研究背景与研究目的：我国2型糖尿病的患病率高、未诊断率高、治疗率低，不仅影响患者个人生活质量，也给社会带来了沉重负担。同时，已诊断的患者药物依从性不佳，之前的研究显示D型人格与个体不良的健康状况以及行为之间存在关联。因此，本研究将对2型糖尿病患者D型人格、药物依从性及血糖控制水平的关系进行探究。 研究方法：在北京市某医院就诊的2500名2型糖尿病患者中，运用系统抽样方法，以135作为抽样的随机起点，以6作为抽样间隔，随机抽取450名患者作为本研究的研究对象。研究采用前瞻性设计，6个月后再次随访。用DS14量表（Type D personality Scale 14）测量研究对象是否具有D型人格，用MMAS-8量表测量研究对象的药物依从性。研究对象的人口学、临床相关因素等基本信息用均值、标准差和数量、百分率表示；运用卡方检验和t检验比较D型人格和非D型人格的2型糖尿病患者在人口学变量和临床相关变量方面的差异；运用Pearson相关分析法探索变量间的相关性；运用多元线性回归分析D型人格对2型糖尿病患者血糖控制水平的影响；用温忠麟提出的中介效应检验流程检验药物依从性是否对D型人格和2型糖尿病患者血糖控制水平之间的关系起中介作用。 研究结果：在基线调查时共有412名患者完成了调查问卷，6个月后有330名患者完成随访，随访率为80.1%。完成随访的330名患者年龄为57.24±11.45，有172名男性（52.1%）、158名女性（47.9%），教育程度为高中及以下的有159名（48.2%）、大专及以上的有171名（51.8%），基线时的HbA1c值为7.22±1.85。失访的2型糖尿病患者与完成了6个月随访的患者在性别、基线时的HbA1c值之间没有明显差异。D型人格的2型糖尿病患者所占比例（27.9%）低于非D型人格的患者（72.1%）。在控制了人口学变量和临床相关变量后，D型人格能够显著预测2型糖尿病患者6个月后的HbA1c水平（β=0.340 P＜0.0001）和药物依从性水平（β=-0.282 P＜0.0001）。药物依从性在D型人格和2型糖尿病患者血糖控制水平之间起部分中介作用，中介效应量为11.94%。 研究结论：D型人格对2型糖尿病患者血糖控制水平的影响路径有两个。一是在控制了人口学和临床相关变量的基础上，D型人格对2型糖尿病患者6个月随访时的药物依从性及血糖控制水平均有直接影响。二是D型人格通过中介变量药物依从性间接对2型糖尿病患者的血糖控制水平产生影响，中介效应量为11.94%。

Research background and objective: The prevalence of type 2 diabetes mellitus in China is high, the rate of undiagnosed diabetes mellitus is high and the rate of treatment is low. It not only affects the quality of life of patients, but also brings heavy burden to society. At the same time, the medication adherence of the diagnosed patients was poor. Previous studies have shown that type D personality is associated with individuals' poor health and behavior. Therefore, this study will explore the relationship between type D personality, medication adherence and blood sugar control level in patients with type 2 diabetes mellitus. Research methods: In a hospital in Beijing, 2500 patients with type 2 diabetes mellitus were randomly selected by systematic sampling. 135 was selected as the starting point, 6 as the sampling interval. 450 patients were randomly selected as the subjects of this study. The study was prospectively designed and followed up six months later. DS14 scale was used to measure the type D personality of the subjects, and MMAS-8 scale was used to measure the medication adherence of the subjects. The demographic and clinical related factors of the subjects were expressed by mean, standard deviation, quantity and percentage; the differences of social demographic variables and clinical related variables between type 2 diabetic patients with type D personality and type 2 diabetic patients with non-type D personality were compared by chi-square test and t-test; the correlation between variables was explored by Pearson correlation analysis; the influence of type D personality on blood sugar control level in patients with type 2 diabetes mellitus by multiple linear regression analysis; Wen Zhonglin's mediating effect test process was used to test whether medication adherence mediated the relationship between type D personality and blood sugar control level in patients with type 2 diabetes mellitus. Research results: At baseline, 412 patients completed the questionnaire, and after 6 months, 330 patients completed the follow-up, with a follow-up rate of 80.1%. The age of 330 patients who completed the follow-up was 57.24±11.45, 172 males (52.1%) and 158 females (47.9%). 159 patients (48.2%) with high school education or below, 171 patients (51.8%) with college education or above, and the HbA1c value at baseline was 7.22±1.85. There was no significant difference in HbA1c level at baseline and gender between the missing type 2 diabetes mellitus patients and those who completed the 6-month follow-up. The proportion of type 2 diabetes mellitus patients with type D personality (27.9%) was lower than that of non-type D personality patients (72.1%). After controlling demographic and clinical variables, type D personality could significantly predict HbA1c level (β= 0.340 P < 0.0001) and medication adherence (β= -0.282 P < 0.0001) in type 2 diabetes mellitus patients 6 months later. Medication adherence partially mediated the level of blood sugar control in patients with type D personality and type 2 diabetes mellitus, with a mediating effect of 11.94%. Research conclusion: Type D personality has two effects on blood sugar control level in patients with type 2 diabetes mellitus. First, on the basis of controlling demographic and clinical variables, type D personality has a direct impact on medication adherence and blood sugar control level of type 2 diabetic patients during 6-month follow-up. Secondly, type D personality indirectly affects blood sugar control level of type 2 diabetes mellitus patients through mediating variable medication adherence, with mediating effect of 11.94%.