Prognosis for residual islet β-cell secretion function in young patients with newly diagnosed type 1 diabetes†
初发年轻1型糖尿病患者残余胰岛功能的预测
Abstract
enBackground
This study investigated possible predictors of residual islet β-cell function (RBF) in young patients with newly diagnosed type 1 diabetes (T1D).
Methods
After analyzing RBF in 443 patients with T1D according to age at diagnosis and disease duration, 110 were followed-up over 18-60 months. A nomogram was developed by logistic regression to explore factors associated with long-term RBF.
Results
Of the 443 T1D patients (mean [±SD] age 20.28 ± 5.50 years; mean [±SD] diabetes duration 28.5 ± 14.6 months), RBF was preserved in 64.3%. Independent predictors for poor RBF outcome among the 110 patients in the follow-up cohort were age at onset (odds ratio [OR] 0.82; 95% confidence interval [CI] 0.73-0.92; P < 0.001), high-risk human leukocyte antigen (HLA) status (OR 4.73; CI 1.28-17.52; P = 0.020), female sex (OR 3.39; CI 1.03-11.22; P = 0.045), and a history of diabetic ketoacidosis (DKA; OR 8.71; CI 2.31-32.83; P < 0.001). Baseline glutamic acid decarboxylase (GAD) antibody, family history of diabetes, body mass index, insulin dosage, and C-peptide and HbA1c levels were not associated with poor RBF outcome. Intensive glycemic control after T1D diagnosis may improve RBF within a mean (±SD) follow-up of 35.1 ± 13.8 months. The calibration plot for the probability of 2-, 3-, and 4-year RBF showed optimal agreement between nomogram-predicted and actual observed probabilities.
Conclusions
Younger age of onset, female sex, higher HLA risk status, and a history of DKA were the main factors predicting long-term poor preserved β-cell function. Glycemic control could improve RBF during the course of diabetes. The nomogram provides an individualized risk estimate of RBF in patients with newly diagnosed T1D within Chinese Han populations.
Abstract
zh摘要
目的
探讨影响初发年轻1型糖尿病(type 1 diabetes, T1D)患者病程进展后残余胰岛功能(Residual β-cell Function, RBF)的预测因素。
方法
对443例T1D患者按诊断年龄、病程进行RBF分析后，入选110例患者进行18-60个月的临床随访研究。基于Logistic回归分析建立Nomogram模型，预测长期RBF的相关因素。
结果
443例T1D患者平均年龄20.28±5.50岁，平均病程28.5±14.6个月，保留RBF者占64.3%。在110例患者的随访研究中，低RBF的独立预测因素有:起病年龄(岁)(OR=0.82，95% CI: 0.73-0.92，P<0.001)、HLA高危表型(OR=4.73, 95%CI:1.28-17.52，P=0.020)、女性(OR=3.39，95%CI: 1.03-11.22，P=0.045)以及酮症酸中毒起病史(OR=8.71，95%CI: 2.31-32.83，P=0.001)。而基线的谷氨酸脱羧酶抗体、糖尿病家族史、体重指数、胰岛素用量、C肽和糖化血红蛋白水平与RBF预后不良无关。在平均35.1±13.8个月的随访中，强化血糖控制可改善RBF。2、3、4年的Calibration曲线显示,Nomogram预测概率和实际观测概率间吻合度较高。
结论
起病年龄小、女性、HLA的高危表型、DKA起病史是较长病程后低RBF的影响因素，病程中的血糖控制可以提升RBF。Nomogram预测模型提供了汉族人群中初发T1D患者RBF的个体化风险评估。
