they had no albuminuria.(2)The progression of estimated GFR decline and its prognostic factors, in the DKD patients diagnosed by supplement standard or the combined standard, were accordant with the characteristics of progression in DKD. The extra-diagnosed patients had no significant differences from those diagnosed by NKF standard, in the progression of estimated GFR and baseline clinical characteristics, except that they had no albuminuria.(3)The combination of DR and GFR decline might be a potential supplement standard of diagnosis for DKD. Its application might be useful for the clinical diagnosis of DKD in the normal albuminuric patients. |
PDF Full Text Request