Service de Pharmacologie Médicale et Clinique, Centre de Pharmacovigilance, de Pharmacoépidémiologie et d'Informations sur le Médicament, UMR 1027 INSERM, CIC 1436, Faculté de Médecine, Centre Hospitalier Universitaire, Université Paul Sabatier, 37 allées Jules-Guesde, 31000, Toulouse, France.
Unité Clinique de Pharmacologie Psychiatrique, Faculté de Médecine, Centre Hospitalier Universitaire, Université Paul Sabatier, Toulouse, France.
Drug Saf. 2018 Nov;41(11):1087-1096. doi: 10.1007/s40264-018-0693-8.
The association between antidepressant exposure and type 2 diabetes mellitus is still debated. Moreover, the pharmacological mechanisms remain unknown.
The objective of this study was to investigate this putative relationship with the role of antidepressant pharmacological targets using the 'pharmacoepidemiological-pharmacodynamic' method.
First, we performed case/non-case analyses in VigiBase (the World Health Organization international database of suspected adverse drug reactions) to examine a signal of increased type 2 diabetes reporting (expressed as the reporting odds ratio and its 95% confidence interval) for antidepressants in general; examine and rank type 2 diabetes signals between the different pharmacological classes of antidepressants and the different antidepressants (58 in total). Second, we performed linear regression analyses to explore the association between the type 2 diabetes signal ranked between antidepressants and their binding affinities for nine targets (serotonin, norepinephrine, dopamine transporters, 5-HT serotonin, D dopamine, α, α adrenergic, M muscarinic and H histamine receptors).
A significant type 2 diabetes signal was found for antidepressants in general, three classes of antidepressants (tricyclic antidepressants, serotonin reuptake inhibitors and "other" antidepressants) and 15 individual antidepressants in particular. Among the antidepressants, three serotonin reuptake inhibitors [escitalopram (adjusted reporting odds ratio 1.15 [1.07-1.25]), paroxetine (1.15 [1.07-1.23]), sertraline (1.23 [1.17-1.31])] and three "other" antidepressants [duloxetine (1.15 [1.07-1.23]), trazodone (1.20 [1.09-1.32]), venlafaxine (1.15 [1.08-1.23])] were the antidepressants most frequently reported with type 2 diabetes. We found a significant correlation between the type 2 diabetes signal and serotonin transporter affinity (slope = 0.14 [0.06-0.23], p = 0.003, R = 0.43) but not the other targets.
The present study suggests a potential role for serotonin transporter in antidepressant-induced type 2 diabetes.
抗抑郁药暴露与 2 型糖尿病之间的关联仍存在争议。此外，其药理学机制尚不清楚。
本研究旨在通过“药物流行病学-药效学”的方法，利用抗抑郁药药理学靶点来研究这种潜在的关系。
首先，我们在世界卫生组织国际药品不良反应监测数据库（VigiBase）中进行病例/非病例分析，以检查抗抑郁药一般情况下 2 型糖尿病报告增加的信号（表示为报告比值比及其 95%置信区间）；检查和排列抗抑郁药不同药物类别和不同抗抑郁药（共 58 种）之间的 2 型糖尿病信号。其次，我们进行线性回归分析，以探讨在抗抑郁药之间排列的 2 型糖尿病信号与其对九个靶点（5-羟色胺、去甲肾上腺素、多巴胺转运体、5-HT 血清素、D 多巴胺、α、α 肾上腺素能、M 毒蕈碱和 H 组胺受体）的结合亲和力之间的关联。
发现抗抑郁药一般、三类抗抑郁药（三环类抗抑郁药、选择性 5-羟色胺再摄取抑制剂和“其他”抗抑郁药）和 15 种特定抗抑郁药存在显著的 2 型糖尿病信号。在抗抑郁药中，三种选择性 5-羟色胺再摄取抑制剂[依他普仑（调整后报告比值比 1.15[1.07-1.25]）、帕罗西汀（1.15[1.07-1.23]）、舍曲林（1.23[1.17-1.31]）]和三种“其他”抗抑郁药[度洛西汀（1.15[1.07-1.23]）、曲唑酮（1.20[1.09-1.32]）、文拉法辛（1.15[1.08-1.23]）]与 2 型糖尿病报告最频繁相关。我们发现 2 型糖尿病信号与 5-羟色胺转运体亲和力之间存在显著相关性（斜率=0.14[0.06-0.23]，p=0.003，R=0.43），但与其他靶点无相关性。
本研究提示 5-羟色胺转运体在抗抑郁药引起的 2 型糖尿病中可能起作用。
