目的：探索基于自我决定理论，从专业、同伴、家庭三个维度提供自主性支持是否能提升糖尿病患者的自我管理行为及血糖控制水平，并分析干预效果的影响因素。方法：采用方便抽样方法，在北京市城区选择6个社区，各社区分别选择条件相似的服务站作为不同干预组别，该站管理的符合纳入标准的糖尿病患者即招募进该组。以社区站为单位分为3组，常规干预组仅发放知识手册并按基本公共卫生服务要求开展健康管理；同伴支持组按社区分为小组，开展医生主导的小组活动；自主性支持组在同伴支持活动的基础上，基于自我决定理论，设计活动材料及流程，并对医生、同伴进行自主性支持的培训，同时以手册形式培训家人，形成三者协同的自主性支持。干预期为3个月，主要评价指标为糖化血红蛋白（HbA1c）及患者自我管理行为、技能、知识、自我效能评分。结果：干预前后，常规干预组的HbA1c分别为7.40%±1.37%、7.30%±1.18%；同伴支持组的HbA1c分别为7.33%±1.15%、7.13%±1.27%；自主性支持组的HbA1c分别为7.42%±1.22%、6.78%±0.80%。干预前后，常规干预组自我管理行为得分分别为10.54±2.28、10.80±2.15，同伴支持组的得分分别为11.09±1.89、11.40±1.78，自主性支持组的得分分别为10.34±1.99、11.10±1.65。自主性支持组的HbA1c均值降低的绝对值高于其余两组，干预后达标率高于其余两组，自我管理行为提升高于其余两组。多因素分析结果显示，干预后HbA1c值与基线HbA1c呈正相关，与干预后自我管理行为得分呈负相关，自主性支持组的HbA1c低于常规干预组，干预后自我管理行为与基线行为得分、干预后自我效能、知识、技能、家庭支持、自主性支持、同伴支持、年龄均呈正相关。结论：基于自我决定理论开展糖尿病自我管理行为干预可以有效促进患者自我管理行为及血糖控制，效果优于单一同伴支持活动。

Objective： To provide autonomy support from three dimensions based on self-determination theory (SDT), i.e. professional support, peer support, family support, and to investigate whether this intervention can improve diabetes self-management behavior and glycemic control of diabetic patients, and to analyze the influencing factors of the effect. Methods: Using convenient sampling method, three communities were selected respectively in Beijing. Each community selected health service stations with similar conditions as different intervention groups. The diabetic patients managed by the station who were eligible for inclusion were recruited into this intervention group. The community stations were divided into three groups. The routine intervention group only issued knowledge manuals and conducted health ma-nagement according to the requirements of basic public health services. Peer support groups were divided into small groups and carried out doctorled group activities. Based on doctor-led peer support activities, the doctors and peers were trained to provide autonomy support based on self-determination theory, and their family members were trained in the form of manuals to provide autonomy support, forming a concerted support of the three dimensions. Activity processes and materials were also designed based on SDT. The intervention duration was 3 months, and the main evaluation indexes were HbA1c and patients’ self-management behaviors, skills, knowledge, and self-efficacy scores. Results: Before and after the intervention, the HbA1c of routine intervention group were 7.40%±1.37%, 7.30%±1.18%. The HbA1c of peer support group before and after the intervention were 7.33%±1.15% and 7.13%±1.27%. The HbA1c of autonomy support group before and after the intervention were 7.42%±1.22% and 6.78%±0.80%. Before and after the intervention, the self-management score in routine intervention group was 10.54±2.28 and 10.80±2.15, the score in peer support group was 11.09±1.89 and 11.40±1.78, the score in autonomy support group was 10.34±1.99 and 11.10±1.65, respectively. The HbA1c and self-management score increased higher in autonomy support group than in the other two groups. After intervention, the control rate in autonomy support group was higher than in the other two groups. According to the multi-factor analysis, the value of HbA1c after intervention was positively related to the baseline HbA1c, and negatively related to self-management behavior. The value in autonomy support group was higher than in routine intervention group. Baseline self-management behavior, self-efficacy, knowledge, skill, family support, autonomy support, peer support and age were positively correlated with the change of behavior. Conclusion: Self-management behavior intervention based on selfdetermination theory can effectively promote selfmanagement behavior and glycemic control of diabetic patients, and the effect is better than single peer support activities.