Department of Psychiatry, University Hospital, Angers, France.
Centre Inserm U1219 Bordeaux Population Health, CIC1401-EC, Institut de Santé Publique, d'Epidémiologie et de Développement, Université de Bordeaux, CHU de Bordeaux, Pôle Santé Publique, Bordeaux, France.
Neuropsychopharmacology. 2022 Apr;47(5):1114-1120. doi: 10.1038/s41386-021-01246-5. Epub 2021 Dec 10.
Recent evidence suggests an association between benzodiazepines (BZDs) use and lower brain amyloid load, a hallmark of AD pathophysiology. Other AD-related markers include hippocampal atrophy, but the effect of BZDs on hippocampal volume remains unclear. We aimed at 1) replicating findings on BZDs use and brain amyloid load and 2) investigating associations between BZDs use and hippocampal volume, in the MEMENTO clinical cohort of nondemented older adults with isolated memory complaint or light cognitive impairment at baseline. Total Standardized Uptake Value Ratio (SUVR) of brain amyloid load and hippocampal volume (HV) were obtained, respectively, from F Florbetapir positron emission tomography (PET) and magnetic resonance imaging (MRI), and compared between BZD chronic users and nonusers using multiple linear regressions adjusted for age, sex, educational level, ApoE ε4 genotype, cognitive and neuropsychiatric assessments, history of major depressive episodes and antidepressant intake. BZD users were more likely to manifest symptoms of depression, anxiety and apathy. In the MRI subgroup, BZD users were also more frequently females with low education and greater clinical impairments as assessed with the clinical dementia rating scale. Short- versus long-acting BZDs, Z-drugs versus non-Z-drugs BZDs, as well as dose and duration of BZD use, were also considered in the analyses. Total SUVR and HV were significantly lower and larger, respectively, in BZD users (n = 38 in the PET subgroup and n = 331 in the MRI subgroup) than in nonusers (n = 251 in the PET subgroup and n = 1840 in the MRI subgroup), with a medium (Cohen's d = -0.43) and low (Cohen's d = 0.10) effect size, respectively. Short-acting BZDs and Z-drugs were more significantly associated with larger HV. We found no effect of dose and duration of BZD use. Our results support the involvement of the GABAergic system as a potential target for blocking AD-related pathophysiology, possibly via reduction in neuronal activity and neuroinflammation. Future longitudinal studies may confirm the causal effect of BZDs to block amyloid accumulation and hippocampal atrophy.
最近的证据表明，苯二氮䓬类药物（BZDs）的使用与大脑淀粉样蛋白负荷降低有关，这是 AD 病理生理学的一个标志。其他与 AD 相关的标志物包括海马体萎缩，但 BZDs 对海马体体积的影响尚不清楚。我们的目的是 1）复制 BZDs 使用与大脑淀粉样蛋白负荷之间的发现，以及 2）在 MEMENTO 临床队列中，对基线时有孤立性记忆主诉或轻度认知障碍的非痴呆老年人，调查 BZDs 使用与海马体体积之间的关联。通过氟[18F]氟比他滨正电子发射断层扫描（PET）和磁共振成像（MRI），分别获得大脑淀粉样蛋白负荷和海马体体积（HV）的总标准化摄取值比（SUVR），并使用多线性回归比较 BZD 慢性使用者和非使用者之间的差异，回归模型调整了年龄、性别、教育水平、ApoE ε4 基因型、认知和神经心理学评估、重性抑郁发作史和抗抑郁药使用情况。BZD 使用者更有可能出现抑郁、焦虑和冷漠的症状。在 MRI 亚组中，BZD 使用者也更常为女性，教育程度较低，临床评估中的认知和神经心理障碍程度更严重。在分析中还考虑了短作用和长作用 BZDs、Z 药物与非 Z 药物 BZDs、BZD 的剂量和使用时间。与非使用者（PET 亚组 n=251，MRI 亚组 n=1840）相比，BZD 使用者（PET 亚组 n=38，MRI 亚组 n=331）的总 SUVR 更低，HV 更大，差异具有中等（Cohen's d=-0.43）和低（Cohen's d=0.10）效应大小。短作用 BZDs 和 Z 药物与更大的 HV 更显著相关。我们没有发现 BZD 使用剂量和时间的影响。我们的结果支持 GABA 能系统作为阻断 AD 相关病理生理学的潜在靶点的作用，可能是通过降低神经元活动和神经炎症。未来的纵向研究可能会证实 BZDs 阻断淀粉样蛋白积累和海马体萎缩的因果效应。
