纵向生存分析揭示糖尿病困扰与抑郁症状的不同预测因素:提示病因路径分歧
该研究对654例1型或2型糖尿病患者每6个月随访24个月,采用Cox比例风险模型分析糖尿病困扰和抑郁症状的发生与缓解预测因素。结果显示两者发病率与缓解率不同,且预测因素存在差异:糖尿病困扰与焦虑障碍相关,抑郁症状与既往抑郁和进食障碍相关,提示病因路径不同。
要点
- 糖尿病困扰24个月累积发生率21%,缓解率70%;抑郁症状发生率33%,缓解率67%。
- 较高HbA1c预测糖尿病困扰和抑郁症状的发生;女性、长期并发症、焦虑障碍史预测糖尿病困扰发生。
- 既往抑郁和进食障碍预测抑郁症状发生;仅较大年龄与抑郁症状缓解相关。
核心结论
- 血糖控制不佳同时预测糖尿病困扰和抑郁症状的发生,但两者的其他预测因素不同,提示病因路径存在分歧。
研究设计与样本
- 纵向观察研究,纳入654例1型或2型糖尿病患者,每6个月收集数据,共24个月,采用Cox比例风险模型分析发生与缓解的预测因素。
主要发现
- 糖尿病困扰24个月累积发生率为21%,缓解率为70%;抑郁症状累积发生率为33%,缓解率为67%。
- 糖尿病困扰中位发生时间为18个月,抑郁症状为24个月;两者中位缓解时间均为6个月。
- 较高HbA1c(HR=1.02,p=0.022)、女性(男性HR=0.55,p=0.043)、长期并发症(HR=2.11,p=0.009)和焦虑障碍史(HR=2.57,p=0.029)显著预测糖尿病困扰发生,无因素与缓解相关。
- 较高HbA1c(HR=1.03,p<0.001)、既往抑郁(HR=2.63,p=0.001)和进食障碍(HR=2.20,p=0.044)预测抑郁症状发生;仅较大年龄(HR=1.02,p=0.045)与缓解显著相关。
临床怎么用
- 对糖尿病患者应同时监测糖尿病困扰和抑郁症状,并关注不同的预测因素,如焦虑障碍史、既往抑郁和进食障碍。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
AIMS/HYPOTHESIS: The aim of the study was to examine the course of diabetes distress and depressive symptoms and their predictors of incidence and remission in individuals with type 1 and type 2 diabetes. METHODS: Data were collected every 6 months over a 24 month period. Participants (n=654) completed measures of diabetes distress (Problem Areas in Diabetes Scale) and depressive symptoms (Patient Health Questionnaire 8). Cox proportional hazards models were applied to examine predictors of incidence and remission, considering demographic, clinical and psychosocial factors. All predictors were assessed at baseline, except for HbA1c, which was modelled as a time-varying covariate. RESULTS: Diabetes distress showed cumulative incident cases in 21% of the sample and a remission rate of 70% across 24 months. For depressive symptoms, cumulative 24 month incidence was 33% and remission was 67%. The median onset time was 18 months for diabetes distress and 24 months for depressive symptoms; the median remission time for both was 6 months. Higher HbA1c (HR=1.02, p=0.022), female gender (male gender HR=0.55, p=0.043), long-term complications (HR=2.11, p=0.009) and a history of anxiety disorders (HR=2.57, p=0.029) significantly predicted the incidence of diabetes distress, while no predictors were associated with remission. For depressive symptoms, higher HbA1c (HR=1.03, p<0.001), prior depression (HR=2.63, p=0.001) and eating disorders (HR=2.20, p=0.044) predicted incidence. Remission was significantly associated only with older age (HR=1.02, p=0.045). CONCLUSIONS/INTERPRETATION: Suboptimal glycaemic outcomes predicted both diabetes distress and depression; however, diabetes distress was associated with anxiety disorders, whereas depressive symptoms were linked to prior depression and eating disorders, hinting at distinct aetiologies.
引用
- Lehmann G, Ehrmann D, Olesen B, et al. Uncovering distinct predictors of diabetes distress and depressive symptoms in a longitudinal survival analysis of incidence and remission: indication for diverging aetiological paths. Diabetologia. 2026. PMID: 41572009. doi:10.1007/s00125-026-06666-w
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。