We read with great interest the article written by Fani et al. on the association between Alzheimer’s disease and Helicobacter pylori infection studied on a population-based cohort (Rotterdam study) [1]. The authors did not find any association between the two diseases, that is, H. pylori infection was not associated with an increased dementia incidence after a mean follow-up of 10 years in more than 4000 participants. As the authors mentioned in their article, we previously carried out a similar study (PAQUID—“Personnes Ag ees QUID” study) on 603 noninstitutionalized individuals aged 65 years and older living in the southwest of France followed from 1989 to 2008 with opposite conclusions [2]. Apart from the plausible causes mentioned by Fani et al. explaining this discrepancy, for example, the difference in infection prevalence, mean age of the study population, or strain virulence, an alternative explanation can be provided. Indeed, although H. pylori infection was diagnosed by ELISA technique in the Rotterdam study, in our study, we combined the same ELISA (Pyloriset EIA-G III ELISA; Orion Diagnostica, Espoo, Finland) with an immunoblot (HELICOBLOT 2.1, Genelabs Diagnostics, Singapore) able to detect long-lasting antibodies [3,4]. Immunoblot is able to detect the presence of antibodies to H. pylori proteins including cytotoxinassociated antigen A [CagA], which is a virulent factor and triggers a strong immune response. Antibodies to CagA may prevail longer after eradication (more than 10 years in some of cases) [5], in contrast to the antibodies detected by ELISAwhich may disappear 1 year or more after an eventual eradication or after development of extensive atrophy (gastric atrophy may be associated with a disappearance of H. pylori) [6,7]. In the PAQUID study, we considered patients with a positive ELISA or immunoblot as infected: 27.6% of “infected patients” were ELISA negative and immunoblot positive. Among immunoblot-positive subjects, 87% were CagA positive. When considering only the subjects with a positive ELISA as “infected” (as in Fani et al.), the risk of developing dementia over the 20-year follow-up dropped from 1.46 (95% confidence interval 1.01–2.11) to 1.27

中文翻译：

---

检测当前和先前的幽门螺杆菌感染对于评估其对痴呆症的影响很重要

我们饶有兴趣地阅读了 Fani 等人撰写的文章。关于阿尔茨海默病与幽门螺杆菌感染之间的关联，基于人群的队列研究（鹿特丹研究）[1]。作者没有发现这两种疾病之间有任何关联，也就是说，在对 4000 多名参与者进行平均 10 年的随访后，幽门螺杆菌感染与痴呆发病率增加无关。正如作者在他们的文章中提到的，我们之前对居住在法国西南部的 603 名年龄在 65 岁及以上的非机构化个人进行了一项类似的研究（PAQUID——“Personnes Ag ees QUID”研究），从 1989 年到 2008 年得出相反的结论[ 2]。除了 Fani 等人提到的可能的原因。解释这种差异，例如，感染率的差异，研究人群的平均年龄或菌株毒力，可以提供另一种解释。事实上，虽然在鹿特丹研究中通过 ELISA 技术诊断了幽门螺杆菌感染，但在我们的研究中，我们将相同的 ELISA（Pyloriset EIA-G III ELISA；Orion Diagnostica，Espoo，Finland）与免疫印迹（HELICOBLOT 2.1，Genelabs Diagnostics , Singapore) 能够检测持久抗体 [3,4]。免疫印迹能够检测幽门螺杆菌蛋白抗体的存在，包括细胞毒素相关抗原 A [CagA]，它是一种毒力因子，可触发强烈的免疫反应。CagA 抗体在根除后可能会持续更长时间（在某些情况下超过 10 年）[5]，相比之下，ELISA 检测到的抗体可能会在最终根除或出现广泛萎缩后 1 年或更长时间消失（胃萎缩可能与幽门螺杆菌消失有关）[6,7]。在 PAQUID 研究中，我们将 ELISA 或免疫印迹阳性的患者视为感染：27.6% 的“感染患者”为 ELISA 阴性和免疫印迹阳性。在免疫印迹阳性的受试者中，87% 为 CagA 阳性。当仅将 ELISA 阳性的受试者视为“感染”（如 Fani 等人）时，20 年随访期间患痴呆症的风险从 1.46（95% 置信区间 1.01-2.11）降至 1.27 6% 的“感染患者”为 ELISA 阴性和免疫印迹阳性。在免疫印迹阳性的受试者中，87% 为 CagA 阳性。当仅将 ELISA 阳性的受试者视为“感染”（如 Fani 等人）时，20 年随访期间患痴呆症的风险从 1.46（95% 置信区间 1.01-2.11）降至 1.27 6% 的“感染患者”为 ELISA 阴性和免疫印迹阳性。在免疫印迹阳性的受试者中，87% 为 CagA 阳性。当仅将 ELISA 阳性的受试者视为“感染”（如 Fani 等人）时，20 年随访期间患痴呆症的风险从 1.46（95% 置信区间 1.01-2.11）降至 1.27