Jones Angus G, Shields Beverley M, Hyde Christopher J, Henley William E, Hattersley Andrew T
NIHR Exeter Clinical Research Facility, University of Exeter Medical School, Exeter, United Kingdom.
Public Health and Epidemiology, University of Exeter Medical School, Exeter, United Kingdom.
PLoS One. 2014 Oct 23;9(10):e111235. doi: 10.1371/journal.pone.0111235. eCollection 2014.
Defining responders to glucose lowering therapy can be important for both clinical care and for the development of a stratified approach to diabetes management. Response is commonly defined by either HbA1c change after treatment or whether a target HbA1c is achieved. We aimed to determine the extent to which the individuals identified as responders and non-responders to glucose lowering therapy, and their characteristics, depend on the response definition chosen.
We prospectively studied 230 participants commencing GLP-1 agonist therapy. We assessed participant characteristics at baseline and repeated HbA1c after 3 months treatment. We defined responders (best quartile of response) based on HbA1c change or HbA1c achieved. We assessed the extent to which these methods identified the same individuals and how this affected the baseline characteristics associated with treatment response.
Different definitions of response identified different participants. Only 39% of responders by one definition were also good responders by the other. Characteristics associated with good response depend on the response definition chosen: good response by HbA1c achieved was associated with low baseline HbA1c (p<0.001), high C-peptide (p<0.001) and shorter diabetes duration (p = 0.01) whereas response defined by HbA1c change was associated with high HbA1c (p<0.001) only. We describe a simple novel method of defining treatment response based on a combination of HbA1c change and HbA1c achieved that defines response groups with similar baseline glycaemia.
The outcome of studies aiming to identify predictors of treatment response to glucose lowering therapy may depend on how response is defined. Alternative definitions of response should be considered which minimise influence of baseline glycaemia.
确定降糖治疗的反应者对于临床护理以及制定分层糖尿病管理方法均具有重要意义。反应通常通过治疗后糖化血红蛋白（HbA1c）的变化或是否达到目标HbA1c来定义。我们旨在确定被认定为降糖治疗反应者和无反应者的个体及其特征在多大程度上取决于所选择的反应定义。
我们前瞻性地研究了230名开始接受胰高血糖素样肽-1（GLP-1）激动剂治疗的参与者。我们在基线时评估了参与者的特征，并在治疗3个月后重复检测了HbA1c。我们根据HbA1c变化或达到的HbA1c定义反应者（反应最佳四分位数）。我们评估了这些方法在多大程度上识别出相同的个体，以及这如何影响与治疗反应相关的基线特征。
不同的反应定义识别出不同的参与者。按一种定义的反应者中，只有39%按另一种定义也是良好反应者。与良好反应相关的特征取决于所选择的反应定义：达到的HbA1c显示良好反应与低基线HbA1c（p<0.001）、高C肽（p<0.001）和较短的糖尿病病程（p = 0.01）相关，而由HbA1c变化定义的反应仅与高HbA1c（p<0.001）相关。我们描述了一种基于HbA1c变化和达到的HbA1c相结合来定义治疗反应的简单新方法，该方法定义了具有相似基线血糖水平的反应组。
旨在识别降糖治疗反应预测因素的研究结果可能取决于反应的定义方式。应考虑替代反应定义，以尽量减少基线血糖的影响。
