Yan Jin-hua, Yang Dai-zhi, Deng Hong-rong, Li Jin, Weng Jian-ping
Department of Endocrinology & Metabolic Disease, Sun Yat-sen University, Guangzhou, China.
Zhonghua Yi Xue Za Zhi. 2013 Mar 26;93(12):897-901.
To determine the incidence and the predictors of diabetes ketoacidosis (DKA) in Chinese type 1 diabetics so as to lay a foundation for better prevention and treatment.
For this cross-sectional study, a total of 611 patients with established type 1 diabetes between August 6, 2010 and March 31, 2012 were recruited from 16 hospitals in Guangdong Province. And 491 of them were over 18 years old. A data entry form was used to collect the patient information on demographics, medical history, acute/chronic complications, smoking/drinking status, diet, exercise, physical examination and treatment, etc. Hemoglobin A1c (HbA1c) and stimulated C peptide levels were centrally measured. The incidence rate of diabetic ketoacidosis (DKA) was calculated at events per 100 patient-years. To determine the predictors of DKA, Poisson's regression model was used for analysis. And backward stepwise logistic regression analysis was performed to identify the predictors of DKA recurrence. The protocol and informed consent form were approved by Ethics Committee of Third Affiliated Hospital, Sun Yat-sen University. Written informed consent was obtained from patients (age > 18 years) or their legal guardians (age < 18 years).
Among them, 53.7% were females. The mean age was 27.8 years (range: 19.5 - 37.3). The age of onset was 22.7 (14.0 - 31.4) years old and disease duration 4.3 (1.7 - 7.9) years. Overweight and obese patients accounted for 10.8% and 1.0% respectively. Among them, the self-monitoring frequency of blood glucose was 0.4 (0.1 - 1.4) times per day. Overall, 26.4% patients reached the target of age-specific HbA1c values. The overall incidence of DKA was 26.4 per 100 patient-years. Significant predictors of DKA in the Poisson regression model were females (RR = 2.12), medical insurance claiming percentage below 50% (RR = 1.84), uncontrolled diet (never controlled diet vs. usually controlled diet, RR = 1.76), smoking (RR = 2.18) as well as worse glycemic control (HbA1c per 1.0% increment, RR = 1.15). Totally, 34.4% of DKA episodes occurred in 3.8% of type 1 diabetics with recurrent events (no less than 2 episodes). The recurrence of DKA was associated with females (RR = 10.56), smoking (RR = 6.99), worse beta cell function (stimulated C peptide per 100 pmol/L decrement, RR = 4.88) and worse glycemic control (HbA1c per 1.0% increment, RR = 1.16).
There is a high incidence of DKA in Chinese type 1 diabetics. And it is recurrent in high-risk patients. Comprehensive management should be offered to control modifiable risk factors in these patients.
确定中国1型糖尿病患者糖尿病酮症酸中毒（DKA）的发病率及其预测因素，为更好地预防和治疗奠定基础。
在这项横断面研究中，2010年8月6日至2012年3月31日期间，从广东省16家医院招募了611例确诊为1型糖尿病的患者。其中491例年龄超过18岁。使用数据录入表收集患者的人口统计学、病史、急/慢性并发症、吸烟/饮酒状况、饮食、运动、体格检查和治疗等信息。糖化血红蛋白（HbA1c）和刺激后C肽水平进行集中检测。糖尿病酮症酸中毒（DKA）的发病率按每100患者年的事件数计算。采用泊松回归模型分析DKA的预测因素。进行向后逐步逻辑回归分析以确定DKA复发的预测因素。该方案和知情同意书经中山大学附属第三医院伦理委员会批准。获得了患者（年龄>18岁）或其法定监护人（年龄<18岁）的书面知情同意。
其中，53.7%为女性。平均年龄为27.8岁（范围：19.5 - 37.3）。发病年龄为22.7（14.0 - 31.4）岁，病程为4.3（1.7 - 7.9）年。超重和肥胖患者分别占10.8%和1.0%。其中，血糖自我监测频率为每天0.4（0.1 - 1.4）次。总体而言，26.4%的患者达到了特定年龄的HbA1c值目标。DKA的总体发病率为每100患者年26.4例。泊松回归模型中DKA的显著预测因素为女性（RR = 2.12）、医疗保险报销比例低于50%（RR = 1.84）、饮食控制不佳（从不控制饮食与通常控制饮食相比，RR = 1.76）、吸烟（RR = 2.18）以及血糖控制较差（HbA1c每增加1.0%，RR = 1.15）。在3.8%的复发事件（不少于2次发作）的1型糖尿病患者中，34.4%的DKA发作发生。DKA的复发与女性（RR = 10.56）、吸烟（RR = 6.99）、β细胞功能较差（刺激后C肽每降低100 pmol/L，RR = 4.88）和血糖控制较差（HbA1c每增加1.0%，RR = 1.16）有关。
中国1型糖尿病患者中DKA的发病率较高。高危患者中DKA易复发。应提供综合管理以控制这些患者中可改变的危险因素。
