基于ROC曲线和决策曲线分析CTRP6和Irisin蛋白对2型糖尿病的预测价值

Analysis of predictive value of CTRP6 and Irisin protein for type 2 diabetes mellitus based on ROC curve and decision curve

目的基于受试者工作特征(ROC)曲线和决策曲线探讨C1q/肿瘤坏死因子6(CTRP6)和鸢尾素(Irisin)蛋白对2型糖尿病的预测价值。方法研究共纳入116例受试者,其中健康体检者61例作为对照组,新诊断2型糖尿病的55例患者作为2型糖尿病组。比较两组患者临床生化指标以及血清CTRP6和Irisin蛋白水平。采用多因素Logistic回归分析2型糖尿病的独立预测因素。分别对患者血清CTRP6和Irisin蛋白水平与部分临床生化指标进行Pearson相关性分析。绘制CTRP6和Irisin蛋白水平预测2型糖尿病的ROC曲线,并进行分析。采用决策曲线分析CTRP6和Irisin蛋白单独以及联合预测2型糖尿病的净获益情况。结果单因素分析结果显示:两组患者体质量指数、甘油三酯、糖化血红蛋白、胰岛素抵抗指数(HOMA-IR)、C反应蛋白、白细胞介素(IL-6)、CTRP6以及Irisin蛋白比较差异有统计学意义(P<0.05)。多因素Logistic回归分析显示:糖化血红蛋白、HOMA-IR、CTRP6以及Irisin蛋白为2型糖尿病的独立预测因素(P<0.05),其中CTRP6每增加1 ng/mL,2型糖尿病风险增加1.178倍(HR:2.178,95%CI:1.245~2.875,P<0.001),Irisin蛋白每减少1μg/mL,2型糖尿病风险增加1.033倍(HR:0.492,95%CI:0.235~0.737,P=0.009)。Pearson相关性分析显示:2型糖尿病患者CTRP6与糖化血红蛋白、HOMA-IR、IL-6均呈正相关(r=0.695、0.748、0.589,P<0.05),Irisin蛋白与糖化血红蛋白、HOMA-IR、C反应蛋白均呈负相关(r=-0.643、-0.732、-0.537,P<0.05)。ROC曲线分析显示:CTRP6和Irisin蛋白预测2型糖尿病的曲线下面积(AUC)分别为0.859和0.793,而两项指标的联合预测的价值最高,AUC为0.939。决策曲线分析显示:在大部分高风险阈值范围内,CTRP6和Irisin蛋白预测有无2型糖尿病均具有良好的净获益,并且联合预测的净获益高于单一指标的净获益。结论CTRP6和Irisin蛋白对2型糖尿病均具有较高的预测价值。

Objective To investigate the predictive value of C1q/tumor necrosis factor-related protein 6(CTRP6)and Irisin protein in type 2 diabetes mellitus based on receiver operator characteristic curve(ROC)curve and decision curve.Methods A total of 116 people were enrolled in the study,including 61 healthy people who had undergone healthy examination in the hospital(control group)and 55 people newly diagnosed with type 2 diabetes mellitus(type 2 diabetes mellitus group).The clinical biochemical parameters and the levels of CTRP6 and Irisin protein of the two groups were compared.Multivariate Logistic regression analysis was used to obtain the independent predictors of type 2 diabetes.The correlation between CTRP6 and Irisin protein and clinical biochemical indicators of the patients were analyzed by Pearson correlation.The ROC curve of CTRP6 and Irisin proteins which were used to predict type 2 diabetes was drawn,and ROC analysis was performed.Results Univariate analysis showed that there were significant differences in body mass index,triglyceride,glycosylated hemoglobin,HOMA-IR,C-reactive protein,IL-6,CTRP6 and Irisin protein between the two groups(P<0.05).Multivariate Logistic regression analysis showed that glycosylated hemoglobin,HOMA-IR,CTRP6 and Irisin protein were independent predictors of type 2 diabetes(P<0.05).The risk of type 2 diabetes increased 1.178 times(HR:2.178,95%CI:1.245—2.875,P<0.001)for every 1 ng/mL increase in CTRP6,and every 1μg/mL decrease in Irisin protein,the risk of type 2 diabetes increased 1.033 times(HR:0.492,95%CI:0.235—0.737,P=0.009).There was a significant positive correlation between CTRP6 and glycosylated hemoglobin,HOMA-IR and IL-6 in type 2 diabetic patients(r=0.695,0.748,0.589,P<0.05),and Irisin protein was significantly negatively correlated with glycosylated hemoglobin,HOMA-IR and C-reactive protein(r=-0.643,-0.732,-0.537,P<0.05).ROC curve analysis showed that the area under the curve(AUC)of CTRP6 and Irisin protein predicted for type 2 diabetes mellitus was 0.859 and 0.793,respectively.The combined prediction by using the 2 indicators had the highest value and AUC was 0.939.The result of decision curve analysis showed that within most high-risk threshold ranges,both CTRP6 and Irisin protein had good net benefits in predicting the presence or absence of type 2 diabetes,and the net benefits of being predicted by using the two indicators jointly were higher than a single indicator.Conclusion CTRP6 and Irisin protein have good predictive value for type 2 diabetes mellitus.